Handbook of Microdialysis, Volume 16: Methods, Applications and Perspectives (Handbook of Behavioral Neuroscience)

List of Contributors Numbers in parenthesis indicate the pages on which the authors’ contribution begins Albert Adell (5...

Author: Ben HC Westerink | Thomas I.F.H. Cremers

34 downloads 727 Views 11MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

List of Contributors Numbers in parenthesis indicate the pages on which the authors’ contribution begins Albert Adell (527), Institut d’Investigacions Biome`diques de Barcelona (CSIC), IDIBAPS, Department of Neurochemistry and Neuropharmacology, Rossello 161, 08036 Barcelona, Spain Malin Andersson (251), Department of Biochemistry, Mass Spectrometry Research Center, Vanderbilt University School of Medicine, 465 21st Ave S, 9160 Medical Research Building III, Nashville, TN 37232, USA Francesc Artigas (527), Institut d’Investigacions Biomediques de Barcelona, Consejo Superior de Investigaciones Cientiﬁcas (CSIC), IDIBAPS, Department of Neurochemistry, Rossello 161, 6th ﬂoor, 08036 Barcelona, Spain Franc- ois Artru (659), Service d’Anesthe´sie-Re´animation, Groupement Hospitalier Est, Hospices Civils de Lyon, 59 Bd Pinel, 69677 Bron Cedex, France Nicole M. Avena (351), Department of Psychology, Princeton University, Princeton, NJ 08544, USA David A. Baker (33), Department of Biomedical Sciences, Marquette University, Schroeder Health Sciences Complex, Suite 426, PO Box 1881, Milwaukee, WI 53233, USA R. Billiras (485), Institut de Recherches Servier, Department of Psychopharmacology, 125 Chemin de Ronde, 78290 Croissy sur Seine, Paris, France Ernst Brodin (473), Pharmacological Pain Research, Department of Physiology and Pharmacology, Karolinska Institutet, SE-17177 Stockholm, Sweden Martin Brunner (625), Department of Clinical Pharmacology, Medical University of Vienna, Waehringer Guertel 18–20, A-1090 Vienna, Austria Peter M. Bungay (131), National Institutes of Health, Division of Bioengineering & Physical Science, Ofﬁce of Research Services, Building 13/3N17, Bethesda, MD 20892-5766, USA Richard M. Caprioli (251), Department of Biochemistry, Mass Spectrometry Research Center, Vanderbilt University School of Medicine, 465 21st Ave S, 9160 Medical Research Building III, Nashville, TN 37232, USA Chandra S. Chaurasia (513), Food and Drug Administration, Division of Bioequivalence, Rockville, MD 20855, USA Vladimir I. Chefer (131), National Institute on Drug Abuse, National Institutes of Health, Integrative Neuroscience Section, Behavioral Neuroscience Branch, Baltimore, MD 21224, USA Kevin C. Chen (47), Department of Chemical & Biomedical Engineering, Florida State University, Rm B337, 2525 Pottsdamer Street, Tallahassee, FL 32310-6046, USA Belinda W.Y. Cheung (601), Department of Pharmaceutics, College of Pharmacy, University of Minnesota, Minneapolis, MN 55455, USA L. Cistarelli (485), Institut de Recherches Servier, Department of Psychopharmacology, 125 Chemin de Ronde, 78290 Croissy sur Seine, Paris, France Thomas I.F.H. Cremers (17), Department of Biomonitoring and Sensoring,University of Groningen, University Center for Pharmacy, A. Deusinglaan 1, 9713 AV Groningen, The Netherlands Fre´de´ric Dailler (659), Service d’Anesthe´sie-Re´animation, Groupement Hospitalier Est, Hospices Civils de Lyon, 59 Bd Pinel, 69677 Bron Cedex, France v

vi

Robert L. Dedrick (131), National Institutes of Health, Division of Bioengineering & Physical Science, Ofﬁce of Research Services, Building 13/3N17, Bethesda, MD 20892-5766, USA Elizabeth C.M. de Lange (545), Leiden/Amsterdam Center for Drug Research, Leiden University, Division of Pharmacology, Gorleaus Laboratories, Einsteinweg 55, PO Box 9502, 2300 RA Leiden, The Netherlands Maria A. de Souza Silva (71), Institute of Physiological Psychology, Heinrich-Heine University of Du¨sseldorf, Universita¨tsstr. 1, 40225 Du¨sseldorf, Germany B. Di Cara (485), Institut de Recherches Servier, Department of Psychopharmacology, 125 Chemin de Ronde, 78290 Croissy sur Seine, Paris, France Matthijs G.P. Feenstra (317), Netherlands Institute for Neurosciences, Meibergdreef 33, 1105 AZ Amsterdam ZO, The Netherlands Marianne Fillenz (455), Department of Physiology, University of Oxford, Parks Road, Oxford, OX1 3PT, UK Raul R. Gainetdinov (399), Department of Cell Biology, Duke University Medical Center, Box 3287, Durham, NC 27710, USA Maria G. Giovannini (399), Department of Pharmacology, University of Florence, Viale Pieraccini 6, 50139 Florence, Italy S. Girardon (485), Institut de Recherches Servier, Department of Psychopharmacology, 125 Chemin de Ronde, 78290 Croissy sur Seine, Paris, France A. Gobert (485), Institut de Recherches Servier, Department of Psychopharmacology, 125 Chemin de Ronde, 78290 Croissy sur Seine, Paris, France Margareta Hammarlund-Udenaes (589), Division of Pharmacokinetics and Drug Therapy, Department of Pharmaceutical Biosciences, Uppsala University, Box 591, 751 24 Uppsala, Sweden Luis Herna´ndez (267, 351), Escuela de Medicina, Laboratorio de Fisiologia de la Conducta, Universidad de los Andes, Me´rida 5101, Venezuela Bart G. Hoebel (351), Department of Psychology, Princeton University, Princeton, NJ 08544, USA K. Hoogenberg (645), Martini Hospital, Department of Internal Medicine, PO Box 30033, 9700 RM Groningen, The Netherlands Kirsten D. Huinink (217), Department of Psychiatry University Medical Centre Groningen, University of Groningen, Hanzeplein 1, PO Box 30.001, 9700 RB Groningen, The Netherlands Joseph P. Huston (71), Institute of Physiological Psychology, Heinrich-Heine University of Du¨sseldorf, Universita¨tsstr. 1, 40225 Du¨sseldorf, Germany Peter W. Kalivas (33), Department of Neurosciences, Medical University of South Carolina, BSB-403, 173 Ashley Ave, Charleston, SC 29425, USA Jan Kehr (111), Karolinska Institutet, Department of Physiology and Pharmacology, Nanna Svartz va¨g 2, 17177 Stockholm, Sweden Robert T. Kennedy (279), Department of Chemistry, University of Michigan, 930 N. University Ave, Ann Arbor, MI 48109-1055, USA Peter T. Kissinger (169), Bioanalytical Systems Inc. (BASi), 2701 Kent Avenue, West Lafayette, IN 479061350, USA Jakob Korf (217), Department of Psychiatry, University Medical Centre Groningen, University of Groningen, Hanzeplein 1, 9713 GZ Groningen, The Netherlands Thomas Lieutaud (659), Service d’Anesthe´sie-Re´animation, Groupement Hospitalier Est, Hospices Civils de Lyon, 59 Bd Pinel, 69677 Bron Cedex, France Astrid C.E. Linthorst (301), Henry Wellcome Laboratories for Integrative Neuroscience and Endocrinology, University of Bristol, Dorothy Hodgkin Building, Whitson Street, Bristol, BS1 3NY, UK

vii

Adrian C. Michael (93), Department of Chemistry, University of Pittsburgh, 219 Parkman Ave, Pittsburgh, PA 15260, USA Yvette Michotte (233, 435), Department of Pharmaceutical Chemistry, Drug Analysis & Drug Information, Research Group Experimental Pharmacology, Vrije Universiteit Brussel, Laarbeeklaan 103, 1090 Brussels, Belgium M.J. Millan (485), Department of Psychopharmacology, Institut de Recherches Servicer, 125 Chemin de Ronde, 78290 Croissy sur Seine, Paris, France Paul F. Morrison (131), National Institutes of Health, Division of Bioengineering & Physical Science, Ofﬁce of Research Services, Building 13/3N17, Bethesda, MD 20892-5766, USA Christian P. Mu¨ller (71), Institute of Physiological Psychology, Heinrich-Heine University of Du¨sseldorf, Universita¨tsstr. 1, 40225 Du¨sseldorf, Germany Markus Mu¨ller (625), Department of Clinical Pharmacology, Medical University of Vienna, Waehringer Guertel 18-20, A-1090 Vienna, Austria George G. Nomikos (183), Amgen Inc., Cambridge Research Center, Neuroscience, One Kendall Square, Bldg 1000, Cambridge, MA 02139, USA T.P. Obrenovitch (201), Pharmacy, School of Life Sciences, University of Bradford, Richmond Road, Bradford, West Yorkshire, UK William T. O’Connor (419), Applied Neurotherapeutics Research Group, School of Biomolecular & Biomedical Sciences, UCD Conway Institute, University College, Dublin Weite H. Oldenziel (17), Department of Biomonitoring and Sensoring, University of Groningen, University Center for Pharmacy, A. Deusinglaan 1, 9713 AV Groningen, The Netherlands Ximena Pa´ez (351), Laboratory of Behavioral Physiology, School of Medicine, Universidad de los Andes, Me´rida 5101, Venezuela F. Panayi (485), Institut de Recherches Servier, Department of Psychopharmacology, 125 Chemin de Ronde, 78290 Croissy sur Seine, Paris, France Daniel Paredes (267), Escuela de Medicina, Laboratorio de Fisiologia de la Conducta, Universidad de los Andes, Me´rida 5101, Venezuela Giancarlo Pepeu (377), Department of Pharmacology, University of Florence, Viale Pieraccini 6, 50139 Florence, Italy Pedro V. Rada (267, 351), Escuela de Medicina, Laboratorio de Fisiologia de la Conducta, Universidad de los Andes, Me´rida 5101, Venezuela Kieran Rea (17), Department of Pharmacology, National University of Ireland, Galway Bernard Renaud (659), Service de Biochimie, Hoˆpital Neurologique, Groupement Hospitalier Est, Hospices Civils de Lyon, 59 Bd Pinel, 69677 Bron Cedex, France and Plateforme de Neurochimie Fonctionnelle, Faculte´ de Pharmacie, Universite´ Claude Bernard et Institut Fe´de´ratif des Neurosciences de Lyon, 8 avenue Rockefeller, 69373 Lyon Cedex 08, France J.M. Rivet (485), Institut de Recherches Servier, Department of Psychopharmacology, 125 Chemin de Ronde, 78290 Croissy sur Seine, Paris, France Hans Rollema (513), Pﬁzer Global Research and Development, Department of Neuroscience, Groton Laboratories, MS 8220-4159, Eastern Point Road, Groton, CT 06340, USA Sergio Rossell (267), Escuela de Medicina, Laboratorio de Fisiologia de la Conducta, Universidad de los Andes, Me´rida 5101, Venezuela Elham Rostami (675), Karolinska Institutet, Department of Physiology and Pharmacology, Nanna Svartz Rd 2, 171 77 Stockholm, Sweden Sophie Sarre (233), Department of Pharmaceutical Chemistry, Drug Analysis & Drug Information, Research Group Experimental Pharmacology, Vrije Universiteit Brussel, Laarbeeklaan 103, B-1090 Brussels, Belgium

viii

Ronald J. Sawchuk (601), Department of Pharmaceutics, College of Pharmacy, University of Minnesota, Weaver-Densford Hall, Room 9-143, 308 Harvard St. SE, Minneapolis, MN 55455, USA A.J.M. Schoonen (645), Biomonitoring & Sensoring, Department of Pharmacy, University of Groningen, Antonius Deusinglaan 1, 9713 AV Groningen, The Netherlands Trevor Sharp (5), Department of Pharmacology, University of Oxford, Mansﬁeld Road, Oxford OX1 3QT, UK Ilse Smolders (435), Department of Pharmaceutical Chemistry, Drug Analysis & Drug Information, Research Group Experimental Pharmacology, Vrije Universiteit Brussel, Laarbeeklaan 103, 1090 Brussels, Belgium Tatyana D. Sotnikova (399), Department of Cell Biology, Duke University Medical Center, Box 3287, Durham, NC 27710, USA Carl-Olav Stiller (473), Clinical Pharmacology, Department of Medicine, Karolinska University Hospital, SE-17177, Stockholm, Sweden Bradley K. Taylor (473), Department of Pharmacology, Tulane University Health Sciences Center, 1430 Tulane Ave, New Orleans, LA 70112-2699, USA Tung-Hu Tsai (573), Institute of Traditional Medicine, School of Medicine, National Yang-Ming University, 155 Li-Nong Street Section 2, Taipei 112, Taiwan Sonia Tucci (267), Escuela de Medicina, Laboratorio de Fisiologia de la Conducta, Universidad de los Andes, Me´rida 5101, Venezuela Urban Ungerstedt (675), Karolinska Institutet, Department of Physiology and Pharmacology, Nanna Svartz Rd 2, 171 77 Stockholm, Sweden J. Urenjak (201), Pharmacy, School of Life Sciences, University of Bradford, Richmond Road, Bradford, West Yorkshire, BD7 1DP, UK Ben H.C. Westerink (17), Department of Biomonitoring and Sensoring, University of Groningen, University Center for Pharmacy, Antonius Deusinglaan 1, 9713 AV Groningen, The Netherlands K.J.C. Wientjes (645), Groningen Research Institute of Pharmacy, University of Groningen, Antonius Deusinglaan 1, 9713 AV Groningen, The Netherlands Agustin Zapata (131), National Institute on Drug Abuse, National Institutes of Health, Integrative Neuroscience Section, Behavioral Neuroscience Branch, Baltimore, MD 21224, USA Tyra Zetterstro¨m (5), Department of Pharmacology, School of Pharmacy, De Montfort University, Leicester, UK

Preface

It is now about 25 years ago that Swedish researchers introduced microdialysis. The method represents an elegant technique based on a simple principle: the extracellular space is sampled via a dialysis membrane in a needle-like probe. Meanwhile, microdialysis has become a routine practice in numerous neuroscience laboratories and over the years many scientists have contributed to the development of the technique. We invited a number of them to summarize the significance of microdialysis for their particular ﬁeld. The fascinating aspect of microdialysis is that it provides a direct look into extracellular compartments of living animals or humans. During the ﬁrst years of its application, various ‘‘classical neurotransmitters’’ such as dopamine, serotonin, acetylcholine, GABA and glutamate were investigated with great enthusiasm and creativity. Throughout this volume much of these efforts are summarized. A critical evaluation of the physiological significance of compounds monitored by microdialysis is constantly needed. The ﬁrst series of chapters in this volume deal with the speciﬁcity and interpretation of microdialysis data. Typical microdialysis issues are not ignored, such as the never ending discussions on the ‘‘recovery’’ and ‘‘true’’ extracellular concentrations of the sampled endogenous or exogenous compounds. Neurotransmitters are very efﬁciently removed after their release, and as a consequence they appear in the chromatograms as the tiniest peaks. Therefore, quantitation of these compounds puts high demands on the detection limits of analytical methods. Without the development of bioanalytical techniques such as liquid chromatography, electrochemical detection, capillary electrophoresis and liquid chromatography/ mass spectrometry, microdialysis would have no practical value. Several chapters in this volume are therefore dedicated to analytical chemistry aspects. Until now liquid chromatography in conjunction with electrochemical or ﬂuorimetric detection has been the most widely used method. Capillary electrophoresis with laser-induced ﬂuorimetric detection seems a very promising novel technique for the fast detection of small amounts of analytes. Note that the world speed record in microdialysis detection is presently in the hands of Dr. Hernandez’ group who were able to detect glutamate in 1 s samples (Chapter 3.3). Microdialysis combined with liquid chromatography/mass spectrometry is still progressing. This powerful analytical method will connect the probe directly to the world of peptidomics. As microdialysis combined with a guide canula is well tolerated in freely behaving animals, many scientists studied the chemistry of behaviour and have made remarkable progress. Animals that feed, run, mate or crave for drugs of abuse are equipped with probes while their neurotransmitter proﬁle is simultaneously being recorded. Several contributions to this topic are included in this volume. Microdialysis is successfully applied in animal models of parkinsonism, stroke, depression, schizophrenia, etc. It is of great significance that the method is also applicable to genetically modiﬁed mice. Seven chapters deal with the role of microdialysis in drug development and PK-PD. Microdialysis is very suitable to study blood–brain transport mechanisms and protein binding. The method originated in brain research is still mostly focused on the central nervous system, but the number of applications in peripheral systems is also rapidly growing. By simultaneously sampling blood and brain tissue a full pharmacokinetic proﬁle is obtained. The last three chapters of this volume are dedicated to clinical applications. Although bedside microdialysis in not yet routinely done, clinical studies have taught us much about the pathology of the brain ix

x

during conditions of certain neurological diseases and trauma. An interesting example of a clinical application is presented by the development of the human glucose sensor. In the near future more sophisticated sampling methods will certainly be introduced. Soon speciﬁc microsensors will deliver real-time chemical signals from the extracellular compartment with the help of telemetric devices. However, due to its simplicity and versatility, microdialysis will remain an efﬁcient research tool for generations of biochemists and neuroscientists for many years to come. We hope that this book might inspire many scientists to further explore the opportunities of this fascinating technique. Ben H.C. Westerink Thomas I.F.H. Cremers

CHAPTER 1.1

What did we learn from microdialysis? Trevor Sharp1, and Tyra Zetterstro¨m2 1 Department of Pharmacology, University of Oxford, Oxford, UK Department of Pharmacology, School of Pharmacy, De Montfort University, Leicester, UK

2

Abstract: At the beginning of 1980s, microdialysis emerged as a novel method for monitoring brain neurochemistry in vivo. This was followed by a rapid, worldwide uptake of microdialysis and an explosion of papers reporting the application of the technique across a broad range of neuroscience research areas. This article discusses the historical background to in vivo neurochemical monitoring, and sets out the techniques available prior to the ﬁrst microdialysis studies. Detail is then given of some of the events and people involved in one of the laboratories that played a central role in the development of microdialysis and the dissemination of this technique to the wider neuroscience community. Finally, some of the early experiments are outlined and used to illustrate a few of the many contributions that microdialysis has made to present-day neuroscience. neuroscience in which microdialysis has had particular value. First, we set out the background of in vivo neurochemical monitoring prior to the ﬁrst microdialysis studies.

I. Introduction Since its inception and development in the early 1980s, microdialysis has arguably become one of the essential tools of modern neuroscience research. A quick search of recent literature databases reveals that the number of citations for ‘microdialysis/brain’ is comparable with those for ‘RT-PCR/brain,’ or indeed ‘microarray/brain’. Why did the neuroscience community take to this technique so keenly, and persist with it today largely in form unchanged from that developed over 20 years ago? Is this faith in the technique justiﬁed? The following is a personal perspective on the beginning of microdialysis and its emergence as a key neuroscience method. A particular emphasis is placed on its application to study CNS drug action. The reader is referred to other chapters in this book for an update on current developments and controversies relating to the technique, and for a broader perspective on the many areas of

II. Early studies of chemical transmission in the brain in vivo The background to microdialysis started with the studies in the 1950–1960s by groups of physiologists and pharmacologists, including Feldberg, Gaddum, Voigt and Mitchell, and their attempts to monitor chemical transmission in the living brain during a physiological event or in response to a drug challenge. Similar pioneering experiments by these workers and others (Feldberg and Gaddum, 1934; Dale et al., 1936) had already laid down the basic principles of chemical transmission at synapses in the periphery, and linked changes in transmitter secretion to functional outputs. The challenge was to carry these concepts forward to the complex neural networks of the brain. Many of these early studies investigated acetylcholine release because there was already a clear

Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

5

DOI: 10.1016/S1569-7339(06)16001-4 Copyright 2007 Elsevier B.V. All rights reserved

6

understanding of this messenger’s properties in peripheral neurones and there were useful bioassays available. Also, at that point of time, evidence of the huge diversity of chemical messengers in the brain was only just beginning to emerge. These early efforts to study chemical transmission in the brain principally utilised three approaches to detect transmitter secretion, all essentially based on perfusion of the extracellular space and chemical analysis of the perfusates: ventricular perfusion or sampling, cortical cup perfusion, and push–pull perfusion. Excellent reviews of these techniques and details of their early application can be found elsewhere (Myers, 1972; Moroni and Pepeu, 1984; Philippu, 1984). In early studies, perfusion of the cat’s ventricular system was achieved using a cannula inserted into one part of the system (typically lateral ventricles) and collection of the efﬂuent using a cannula inserted into another part (e.g. aqueduct) (Bhattacharya and Feldberg, 1958; Carmichael et al., 1964). Chemical analysis of the ventricular contents, sometimes simply through repeated collection of the cerebrospinal ﬂuid, remained in common use until the 1980s and extended to monoamines and other molecules (Tilson and Sparber, 1972; Elghozi et al., 1983; Sarna et al., 1984). Ventricular perfusion/sampling is a relatively non-invasive method to study brain neurochemistry in vivo, it has utility in large laboratory animals as well as small ones, and it has a clear translational value in that analogous measurements (cerebrospinal ﬂuid sampling) can be carried out in humans. However, by its very nature this technique provides little direct information about chemical transmission in speciﬁc brain areas, which is obviously important in the context of understanding the complexities of brain function. The cortical cup technique enables sampling of chemicals released on the surface of the cerebral cortex (Moroni and Pepeu, 1984). This was achieved by placing a cylinder over the exposed cerebral cortex to form a small cup that was ﬁlled with physiological ﬂuid, which was replaced and then analysed at regular intervals. Early studies demonstrated the utility of this method in sheep, cats and rabbits (Mitchell, 1963; Collier and Mitchell, 1966), largely in anaesthetised

preparations but also under non-anaesthetised conditions. Careful steps had to be taken to exclude contamination of the perfusion ﬂuid by chemicals arising from blood, and methodological reﬁnements allowed application of the technique in rats (Aquilonius et al., 1972). However, as with ventricular perfusion/sampling, a drawback of the cortical cup technique is the limited scope for regional brain analysis. Perfusion of speciﬁc brain areas was achieved using an open-ended arrangement of two concentric cannulae that allowed perfusion ﬂuid to be ‘pushed’ into a particular brain area and then ‘pulled’ out to allow chemical analysis of the perfusates (see Philippu, 1984). The application of ‘push–pull’ perfusion to the brain has its origins in the works of Gaddum (1961) and others, in the 1960s, who adapted a system applied to subcutaneous tissue (Fox and Hilton, 1958), and the method played a dominant role in the ﬁeld of in vivo neurochemistry for the next 25 years. As with ventricular perfusion and the cortical cup technique, studies with push–pull perfusion were focused on larger laboratory animals, often anaesthetised. Application in smaller animals was reported, but careful technical reﬁnements of the cannula construction and perfusion conditions were necessary to minimise tissue disturbance from the open-ended system of perfusion (Myers et al., 1998). Studies utilising push–pull perfusion laid the foundations in many areas of in vivo neurochemical research that subsequently became a focus of microdialysis studies, including basic properties of transmitter release (Cheramy et al., 1981) and functional interactions between transmitter pathways (Glowinski et al., 1978; Philippu, 1984; Soubrie et al., 1984). The application of the above techniques was, however, often hampered by the challenge of detecting the trace amounts of transmitters in the perfusates using the relatively insensitive analytical methods available at the time. Transmitter detection was often at the limits of traditional bioassays, and reliable analysis was sometimes possible only through use of preloaded radio-labelled precursors or time-consuming radio-enzymatic methods. The eventual combination of push–pull perfusion with gas chromatography coupled to a

7

mass spectrometer provided the necessary sensitivity and speciﬁcity (e.g. Wolfensberger, 1984), but the application of this analytical technique clearly required specialist know-how and high start-up costs, which were available to only a few. Commencing in the 1970s, largely through the pioneering work of Adams and colleagues (McCreery et al., 1974; Conti et al., 1978), voltammetry offered a quite different approach to in vivo measurement of chemical transmission in the brain. This approach utilises various voltammetric scanning techniques to monitor levels of transmitter in the close vicinity of an implanted microelectrode. Although these early studies were restricted to measurement of those transmitter species that are electroactive, and specifically the monoamines (dopamine, noradrenaline and 5-HT), this technique held the promise of continuous online monitoring of transmitter secretion with high time resolution and anatomical speciﬁcity, and without the tissue disturbance associated with perfusion methods. However, despite initial enthusiasm it soon became apparent that the early voltammetric signals monitored in vivo comprised several chemical species, mostly monoamine metabolites and ascorbic acid. Therefore, at the end of the 1970s, it was recognised that significant advances in electrode design and voltammetric methodology would be needed to detect the monoamine transmitters themselves. As an alternative to direct in vivo monitoring of chemical transmission, neurochemical measurements of transmitter utilisation in post-mortem brain evolved during the 1960s to provide an index of transmitter activity immediately prior to removal of the tissue. This approach included ex vivo measurement of steady-state levels of transmitters and their metabolites, activity of key enzymes and rate of transmitter ﬂux through biosynthetic or metabolic pathways. Measurements of transmitter utilisation were obtained within discrete brain regions across a range of species, and for a range of transmitters including catecholamines (Carlsson and Lindqvist, 1963; Roth et al., 1974; Westerink and Korf, 1976), 5-HT (Neff and Tozer, 1968; Fuller et al., 1974), GABA (Bertilsson et al., 1977) and acetylcholine (Racagni et al., 1976).

Studies of this sort, which were especially popular in the 1970s, laid down the ﬁrst understanding of neurochemical mechanisms involved in transmitter utilisation in the brain and were applied to great effect to investigate the effect of physiological and pharmacological manipulations. However, measurement of transmitter utilisation, by its very nature, provides an indirect measure of transmitter released from the neurones. Also, complex experimental protocols and large numbers of animals were required to obtain a time resolution of the order of 10–20 min. Therefore, by the end of the 1970s, in vivo neurochemical techniques were available to study transmitter function in key brain structures of laboratory animals, but researchers were unable to approach many important questions because of the difﬁculties in routinely detecting the small amounts of transmitter released. In addition, the perfusion techniques were generally too invasive or too cumbersome for application in small laboratory animals (rats and mice), which at the time were the model of choice of many of the rapidly growing areas of in vivo neuroscience including neuropharmacology, functional neuroanatomy and especially behavioural science (for which there was also an obvious need to work without anaesthesia). On the other hand, voltammetry was much less invasive and could be used in small animals but the use of this technique was limited to a small number of electroactive transmitters, and at the time there were major uncertainties about the identity of the chemical species being monitored. In short, at the beginning of the 1980s, there was a strong need for an in vivo neurochemical technique that had the broad application and regional selectivity of perfusion methods, that could be linked to analytical methods with high sensitivity and selectivity and that was applicable to small laboratory animals.

III. The introduction of microdialysis There are various early descriptions of techniques in which the content of the brain extracellular space was sampled using probes constructed from semipermeable dialysis membranes. One of the earliest

8

involved a primitive dialysis bag that was temporarily implanted into the ventricular system of the dog, and then withdrawn for analysis of the bag contents (Bito et al., 1966). Of the many types of brain perfusion cannulae described, one consisted of two parallel cannulae and two electrodes called the ‘chemitrode’ developed by Delgado in Madrid (Delgado, 1962), and a variant of this, ‘the dialytrode’, which had a tip covered by a semipermeable membrane (Delgado et al., 1972; Cornish and Hall, 1979). The latter was not strictly a perfusion probe in that samples were withdrawn from the probe at regular intervals. In Stockholm, Ungerstedt had also picked up on the idea of designing perfusion probes constructed from a tube of semipermeable membrane (Ungerstedt and Pycock, 1974), and was specifically attracted by the concept of using such probes to act as an ‘artiﬁcial blood vessel’ to sample the brain extracellular space. Ungerstedt had already made seminal contributions to the anatomical mapping of dopamine pathways and modelling their function in behavioural paradigms (Ungerstedt, 1971), and he was now pressed by the idea of developing a technique capable of detecting dopamine release in the living brain. The attraction of perfusion probes based on a dialysis membrane was that there would be no direct contact between the brain tissue and the perfusion ﬂuid. In theory this would circumvent any tissue damage caused by constant washing of tissue by the perfusion ﬂuid, which occurred with openended perfusion cannulae, and at the same time provided an online clean-up of the samples of extracellular ﬂuid. It was also hoped that the neurones and nerve terminals in the vicinity of the dialysis membrane remained bathed in ‘normal’ extracellular ﬂuid rather than exogenous artiﬁcial media. Ungerstedt’s initial studies with probes constructed from thin dialysis membranes were hampered by difﬁculties in probe reliability and reproducibility, as well as difﬁculties in dopamine detection (Ungerstedt and Pycock, 1974). However, two events dramatically changed the situation. First, an intake of new graduate students (Tyra Zetterstro¨m and Ulf Tossman) in 1979 stimulated advances in probe design. Although simple in construction, the ‘dialysis-loop probe’ and ‘transcranial probe’ allowed problem-free perfusion of discrete

brain areas in a small laboratory animal (speciﬁcally the rat). Moreover, the Stockholm group found that animals could be perfused in the absence of anaesthesia by connecting the probes to a liquid swivel and using a harness to restrain the animal within the conﬁnes of a circular Perspex chamber. This step forward in probe construction was coincident with a second key development, the introduction into the Stockholm laboratory of a newly available HPLC with electrochemical detection assay, which had sufﬁcient sensitivity to detect dopamine in the ‘dialysates’. This analytical technique had developed spectacularly in just a few years since its initial description (Kissinger et al., 1973), and was commercially available by the mid-1970s. As a consequence of these advances, within a short period of time, the group began to accumulate dialysis data on dopamine. In addition, as the group developed and expanded with additional students (including Nils Lindefors and Lars Sta˚hle) and technicians (including Agneta Eliasson and Anna Karin Collin who joined A˚se Hallstro¨m), the group’s interests diversiﬁed to include studies of other chemical messengers and also the simultaneous monitoring of transmitter release during behaviour (Tomas Ljungberg and Mario HerreraMarschitz). Although little data had been published at this stage (a manuscript submitted to Nature was returned with a report from a referee which doubted whether one could do useful neuroscience ‘in a salad bowl’), news of the work in Stockholm began to spread, and in 1981 a graduate student of Gaetano Di Chiara in Caligary (Assunta Imperato) visited Stockholm to learn the technique. Data on the brain dialysis technique were presented to an international audience in 1981, ﬁrst at a Congress on dopamine receptor pharmacology in Japan in June (Ungerstedt et al., 1982), followed by an International Summer School of Brain Research at the Royal Netherlands Academy of Sciences in Amsterdam in September, and a few weeks later at a meeting of the European Neuroscience Association in Liege. Also present at the Amsterdam meeting was a graduate student (Trevor Sharp) from Charles Marsden’s group in Nottingham, which was pioneering the neurochemical application of

9

electrochemical detection methods (HPLC and in vivo voltammetry). As a consequence, interactions between Nottingham and Stockholm commenced leading to exchanges in people (Sharp joined the Ungerstedt group), transfer of HPLC know-how (poor-resolution cation-exchange liquid chromatography columns were replaced by high-efﬁciency reverse-phase HPLC columns), the setting up of the dialysis technique in the UK laboratory, and a joint paper reporting initial microdialysis measurements of dopamine (Zetterstro¨m et al., 1983). Numerous interactions between Stockholm and other groups in Sweden (Gothenberg, Lund) and around the world very quickly followed, and Ungerstedt established a spin-off company (Carnegie Medicine from which grew CMA) to facilitate development of the method and to further increase the accessibility. During this period the brain dialysis technique went through several name changes including ‘intracerebral dialysis,’ ‘intracranial dialysis’ and ‘transcranial dialysis’, before general opinion fell in favour of ‘microdialysis’. Ungerstedt was quick to realise the potential of the microdialysis method for studying not only dopamine but also other transmitter systems, and in the early 1980s his group (together with several local collaborators including Ernst Brodin and Bertil Freedholm) was utilising a range of analytical techniques (HPLC with electrochemical and ﬂuorescent detection, radioimmunoassays) to demonstrate the presence in the dialysates of detectable amounts of numerous other chemical messengers including purines (adenosine; Zetterstro¨m et al., 1982), amino acids (GABA, glutamate and aspartate; Tossman et al., 1983) and neuropeptides (Substance P and other tachykinins; Brodin et al., 1983). This interest expanded to include markers of metabolism and hypoxia such as lactate (Hallstro¨m et al., 1989), which provided the basis for the clinical application of microdialysis that Ungerstedt and others developed over the ensuing years (Hillered et al., 1990; Meyerson et al., 1990).

IV. Early microdialysis studies in Stockholm Many of the initial microdialysis studies carried out by the Ungerstedt group focused on basic

aspects of dopamine pharmacology and neurochemistry. A succession of the group’s publications in the early 1980s demonstrated that extracellular dopamine showed dramatic changes in response to administration of a range of drugs. With today’s perspective, most of the observations accorded with predictions but at the time each result was a source of amazement because in most instances it was the ﬁrst occasion that such measurements were possible. The ﬁndings included increases in dopamine in response to amphetamines, inhibitors of dopamine reuptake and metabolism, dopamine precursor loading, and dopamine receptor antagonists, and decreases in response to dopamine receptor agonists and dopamine neurotoxins (Zetterstro¨m et al., 1983, 1984, 1986a, c; Zetterstro¨m and Ungerstedt, 1984; Sharp et al., 1986b). Other emerging microdialysis groups independently obtained similar data (Imperato and Di Chiara, 1984; Kato et al., 1986; Kito et al., 1986; Westerink and Tuinte, 1986; Church et al., 1987; Butcher et al., 1988; Imperato et al., 1988). Many of the early microdialysis experiments in Stockholm were carried out in conscious animals, and this allowed a side-by-side comparison of changes in dopamine with alterations in behaviour, something that had not been achievable with earlier in vivo monitoring methods. An obvious starting point for such experiments was to study amphetamine, a drug with behavioural effects that had long been associated with increased release of dopamine. Using a simple behavioural chamber and home-made harness for securing perfusate collecting tubes (Sharp et al., 1986a), experiments demonstrated a close correlation between, on the one hand, amphetamine-induced behavioural stereotypes and dopamine released in striatum, and on the other hand, amphetamine-induced locomotor stimulation and dopamine released in the nucleus accumbens (Sharp et al., 1987a). These ﬁndings ﬁtted closely the predictions of an earlier literature based on studies of the behavioural effects of dopamine lesions and intracerebral drug injection (Kelly, 1977). At this time some of the ﬁrst microdialysis experiments correlating the behavioural effects of cocaine with changes in extracellular dopamine were also reported (Pettit and Justice, 1989).

10

Separate experiments found a strong correlation between the increase of striatal dopamine induced by amphetamine as well as L-DOPA, and rotation in rats with unilateral dopamine denervation (the ‘Ungerstedt rotation model’) (Zetterstro¨m et al., 1986b). The latter ﬁnding was the ﬁrst direct evidence that L-DOPA increases dopamine output in dopamine denervated tissue, which of course was predicted from the known therapeutic efﬁcacy of L-DOPA in Parkinson’s disease. The Stockholm laboratory in collaboration with Anders Bjorklund’s group in Lund also demonstrated that transplantation of rat foetal mesencephalic neurones into the striatum of dopamine denervated rats restored extracellular dopamine to near normal levels (Zetterstro¨m et al., 1986a; Strecker et al., 1987). Such data played a valuable role in the work by the Lund group and many others that provided a rational basis for the subsequent clinical trials of dopaminergic grafts in Parkinson’s patients. Although attempts in Stockholm to detect changes in dopamine release during the performance of conditioned and non-conditioned behavioural tasks were hampered by technical problems, many groups reported success in subsequent years (Westerink, 1995; see Feenstra, this volume). In the early experiments, it was immediately apparent that dialysate levels of dopamine metabolites (DOPAC and HVA) were up to 100 times greater than dopamine (Zetterstro¨m et al., 1983). This ﬁnding had clear implications for voltammetry studies using electrodes that could not readily discriminate between dopamine and the metabolites. Indeed, a side-by-side comparison of voltammetry (specifically, differential pulse voltammetry) and microdialysis helped reveal that DOPAC and not dopamine was a major contributor to the voltammetric signal for catechols (Sharp et al., 1984). Findings of high levels of uric acid in the dialysates (Zetterstro¨m et al., 1983) led to the discovery that another voltammetric signal, previously attributed to indoles (5-HT and its metabolites), was in fact contaminated by this purine metabolite (Crespi et al., 1983). Work of this sort, alongside efforts by many others, helped to stimulate advances in voltammetric methodology to increase monoamine selectivity, including the application of surface-modiﬁed electrodes

(Crespi et al., 1988) and the shift towards fast scanning voltammetry with its greater chemical and temporal resolution (Marsden et al., 1988; Kawagoe et al., 1993). Another issue raised by the early studies comparing dopamine and dopamine metabolites was that in many experiments there was a poor correlation between the two. These results were a surprise because the prevailing view was that the metabolites were a good marker of dopamine release as they were produced from recently released and reuptaken dopamine. To account for the ﬁndings a different view of dopamine neurochemistry emerged, in which the metabolites were seen to have originated from the metabolism of cytoplasmic newly synthesised dopamine, before it is released (Commisong, 1985; Zetterstro¨m et al., 1986c, 1988; Butcher et al., 1988). While this theory predicted the metabolism of large amounts of unused dopamine, the dopamine nerve terminal would have access to large amounts of dopamine derived directly from tyrosine hydroxylase under conditions of high demand (i.e. increased dopaminergic activity). The situation envisaged directly paralleled that independently proposed for 5-HT neurones (Grahame-Smith, 1974; Kuhn et al., 1985). These ﬁndings emphasised that the monoamine metabolites provided an indirect measure of monoamine release, and today the metabolites are rarely used for this purpose.

V. Microdialysis and neuropharmacology On the basis of the early microdialysis studies of the effects of pharmacological manipulation of the dopamine system, it was clear that the technique had great potential for the study of drug action and the development of novel therapeutic agents. As the range of transmitters accessible to study through microdialysis rapidly increased, the areas of neuropharmacology in which microdialysis had value quickly broadened to include the study of antidepressant and anxiolytic agents among many others (L’Heureux et al, 1986; Kalen et al, 1988; Pei et al., 1989; Sharp et al., 1989b; Adell and Artigas, 1991; Wright et al., 1992; Tanda et al., 1994). Today microdialysis is a frontline technique

11

in research aimed at the identiﬁcation and characterisation of novel pharmacological strategies for the treatment of common, disabling psychiatric conditions such as mood disorder, anxiety and schizophrenia. It is very likely that the next generations of therapeutic drugs for these illnesses will be strongly supported by microdialysis data. Some of the current exciting areas of neuropharmacology in which microdialysis is playing an important role are reviewed by Millan and Artigas in other chapters of this book. Another area of neuropharmacology in which microdialysis has had a large impact is the study of psychostimulants and other classes of addictive drugs. Indeed, microdialysis data are a cornerstone of current biological theories implicating mesolimbic dopamine (Di Chiara and Imperato, 1988) and more recently cortical amino acid systems in the processes leading up to drug addiction (McFarland and Kalivas, 2001; Hotsenpiller and Wolf, 2002) (see the chapter by Di Chiara). Interestingly, microdialysis studies of the effects of drug challenges in animals are currently assuming a special importance in the guidance and development of exciting new non-invasive PET imaging techniques to monitor transmitter release in humans (Laruelle, 2000). This novel approach is based on radioligand displacement methodology, and studies are using transmitter release data derived from microdialysis studies to interpret displaceability (or lack of it) of speciﬁc radioligands by competing transmitter. The most successful example of this research to date is the use of PET radioligand displacement as a measure of dopamine release, and this approach has translated from animal studies through to investigations in patients with psychiatric and neurological disorders (Laruelle and Huang, 2001). Studies combining microdialysis and PET are ongoing in an attempt to guide the development of radioligand displacement methods to monitor the release of many other transmitters including 5-HT (e.g. Hirani et al., 2003). In addition to PET imaging, microdialysis is likely to play an increasing role in the development and validation of other imaging modalities (functional magnetic resonance imaging, magnetic resonance spectroscopy) to monitor transmitter release.

VI. Microdialysis problems The initial popularity of microdialysis was not without consideration for the shortcomings of the technique, and especially the likely tissue damage induced by probe implantation. Prior to the 1980’s there was an existing literature on the deleterious effects on brain tissue of implanted probes (Edvinsson et al., 1971), and there was a realistic prospect that a proportion of transmitter in the dialysates derived from blood leaked from damaged nerve terminals or arose from non-neuronal cells. A series of studies in the mid-1980s addressed the crucial question of whether, and to what extent, transmitters in the dialysates derived from normally functioning nerve terminals by the application of a range of ‘transmitter criteria’ (Di Chiara, 1990; Westerink, 1995). For the monoamines, dopamine, noradrenaline and 5-HT, and acetylcholine it became clear that under a wide variety of experimental conditions, and with few exceptions, dialysate contents of these transmitters indeed arose predominantly from neuronal secretion (Imperato and Di Chiara, 1984; L’Heureux et al., 1986; Consolo et al., 1987; Kalen et al., 1988; Sharp et al., 1989a; Sharp and Zetterstro¨m, 1992). Specifically, dialysate levels of these transmitters decreased when excitation-secretion coupled release mechanisms were inhibited (low calcium, blockade of voltage-gated sodium channels), and increased when neurones were stimulated. The case was less clear, however, for the principal amino acid transmitters, glutamate and GABA (Bourdelais and Kalivas, 1992; Kehr and Ungerstedt, 1988; Osborne et al., 1990; Smith and Sharp, 1994; Timmerman and Westerink, 1997). Indeed, it is only in recent years that a more complete picture of the origins of these transmitters has emerged, as discussed in detail elsewhere in this volume (Baker et al., 2002; Rea et al., 2005; see also Westerink, this volume). By its very nature, the time resolution of neurotransmission of the brain perfusion technique cannot approach that of neurophysiological recording methods. Nethertheless, even early studies demonstrated that microdialysis sampling intervals of

12

the order of a minute were achievable through application of microbore HPLC (Church and Justice, 1987; Sharp et al., 1987b). Arguably one of the main drawbacks of microdialysis, is the narrow time window of a few days in which transmitter measurements can be made before data interpretation becomes complicated by the proliferation of glial cells around the implanted probe (Imperato and Di Chiara, 1984; Stro¨mberg et al., 1985; Benveniste et al., 1987). This reaction probably alters the chemical milieu of the extracellular space surrounding the microdialysis probe, particularly in relation to amino acid reuptake, synthesis, metabolism and release in which glial cells play important roles. Moreover, the proliferation of glial cells likely acts as a barrier-limiting movement of molecules to and from the microdialysis probe. It remains to be seen whether these problems can be circumvented through further miniaturization of microdialysis probes or alterations in probe materials and perfusion media. Until this situation improves, experiments requiring repeated measurements over a period of time from the same animal (e.g. during behavioural training, chronic drug administration or a progressive neuronal lesion) will remain problematic. Details of the effects of microdialysis probe implantation on brain tissue continue to be documented. One current theory is that monoamine nerve terminals in the immediate vicinity of the probe are damaged, resulting in underestimation of basal extracellular dopamine levels (Borland et al., 2005). This might explain the long-standing ‘thorny issue’ of why basal extracellular levels of dopamine and the other monoamines estimated by microdialysis (nanomolar) are several orders of magnitude lower than those estimated by fast cyclic voltammetry (micromolar). In contrast, other voltammetric modalities (differental pulse voltammetry) produce estimates similar to microdialysis (Crespi et al., 1988), suggesting that we have not yet arrived at the full answer. Perhaps Adams (1990) had the right one, ‘‘There is little merit in arguing as to which technique is giving the more correct values – both have questions about the suitability of their in vitro calibration proceduresy’’.

VII. Conclusions Since its inception in the early 1980s, quite remarkable progress has been made through the application of microdialysis to monitor the neurochemistry of the brain in vivo. On the basis of early experiments in Stockholm and other cities around the world, several things quickly became apparent. Firstly, the technique could be applied in small laboratory animals (rats and eventually mice), which were emerging as a key model system for many areas of neuroscience. Secondly, the number of neurochemicals in the extracellular space that were accessible to microdialysis stretched well beyond dopamine to encompass most other major transmitters and many other key brain metabolites. This offered the possibility to investigate not only chemical aspects of neurotransmission but also many fundamental neurochemical mechanisms in the brain including those relating to cerebral metabolism, haemodynamic control and other energy-dependent processes. Thirdly, neurochemical measurements could be made in conscious animals, which opened up avenues for the investigation of the chemical basis of behaviour that had not existed previously. Finally, and importantly, at least at face value the method had a technical and theoretical simplicity, and reasonably low costs. All of this put the technology well within the grasp of a very wide community of neuroscientists, and unlocked areas of research that previously had been difﬁcult to penetrate and continue to be investigated today. Just a few of the many areas of neuroscience that have gained from this technological advance, and the people that have contributed, are covered herein. References Adams, R.N. (1990) In vivo electrochemical measurements in the CNS. Prog. Neurobiol., 35: 297–311. Adell, A. and Artigas, F. (1991) Effects of clomipramine on extracellular serotonin in the rat frontal cortex. Adv. Exp. Med. Biol., 294: 451–454. Aquilonius, S.M., Lundholm, B. and Windbladh, B. (1972) Effects of some anticholinergics on cortical acetylcholine release and motor activity in rats. Eur. J. Pharmacol., 20: 224–230.

13 Baker, D.A., Xi, Z.X., Shen, H., Swanson, C.J. and Kalivas, P.W. (2002) The origin and neuronal function of in vivo nonsynaptic glutamate. J. Neurosci., 22: 9134–9141. Benveniste, H., Drejer, A., Schousboe, A. and Diemer, N.H. (1987) Regional cerebral glucose phosphorylation and blood ﬂow after insertion of a microdialysis ﬁber through the dorsal hippocampus in the rat. J. Neurochem., 49: 729–734. Bertilsson, L., Mao, C.C. and Costa, E. (1977) Application of principles of steady-state kinetics to the estimation of gamma-aminobutyric acid turnover rate in nuclei of rat brain. J. Pharmacol. Exp. Ther., 200: 277–284. Bhattacharya, B.K. and Feldberg, W. (1958) Perfusion of cerebral ventricles: effects of drugs on outﬂow from the cisterna and the aqueduct. Br. J. Pharmacol. Chemother., 13: 156–162. Bito, L., Davson, H., Levin, E., Murray, M. and Snider, N. (1966) The concentrations of free amino acids and other electrolytes in cerebrospinal ﬂuid, in vivo dialysate of brain, and blood plasma of the dog. J. Neurochem., 13: 1057–1067. Borland, L.M., Shi, G., Yang, H. and Michael, A.C. (2005) Voltammetric study of extracellular dopamine near microdialysis probes acutely implanted in the striatum of the anesthetized rat. J. Neurosci. Methods, 146: 149–158. Bourdelais, A.J. and Kalivas, P.W. (1992) Modulation of extracellular gamma-aminobutyric acid in the ventral pallidum using in vivo microdialysis. J. Neurochem., 58: 2311–2320. Brodin, E., Lindefors, N. and Ungerstedt, U. (1983) Potassium evoked in vivo release of substance P in rat caudate nucleus measured using a new technique of brain dialysis and an improved substance P-radioimmunoassay. Acta Physiol. Scand. Suppl., 515: 17–20. Butcher, S.P., Fairbrother, I.S., Kelly, J.S. and Arbuthnott, G.W. (1988) Amphetamine-induced dopamine release in the rat striatum: an in vivo microdialysis study. J. Neurochem., 50: 346–355. Carlsson, A. and Lindqvist, M. (1963) Effect of chlorpromazine or haloperidol on formation of 3-methoxytyramine and normetanephrine in mouse brain. Acta Pharmacol. Toxicol. (Copenh.), 20: 140–144. Carmichael, E.A., Feldberg, W. and Fleischhauer, K. (1964) Methods for perfusing different parts of the cat’s cerebral ventricles with drugs. J. Physiol., 173: 354–367. Cheramy, A., Leviel, V. and Glowinski, J. (1981) Dendritic release of dopamine in the substantia nigra. Nature, 289: 537–542. Church, W.H. and Justice, J.B. (1987) Rapid sampling and determination of extracellular dopamine. Anal. Chem., 59: 712–716. Church, W.H., Justice, J.B. Jr. and Byrd, L.D. (1987) Extracellular dopamine in rat striatum following uptake inhibition by cocaine, nomifensine and benztropine. Eur. J. Pharmacol., 139: 345–348. Collier, B. and Mitchell, J.F. (1966) Release of acetylcholine from the cerebral cortex during stimulation of the optic pathway. Nature, 210: 424–425. Commisong, J.W. (1985) Monoamine metabolites: their relationship and lack of relationship to monoaminergic neuronal activity. Biochem. Pharmacol., 34: 1127–1131.

Consolo, S., Fu Wu, C., Fiorentini, F., Ladinsky, H. and Vezzani, A. (1987) Determination of endogenous acetylcholine release in freely moving rats by transtriatal dialysis coupled to a radioenzymatic assay: effect of drugs. J. Neurochem., 48: 1459–1465. Conti, J.C., Strope, E., Adams, R.N. and Marsden, C.A. (1978) Voltammetry in brain tissue: chronic recording of stimulated dopamine and 5-hydroxytryptamine release. Life Sci., 23: 2705–2715. Cornish, K.G. and Hall, R.E. (1979) A modiﬁcation of the dialytrode for simultaneous CNS recording and chemical stimulation. Pharmacol. Biochem. Behav., 10: 389–392. Crespi, F., Martin, K.F. and Marsden, C.A. (1988) Measurement of extracellular basal levels of serotonin in vivo using nafion-coated carbon ﬁbre electrodes combined with differential pulse voltammetry. Neuroscience, 27: 885–896. Crespi, F., Sharp, T., Maidment, N. and Marsden, C. (1983) Differential pulse voltammetry in vivo–evidence that uric acid contributes to the indole oxidation peak. Neurosci. Lett., 43: 203–207. Dale, H.H., Feldberg, W. and Voigt, M. (1936) Release of acetylcholine at voluntary motor nerve endings. J. Physiol., 86: 353–380. Delgado, J.M.R. (1962) Pharmacological modiﬁcations of social behaviour. In: Paton, W.D.M. (Ed.), Pharmacological Analysis of Central Nervous System Action. Pergamon Press, Oxford, pp. 265–292. Delgado, J.M., DeFeudis, F.V., Roth, R.H., Ryugo, D.K. and Mitruka, B.M. (1972) Dialytrode for long term intracerebral perfusion in awake monkeys. Arch. Int. Pharmacodyn Ther., 198: 9–21. Di Chiara, G. (1990) In-vivo brain dialysis of neurotransmitters. Trends Pharmacol. Sci., 11: 116–121. Di Chiara, G. and Imperato, A. (1988) Drugs abused by humans preferentially increase synaptic dopamine concentrations in the mesolimbic system of freely moving rats. Proc. Natl. Acad. Sci. U.S.A., 85: 5274–5278. Edvinsson, L., Nielsen, K.C., Owman, C. and West, K.A. (1971) Alterations in intracranial pressure, blood-brain barrier, and brain oedema after sub-chronic implantation of a cannula into the brain of conscious animals. Acta Physiol. Scand, 82: 527–531. Elghozi, J.L., Mignot, E. and Le Quan-Bui, K.H. (1983) Probenecid sensitive pathway of elimination of dopamine and serotonin metabolites in CSF of the rat. J. Neural Transm., 57: 85–94. Feldberg, W. and Gaddum, J.H. (1934) The chemical transmitter at synapses in a sympathetic ganglion. J. Physiol., 81: 305–319. Fox, R.H. and Hilton, S.M. (1958) Bradykinin formation in human skin as a factor in heat vasodilatation. J. Physiol., 142: 219–232. Fuller, R.W., Perry, K.W., Wong, D.T. and Molloy, B.B. (1974) Effects of some homologues of 4-chloroamphetamine on brain serotonin metabolism. Neuropharmacology, 13: 609–614. Gaddum, J.H. (1961) Push–pull cannulae. J. Physiol., 155: 1P. Glowinski, J., Niecoullon, A. and Cheramy, A. (1978) Regulations of the activity of the nigrostriatal dopaminergic

14 pathways by cortical, cerebellar, and sensory neuronal afferences. Adv. Biochem. Psychopharmacol., 19: 75–87. Grahame-Smith, D.G. (1974) How important is the synthesis of brain 5-hydroxytryptamine in the physiological control of its central function? Adv. Biochem. Psychopharmacol., 10: 83–91. Hallstro¨m, A., Carlsson, A., Hillered, L. and Ungerstedt, U. (1989) Simultaneous determination of lactate, pyruvate, and ascorbate in microdialysis samples from rat brain, blood, fat, and muscle using high-performance liquid chromatography. J. Pharmacol. Methods, 22: 113–124. Hillered, L., Persson, L., Ponten, U. and Ungerstedt, U. (1990) Neurometabolic monitoring of the ischaemic human brain using microdialysis. Acta Neurochir. (Wien), 102: 91–97. Hirani, E., Sharp, T., Sprakes, M., Grasby, P. and Hume, S. (2003) Effect of endogenous 5-HT on [11C]MICRODIALYSISL 100907 binding investigated in rat brain using PET and immediate early gene expression. Synapse, 50: 251–260. Hotsenpiller, G. and Wolf, M.E. (2002) Extracellular glutamate levels in prefrontal cortex during the expression of associative responses to cocaine related stimuli. Neuropharmacology, 43: 1218–1229. Imperato, A. and Di Chiara, G. (1984) Trans-striatal dialysis coupled to reverse phase high performance liquid chromatography with electrochemical detection: a new method for the study of the in vivo release of endogenous dopamine and metabolites. J. Neurosci., 4: 966–977. Imperato, A., Tanda, G., Frau, R. and Di Chiara, G. (1988) Pharmacological proﬁle of dopamine receptor agonists as studied by brain dialysis in behaving rats. J. Pharmacol. Exp. Ther., 245: 257–264. Kalen, P., Strecker, R.E., Rosengren, E. and Bjo¨rklund, A. (1988) Endogenous release of neuronal serotonin and 5-hydroxyindoleacetic acid in the caudate-putamen of the rat as revealed by intracerebral dialysis coupled to high-performance liquid chromatography with ﬂuorimetric detection. J. Neurochem., 51: 1422–1435. Kato, T., Dong, B., Ishii, K. and Kinemuchi, H. (1986) Brain dialysis: in vivo metabolism of dopamine and serotonin by monoamine oxidase A but not B in the striatum of unrestrained rats. J. Neurochem., 46: 1277–1282. Kawagoe, K.T., Zimmerman, J.B. and Wightman, R.M. (1993) Principles of voltammetry and microelectrode surface states. J. Neurosci. Methods, 48: 225–240. Kehr, J. and Ungerstedt, U. (1988) Fast HPLC estimation of gamma-aminobutyric acid in microdialysis perfusates: effect of nipecotic and 3-mercaptoproprionic acids. J. Neurochem., 51: 1308–1310. Kelly, P.H. (1977) Drug-induced motor behaviour. In: Iversen, L.L., Iversen, S.D. and Snyder, S.H. (Eds.), Handbook in Psychopharmacology. Plenum Press, New York, pp. 295–332. Kissinger, P.T., Refshuage, C.J., Dreiling, R., Blank, L., Freman, R. and Adams, R.N. (1973) An electrochemical detector for liquid chromatography with picogram sensitivity. Anal. Lett., 6: 465–477.

Kito, S., Shimoyama, M. and Arakawa, R. (1986) Effects of neurotransmitters or drugs on the in vivo release of dopamine and its metabolites. Jpn. J. Pharmacol., 40: 57–67. Kuhn, D.M., Wolf, W.A. and Youdim, M.B. (1985) 5-Hydroxytryptamine release in vivo from a cytoplasmic pool: studies on the 5-HT behavioural syndrome in reserpinized rats. Br. J. Pharmacol., 84: 121–129. Laruelle, M. (2000) Imaging synaptic neurotransmission with in vivo binding competition techniques: a critical review. J. Cereb. Blood Flow Metab., 20: 423–451. Laruelle, M. and Huang, Y. (2001) Vulnerability of positron emission tomography radiotracers to endogenous competition. New insights. Q. J. Nucl. Med., 45: 124–138. L’Heureux, R., Dennis, T., Curet, O. and Scatton, B. (1986) Measurement of endogenous noradrenaline release in the rat cerebral cortex in vivo by transcortical dialysis: effects of drugs affecting noradrenergic transmission. J. Neurochem., 46: 1794–1801. Marsden, C.A., Joseph, M.H., Kruk, Z.L., Maidment, N.T., O’Neill, R.D., Schenk, J.O. and Stamford, J.A. (1988) In vivo voltammetry–present electrodes and methods. Neuroscience, 25: 389–400. McCreery, R.L., Dreiling, R. and Adams, R.N. (1974) Voltammetry in brain tissue: quantitative studies of drug interactions. Brain Res., 73: 23–33. McFarland, K. and Kalivas, P.W. (2001) The circuitry mediating cocaine-induced reinstatement of drug-seeking behavior. J. Neurosci., 21: 8655–8663. Meyerson, B.A., Linderoth, B., Karlsson, H. and Ungerstedt, U. (1990) Microdialysis in the human brain: extracellular measurements in the thalamus of parkinsonian patients. Life Sci., 46: 301–308. Mitchell, J.F. (1963) The spontaneous and evoked release of acetylcholine from the cerebral cortex. J. Physiol., 165: 98–116. Moroni, F. and Pepeu, G. (1984) The cortical cup technique. In: Marsden, C. (Ed.), Measurement of Transmitter Release In Vivo. Wiley, Chichester, pp. 63–80. Myers, R.D. (1972) Methods for perfusing different structures of the brain. In: Myers, R.D. (Ed.), Methods in Psychobiology. Academic Press, London, pp. 169–211. Myers, R.D., Adell, A. and Lankford, M.F. (1998) Simultaneous comparison of cerebral dialysis and push–pull perfusion in the brain of rats: a critical review. Neurosci. Biobehav. Rev., 22: 371–387. Neff, N.H. and Tozer, T.N. (1968) In vivo measurement of brain serotonin turnover. Adv. Pharmacol., 6: 97–109. Osborne, P.G., O’Connor, N.T., Drew, K.L. and Ungerstedt, U. (1990) An in vivo microdialysis characterization of extracellular dopamine and GABA in dorsolateral striatum of awake freely moving and halothane anaesthetised rats. J. Neurosci. Methods, 34: 99–105. Pei, Q., Zetterstrom, T. and Fillenz, M. (1989) Both systemic and local administration of benzodiazepine agonists inhibit the in vivo release of 5-HT from ventral hippocampus. Neuropharmacology, 28: 1061–1066. Pettit, H.O. and Justice, J.B. (1989) Dopamine in the nucleus accumbens during cocaine self-administration as studie by in

15 vivo microdialysis. Pharmacol. Biochem. Behav., 34: 899–904. Philippu, A. (1984) Use of push–pull perfusion cannulae to determine the release of endogenous neurotransmitters in distinct brain areas of anaesthetised and freely moving animals. In: Marsden, C. (Ed.), Measurement of Transmitter Release In Vivo. Wiley, Chichester, pp. 3–38. Racagni, G., Cheney, D.L., Zsilla, G. and Costa, E. (1976) The measurement of acetylcholine turnover rate in brain structures. Neuropharmacology, 15: 723–736. Rea, K., Cremers, T.I. and Westerink, B.H. (2005) HPLC conditions are critical for the detection of GABA by microdialysis. J. Neurochem., 94: 672–679. Roth, R.H., Walters, J.R. and Morgenroth, V.H. III. (1974) Proceedings: effects of alterations in impulse ﬂow on transmitter metabolism in central dopaminergic neurons. Psychopharmacol. Bull., 10: 40. Sarna, G., Hutson, P.H. and Curzon, G. (1984) A technique for repeated sampling of cerebrospinal ﬂuid in freely moving rats and its uses. J. Physiol. (Paris), 79: 536–537. Sharp, T., Bramwell, S.R., Clark, D. and Grahame-Smith, D.G. (1989a) In vivo measurement of extracellular 5-hydroxytryptamine in hippocampus of the anaesthetized rat using microdialysis: changes in relation to 5-hydroxytryptaminergic neuronal activity. J. Neurochem., 53: 234–240. Sharp, T., Bramwell, S.R. and Grahame-Smith, D.G. (1989b) 5-HT1 agonists reduce 5-hydroxytryptamine release in rat hippocampus in vivo as determined by brain microdialysis. Br. J. Pharmacol., 96: 283–290. Sharp, T., Ljungberg, T., Zetterstrom, T. and Ungerstedt, U. (1986a) Intracerebral dialysis coupled to a novel activity box–a method to monitor dopamine release during behaviour. Pharmacol. Biochem. Behav., 24: 1755–1759. Sharp, T., Maidment, N.T., Brazell, M.P., Zetterstro¨m, T., Ungerstedt, U., Bennett, G.W. and Marsden, C.A. (1984) Changes in monoamine metabolites measured by simultaneous in vivo differential pulse voltammetry and intracerebral dialysis. Neuroscience, 12: 1213–1221. Sharp, T. and Zetterstro¨m, T. (1992) In vivo measurement of monoamine neurotransmitter release and metabolism using brain microdialysis. In: Stamford, J.A. (Ed.), Monitoring Neuronal Activity: A Practical Approach. Oxford, University Press, pp. 147–179. Sharp, T., Zetterstro¨m, T., Ljungberg, T. and Ungerstedt, U. (1987a) A direct comparison of amphetamine-induced behaviours and regional brain dopamine release in the rat using intracerebral dialysis. Brain Res., 401: 322–330. Sharp, T., Zetterstro¨m, T., Series, H.G., Carlsson, A., Yrahame-Smith, D.G. and Ungerstedt, U. (1987b) HPLC-EC analysis of catechols and indoles in rat brain dialysates. Life Sci., 41: 869–872. Sharp, T., Zetterstro¨m, T. and Ungerstedt, U. (1986b) An in vivo study of dopamine release and metabolism in rat brain regions using intracerebral dialysis. J. Neurochem., 47: 113–122. Smith, S.E. and Sharp, T. (1994) An investigation of the origin of extracellular GABA in rat nucleus accumbens measured in

vivo by microdialysis. J. Neural Transm. Gen. Sect., 97: 161–171. Soubrie, P., Reisine, T.D. and Glowinski, J. (1984) Functional aspects of serotonin transmission in the basal ganglia: a review and an in vivo approach using the push–pull cannula technique. Neuroscience, 13: 605–625. Strecker, R.E., Sharp, T., Brundin, P., Zetterstro¨m, T., Ungerstedt, U. and Bjo¨rklund, A. (1987) Autoregulation of dopamine release and metabolism by intrastriatal nigral grafts as revealed by intracerebral dialysis. Neuroscience, 22: 169–178. Stro¨mberg, I., Herrera-Marschitz, M. and Ungerstedt, U. (1985) Chronic implants of chromafﬁn tissue into the dopamine-denervated striatum. Effects of NGF on graft survival, ﬁber growth and rotational behaviour. Exp. Brain Res., 60: 335–349. Tanda, G., Carboni, E., Frau, R. and Di Chiara, G. (1994) Increase of extracellular dopamine in the prefrontal cortex: a trait of drugs with antidepressant potential? Psychopharmacology (Berl.), 115: 285–288. Tilson, H.A. and Sparber, S.B. (1972) Studies on the concurrent behavioral and neurochemical effects of psychoactive drugs using the push–pull cannula. J. Pharmacol. Exp. Ther., 181: 387–398. Timmerman, W. and Westerink, B.H. (1997) Brain microdialysis of GABA and glutamate: what does it signify? Synapse, 27: 242–261. Tossman, U., Eriksson, S., Delin, A., Hagenfeldt, L., Law, D. and Ungerstedt, U. (1983) Brain amino acids measured by intracerebral dialysis in portacaval shunted rats. J. Neurochem., 41: 1046–1051. Ungerstedt, U. (1971) Stereotaxic mapping of the monoamine pathways in the rat brain. Acta Physiol. Scand. Suppl., 367: 1–48. Ungerstedt, U. and Pycock, C. (1974) Functional correlates of dopamine neurotransmission. Bull. Schweiz. Akad. Med. Wiss., 30: 44–55. Ungerstedt, U., Herrera-Marschitz, M., Jungnelius, U., Sta˚hle, L., Tossman, U. and Zetterstro¨m, T. (1982) Dopamine synaptic mechanisms reﬂected in studies combining behavioural recordings and brain dialysis. In: Kohsaka, H. (Ed.), Advances in Dopamine Research. Pergamon Press, Oxford and New York, pp. 219–231. Westerink, B.H. (1995) Brain microdialysis and its application for the study of animal behaviour. Behav. Brain Res., 70: 103–124. Westerink, B.H. and Korf, J. (1976) Regional rat brain levels of 3,4-dihydroxyphenylacetic acid and homovanillic acid: concurrent ﬂuorometric measurement and inﬂuence of drugs. Eur. J. Pharmacol., 38: 281–291. Westerink, B.H. and Tuinte, M.H. (1986) Chronic use of intracerebral dialysis for the in vivo measurement of 3,4-dihydroxyphenylethylamine and its metabolite 3,4-dihydroxyphenylacetic acid. J. Neurochem., 46: 181–185. Wolfensberger, M. (1984) Gas-chromatographic and massfragmentographic measurement of amino acids released into brain perfusates collected in vivo by push–pull cannula techniques. In: Marsden, C. (Ed.), Measurement of Transmitter Release In Vivo. Wiley, Chichester, pp. 39–62.

16 Wright, I.K., Upton, N. and Marsden, C.A. (1992) Effect of established and putative anxiolytics on extracellular 5-HT and 5-HIAA in the ventral hippocampus of rats during behaviour on the elevated X-maze. Psychopharmacology (Berl.), 109: 338–346. Zetterstro¨m, T. and Ungerstedt, U. (1984) Effects of apomorphine on the in vivo release of dopamine and its metabolites, studied by brain dialysis. Eur. J. Pharmacol., 97: 29–36. Zetterstro¨m, T., Brundin, P., Gage, F.H., Sharp, T., Isacson, O., Dunnett, S.B., Ungerstedt, U. and Bjorklund, A. (1986a) In vivo measurement of spontaneous release and metabolism of dopamine from intrastriatal nigral grafts using intracerebral dialysis. Brain Res., 362: 344–349. Zetterstro¨m, T., Herrera-Marschitz, M. and Ungerstedt, U. (1986b) Simultaneous measurement of dopamine release and rotational behaviour in 6-hydroxydopamine denervated rats using intracerebral dialysis. Brain Res., 376: 1–7. Zetterstro¨m, T., Sharp, T., Collin, A.K. and Ungerstedt, U. (1988) In vivo measurement of extracellular dopamine and

DOPAC in rat striatum after various dopamine-releasing drugs; implications for the origin of extracellular DOPAC. Eur. J. Pharmacol., 148: 327–334. Zetterstro¨m, T., Sharp, T., Marsden, C.A. and Ungerstedt, U. (1983) In vivo measurement of dopamine and its metabolites by intracerebral dialysis: changes after d-amphetamine. J. Neurochem., 41: 1769–1773. Zetterstro¨m, T., Sharp, T. and Ungerstedt, U. (1984) Effect of neuroleptic drugs on striatal dopamine release and metabolism in the awake rat studied by intracerebral dialysis. Eur. J. Pharmacol., 106: 27–37. Zetterstro¨m, T., Sharp, T. and Ungerstedt, U. (1986c) Further evaluation of the mechanism by which amphetamine reduces striatal dopamine metabolism: a brain dialysis study. Eur. J. Pharmacol., 132: 1–9. Zetterstro¨m, T., Vernet, L., Ungerstedt, U., Tossman, U., Jonzon, B. and Fredholm, B.B. (1982) Purine levels in the intact rat brain. Studies with an implanted perfused hollow ﬁbre. Neurosci. Lett., 29: 111–115.

CHAPTER 1.2

Microdialysis of glutamate and GABA in the brain: analysis and interpretation Ben H.C. Westerink1,, Kieran Rea2, Weite H. Oldenziel1 and Thomas I.F.H. Cremers1 1

Department of Biomonitoring and Sensoring, University Center for Pharmacy, University of Groningen, Groningen, The Netherlands 2 Pharmacology Department, National University of Ireland, Galway

Abstract: Numerous studies have analyzed glutamate and GABA in microdialysis samples. It is a common practice among most researchers to relate changes in extracellular levels of glutamate and GABA directly to the activity of the corresponding neurons. However, several authors have expressed concern about this practice as basal values of glutamate and (to a certain extent) GABA do not fulﬁll the criteria for exocytotic release. In the present chapter, the signiﬁcance and origin of glutamate and GABA as detected in microdialysates is discussed. In particular, the role of glial cells in the release and metabolism of the amino acids is emphasized. Special attention is given to the chemical analysis of GABA. productive and as many as 1,000 papers have been produced over the past 15 years. However, in contrast to neurotransmitters, such as the catecholamines, serotonin, and acetylcholine, it has been difﬁcult to prove that levels of GABA and glutamate in dialysates originate from neuronal activity as the basal levels of these transmitters do not fulﬁll the criteria of exocytosis (calcium- and sodium-channel dependency) (Westerink et al., 1987; Millan et al., 1991; Herrera-Marschitz et al., 1996; Miele et al., 1996; Timmerman and Westerink, 1997). Despite the uncertainty of the origin of glutamate and GABA in dialysates, it is a common practice in most microdialysis studies to relate changes in extracellular levels directly to the activity of the corresponding neurons. In this chapter, we try to ﬁnd explanations for the typical properties of GABA and glutamate in microdialysis experiments.

I. Introduction GABA and glutamate are the main inhibitory and excitatory neurotransmitters, respectively, in the central nervous system (CNS). Both neurotransmitters play an important role not only in the physiology of the brain, but also in various pathophysiological concepts, such as depression, schizophrenia, and epilepsy. In addition, the transmitters are implicated in the mechanism of action of a variety of centrally acting drugs. Therefore, detection of the synaptic release of glutamate and GABA directly into the brains of freely moving animals is of great signiﬁcance to help us better understand of the role that these transmitters play in the homeostasis, as well as the pathology, of the CNS. Synaptic release of GABA and glutamate is often estimated by analysis of their extracellular concentrations in the brain, and in this regard, the microdialysis method is frequently used to monitor these transmitters. Research along this line has been very

II. Analysis of GABA and glutamate Microdialysis samples are relatively clean, as the dialysis membrane excludes cell fragments and

Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

17

DOI: 10.1016/S1569-7339(06)16002-6 Copyright 2007 Elsevier B.V. All rights reserved

18

large molecules, such as proteins. Amino acids in these samples are currently analyzed with chromatographic methods, such as HPLC or capillary electrophoresis (CE) without prior puriﬁcation. II.A. HPLC analysis of GABA and glutamate To improve the chromatographic behavior and sensitivity of detection, amino acids are ﬁrst derivatized with ﬂuorogenic reagents, which are speciﬁc for primary amines, such as o-phthalaldehyde (OPA), ﬂuorescamine, or naphthalene-2,3-dicarboxaldehyde (NDA) (Stobaugh et al., 1983; Jacobs et al., 1986). These ﬂuorophores react with amines in the presence of a nucleophile (usually CN for NDA and a alkylthiol for OPA) to form electroactive and ﬂuorescent derivatives. In microdialysis studies, OPA is most frequently used. The derivatization reaction occurs within minutes at room temperature. The reaction scheme is depicted in Fig. 1. For alkylthiol, mercaptoethanol or 3-mercaptopropionic acid are often used. Some authors use 3-mercaptopropionic acid and alkaline hypochlorite as these conditions increase the ﬂuorescence of the derivatives and improve the stability of the ﬂuorophores toward oxidation (Fiorino et al., 1989). Automated mixing followed by injection is often applied to perform the coupling reaction followed by online HPLC separation. For GABA, there is some concern in the literature about the stability of the reaction product with OPA. The poor stability (and also the stench of the thiol compound) was overcome by using tert-butylthiol (Allison et al., 1984) or by substitution of the OPA-alkylthiol reaction with an OPA-sulphite reaction (Smith and Sharp, 1994). However, it is our experience that the OPA-mercaptoethanol reaction product, once injected onto O

S R' H H

R'-SH

+

R-NH2

N R

O Fig. 1. The scheme for the reaction between amino acids, ophthalaldehyde (OPA) and alkylthiol.

the separation column, is very stable even at retention times as long as 60 min (Rea et al., 2005). A typical analysis of GABA and glutamate in the dialysates is carried out as follows. The derivative reagent consists of 100 mg OPA dissolved in 2-mL methanol and added to 200-mL 0.5-mol/L NaHCO3 containing 20-mL 2-mercaptoethanol. The reagent is freshly prepared daily. The pH is adjusted to 9.5 with NaOH. Next, 30 mL microdialysate samples is derivatized with 50-mL OPA/mercaptoethanol reagent, mixed, and allowed to react for 2 min. Fifty microliters of the reaction mixture is then injected by an autosampler onto the HPLC apparatus. The derivatization mixture is separated using an isocratic mobile phase and measured by ﬂuorometric or electrochemical detection. The mobile phase consists of 70 mM di-sodium hydrogen phosphate, 400 mM EDTA, 0.15% (v/v) tetrahydrofurane, and methanol. To determine the optimal separation conditions, the organic modiﬁer concentration can be varied between 30 and 40% (v/v) methanol. For optimal separation, the pH of the mobile phase is adjusted between 4.0 and 6.0 with phosphoric acid. The retention time of most of the derivatized amino acids is not very dependent on the pH; however, the GABA derivative is an exception in this respect. In our hands, there is only a small pH window in which GABA is reliably separated from the interfering peaks by ﬂuorescence detection (Rea et al., 2005). A representative HPLC system consists of a reverse-phase column (e.g., 150 mm 4.6 mm; particle size, 3 mm), a high-precision pump (ﬂow rate, 1.0 mL/min), and a ﬂuorimetric (excitation l ¼ 350 nm, emission l ¼ 450 nm) or electrochemical detector. II.B. Capillary electrophoresis of GABA and glutamate Microdialysis produces small samples of volumes often less than 20 mL, making them of interest for separation by CE. Similarl to the HPLC methods, NDA or OPA are used as derivatizing agents and the reaction products are quantiﬁed by ﬂuorimetric or electrochemical detection (O’Shea et al., 1992). Recently, various authors have demonstrated that the detection limit, and therefore, the temporal resolution of the microdialysis

19

experiment can be further improved, when CE is coupled with laser-induced ﬂuorescence detection (CE–LIFD) (see references in Sauvinet et al., 2003). With such an approach, a sampling time of 1–30 s can be performed (see also Chapter 3.3). In a typical analysis of GABA and glutamate on CE, a 0.5-mL microdialysis sample is derivatized for 30 s with ﬂuorescein isothiocyanate (14 h at room temperature) (Rada et al., 2003). Glutamate and GABA are then separated on a CE system equipped with an argon laser tuned at 488 nm. A disadvantage of the CE method is that the handling of nanoliter samples requires special care and specialized equipment. II.C. Online sensors for glutamate in microdialysates In the case of sampling glutamate by microdialysis, samples have been analyzed by continuous online detection ex vivo using enzyme-based miniaturized devices. One method makes use of the ﬂuorescent properties of reduced nicotinamide adenine dinucleotide (NADH) formed by the metabolism of glutamate to a-ketoglutarate by glutamate oxidase, in the presence of NAD+ (Obrenovitch et al., 1993). In another approach, glutamate is electrochemically detected after conversion to H2O2 by glutamate dehydrogenase. To that end, amperometric electrodes have been constructed that detect H2O2 at glassy carbon surfaces modiﬁed by osmium-poly(4-vinylpyridine)gel and horseradish peroxidase (Niwa et al., 1996; Shi et al., 2003). As the samples are not separated by chromatography, the speciﬁcity of the electrode for glutamate is a matter of concern. Preventing interference of the detection by reducing compounds, such as ascorbic acid requires special pretreatment of the electrodes with ascorbic acid oxidase.

III. Microdialysis of glutamate: interpretation III.A. The origin of glutamate in the extracellular fluid questioned When neurotransmitters such as dopamine, noradrenaline, serotonin, and acetylcholine are sampled

Fig. 2. Effect of TTX infusion on extracellular levels of GABA and dopamine in the brain. Closed circles: dopamine, results obtained from Feenstra and Botterblom (1996). Open symbols: various studies on microdialysis of GABA, summarized in Timmerman and Westerink (1997).

by microdialysis, infusion of a sodium-channel blocker (e.g., tetrodotoxin TTX) or calcium-free Ringer leads to a rapid decline of the transmitter levels in the dialysis samples (Westerink et al., 1987). An example of this type of experiment is shown in Fig. 2. By using a 5-min sampling time, infusion of TTX resulted in a rapid and almost complete disappearance of dopamine from the extracellular ﬂuid (ECF) (Feenstra and Botterblom, 1996). This result clearly demonstrates that dopamine sampled by microdialysis is very likely of synaptic origin. However, when similar experiments are performed for glutamate, the transmitter output does not respond. Therefore, various authors have expressed serious doubts whether glutamate sampled by microdialysis is derived from synaptic release (Westerink et al., 1987; Millan et al., 1991; Herrera-Marschitz et al., 1996; Miele et al., 1996; Timmerman and Westerink, 1997). An explanation for the peculiar behavior of glutamate in the ECF might be found in the speciﬁc role that this transmitter plays in neurotransmission. To assure a high signal-to-noise ratio of neurotransmission and to avoid excitotoxic actions of glutamate, the extracellular concentrations should be kept low (Katayama et al., 1990; Bergles

20

et al., 1999; Amara and Fontana, 2002). Released glutamate immediately needs to be removed from the synaptic cleft. For that reason, high-afﬁnity glutamate transporters are localized in the vicinity of the synapse (Diamond and Jahr, 1997; Lehre and Danbolt, 1998). Glutamate concentrations in the synaptic cleft may reach concentrations of several millimoles and it has been calculated that glutamate transporters have the potency to bring concentrations in the ECF locally down to 20 nmol. Although these transporters are also present on glutamate neurons, the astroglial cells seem to play a dominant role in this respect. The intracellular concentration of glutamate in neurons is in the order of 10 mM, whereas in astrocytes, the concentration does not exceed 50 mM (due to the fact that glutamate in astrocytes is metabolized to glutamine). Several authors assume that (extrasynaptic) extracellular concentration of glutamate has to be maintained at 1–3 mM to assure a high signal-to-noise ratio for neurotransmission and to avoid excitotoxic actions of glutamate on neurons. Indeed, glutamate concentrations in the ECF determined by microdialysis are in the range of 1–4 mM (Zilkha et al., 1995; Niwa et al., 1996; Shiraishi et al., 1997; Lada et al., 1998); however, microsensor experiments have indicated concentrations around 25 mM (Kulagina et al., 1999; Oldenziel et al., submitted). As the glutamate transporters are not evenly distributed over the glial membranes, it is to be expected that the glutamate concentration will also display a heterogeneous concentration in the ECF. Moreover, there is accumulating evidence that astrocytes also release glutamate, which means that at certain sites, glutamate release will dominate over uptake (see further discussion below). Therefore, different extracellular compartments of glutamate might exist in the ECF. III.B. Is glutamate able to escape from the synaptic cleft? The ﬁrst task of released glutamate is to mediate rapid point-to-point transmission within the synaptic cleft implicating that within a time frame of milliseconds, concentrations of several millimoles are reached and again removed. However,

under certain conditions, the transmitter seems to be able to escape from the synaptic cleft to stimulate extra- and heterosynaptic receptors in the perisynaptic environment (Arnth-Jensen et al., 2002; Rusakov and Lehre, 2002). This process is referred to as ‘‘spill-over’’. The extent of this glutamate spill-over depends on the amount and rate of escape of glutamate, the distance between adjacent synapses, and the location and abundance of glutamate transporters (Bergles et al., 1999). As the amount of transmitter spill-over is determined by the activity of the glutamate neurons, it has been hypothesized that certain perisynaptic receptors are only reached by glutamate during intense stimulus patterns. As the degree of astroglial wrapping of the synapse determines the distance that glutamate is able to migrate out of the synaptic cleft (Hawkins et al., 1995), the transporters appear to have a strategic role in the modulation of glutamate neurotransmission. The transporters may modify the time course of synaptic events, as well as the extent and pattern of activation, and desensitization of receptors outside the synaptic cleft and at neighboring synapses. Glutamate high-afﬁnity transport is now recognized as a dynamic process. A variety of soluble compounds, for example, glutamate itself, cytokines, and growth factors, can inﬂuence the expression and activity of the glutamate transporters, and the extent to which astrocytic processes wrap around synapses (Danbolt, 2001; Amara and Fontana, 2002). These mechanisms allow a time-dependent modiﬁcation of the efﬁcacy of glutamate uptake, which in turn can shape extrasynaptic glutamergic transmission (Oliet et al., 2001; Nedergaard et al., 2002). Taken together, evidence is accumulating that under certain conditions, glutamate is able to escape from the synaptic cleft. During these conditions, glutamate might appear in ECF sampled by microdialysis. III.C. Glutamate released from astroglial cells: gliotransmission and other mechanisms For a long time, it was thought that astrocytes were passive cells with only an intermediate function between blood vessels and neurons. As astrocytes show only few responses to

21

electrophysiological recording techniques, the absence of electrical activity was interpreted as functional inactivity (Nedergaard et al., 2002; Volterra and Meldolesi, 2005). Nowadays, it is recognized that in many aspects, astrocytes are similar to neurons. Astrocytes communicate with each other by changing intracellular calcium, and gap junctions connect them with one another. Ca2+ signals may spread between astrocytes in the form of Ca2+ waves. Spontaneous calcium oscillations in astrocytes can excite neighboring neurons, but neurons can also activate astrocytes. In response to these changes in intracellular calcium, astrocytes release neurotransmitters, often described as gliotransmitters, whereas the process is called gliotransmission. This mechanism was ﬁrst revealed in 1994 when increases in [Ca2+]i were shown to induce glutamate release followed by neuronal activation (Parpura et al., 1994). It is interesting to note that one of the ﬁrst gliotransmitters described was glutamate. Over the years, the number of proposed gliotransmitters has increased (e.g., ATP, adenosine, D-serine, proteins, and peptides). The ﬁnding that astrocytes communicate chemically has triggered research on their functional role. Although many functions of astrocytes still need to be revealed, it is evident that they play a key role in many processes in the CNS, for example, in the glutamate–glutamine cycle, in the regulation of the cerebral blood ﬂow, in the regulation, modulation, and synchronization of neuronal activities in both neurogenesis and synaptogenesis, in learning and memory, and in pathological processes in the brain (Nedergaard et al., 2002; Volterra and Meldolesi, 2005). However, a profound difference between the two cell types is the timescale: glial Ca2+ signals propagate at rates of micrometers per second, whereas action potentials propagate at rates of meters per second (Haydon, 2001). Several mechanisms of glutamate release from astroglia have now been described. Glutamate is released from astrocytes across the plasma membrane through specialized proteins, such as channels and transporters (Volterra and Meldolesi, 2005). Some mechanisms act during control conditions; others are only activated during

pathological states. Here, the different types of glutamate release that have been proposed are summarized. (1) Glutamate release regulated by volume-sensitive organic anion channels (VSOACs), also referred as volume-regulated anion channels (VRACs). Hypo-osmotic solutions activate these channels as part of the volume regulation and allow the efﬂux of aspartate, taurine, glutamate, chloride, and other anions in a calcium-independent manner. Activation of metabotropic glutamate receptors is also associated with astrocytic swelling (Hansson et al., 2000). (2) Glutamate released by GAP-junction hemichannels: they release both ATP and glutamate, and their opening probability is controlled by changes in intracellular Ca2+ concentrations (Hansson et al., 2000; Volterra and Meldolesi, 2005). (3) Glutamate release controlled by the purinergic P2X7 receptor: this receptor is gated by ATP and has been proposed to be involved in the release of glutamate as well as D-aspartate. It shares some properties with the GAP-junction hemichannels, including the increased opening probability at low Ca2+ concentration, but their pharmacology is different (Wang et al., 2004). (4) Sodium-dependent high-affinity heteroexchange mechanisms for glutamate and ascorbate, for glutamate and GABA, and for glutamate and glycine have been reported, but are poorly characterized (Raiteri et al., 2002). (5) Glutamate release mediated by a prostaglandin-dependent mechanism (Bezzi et al., 1998). (6) Reversal of sodium-dependent uptake by glutamate transporters: this does not seem to occur during normal brain functioning. It relates the massive release of glutamate from neurons and astrocytes after traumatic or ischemic injury to brain tissue (Benveniste et al., 1984; Katayama et al., 1990). (7) Glutamate release coupled to the cystineglutamate exchanger: this is a sodiumdependent glutamate transporter, which

22

can operate in the reverse mode, that is, cystine is exchanged for glutamate by transporting one molecule of cystine intracellularly for one molecule of glutamate out of the cell. The accumulated cystine is reduced to cysteine and in turn converted to glutathione (Warr et al., 1999; Baker et al., 2002; Chen and Swanson, 2003; Xi et al., 2003). (See Chapter 1.3 for further discussion.) III.D. Is glutamate released from astrocytes detectable in microdialysates? It is of interest to know whether the aforementioned mechanisms of glutamate release from astrocytes are detectable in microdialysates. Glutamate release from astrocytes is described as nonvesicular, although recently, a sort of vesicular release has been observed in the hippocampus (Bezzi et al., 2004). Nonexocytotic release of glutamate may occur via plasma membrane channels. Therefore, astrocytic release of glutamate will not likelyrespond to infusion of sodium-channel blockers (TTX). Similarly, most of the described release mechanisms from astrocytes are calcium independent or make use of mobilization of intracellular calcium stores. In this regard, depletion of calcium during microdialysis experiments will probably not affect glutamate release from astrocytes. It is hypothesized that astrocytic mGlu receptors sense glutamate that is released during synaptic transmission to adjust its extracellular concentration by modulating uptake activity. Interestingly, stimulation or inhibition of mGlu receptors by infusion of appropriate agonists/ antagonists was detectable in glutamate levels in microdialysates (Xi et al., 2003). As expected, this type of glutamate release was found to be TTX independent. Baker et al. (2002) provided evidence that by using speciﬁc agonists and antagonists, the cystine–glutamate antiporter is detectable in microdialysates (see Chapter 1.3). At present, it is a matter of debate whether this exchanger contributes to extracellular glutamate release during normal brain function or whether it occurs when the

energy metabolism of astrocytes is compromised (Nedergaard et al., 2002; Hertz and Zielke, 2004; Cavelier and Atwell, 2005). However, Baker et al. (2002) demonstrated that glutamate released by the antiporter was able to modify dopamine transmission. Other types of glutamate releases from astrocytes, for example, related to VSOACs, GAPjunction hemichannels, and the prostaglandin dependent as well as the purinergic P2X7 receptor mechanisms have poorly been studied so far in microdialysis studies. Presently, it is a matter of debate whether the different release mechanisms from astrocytes to operate under normal or pathological conditions. It is emphasized that implantation of a microdialysis probe causes extensive damage to brain, which implicates that this method represents a pathological rather than a physiological condition (see further discussion in Section III.H). III.E. Other sources of extracellular glutamate Apart from its role in neurotransmission, glutamate fulﬁlls various other functions in the CNS. It serves as a precursor for peptide and protein synthesis, is involved in fatty acid synthesis, contributes to the regulation of ammonia levels, serves as a precursor for GABA, and is involved in carbohydrate metabolism. It is of interest to note that in terms of energy balance, the majority of glutamate metabolism in the brain is directed to glutamate release and uptake (see Chapter 1.3). As blood levels of glutamate are around 100 mmol/L, the general circulation should be considered as a possible source of extracellular glutamate. However, studies based on infusion of glutamate into the bloodstream indicated that the neurotransmitter penetrates the brain only after a considerable delay (Hawkins et al., 1995). Artiﬁcial sources of glutamate produced during microdialysis experiments should not be ruled out. It was recently shown that microbial contamination of the microdialysis system might confound glutamate detection because microorganisms can metabolize glutamate from glutamine. Sterilization of probes and tubing is necessary for implantation periods longer than 24 h (Zhou et al., 2002).

23

In our laboratory, we have observed that simply ‘handling’ of tubing that has been used before, can produce glutamate. These sources of glutamate are obviously not TTX dependent. Recently, the presence of D-glutamate in rat brain samples was reported (Quan and Liu, 2003). As D-glutamate is only transported by the lowafﬁnity glutamate uptake system and not converted inside astroglia to glutamine (Pow and Crook, 1996), its levels might accumulate in the ECF. As HPLC separations do not differentiate between the enantiomers, the possible contributions of D-glutamate to glutamate levels in the ECF need further investigation. However, preliminary experiments in our laboratory have shown that D-glutamate levels are less than 10% of Lglutamate and that this enantiomer does not respond to infusion of elevated potassium. III.F. Is synaptic glutamate detectable in microdialysates during excessive neuronal stimulation? Although many authors agree that basal levels of glutamate in microdialysis samples are not derived from synaptic processes, various attempts have been made to detect synaptic glutamate during excessive neuronal stimulation. To achieve these conditions, different strategies were applied. Chemical stimulation of glutamate release by infusions of excitatory agents via the microdialysis probe has been performed in various studies. Infusion of high potassium concentration (60–100 mM) increases the release of glutamate drastically. This increase is easily detectable in microdialysates and was characterized as calcium dependent. The evoked glutamate is supposed to be of synaptic origin, although release from astrocytes as part of the internal cell volume regulation (Kimelberg et al., 1995) cannot be excluded. Another way to enhance transmitter release is to prolongate the opening of potassium channels by infusing the potassium-channel blocker, 4-aminopyridine. This resulted in an increase in glutamate levels in microdialysates, which was largely TTX dependent (Pena and Tapia, 2000). Infusion of the glutamate high-afﬁnity transport blockers, TBOA, resulted in a pronounced

increase of glutamate (to 300% of controls). However, this increase in glutamate was not TTX dependent (Xi et al., 2003) which indicates that glutamate accumulating under these conditions in microdialysates is primarily derived from astrocytes. A different but rather successful approach to detect synaptic glutamate is to activate the cellbody area of glutamatergic neurons by chemical stimulation and to monitor simultaneously, the transmitter in the terminal area. In this regard, application of substance P in the substantia nigra increased extracellular levels of glutamate in the striatum (Reid et al., 1990). In comparable experiments, local administration of NMDA in the frontal cortex increased glutamate levels in the striatum (Palmer et al., 1989; Dijk et al., 1995). In several studies, electrical stimulation of the cell-body areas of glutamate pathways has been successfully applied to increase extracellular glutamate levels in microdialysates of the corresponding nerve terminal areas. For example, stimulation of the medulla, sciatic nerve or raphe magnus increased glutamate in areas of the spinal cord (Kapoor et al., 1990; Paleckova et al., 1992; Sorkin and McAdoo, 1993). Electrical stimulation of the nervus vagus increased glutamate levels in the nucleus tractus solitarius (Allichin et al., 1994), whereas stimulation of the nucleus paragigantocellularis increased glutamate in the ipsilateral locus coeruleus (Liu et al., 1999). Electrical stimulation of the prefrontal cortex induced an increase (in 5 s samples) of extracellular glutamate in the striatum of anesthetized rats (Lada et al., 1997; Arnth-Jensen et al., 2002). Application of a 10-s train of pulses induced a rapid increase in glutamate to 200–300% of controls that returned to basal values within 60 s. The increase in glutamate was calcium and TTX dependent. The stimulation of glutamate was suppressed by the m-glutamate receptor agonist, 1-aminocyclopentane-trans-1,3-dicarboxylate. Tucci et al. (2000) demonstrated that electrical stimulation of the prefrontal cortex increased glutamate in the nucleus accumbens. Some of the results of the chemical or electrical stimulation experiments were shown to be TTX dependent, but it is emphasized that TTX infusion

24

is not always the right control experiment as infusion of sodium-channel blockers also blocks the transmission of the applied (electrical or neuronal) stimulation and the subsequent physiological and behavioral effects. Certain types of behavior stimulations are accompanied by characteristic changes (often an increase) in extracellular levels of glutamate in microdialysis experiments. In several studies, extracellular glutamate was recorded during restraint stress. Substantial increases in glutamate levels to 200–450% of basal values lasting for 10–20 min were reported for the prefrontal cortex, hippocampus, nucleus accumbens, and striatum (see references in Timmerman and Westerink, 1997). Photic stimulation of the conscious rat increased extracellular glutamate in the visual cortex to 200% of controls for at least 3 min (Reyes et al., 2002), and formalin injection into the hind paw resulted in an increase of extracellular glutamate that lasted for 2 min in certain subareas of the hypothalamus (preoptic area but not in lateral and ventromedial hypothalamus) (Silva et al., 2004). Results with a different time scale were presented by Rosell et al. (2003), who stimulated a whisker of the rat and sampled glutamate in the motor cortex. By using CE coupled to laser-induced ﬂuorescence detection, they were able to determine glutamate in 1 s microdialysis samples (see also Chapter 3.3). Upon whisker stimulation, glutamate was increased in the 1 s sample only; in the 2 s sample, glutamate again fell back to control values, although the whisker stimulation was continued. The results of the 1-s lasting increase in glutamate content contrasts with the 2–20 min lasting increases in glutamate that were observed in the behavioral experiments discussed above. These data demonstrate the importance of the time scale in detecting physiological changes in extracellular glutamate. It is likely that glutamate displaying subsecond increases reﬂects different origin than glutamate that was enhanced in behavioral experiments lasting several minutes. An intriguing question is whether the above-described behavior-induced changes are calcium and TTX dependent. It was found that handling stressinduced increases in glutamate that persisted in the presence of TTX (Timmerman and Westerink,

1997). Even mild activation of rats induces changes in glutamate levels in the ECF. Grooming–induced by dropping water onto the rat’s snout–increased extracellular glutamate in the striatum to 200% of controls (Miele et al., 1996). However, a similar increase was seen when the experiment was performed in the presence of TTX. In contrast, there is a report about feedinginduced increase in extracellular glutamate in the nucleus accumbens that was found to be TTX dependent (Saulskaya and Mikhailova, 2004). The study that used formalin injection into the hind paw described increases in glutamate that were both TTX and calcium dependent (Silva et al., 2004). However, the speciﬁcity of this observation is questioned by the fact that other amino acids, such as arginine and aspartate also responded in a calcium- and TTX-dependent manner. In conclusion, evidence is provided that glutamate from synaptic sources is detectable in microdialysates collected during excessive stimulation of glutamate pathways. The behavioral experiments have so far produced most interesting but also complex results. A rapid increase in glutamate with a time scale of 1 s was detected apart from changes in glutamate that were lasting during several minutes. It is unclear whether these different processes relate to synaptic release at the present time, as in most of these experiments, the TTX and calcium dependency of the observed effects was not consistently studied or demonstrated. It is tempting to speculate that changes of glutamate in subseconds time frame are related to synaptic activity, whereas changes in minutes might represent astrocytic (nonexocytotic) release. III.G. A new approach detecting extracellular glutamate: the use of microsensors Carbon ﬁbers (diameter 10 mm) have been used to study the dynamics of the release of neurotransmitters, such as dopamine and serotonin. As these compounds are directly oxidized at the carbon surface, changes in transmitter level in the subsecond time frame can be monitored. Carbon ﬁber-based glutamate microsensors have been constructed by applying electroactive polymers and speciﬁc enzymes that oxidize glutamate in vivo

25

(Kulagina et al., 1999). However, the use of these microsensors is limited by a number of points. First, the response time of these sensors is rather slow (10 s) due to slow diffusion of glutamate through the polymer. Second, the sensors are sensitive to ascorbate and uric acid, and the levels of these compounds in the brain are high and difﬁcult to control (Oldenziel et al., in press). However, a most interesting observation was made by Kulagina et al. (1999), who reported that glutamate detected in the brain by microsensors displayed TTX dependency. Recently, these observations were conﬁrmed by two different laboratories: Day et al. (2006) by using keramic-based glutamate sensors and Oldenziel et al. (submitted) who optimized the microsensor concept of Kulagina et al. (1999). The observation that microsensors can detect TTX-dependent glutamate whereas microdialysis probes do might not have a signiﬁcant impact on future research on extracellular glutamate. III.H. Interpretation of extracellular glutamate: two recording methods, two different sources? Recent evidence indicates that microsensors are able to sample TTX-sensitive glutamate that may be derived from synaptic and/or perisynaptic sites. The absence of effects of calcium depletion or sodium-channel blockers on glutamate concentrations, as determined by microdialysis, suggests that this source of glutamate may represent a different pool of origin, such as astrocytic glutamate released by gliotransmission. This would mean that both methods are complementary. However, it is emphasized that the implantation of the microdialysis probe creates an artiﬁcial cavity in the brain, where active proliferating astroglial cells and microcytes will accumulate. Soon after implantation, this cavity will be surrounded by scar tissue. It should also be pointed out that implantation of a microdialysis probe causes extensive damage to brain. It has been shown that the dialysis probe creates a disrupted zone of tissue that extends as far as 1.4 mm from the implantation site. In contrast, electron microscopy of the track created by a carbon ﬁber revealed that the tissue damage was conﬁned to a distance of 3 mm (see also Chapter

1.6). Therefore, the microdialysis method might represent a pathological rather than a physiological condition. This effect certainly complicates the interpretation of glutamate sampled by microdialysis.

IV. Microdialysis of GABA: interpretation Similar to glutamate, there is a debate in the literature about the neuronal origin of GABA in microdialysis samples. Studies that investigated the effect of infusion of TTX or calcium-free Ringer solutions on GABA levels in dialysates produced inconsistent results. Calcium omission and TTX infusion have been reported to be partly effective (Osborne et al., 1990; Drew and Ungerstedt, 1991; Campbell et al., 1993; Singewald et al., 1993; Rakovska et al., 1998), slightly effective (Stengard and O’Connor, 1994), or nonresponsive (Morari et al., 1993; Bourdelais and Deutch, 1994; Timmerman and Westerink, 1995; Ferraro et al., 2000) in reducing basal GABA levels. In the studies that reported effects of TTX (summarized in Fig. 2), GABA levels decreased only slowly, reaching a statistical signiﬁcance after 30–60 min of the start of infusion. These results are in strong contrast with the fast-transmitter role that GABA is supposed to play in the CNS. Fig. 2 shows that the response of GABA to TTX is incomplete and much slower when compared with dopamine. Another controversy in the literature is the inability to detect effects of GABA autoreceptor agonists in microdialysis experiments. It is well known from in vitro studies that activation of GABAB autoreceptors by speciﬁc agonists, such as baclofen, strongly decrease GABA release as result of an inhibitory autoreceptor feedback mechanism (Waldmeier et al., 1988; Baumann et al., 1990; Lanza et al., 1993; Lambert and Wilson, 1994). However, in microdialysis studies, baclofen was found to have little (Bourdelais and Kalivas, 1992) or no effect (Waldmeier et al., 1992; Richards et al., 1995; Timmerman and Westerink, 1995) on basal GABA levels. Because of the limited response of GABA to TTX, Ca++ omission, and baclofen, some

26

investigators have questioned the neuronal origin of GABA in microdialysates. In this regard, it has been postulated that GABA measured by microdialysis may be derived from nonclassical neurotransmission, such as reversed uptake by carrier-mediated processes (Bernath and Zigmond, 1988; Pin and Bockaert, 1989) or from nonneuronal or cytoplasmic pools (Bernath and Zigmond, 1989), such as glial cells (Campbell et al., 1993; Timmerman and Westerink, 1997). IV.A. A critical comment on the analysis of GABA Because of the discrepancies listed above, we have recently investigated whether the chromatographic separation of GABA may have contributed to the controversial ﬁndings of numerous microdialysis studies on this neurotransmitter. To that end, we critically evaluated a method that was most frequently used by us and others: derivatization of dialysates with OPA-mercaptoethanol followed by HPLC and ﬂuorescence detection. To delay the retention time of the GABA derivative, the methanol content of the mobile phase was reduced. When the retention time of GABA was increased to 60 min, we noticed that there were several peaks coeluting close to GABA. These peaks were nondistinguishable from GABA, when the retention time was adjusted to values less than 30 min.In contrast to other amino acids and many interfering compounds, the retention time of the GABA derivative appeared to be very sensitive to the pH of the mobile phase. Fig. 3 shows that a small pH change in the order of 0.03 units moved the GABA peak over three other unknown peaks. It is emphasized that pH dependence of the retention of GABA derivative is often overlooked in the literature as the

applied pH of the mobile phase differs strongly among the various methods (pH 3.6–6.1). After having analyzed pooled microdialysis samples, using a large range of chromatographic conditions (pHs ranging from 6.00, 5.95, and so on; decreasing by 0.05 pH units, to 4.0) at the various methanol concentrations (30, 35, 40, 45, and 50%), it was concluded that stringent pH and methanol conditions are required. A pH of 5.2570.02 was the only window in the pH range that allowed a reliable separation. At our end, a retention time of 60 min is a prerequisite for a reliable chromatographic separation of GABA. With the adapted method, low basal GABA levels were obtained in microdialysis samples: 2.45, 2.95, and 3.6 fmol/mL sample/mm dialyzing membrane in dialysates from hippocampus, striatum, and prefrontal cortex, respectively. These results are signiﬁcantly lower than many values reported in the literature (range 1. 6–300 fmol/mL sample/ min; n ¼ 39). The low GABA levels that were obtained support the speciﬁcity of the present approach. Under the modiﬁed chromatographic conditions, it was observed for the ﬁrst time in our laboratory that TTX infusion (60 min) resulted in a consistent decrease of GABA in all brain regions studied. GABA decreased to 30% (hippocampus), 45% (striatum), and 55% (prefrontal cortex) of controls. Even so, the omission of Ca++ (60 min) caused a decrease to 55, 60, and 67% of control GABA basal levels in microdialysates from hippocampus, striatum, and prefrontal cortex, respectively (see for hippocampus, Fig. 4). This indicates that a signiﬁcant portion of GABA measured by the modiﬁed analytical approach is indeed of neuronal origin.

Fig. 3. The effect of minor changes in pH on the chromatographic behavior of GABA in microdialysates.

27

Fig. 4. Effect of infusion of TTX or calcium-free Ringer on extracellular levels of GABA in the hippocampus.

Fig. 5. Effect of infusion of baclofen on extracellular levels of GABA in the hippocampus.

Next, we investigated the controversy in the literature about the inability to detect effects of autoreceptor agonists, such as baclofen. With the adapted analytical approach, infusion of baclofen decreased the GABA levels to 35% of controls in various brain regions (see for hippocampus, Fig. 5). These results are in good agreement with data from in vitro studies (Waldmeier et al., 1988; Baumann et al., 1990; Lanza et al., 1993; Lambert and Wilson, 1994). It is unclear to what extent GABA values reported in the literature were inﬂuenced by an incomplete chromatographic separation. For example, we have not studied the chromatographic conditions in conjunction with electrochemical detection. But the fact that we could easily and consistently demonstrate effects of TTX, calcium omission, and baclofen might inspire other workers to inspect critically their chromatographic conditions. It is unclear at the present time, whether the methods based on CE provides the required speciﬁcity of GABA in microdialysates. The best way to investigate this issue is to perform microdialysis experiments during which extracellular levels of GABA are supposed to decline (e.g., administration of TTX, depletion of calcium or infusion of baclofen). As far as we know, these experiments have not been published till date. It is interesting to note that in a recent study that used

CE, GABA levels decreased to 40% of controls after infusion of a GABA synthesis inhibitor (Sauvinet et al., 2003). IV.B. On the origin of GABA in microdialysates Although we have demonstrated that GABA levels established during the infusion of TTX, baclofen, or calcium omission consistently decreased, its response is still relatively slow and does not go further down to 40% of controls. It is evident that GABA levels in microdialysates are less responsive when compared with the ‘‘slow-acting’’ neurotransmitters, such as dopamine (compare dopamine in Fig. 2 with GABA in Fig. 4). Here, we refer to the discussion on the origin of glutamate in dialysates (see above). The fact that GABA–similar to glutamate–after being released is taken up by high-afﬁnity transporters into glial cells that might lead to a similar explanation as given for glutamate. It is likely that the microdialysis probe is in much closer contact with the overwhelming majority of glial cells and subsequently, with the glial GABA pool rather than with the released transmitter in the synaptic cleft. In other words, microdialysis probes can only indirectly detect synaptic release by sampling the accumulating GABA in glial cells. We hypothesize that the GABA source in glial cells acts like a buffer that controls the GABA concentration in

28

the ECF. Dilution of the glial pool buffer may explain the time delay in the changes of extracellular GABA shown in Figs. 4 and 5. V. Conclusions It is unlikely that basal values of glutamate in dialysates represent neuronal release. Evidence is accumulating that basal glutamate is derived from different astrocytic pools. More research is indicated to resolve the relation between glutamate in dialysis samples and glutamate neurotransmission. During excessive neuronal activity, glutamate might escape from the synaptic cleft and might have access to the dialysis probe. Pathological release processes, resulting from tissue damage caused by the dialysis probe, might contribute signiﬁcantly to glutamate levels in microdialysis samples. The latter assumption is supported by the ﬁnding that microsensors are able to detect TTXsensitive glutamate. The chromatographic separation of the OPA derivative of GABA is crucial for reliable detection of GABA in dialysates. We hypothesize that the transfer of GABA from synaptic sites to the dialysis probe is delayed because the probe is sampling GABA partly derived from glial cells. References Allichin, R.E., Batten, T.F., McWilliams, P.N. and Vaughan, P.F. (1994) Electrical stimulation of the vagus increases extracellular glutamate recovered from the nucleus tractus solitarii of the cat by in vivo microdialysis. Exp. Physiol., 79: 265–268. Allison, L.A., Mayer, G.S. and Shoup, R.E. (1984) o-Phthalaldehyde derivatives of amines for high-speed liquid chromatography/electrochemistry. Anal. Chem., 56: 1089–1096. Amara, S.G. and Fontana, A.C.K. (2002) Excitatory amino acid transporters: keeping up with glutamate. Neurochem. Int., 41: 313–318. Arnth-Jensen, N., Jabaudon, D. and Scanziani, M. (2002) Cooperation between independent hippocampal synapses is controlled by glutamate uptake. Nat. Neurosci., 5: 325–331. Baker, D.A., Xi, Z.X., Shen, H., Swanson, C.J. and Kalivas, P.W. (2002) The origin and neuronal function of in vivo nonsynaptic glutamate. J. Neurosci., 22: 9134–9141. Baumann, P.A., Wicki, P., Stierlin, C. and Waldmeier, P.C. (1990) Investigations on GABAB receptor-mediated autoinhibition of GABA release. Naunyn Schmiedebergs Arch. Pharmacol., 341: 88–93.

Benveniste, H., Dreje, J., Schousbroe, A. and Diemer, H. (1984) Elevation of the extracellular concentrations of glutamate and aspartate in rat hippocampus during transient cerebral ischemia monitored by intracerebral microdialysis. J. Neurochem., 43: 1369–1374. Bergles, D.E., Diamond, J.S. and Jair, C.E. (1999) Clearance of glutamate inside the synapse and beyond. Curr. Opin. Neurobiol., 9: 293–298. Bernath, S. and Zigmond, M.J. (1988) Characterization of [3H]GABA release from striatal slices: evidence for a calciumindependent process via the GABA uptake system. Neuroscience, 27: 563–570. Bernath, S. and Zigmond, M.J. (1989) Dopamine may inﬂuence striatal GABA release via three separate mechanisms. Brain Res., 476: 373–376. Bezzi, P., Carmignoto, G., Pasti, L., Vesce, S., Rossi, D., Rizzini, B.L., Pozzan, T. and Volterra, A. (1998) Prostaglandins stimulate calcium-dependent glutamate release in astrocytes. Nature, 391: 281–285. Bezzi, P., Gundersen, V., Galbete, J.L., Seifert, G., Steinhauser, C., Pilati, E. and Volterra, A. (2004) Astrocytes contain a vesicular compartment that is competent for regulated exocytosis of glutamate. Nat. Neurosci., 7: 613–620. Bourdelais, A.J. and Deutch, A.Y. (1994) The effects of haloperidol and clozapine on extracellular GABA levels in the prefrontal cortex of the rat: an in vivo microdialysis study. Cereb. Cortex, 6: 69–77. Bourdelais, A.J. and Kalivas, P.W. (1992) Modulation of extracellular gamma-aminobutyric acid in the ventral pallidum using in vivo microdialysis. J. Neurochem., 58: 2311–2320. Campbell, K., Kalen, P., Wictorin, K., Lundberg, C., Mandel, R.J. and Bjorklund, A. (1993) Characterization of GABA release from intrastriatal striatal transplants: dependence on host-derived afferents. Neuroscience, 53: 403–415. Cavelier, P. and Atwell, D. (2005) Tonic release of glutamate by a DIDS-sensitive mechanism in rat hippocampal slices. J. Physiol., 564: 397–410. Chen, Y. and Swanson, R.A. (2003) The glutamate transporters EAAT2 and EAAT3 mediate cysteine uptake in cortical neuron cultures. J. Neurochem., 84: 1332–1339. Danbolt, N.C. (2001) Glutamate uptake. Prog. Neurobiol., 65: 1–105. Day, B.K., Pomerleau, F., Burmeister, J.J., Huettl, P. and Gerhardt, G.A. (2006) Microelectrode array studies of basal and potassium-evoked release of L-glutamate in the anesthetized rat brain. J. Neurochem., 96: 1626–1635. Diamond, J.S. and Jahr, C.E. (1997) Transporters buffer synaptically released glutamate on a submillisecond time scale. J. Neurosci., 17: 4672–4687. Dijk, S.N., Francis, P.T., Stratmann, G.C. and Bowen, D.M. (1995) NMDA-induced glutamate and aspartate release from rat cortical pyramidal neurones: evidence for modulation by a 5-HT1A antagonist. Br. J. Pharmacol., 115: 1169–1174. Drew, K.L. and Ungerstedt, U. (1991) Pergolide presynaptically inhibits calcium-stimulated release of gamma-aminobutyric acid. J. Neurochem., 57: 1927–1930.

29 Feenstra, M.G.P. and Botterblom, M.H.A. (1996) Rapid sampling of extracellular dopamine in the rat prefrontal cortex during food consumption, handling and exposure to novelty. Brain Res., 742: 17–24. Ferraro, L., O’Connor, W.T., Glennon, J., Tomasini, M.C., Bebe, B.W., Tanganelli, S. and Antonelli, T. (2000) Evidence for a nucleus accumbens CCK2 receptor regulation of rat ventral pallidal GABA levels: a dual probe microdialysis study. Life Sci., 68: 483–496. Fiorino, A., Frigo, G. and Cucchetti, E. (1989) Liquid chromatographic analysis of amino and imino acids in protein hydrolysates by post-column derivatization with o-phthalaldehyde and 3-mercaptopropionic acid. J. Chromatogr., 476: 83–92. Hansson, E., Muyderman, H., Leonova, J., Allansson, L., Sinclair, J., Blomstrand, F., Thorlin, T., Nilsson, M. and Ronnback, L. (2000) Astroglia and glutamate in physiology and pathology: aspects on glutamate transport, glutamateinduced cell swelling and gap-junction communication. Neurochem. Int., 37: 317–329. Hawkins, R.A., DeJoseph, M.R. and Hawkins, P.A. (1995) Regional brain glutamate transport in rats at normal and raised concentrations of circulating glutamate. Cell Tissue Res., 281: 207–214. Haydon, P.G. (2001) Glia: listening and talking to the synapse. Nat. Rev. Neurosci., 2: 185–193. Herrera-Marschitz, M., You, Z.B., Goiny, M., Meana, J., Silveira, S., Godukhin, O.V., Chen, Y., Espinoza, S., Petterson, E., Loidl, C.F., Lubec, G., Andersson, K., Nylander, I., Terenius, L. and Ungerstedt, U. (1996) On the origin of extracellular glutamate levels monitored in the basal ganglia of the rat by in vivo microdialysis. J. Neurochem., 66: 1726–1735. Hertz, L. and Zielke, H.R. (2004) Astrocytic control of glutamatergic activity: astrocytes as stars of the show. Trends Neurosci., 27: 735–743. Jacobs, W.A., Leburg, M.W. and Madaj, E.J. (1986) Stability of o-phthalaldehyde-derived isoindoles. Anal. Biochem., 156: 334–340. Kapoor, K., Nakahara, D., Blood, R.J. and Chalmers, J.P. (1990) Preferential release of neuroactive amino acids from the ventrolateral medulla of the rat in vivo as measured by microdialysis. Neuroscience, 37: 187–191. Katayama, Y., Becker, P., Tamura, T. and Hovda, D. (1990) Massive increases in extracellular potassium and the indiscriminate release of glutamate following concussive brain injury. J. Neurosurg., 73: 889–900. Kimelberg, H.K., Rutledge, E., Goderie, S. and Charniga, C. (1995) Astrocytic swelling due to hypotonic or high K+ medium causes inhibition of glutamate and aspartate uptake and increases their release. J. Cereb. Blood Flow Metab., 15: 409–416. Kulagina, N.V., Shankar, L. and Michael, A.C. (1999) Monitoring glutamate and ascorbate in the extracellular space of brain tissue with electrochemical microsensors. Anal. Chem., 71: 5093–5100. Lada, M.W., Vickroy, T.W. and Kennedy, R.T. (1997) High temporal resolution monitoring of glutamate and aspartate in vivo using microdialysis on-line with capillary electrophoresis with laser-induced ﬂuorescence detection. Anal. Chem., 69: 4560–4565.

Lada, M.W., Vickroy, T.W. and Kennedy, R.T. (1998) Evidence for neuronal origin and metabotropic receptor-mediated regulation of extracellular glutamate and aspartate in rat striatum in vivo following electrical stimulation of the prefrontal cortex. J. Neurochem., 70: 617–625. Lambert, N.A. and Wilson, W.A. (1994) Temporally distinct mechanisms of use-dependent depression at inhibitory synapses in the rat hippocampus in vitro. J. Neurophysiol., 72: 121–130. Lanza, M., Fassio, A., Gemignani, A., Bonanno, G. and Raitiri, M. (1993) CGP 52432: a novel potent and selective GABAB autoreceptor antagonist in rat cerebral cortex. Eur. J. Pharmacol., 24: 191–195. Lehre, K.P. and Danbolt, N.C. (1998) The number of glutamate transporter subtype molecules at glutamatergic synapses: chemical and stereological quantiﬁcation in young adult rat brain. J. Neurosci., 18: 8751–8757. Liu, N., Ho, I. and Rockhold, R.W. (1999) Contribution of glutamatergic systems in locus coeruleus to nucleus paragigantocellularis stimulation-evoked behavior. Pharmacol. Biochem. Behav., 63: 555–567. Miele, M., Boutelle, M.G. and Fillenz, M. (1996) The source of physiologically stimulated glutamate efﬂux from the striatum of conscious rats. J. Physiol., 497: 745–751. Millan, M.H., Obrenovitch, T.P., Sarna, G.S., Lok, S.Y., Symon, L. and Meldrum, B.S. (1991) Changes in rat brain extracellular glutamate concentration during seizures induced by systemic picrotoxin or focal bicuculline injection: an in vivo dialysis study with on-line enzymatic detection. Epilepsy Res., 9: 86–91. Morari, M., O’Connor, W.T., Ungerstedt, U. and Fuxe, K. (1993) N-Methyl-D-aspartic acid differentially regulates extracellular dopamine, GABA, and glutamate levels in the dorsolateral neostriatum of the halothane-anesthetized rat: an in vivo microdialysis study. J. Neurochem., 60: 1884–1893. Nedergaard, M., Takano, T. and Hansen, J. (2002) Beyond the role of glutamate as a neurotransmitter. Nat. Rev. Neurosci., 3: 748–755. Niwa, O., Torimitsu, K., Morita, K., Osborne, P.G. and Yamamoto, K. (1996) Concentration of extracellular L-glutamate released from cultured nerve cells measured with a small-volume online sensor. Anal. Chem., 68: 1865–1870. Obrenovitch, T.P., Richards, D.A., Sarna, G.S. and Symon, L. (1993) Combined intracerebral microdialysis and electrophysiological recording: methodology and applications. J. Neurosci. Methods, 47: 139–145. Oldenziel, W.H., Beukema, W. and Westerink, B.H.C. (2004) Improving the reproducibility of hydrogel-coated glutamate microsensors by using an automated dipcoater. J. Neurosci. Methods, 140: 117–126. Oldenziel, W.H., Dijkstra, G., Cremers, T.I.F.H. and Westerink, B.H.C. (in press) In vivo monitoring of extracellular glutamate in the brain with a microsensor. Brain Res. Oliet, S.H.R., Piet, R. and Poulain, D.A. (2001) Control of glutamate clearance and synaptic efﬁcacy by glial coverage of neurons. Science, 292: 923–936.

30 O’Shea, T.J., Weber, P.L., Bammel, B.P., Lunte, C.E., Lunte, S.M. and Smyth, M.R. (1992) Monitoring excitatory amino acid release in vivo by microdialysis with capillary electrophoresis-electrochemistry. J. Chromatogr., 608: 189–195. Osborne, P.G., O’Connor, W.T., Drew, K.L. and Ungerstedt, U. (1990) An in vivo microdialysis characterization of extracellular dopamine and GABA in dorsolateral striatum of awake freely moving and halothane anaesthetised rats. J. Neurosci. Methods, 34: 99–105. Paleckova, V., Palecek, J., McAdoo, D.J. and Willis, W.D. (1992) The non-NMDA antagonist CNQX prevents release of amino acids into the rat spinal cord dorsal horn evoked by sciatic nerve stimulation. Neurosci. Lett., 148: 19–22. Palmer, A.M., Hutson, P.H., Lowe, S.L. and Bowen, D.M. (1989) Extracellular concentrations of aspartate and glutamate in rat neostriatum following chemical stimulation of frontal cortex. Exp. Brain Res., 75: 659–663. Parpura, V., Basarsky, T.A., Liu, F., Jeftinija, K., Jeftinija, S. and Haydon, P.G. (1994) Glutamate-mediated astrocyteneuron signalling. Nature, 369: 744–747. Pena, F. and Tapia, R. (2000) Seizures and neurodegeneration induced by 4-aminopyridine in rat hippocampus in vivo: role of glutamate- and GABA-mediated neurotransmission and of ion channels. Neuroscience, 101: 547–561. Pin, J.P. and Bockaert, J. (1989) Two distinct mechanisms, differentially affected by excitatory amino acids, trigger GABA release from fetal mouse striatal neurons in primary culture. J. Neurosci., 9: 648–656. Pow, D.V. and Crook, D.R. (1996) Direct immunocytochemical evidence for the transfer of glutamine from glial cells to neurons: use of speciﬁc antibodies directed against the D-stereoisomers of glutamate and glutamine. Neuroscience, 70: 295–302. Quan, Z. and Liu, Y.-M. (2003) Capillary electrophoretic separation of glutamate enantiomers in neural samples. Electrophoresis, 24: 1092–1096. Rada, P., Mendialdua, A., Hernandez, L. and Hoebel, B.G. (2003) Extracellular glutamate increases in the lateral hypothalamus during meal initiation, and GABA peaks during satiation: microdialysis measurements every 30 s. Behav. Neurosci., 117: 222–227. Rakovska, A., Giovannini, M.G., Della Corte, L., Kalﬁn, R., Bianchi, R. and Pepeu, G. (1998) Neurotensin modulation of acetylcholine and GABA release from the rat hippocampus: an in vivo microdialysis study. Neurochem. Int., 33: 335–340. Raiteri, L., Raiteri, M. and Bonanno, G. (2002) Coexistence and function of different neurotransmitter transporters in the plasma membrane of CNS neurons. Prog. Neurobiol., 68: 287–309. Rea, K., Cremers, T.I. and Westerink, B.H. (2005) HPLC conditions are critical for the detection of GABA by microdialysis. J. Neurochem., 94: 672–679. Reid, M.S., Herrera-Marschitz, M., Kehr, J. and Ungerstedt, U. (1990) Striatal dopamine and glutamate release: effects of intranigral injections of substance P. Acta Physiol. Scand., 140: 527–537.

Reyes, E., Rosell, S., Paredes, D., Rada, P., Tucci, S., Gonzalez, L.E. and Hernandez, L. (2002) Haloperidol abolished glutamate release evoked by photic stimulation of the visual cortex in rats. Neurosci. Lett., 327: 149–152. Richards, D.A., Lemos, T., Whitton, P.S. and Bowery, N.G. (1995) Extracellular GABA in the ventrolateral thalamus of rats exhibiting spontaneous absence epilepsy: a microdialysis study. J. Neurochem., 65: 1674–1680. Rosell, S., Gonzalez, L.E. and Hernandez, L. (2003) One-second time resolution brain microdialysis in fully awake rats. Protocol for the collection, separation and sorting of nanoliter dialysate volumes. J. Chromatogr. B, 784: 385–393. Rusakov, D.A. and Lehre, K.P. (2002) Perisynaptic asymmetry of glia: new insights into glutamate signalling. Trends Neurosci., 25: 492–494. Saulskaya, N.B. and Mikhailova, M.O. (2004) Increased glutamate release into the intercellular space of the nucleus accumbens (N. accumbens) during substitution of food reinforcement with aversive or neutral stimuli. Behav. Physiol., 34: 109–113. Sauvinet, V., Parrot, S., Benturquia, N., Bravo-Moraton, E., Renaud, B. and Denoroy, L. (2003) In vivo simultaneous monitoring of gamma-aminobutyric acid, glutamate, and L-aspartate using brain microdialysis and capillary electrophoresis with laser-induced ﬂuorescence detection: analytical developments and in vitro/in vivo validations. Electrophoresis, 24: 3187–3196. Shi, G., Yamamoto, K., Zhou, T., Xu, F., Kato, T., Ji-ye, J. and Jin, J. (2003) On-line biosensors for simultaneous determination of glucose, choline, and glutamate integrated with a microseparation system. Electrophoresis, 24: 3266–3272. Shiraishi, M., Kamiyama, Y., Huttemeier, O.C. and Benveniste, H. (1997) Extracellular glutamate and dopamine measured by microdialysis in the rat striatum during blockade of synaptic transmission in anesthetized and awake rats. Brain Res., 759: 221–227. Silva, E., Hernandez, L., Quinonez, B., Gonzalez, L.E. and Colasante, C. (2004) Selective amino acids changes in the medial and lateral preoptic area in the formalin test in rats. Neuroscience, 124: 395–404. Singewald, N., Guo, L. and Phillipu, A. (1993) Release of endogenous GABA in the posterior hypothalamus of the conscious rat; effects of drugs and experimentally induced blood pressure changes. Naunyn Schmiedenergs Arch. Pharamcol., 347: 402–406. Smith, S. and Sharp, T. (1994) Measurement of GABA in rat brain microdialysates using o-phthaldialdehyde-sulphite derivatization and high-performance liquid chromatography with electrochemical detection. J. Chromatogr. B, 652: 228–233. Sorkin, L.S. and McAdoo, D.J. (1993) Amino acids and serotonin are released into the lumbar spinal cord of the anesthetized cat following intradermal capsaicin injections. Brain Res., 607: 89–98. Stengard, K. and O’Connor, W.T. (1994) Acute toluene exposure decreases extracellular gamma-aminobutyric acid in the globus pallidus but not in striatum: a microdialysis study in awake, freely moving rats. Eur. J. Pharmacol., 292: 43–46.

31 Stobaugh, J.F., Repta, A.J., Sternson, L.A. and Garren, K.W. (1983) Factors affecting the stability of ﬂuorescent isoindoles derived from reaction of o-phthalaldehyde and hydroxyalkylthiols with primary amines. Anal. Biochem., 135: 495–504. Timmerman, W. and Westerink, B.H.C. (1995) Extracellular gamma-aminobutyric acid in the substantia nigra reticulata measured by microdialysis in awake rats: effects of various stimulants. Neurosci. Lett., 197: 21–24. Timmerman, W. and Westerink, B.H.C. (1997) Brain microdialysis of GABA and glutamate: what does it signify? Synapse, 27: 242–261. Tucci, S., Contreras, Q., Paez, X., Gonzalez, L., Rada, P. and Hernandez, L. (2000) Medial prefrontal transection enhances social interaction. II: neurochemical studies. Brain Res., 887: 259–265. Volterra, A. and Meldolesi, J. (2005) Astrocytes, from brain glue to communication elements: the revolution continues. Nat. Rev. Neurosci., 6: 626–640. Waldmeier, P.C., Stocklin, K. and Feldtrauer, J.J. (1992) Weak effects of local and systemic administration of the GABA uptake inhibitor, SK&F 89976, on extracellular GABA in the rat striatum. Naunyn Schmiedebergs Arch. Pharmacol., 345: 544–547. Waldmeier, P.C., Wicki, P., Feldtrauer, J.J. and Baumann, P.A. (1988) Potential involvement of a baclofen-sensitive autoreceptor in the modulation of the release of endogenous

GABA from rat brain slices in vitro. Naunyn Schmiedebergs Arch. Pharmacol., 337: 289–295. Wang, X., Arcuino, G., Takano, T., Lin, J., Peng, W.G., Wan, P., Li, P., Xu, Q., Liu, Q.S., Goldman, S.A. and Nedergaard, M. (2004) P2X7 receptor inhibition improves recovery after spinal cord injury. Nat. Med., 10: 821–827. Warr, O., Takahashi, M. and Attwell, D. (1999) Modulation of extracellular glutamate concentration in rat brain slices by cystine-glutamate exchange. J. Physiol., 514: 783–793. Westerink, B.H.C., Tuntler, J., Damsma, G., Rollema, H. and de Vries, J.B. (1987) The use of tetrodotoxin for the characterization of drug-enhanced dopamine release in conscious rats studied by brain dialysis. Naunyn Schmiedebergs Arch. Pharmacol., 336: 502–507. Xi, Z.X., Shen, H., Baker, D.A. and Kalivas, P.W. (2003) Inhibition of non-vesicular glutamate release by group III metabotropic glutamate receptors in the nucleus accumbens. J. Neurochem., 87: 1204–1212. Zhou, F., Braddock, J.F., Hu, Y., Zhu, X., Castellani, R.J., Smith, M.A. and Drew, K.L. (2002) Microbial origin of glutamate, hibernation and tissue trauma: an in vivo microdialysis study. J. Neurosci. Methods,, 119: 121–128. Zilkha, E., Obrenovitch, T., Koshy, A., Kusakabe, J. and Bennetto, H. (1995) Extracellular glutamate: on-line monitoring using microdialysis coupled to enzyme-amperometric analysis. J. Neurosci. Methods,, 60: 1–9.

CHAPTER 1.3

Insights into glutamate physiology: contribution of studies utilizing in vivo microdialysis David A. Baker1 and Peter W. Kalivas2, 2

1 Department of Biomedical Sciences, Marquette University, Milwaukee, WI, USA Department of Physiology and Neuroscience, Medical University of South Carolina, Charleston, SC, USA

Abstract: Glutamate as the primary excitatory neurotransmitter is critical to most aspects of brain functioning in the normal and diseased states and as a result, there is great interest in monitoring in vivo ﬂuctuations. Microdialysis is the most commonly used technique to sample in vivo glutamate levels, however, the signiﬁcance of dialysate glutamate have been questioned because of the myriad of physiological functions involving glutamate as well as the widespread observation that basal levels of glutamate are primarily maintained by nonvesicular release mechanisms. The present chapter reviews the cycling of glutamate and synthesis-precursors such as glutamine and demonstrates that it is difﬁcult to delineate a separate metabolic and neurotransmitter pool of glutamate in the extrasynaptic space. Further, evidence is presented indicating that synaptic and extrasynaptic pools of glutamate are capable of modulating neurotransmission by stimulating glutamate receptors. Potential cellular mechanisms underlying vesicular and nonvesicular glutamate release are reviewed regarding the extent to which each may contribute to glutamate signaling. It is concluded that studies utilizing in vivo microdialysis to monitor extrasynaptic pools of glutamate are complimentary to electrophysiological approaches that monitor synaptic glutamate, and that in vivo microdialysis can be used to advance our understanding of glutamate signaling in the normal and diseased states; indeed, future studies utilizing in vivo microdialysis have the potential to identify novel therapeutic targets for a variety of neurological disorders. neurotransmitter in the brain are complicated by the various physiological functions involving glutamate that may be unrelated to its role as a neurotransmitter. The relevance of dialysate glutamate levels have been questioned due to the various metabolic functions of glutamate coupled with the widespread observation that basal levels of glutamate sampled using in vivo microdialysis are primarily maintained by nonvesicular release mechanisms (Bradford et al., 1987; Miele et al., 1996; Timmerman and Westerink, 1997; Baker et al., 2003). The scrutiny regarding the relevance of nonvesicular glutamate has contributed, in part, to studies seeking to further characterize cellular processes regulating synaptic glutamate. The beneﬁt of these efforts is that identifying the relevance

Efforts to unravel the pathology underling a myriad of neurological disorders, including ALS, ischemia, drug addiction, and schizophrenia will likely require a complete understanding of the cellular processes contributing to glutamate neurotransmission in the normal and diseased states. Glutamate is ubiquitously expressed throughout the mammalian central nervous system where it functions as the primary excitatory neurotransmitter. Given its vital role in most aspects of brain functioning, there is great interest in monitoring in vivo glutamate ﬂuctuations. Attempts to sample glutamate functioning as the primary excitatory Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

33

DOI: 10.1016/S1569-7339(06)16003-8 Copyright 2007 Elsevier B.V. All rights reserved

34

of extrasynaptic glutamate has expanded our understanding of glutamate signaling and may have identiﬁed novel therapeutic targets in the treatment of addiction and schizophrenia which is reminiscent of Eugene Ionesco’s sentiment, ‘‘It is not the answer that enlightens but the question’’. Collectively, these studies have provided a mere hint at the fascinating, complex regulation of the cellular processes underlying glutamate signaling and indicates that the uniqueness of these processes precludes constraints that might deﬁne the relevance of traditional neurotransmitters. The mammalian central nervous system contains abundant glutamate, with neuronal intracellular concentrations in the millimolar range and extrasynaptic levels ranging between 0.2 and 7 mM in rats (Benveniste et al., 1984; Ronne-Engstrom et al., 1995; Baker et al., 2003). The mean distance between synapses, as estimated in the hippocampal dentate gyrus, is 0.5 mm (Rusakov and Kullmann, 1998; Ventura and Harris, 1999). Within this space, there are numerous cellular processes capable of regulating glutamate signaling; the complexity of the cellular machinery regulating glutamate neurotransmission is evident from the sheer number of receptors, uptake transporters, and release mechanisms (for review see Danbolt, 2001). The identiﬁcation of these processes is critical to our understanding of how glutamate signaling contributes to brain functioning in the normal and diseased states and must be appreciated when examining the relevance of extrasynaptic glutamate.

I. Glutamate sampling techniques A variety of techniques are available to sample in vivo levels of glutamate, including microdialysis and voltammetry (Hutchinson et al., 2002; Fillenz, 2005). The noninvasive procedures nuclear magnetic resonance spectroscopy (Had-Aissouni et al., 2002) represents a newer approach, however, few studies have used this technique, and as a result, it is difﬁcult to compare ﬁndings obtained with this technique to microdialysis and voltammetry. Microdialysis and voltammetry represent complementary rather than redundant procedures as each makes a unique contribution to the ﬁeld of in

vivo neurochemistry on the basis of respective advantages and disadvantages. Advantages of voltammetry include temporal resolution and vastly reduced tissue damage. The signiﬁcance of reduced tissue damage has been suggested to contribute to the observation that glutamate measured using carbon ﬁber electrodes coated with glutamate oxidase detects a much higher basal value (29 mM; Kulagina et al., 1999) than what is obtained with microdialysis (0.2–7.0 mM; Benveniste et al., 1984; Ronne-Engstrom et al., 1995; Baker et al., 2003). Further, basal levels sampled using in vivo voltammetry are at least partially reﬂective of synaptic release since levels are reduced by 25–80% following infusion of 100 mM TTX (Kulagina et al., 1999). This is in contrast to the ﬁndings using microdialysis in which basal levels are typically insensitive to TTX (Bradford et al., 1987; Miele et al., 1996; Timmerman and Westerink, 1997; Baker et al., 2003), although much lower concentrations of TTX have been used (e.g., 10 mM; Verma and Moghaddam, 1998; Baker et al., 2002). It is important to note that reductions in basal glutamate levels in the striatum sampled using microdialysis have been obtained using inhibitors of voltage-gated calcium channels (Baker et al., 2002; Xi et al., 2002) or following calcium depletion (Dawson et al., 1995), but see Westerink et al. (1989) and You et al. (1994). These data indicate that both in vivo voltammetry and microdialysis sample vesicular and nonvesicular glutamate. A second advantage of voltammetry relates to temporal resolution; voltammetry permits rapid sampling as short as 100 ms. Microdialysis typically involves sampling times of 5–20 min, although this is dictated by the analytical technique used to determine analyte concentrations. Indeed, the use of capillary electrophoresis coupled with laserinduced ﬂuorescence permits rapid microdialysis sampling times of p5 s (Lada et al., 1998; Rossell et al., 2003). Voltammetry also results in less tissue damage since the diameter of the electrodes range between 1 and 30 mm whereas microdialysis probes are typically X200 mm. Advantages of microdialysis include the range of chemicals that can be sampled, ability to perfuse surrounding tissue with various compounds, and enhanced chemical resolution. As a result, in vivo microdialysis has seen

35

a wider range of applications than in vivo voltammetry. Owing to the physical limitations, in vivo microdialysis and voltammetry are incapable of directly measuring neurotransmitter levels released into the synapse. The diameter of electrodes used for voltammetry (1–30 mm) or microdialysis probes (200–300 mm; see Drew et al., 2004) greatly exceed the size of the synaptic cleft (15 nm), at least as estimated in the CA1 region of the hippocampus (Glavinovic´, 1999). As a result, both techniques sample extrasynaptic pools of neurotransmitters. The extrasynaptic pool of classical neurotransmitters, such as dopamine, largely consists of synaptic overﬂow; thereby permitting a measure of synaptic ﬂuctuations. This is evident from studies that have shown that blockade of voltage-gated sodium and calcium channels, which prevents exocytosis, drastically reduces extrasynaptic levels of monoamines (Westerink et al., 1987; Osborne et al., 1991; Kato et al., 1992; Morari et al., 1993). However, in the case of amino acid neurotransmitters, it is evident that dialysate levels are maintained by both vesicular and nonvesicular release since levels are only partially sensitive to blockade of voltage-dependent Na+ and Ca2+ channels (Bradford et al., 1987; Miele et al., 1996; Timmerman and Westerink, 1997; Baker et al., 2003). Nonvesicular glutamate release has also been detected in hippocampus and prefrontal cortex using electrophysiological recordings (Jabaudon et al., 1999; Angulo et al., 2004; Cavelier and Attwell, 2005), indicating that nonvesicular release is not unique to the nucleus accumbens nor is it an artifact of microdialysis. The capacity of in vivo monitoring techniques to sample extrasynaptic glutamate provides complimentary measures to the capacity of electrophysiological techniques to monitor synaptic glutamate.

II. Astrocytes maintain neuronal glutamate The signiﬁcance of dialysate glutamate have been questioned because of the myriad of physiological functions involving glutamate as well as the widespread observation that basal levels of glutamate are primarily maintained by nonvesicular release

mechanisms. Metabolic functions range from protein or peptide synthesis to oxidative degradation yielding carbon dioxide and water for energy metabolism (Fonnum, 1993; Had-Aissouni et al., 2002). The compartmentalization of glutamate is necessary to permit the various physiological functions. The largest of these pools is known as the metabolic pool. A smaller neuronal pool is releasable from the nerve endings during neurotransmission and is thought to represent the glutamate neurotransmitter pool. The separate, glial pool is believed to serve the recycling of transmitter glutamate. Thus, the metabolic pool of glutamate is restricted to the intracellular domain of neurons, whereas the latter two pools involve the trafﬁcking of glutamate and precursor substrates that can be used by neurons to replenish neurotransmitter levels (Danbolt, 2001; Rothman et al., 2003). Indeed, studies using 13C nuclear magnetic resonance spectroscopy measuring the ﬂow of 13C label from glucose to glutamate to glutamine suggest that glutamate release and metabolic recycling cannot be distinguished from glutamate neurotransmission (Rothman et al., 2003). Thus, it has been argued that it is inaccurate to view metabolic and neurotransmitter glutamate as functionally distinct pools, but rather evidence of a highly coordinated interaction between neurons and astrocytes needed to maintain glutamate signaling (Rothman et al., 2003). It has long been appreciated that clearance of synaptically released glutamate represents an important contribution to glutamate signaling by astrocytes. Glutamate that is synaptically released by neurons is primarily cleared by diffusion and the activity of sodium-dependent glutamate transporters (XAG). The family of XAG transporters are located on both astrocytes (e.g., GLAST and GLT) and neurons (EAAC and EAAT4); however, the glial transporters GLT and GLAST are more active in clearing glutamate relative to the cloned neuronal transporters (for review see Danbolt, 2001). The capacity of glutamate uptake by astrocytes is evident in estimates (Lehre and Danbolt, 1998) that glial cells have 15,000–21,000 transporters per mm3 tissue, compared with 2,000 for the neuronal transporter EEAT4 (Dehnes et al., 1998). The need for this capacity to clear

36

glutamate is evident when considering that individual vesicles may contain 400–4,300 molecules of glutamate and that the average synapse has 20 release sites that can be reﬁlled every 10 s (Stevens and Tsujimoto, 1995). It is interesting to speculate that there is an evolutionary advantage underlying the predominant role for astrocytes in glutamate uptake since glutamate uptake also results in sodium uptake. The uptake of each glutamate molecule requires the cotransport of 2–3 Na2+ ions glutamate transport; thus the potential to transport up to 8,600 molecules of glutamate, and consequently up to 25,800 sodium molecules, per second would greatly increase neuronal metabolic demands to maintain sodium concentration gradients and membrane potentials. Regardless, the result is a high rate of glutamate trafﬁcking from neurons to astrocytes. Further, neurons, unlike astrocytes, cannot synthesis glutamate from glucose (Hertz et al., 1999) nor does glutamate easily penetrate the blood-brain barrier (McCall et al., 1979; Smith, 2000). Thus, neurons are dependent on astrocytes to maintain glutamate levels needed to support synaptic release. Astrocytes do not return glutamate to neurons, but instead supply precursors that can be used for glutamate synthesis. Precursors that can be utilized by astrocytes to synthesis glutamate include glutamine as well as a-ketoglutarate and tricarboxylic acid (TCA) cycle intermediates including lactate or pyruvate (Waagepetersen et al., 1998; Garcia-Espinosa et al., 2004). Glutamate-glutamine is thought to be the primary metabolic fate of synaptic glutamate taken into astrocytes. The cycle occurs when glutamate in astrocytes is amidated to glutamine by the addition of ammonia, a reaction catalyzed by the glial enzyme glutamine synthetase. Glutamine is then released into the extracellular space where it can be taken up into neurons and converted to glutamate by phosphatase-activated glutaminase (Kvamme et al., 1988; Danbolt, 2001); the advantage of using glutamine to restore neuronal glutamate levels is that glutamine will not produce excitotoxicity because it does not stimulate glutamate receptors. Evidence that glutamine is used to synthesize the neurotransmitter pool of glutamate is provided by studies indicating that blockade of

glutamine synthetase reduces glutamate release detected in vivo in the frontal cortex or hippocampal formation (Bo¨ttcher et al., 2003) or in the striatum (Paulsen and Fonnum, 1989). However, glutamine is not solely used as a precursor to maintain the neurotransmitter pool of glutamate; astrocytes as well as nonglutamatergic neurons take up glutamine from the extracellular space; additional uses of glutamine include the synthesis of GABA or as a metabolic fuel. Further, studies have demonstrated that at least some glutamate neurons lack the enzyme needed to convert glutamine into glutamate (Ottersen et al., 1998). Thus, additional substrates must be utilized by neurons to maintain glutamate levels since there is less than a 1:1 relationship between the precursor glutamine and synthesized glutamate. Neuronal glutamate levels can also be maintained by utilizing a-ketoglutarate or a-ketoglutarate products (Schousboe et al., 1997). a-Ketoglutarate in astrocytes is ultimately converted to pyruvate and subsequently lactate; all of these products can be released from astrocytes and taken up by neurons for the synthesis of glutamate (Waagepetersen et al., 1998; Hertz et al., 1999; Garcia-Espinosa et al., 2004). Interestingly, blocking glutamine synthetase also results in an increase in extracellular levels of lactate and pyruvate (Bo¨ttcher et al., 2003). These data indicate that the same pool of glutamate, most likely synaptically released glutamate, serves as a substrate for both glutamine and a-ketoglutarate metabolic pathways. In vivo microdialysis or voltammetry are unlikely to sample pools of glutamate utilized in the metabolic pathways glutamine and a-ketoglutarate metabolic pathways. The metabolic fate of glutamate in either astrocytes or neurons is restricted to intracellular pools of glutamate, and trafﬁcking of substrate for metabolic functions does not involve the transfer of glutamate, but rather precursors used for subsequent synthesis of glutamate. It is important to consider that the only point of extracellular transfer of glutamate to support either metabolic pathway is the synaptic release of glutamate from neurons which is then cleared by glial XAG transporters. This is conﬁrmed in studies using nuclear magnetic resonance spectroscopy to monitor the fate of 13C isotope

37

from a [1-13C] glucose precursor in terms of subsequent labeling of glutamine, lactate, and glutamate (for review see Rothman et al., 2003). These studies have demonstrated that metabolic fate of glutamate in astrocytes is inexplicablely linked to large neurotransmitter pools in neurons. The need to parse out metabolic glutamate even in the extrasynaptic space from glutamate contributing to neurotransmission is apparent from observations that basal, extrasynaptic glutamate can inﬂuence signaling by stimulating extrasynaptic glutamate receptors.

III. Glutamate receptors Glutamate signaling involves activation of ionotropic (NMDA, AMPA, and kainate) and metabotropic (mGluR1–7) receptors; stimulation of these receptors underlies most of the excitatory neurotransmission in the mammalian brain. Given the importance of the activity of glutamate receptors to brain functioning, an important point to consider regarding the signiﬁcance of glutamate sampled using in vivo microdialysis is whether this pool of glutamate is stimulating glutamate receptors. Interestingly, ionotropic and metabotropic glutamate receptors are located both within the synaptic cleft as well as extrasynaptically, and extrasynaptic glutamate receptors can modulate the release of an impressive array of neurotransmitters including glutamate, GABA, substance P, dopamine, and norepinephrine (Ozawa et al., 1998; Cartmell and Schoepp, 2000; Gomes et al., 2003). This indicates the possibility that extrasynaptic glutamate maintained primarily by nonvesicular release contributes to signaling of a number of neurotransmitter systems. Excitatory postsynaptic currents (EPSCs) in most mammalian central neurons have a fast a-amino-3-hydroxy-5-methyl-4-isoazole-proprionic acid (AMPA) receptor-mediated component, lasting a few milliseconds, and a slow N-methyl-Daspartic acid (NMDA)-receptor-mediated component, lasting hundreds of milliseconds. Therefore, voltage-clamped glutamate EPSCs appear biphasic, with a very fast AMPA transient, and a much slower NMDA tail. As a result, both NMDA and

AMPA receptors are indispensable to brain functioning in the normal and diseased state. An interesting, puzzling attribute is that AMPA receptors are not only located within the synapse, but also are found on extrasynaptic neuronal and glial processes. It is unclear that AMPA receptors, which have relatively low afﬁnity for glutamate (EC50 ¼ 0.1–0.5 mM; Patneau and Mayer, 1990), would be directly stimulated by extrasynaptic glutamate levels which are routinely estimated in the low micromolar. However, AMPA receptors located on glial cells in the hippocampus or cerebellum appear to directly appose presynaptic neurons and recent data indicates that these receptors may be stimulated by quantal release of glutamate resulting in concentrations within the afﬁnity range of AMPA receptors (Bergles et al., 2000; Matsui and Jahr, 2003; Matsui et al., 2005). As a result of the poor temporal resolution associated with microdialysis coupled with HPLC analysis of analyte concentration, in vivo characterization of this effect will likely require the use of microdialysis coupled with capillary electrophoresis/laser-induced ﬂuorescence or in vivo voltammetry. Various attributes of the NMDA receptor poses the possibility that these receptors may be stimulated by extrasynaptic glutamate. First, in contrast AMPA receptors, NMDA receptors exhibit very high afﬁnity for glutamate that is within the range of extracellular glutamate. Extant data indicate that NMDA receptors are present in both the synaptic cleft and the extrasynaptic space and that the subcellular distribution may be determined by subunit composition; these receptors comprised NR1 and NR2 subunit, including NR2A, NR2B, NR2C, and NR2D. NMDA receptors comprised NR1/NR2A subunits localized in the synapse and are characterized by rapid offset kinetics while receptors comprised NR1/NR2B subunits which are extrasynaptic and exhibit slow kinetics (Stocca and Vicini, 1998; Rumbaugh and Vicini, 1999; Tovar and Westbrook, 1999). The offset kinetics may reﬂect separate sources of stimulation; the slow offset kinetics of extrasynaptic receptors may indicate stimulation from a pool of glutamate that is not rapidly cleared. However, these receptors, which are characterized by high permeability to

38

calcium, are unlikely to receive continual endogenous stimulation, in part because these receptors are subjected to voltage-dependent blockade by magnesium. Several studies have demonstrated that NMDA receptors can be stimulated by both vesicular and nonvesicular glutamate (Jabaudon et al., 1999; Angulo et al., 2004; Cavelier and Attwell, 2005). Some of these studies found that the stimulation of NMDA receptors by nonvesicular release was not apparent until glutamate uptake by XAG transporters was removed by the nonspeciﬁc glutamate uptake inhibitor DL-threo-beta-benzyloxyaspartate (TBOA; Jabaudon et al., 1999; Angulo et al., 2004; Cavelier and Attwell, 2005). Interestingly, NMDA receptors that were stimulated with glutamate uptake intact have been suggested to be extrasynaptic (Breukel et al., 1998; Angulo et al., 2004). Further, in the absence of magnesium blockade, these receptors appear to contribute to background current levels detected using patch clamp recordings indicating that extrasynaptic glutamate receptors may be subject to tonic stimulation (Gottesman and Miller, 2003; Angulo et al., 2004). Collectively, these studies pose the possibility that NMDA receptors may be stimulated by extrasynaptic glutamate, and that XAG transporters actively compartmentalize extrasynaptic and synaptic glutamate. The subcellular localization is an important determinant of the physiological function of the NMDA receptor. In support, several studies implicate extrasynaptic NMDA receptors as auto- or heteroreceptors capable of regulating the release of glutamate, GABA, and monoamines (Pittaluga and Raiteri, 1990; Krebs et al., 1991; Wang et al., 1992; Malva et al., 1994). Interestingly, the release of glutamate by extrasynaptic NMDA receptors can involve vesicular (Wang et al., 1992; Malva et al., 1994) or nonvesicular glutamate (Breukel et al., 1998). Given that NMDA receptors are not located on glia, this indicates that in addition to astrocytes, neurons are capable of ligand-mediated nonvesicular release. Metabotropic glutamate receptors are perhaps the most likely of all receptors to receive tonic stimulation by extrasynaptic glutamate levels. These receptors classiﬁed as groups I (mGluR1 and 5), II

(mGluR2 and 3), and III (mGluR4, 6, 7, and 8) exhibit a very high afﬁnity for glutamate that is comparable with NMDA receptors (Conn and Pin, 1997; Anderson and Swanson, 2000). Unlike NMDA receptors, however, these receptors are not subjected to voltage-dependent magnesium blockade. With the exception of mGluR1 and 6, each of the receptors are expressed either on presynaptic neurons (mGluR2, 4, 7, and 8) or on astrocytes (mGluR3 and 5) indicating that most receptors are in the extrasynaptic domain. Group I receptors are positively linked to phospholipase C and largely thought to augment ionotropic receptor function glutamate neurotransmission (Bordi and Ugolini, 1999). Groups II and III are negatively coupled to adenylyl cyclase and function as auto- and heteroreceptors capable of negative modulation of glutamate neurotransmission. As a result, metabotropic glutamate receptors have been shown to modulate synaptic glutamate signaling. Extant data utilizing in vivo microdialysis indicate that group II mGluRs receive tonic stimulation. The existence of endogenous stimulation of these receptors is evident from studies demonstrating that infusion of the group II mGluR antagonist (RS)-1-amino-5-phosphonoindan-1carboxylic acid (APICA) produces a signiﬁcant increase in extracellular glutamate (Baker et al., 2002), which is Ca2+-dependent (Xi et al., 2002). These data are consistent with the notion that group II mGluRs function as autoreceptors to regulate vesicular release of glutamate (Baskys and Malenka, 1991; Cochilla and Alford, 1998; Hu et al., 1999; Xi et al., 2002) or as heteroreceptors capable of regulating synaptic release of other neurotransmitters including glutamate and GABA (Conn and Pin, 1997; Cartmell and Schoepp, 2000). Interestingly, tonic stimulation of group II mGluR originates from nonvesicular glutamate release from cystine–glutamate exchange. In support, infusion of the cystine–glutamate exchange inhibitor (S)-4-carboxyphenylglycine (CPG) prevents the APICA-induced rise in glutamate. A similar conclusion can be drawn from the observation that cystine application to cortical or accumbens tissue slices produces a decrease in the frequency of mini EPSCs (Moran et al., 2005). Further, blockade of this effect by CPG

39

demonstrates that the reduction in the frequency of mini EPSCs depends on nonvesicular glutamate release from cystine–glutamate antiporters, and blockade of this effect by the group II mGluR antagonist LY341495 shows that the reduction in the frequency of mini EPSCs is also dependent on stimulation of group II mGluRs. Collectively, these data indicate that nonvesicular glutamate released from cystine–glutamate antiporters regulates glutamate neurotransmission by stimulating group II mGluRs. In addition, these data imply that the contribution of cystine–glutamate antiporters to glutamate neurotransmission occurs in both the prefrontal cortex and nucleus accumbens (Baker et al., 2002; Melendez et al., 2005), posing the possibility that cystine–glutamate antiporters may function in this capacity throughout the brain. These are signiﬁcant ﬁndings, in part, because glutamate released from cystine–glutamate exchange is sampled by in vivo microdialysis (Baker et al., 2002). Collectively, these data strongly indicate that microdialysis is sampling glutamate that is nonvesicular yet capable of regulating glutamate signaling. The extant data regarding receptor distribution, function, and kinetics support pharmacological data indicating that nonvesicular glutamate modulates neural signaling involving a wide-array of neurotransmitters by stimulating both ionotropic and metabotropic glutamate receptors. Fluctuations in nonvesicular glutamate may contribute to a variety of pathologies arising through altered signaling of these systems. As a result, it is important to characterize the cellular mechanisms underlying basal, extrasynaptic glutamate sampled using in vivo microdialysis because these mechanisms likely modulate glutamate signaling. As a result, these processes may have profound therapeutic signiﬁcance for a myriad of disorders.

IV. Cellular processes capable of glutamate release Synaptic, vesicular release from neurons is widely studied, however, as indicated earlier, measures of glutamate by in vivo microdialysis are maintained largely by nonvesicular release. Interestingly, there are number of cellular mechanisms capable of

glutamate release that may involve both neurons and astrocytes. These mechanisms include vesicular release from astrocytes (Bezzi et al., 2004; Kreft et al., 2004; Zhang et al., 2004), as well as nonvesicular release from cystine–glutamate exchange (Baker et al., 2002; Moran et al., 2005), astrocytic gap junction hemichannels (Ye et al., 2003), volume-sensitive organic anion channels (VSOACs; Kimelberg et al., 1990; Strange et al., 1996; Basarsky et al., 1999), hydrolysis of N-acetylaspartylglutamate (Blakely et al., 1988; Zhou et al., 2005), or reversal of sodium-dependent transporters (Levi and Raiteri, 1993). Further, there are recent reports indicating that glutamate may be synthesized from glutamine hydrolysis in the extrasynaptic space (Mena et al., 2005). Despite the importance of glutamate in both the synaptic and extrasynaptic space, surprisingly little is known regarding the relative importance of these mechanisms to glutamate signaling. However, nonvesicular release is detected using electrophysiological approaches indicating that it is not an artifact of microdialysis and that this pool of glutamate can stimulate glutamate receptors (Jabaudon et al., 1999; Angulo et al., 2004; Cavelier and Attwell, 2005; Moran et al., 2005). As a result, microdialysis is a critical tool in determining the contribution of these release mechanisms unique to the role that glutamate plays in vivo in normal and pathological brain functioning. Evidence that astrocytes are a key component of glutamate signaling stems from observations that these cells are capable of glutamate release involving exocytosis (for review see Haydon, 2001; Newman, 2003). Recently, a synaptic-like microvesicle capable of uptake, storage, and release of glutamate has been identiﬁed in hippocampal astrocytes (Bezzi et al., 2004; Kreft et al., 2004; Zhang et al., 2004). The loading of these vesicles, similar to neurons, involves the activity of vesicular glutamate transporters; the fusion of these vesicles to the plasma membrane has been shown to be calcium-dependent. Unlike neurons, the process is not dependent on the opening of voltage-gated calcium channels, but instead by IP3-mediated calcium release from endoplasmic reticulum following activation of a variety of G-coupled receptors (Volterra and Meldolesi,

40

2005). Receptors that may be capable of elevating intracellular levels of calcium that have been found on astrocytes include metabotropic glutamate, muscarinic, histamine, substance P, and noradrenergic receptors (for review see Verkhratsky et al., 1998). Stimulation of NMDA receptors on neurons contributes to vesicular glutamate release from neurons. In support, data obtained using immunogold electron microscopy indicate that extrasynaptic NMDA receptors on dendrites of cortical pyramidal cells face glial processes (Kharazia and Weinberg, 1999). Further, glutamate released from astrocytes results in inward currents that are susceptible to NMDA receptor inhibition (Parpura et al., 1994; Araque et al., 1998; Parpura and Haydon, 2000). The contribution of vesicular glutamate release from astrocytes to dialysate glutamate levels is unknown because traditional attempts to inhibit vesicular release in neurons, including blockade of voltage-dependent sodium and calcium channels, would not alter vesicular glutamate release from astrocytes. Given that this source of glutamate is released into the extrasynaptic space, it seems likely that dialysate glutamate concentrations sampled in vivo microdialysis may originate, in part, from vesicular glutamate release from astrocytes. These studies indicate the importance of microdialysis to contribute to investigate the contribution of these unique release mechanisms to in vivo glutamate signaling. Glutamate may also be released through hemichannels located on astrocytes. The opening of hemichannels resulting in glutamate release can be demonstrated in divalent cation-free solution (Ye et al., 2003) indicating that is most likely to contribute to glutamate release under pathological conditions. In support, ambient levels of calcium are thought to maintain hemichannels in the closed state, however, reductions in calcium may result in a small portion of hemichannel opening during normal physiological conditions (Liu et al., 1995; Hofer and Dermietzel, 1998; Kamermans et al., 2001). As a result, additional studies examining changes in basal glutamate levels following blockade of hemichannels has the potential to contribute to our understanding of glutamate signaling.

The cystine–glutamate antiporter represents a nonvesicular release mechanism of glutamate release that appears to be critical for synaptic glutamate transmission (Baker et al., 2002; Moran et al., 2005). The antiporter is a plasma membrane bound, Na+-independent, anionic amino acid transporter that exchanges extracellular cystine for intracellular glutamate (Bannai et al., 1986; Danbolt, 2001) and exists as two separate proteins, the light chain xCT unique to cystine–glutamate antiporters and the heavy chain 4F2 that is common to many amino acid transporters (Sato et al., 1999; Bridges et al., 2001). Similar to Na+-dependent glutamate transporters, the antiporter is ubiquitously distributed throughout the body; and in the brain, the antiporter is thought to be located on both neurons and glia (Cho and Bannai, 1990; Murphy et al., 1990; Danbolt, 2001). Evidence suggests that nonvesicular glutamate release via cystine–glutamate exchange is a fundamental component of glutamate neurotransmission in the nucleus accumbens (see Fig. 1). First, nonvesicular glutamate release is the primary source of glutamate in the nucleus accumbens, as assessed using in vivo microdialysis (Baker et al., 2002). This report indicated that reverse dialysis of cystine–glutamate exchange inhibitors reduced

Fig. 1. Schematic illustrates the regulation of vesicular glutamate release by cystine–glutamate antiporters. Extracellular cystine is exchanged for intracellular glutamate by cystine/ glutamate antiporters. Glutamate released from these antiporters has been shown to provide tonic stimulation to group II mGluRs, which regulate vesicular glutamate release through a PKA-dependent mechanism.

41

basal extracellular levels by 60%. In this same report, blockade of voltage-gated sodium or calcium channels produced a 10–20% decrease in dialysate glutamate levels (Baker et al., 2002). Second, glutamate released from the antiporter provides endogenous stimulation to group II mGluRs which is evident from the observation that blockade of group II mGluRs produces a signiﬁcant increase in extracellular glutamate levels in the nucleus accumbens (Hu et al., 1999; Baker et al., 2002; Xi et al., 2002). However, prior blockade of cystine–glutamate antiporters prevents this increase in glutamate, which is consistent with the removal of endogenous stimulation of group II mGluRs (Baker et al., 2002). These ﬁndings obtained using in vivo microdialysis were replicated using patch clamp recordings in frontal cortical and accumbens tissue slices. Brieﬂy, cystine applications produced a decrease in the frequency of mini EPSCs, spontaneous EPSCs, as well as evoked EPSCs. These effects were blocked by cystine–glutamate antiporter inhibitor or a group II mGluR antagonist (Moran et al., 2005). Collectively, these studies imply that, at least in the prefrontal cortex and nucleus accumbens, nonvesicular glutamate release from cystine–glutamate antiporters regulates glutamate neurotransmission. Group II mGluRs also function as heteroreceptors regulating dopamine release (Hu et al., 1999) indicating that nonvesicular glutamate released from cystine/glutamate exchange and sampled using in vivo microdialysis also regulates synaptic levels of neurotransmitters other than glutamate. In support, blockade of cystine–glutamate antiporters produces an increase in extracellular dopamine in the nucleus accumbens which can be reversed by restoring tone to group II mGluRs (Baker et al., 2002). Volume-sensitive organic anion channels (VSOAC) represent another mechanism contributing to nonvesicular pools of extracellular glutamate. Activation of VSOACs occurs to maintain cell volume through a process called regulatory volume decrease (RVD). VSOACs are permeable to a variety of organic anions including glutamate (Jackson and Strange, 1993; Strange et al., 1996; Basarsky et al., 1999). 5-Nitro-2-(3-phenylpropyamino) benzoic acid (NPPB) or tamoxifen can be

used to block VSOACs. However, while the contribution of VSOACs to in vivo basal levels has not been determined, VSOACs are known to contribute to glutamate neurotransmission in a pathological state. Spreading depression is a wave of glial and neuronal depolarization that propagates through the CNS and can lead to excitotoxic damage (Gorji, 2001). Glutamate release through VSOACs occurs during spreading depression, and blockade of these channels can depress the rate of spreading depression (Basarsky et al., 1999). N-acetyl aspartylglutamate (NAAG) is endogenous and abundantly expressed, and can be metabolized by NAALAdase inhibitors in the extracellular space to N-acetylaspartate and glutamate. This poses the possibility that NAAG hydrolysis may contribute to basal glutamate levels detected using microdialysis. Interestingly, NAAG also functions as a group II mGluR agonist and a partial agonist at NMDA receptors (Westbrook et al., 1986; Wroblewska et al., 1997; Bruno et al., 1998). It is unclear whether the NAALAdase inhibitor 2-(phosphonomethyl)pentanedioic acid (2-PMPA) decreases basal glutamate levels, but 2-PMPA signiﬁcantly reduces elevated glutamate following transient middle cerebral artery occlusion (Slusher et al., 1999). These data illustrate that NAAG metabolism may contribute to pathogenic glutamate signaling during an ischemic event.

V. Synaptic origin of stimulated release The preceding sections detail that extrasynaptic glutamate sampled by in vivo microdialysis is likely maintained by nonvesicular and vesicular release from both neurons and astrocytes. However, an interesting question is whether microdialysis can sample synaptic glutamate. First, striatal dialysis levels are at least partially maintained by vesicular release from neurons since dialysate levels are partially reduced following calcium depletion (Dawson et al., 1995) or blockade of voltage-gated calcium channels (Baker et al., 2002; Xi et al., 2002), but see Westerink et al. (1989) and You et al. (1994). Further, dialysate

42

glutamate levels are sensitive to voltage-gated sodium and calcium channels following a number of manipulations, including group II/III mGluR blockade (Xi et al., 2002), electrical stimulation of the prefrontal cortex (Lada et al., 1998), presentation of an appetitive stimulus (Saulskaya and Soloviova, 2004), or cocaine administration (Pierce et al., 1996; McFarland et al., 2003). Collectively, these studies indicate that microdialysis can sample synaptic glutamate and many of these studies have advanced our understanding of the contribution of glutamate to pathological states including ischemia-induced toxicity and cocaine addiction.

VI. Therapeutic targets detected by microdialysis Studies utilizing microdialysis have been instrumental in efforts to reveal the pathology and identify novel therapeutic targets for a variety of neurological disorders including ischemia, epilepsy, and schizophrenia; since these advances are the subject of latter chapters in this book they will not be described here. However, microdialysis has been particularly useful in identifying pathogenic changes in glutamate neurotransmission contributing to compulsive cocaine-seeking. Repeated cocaine administration produces a persistent reduction in basal glutamate levels in the nucleus accumbens (Pierce et al., 1996; Baker et al., 2003) that persists for at least 60 days (Baker et al., unpublished data). The reduction in basal glutamate levels contributes to a phasic elevation of glutamate observed following a cocaine challenge (Pierce et al., 1996; Reid and Berger, 1996; Reid et al., 1997; Baker et al., 2003), and this rise in glutamate is necessary for primed drug-seeking behavior (Baker et al., 2003; McFarland et al., 2003). Interestingly, the reduction in basal glutamate levels in the nucleus accumbens involves cocaine-induced blunting of cystine–glutamate exchange and this effect can be masked or reversed using cysteine prodrugs including N-acetylcysteine (Baker et al., 2003). Further, extant clinical data supports the anticraving efﬁcacy of N-acetylcysteine as a novel pharmacotherapeutic agent for cocaine addiction (LaRowe et al., 2006).

VII. Conclusion The study of glutamate homeostasis has produced a number of seminal ﬁndings that have led many to challenge dogma regarding how the brain functions. At the very least, it is now evident that the regulation of glutamate homeostatis is vastly more complex than traditional neurotransmitters and that constraints applicable to traditional neurotransmitters may be inappropriate for glutamate. Unlike most neurotransmitters, in vivo levels of glutamate levels monitored using microdialysis are only partially maintained by vesicular release; instead extrasynaptic glutamate levels are primarily maintained by nonvesicular release mechanisms likely originating in astrocytes. The existence of vesicular and nonvesicular release mechanisms has also been detected in electrophysiological studies which have demonstrated that vesicular and nonvesicular release are capable of stimulating glutamate receptors, especially when the receptors are located in the extrasynaptic domain. As a result, it appears that extrasynaptic glutamate levels, unlike traditional neurotransmitters, do not simply reﬂect neurotransmitter overﬂow, but may be maintained by a variety of release mechanisms that are unique to glutamate. Potential cellular processes include vesicular and nonvesicular release from astrocytes, voltage-sensitive organic anion channels, hemichannels, NAAG metabolism, and cystine–glutamate exchange; all of which may ultimately represent novel therapeutic targets for a number of neurological disorders involving altered glutamate signaling. Interestingly, sodium-dependent transporters appear to compartmentalize synaptic and extrasynaptic pools of glutamate. This poses the scenario that there are at least two distinct extracellular pools of glutamate, each with respective receptors. The contribution of most release mechanisms to the levels of glutamate measured by in vivo microdialysis has yet to be determined, however, many of these release mechanisms have been implicated in pathological states, underscoring the need to further investigate these cellular processes. One mechanism that has been linked to in vivo basal levels of glutamate, at least in the nucleus accumbens, is cystine–glutamate exchange. Cystine–glutamate exchange was

43

identiﬁed as a novel therapeutic target in the treatment of cocaine addiction largely through the use of in vivo microdialysis. As a result, microdialysis is a critical tool that has the potential to advance our understanding of glutamate signaling in the normal and diseased states. Further, the capacity of microdialysis to identify cellular processes regulating extrasynaptic glutamate provides a complimentary approach to the use of electrophysiological studies to monitor processes regulating synaptic glutamate. References Anderson, C.M. and Swanson, R.A. (2000) Astrocyte glutamate transport: review of properties, regulation, and physiological functions. Glia, 32: 1–14. Angulo, M.C., Kozlov, A.S., Charpak, S. and Audinat, E. (2004) Glutamate released from glial cells synchronizes neuronal activity in the hippocampus. J. Neurosci., 24: 6920–6927. Araque, A., Sanzgiri, R.P., Parpura, V. and Haydon, P.G. (1998) Calcium elevation in astrocytes causes an NMDA receptor-dependent increase in the frequency of miniature synaptic currents in cultured hippocampal neurons. J. Neurosci., 18: 6822–6829. Baker, D.A., McFarland, K., Lake, R.W., Shen, H., Tang, X.C., Toda, S. and Kalivas, P.W. (2003) Neuroadaptations in cystine–glutamate exchange underlie cocaine relapse. Nat. Neurosci., 6: 743–749. Baker, D.A., Xi, Z.X., Shen, H., Swanson, C.J. and Kalivas, P.W. (2002) The origin and neuronal function of in vivo nonsynaptic glutamate. J. Neurosci., 22: 9134–9141. Bannai, S., Takada, A., Kasuga, H. and Tateishi, N. (1986) Induction of cystine transport activity in isolated rat hepatocytes by sulfobromophthalein and other electrophilic agents. Hepatology, 6: 1361–1368. Basarsky, T.A., Feighan, D. and MacVicar, B.A. (1999) Glutamate release through volume-activated channels during spreading depression. J. Neurosci., 19: 6439–6445. Baskys, A. and Malenka, R.C. (1991) Agonists at metabotropic glutamate receptors presynaptically inhibit EPSCs in neonatal rat hippocampus. J. Physiol., 444: 687–701. Benveniste, H., Drejer, J., Schousboe, A. and Diemer, N.H. (1984) Elevation of the extracellular concentrations of glutamate and aspartate in rat hippocampus during transient cerebral ischemia monitored by intracerebral microdialysis. J. Neurochem., 43: 1369–1374. Bergles, D.E., Roberts, J.D., Somogyi, P. and Jahr, C.E. (2000) Glutamatergic synapses on oligodendrocyte precursor cells in the hippocampus. Nature, 405: 187–191. Bezzi, P., Gundersen, V., Galbete, J.L., Seifert, G., Steinhauser, C., Pilati, E. and Volterra, A. (2004) Astrocytes contain a vesicular compartment that is competent for regulated exocytosis of glutamate. Nat. Neurosci., 7: 613–620.

Blakely, R.D., Robinson, M.B., Thompson, R.C. and Coyle, J.T. (1988) Hydrolysis of the brain dipeptide N-acetyl-L-aspartyl-L-glutamate: subcellular and regional distribution, ontogeny, and the effect of lesions on N-acetylated-alpha-linked acidic dipeptidase activity. J. Neurochem., 50: 1200–1209. Bordi, F. and Ugolini, A. (1999) Group I metabotropic glutamate receptors: implications for brain diseases. Prog. Neurobiol., 59: 55–79. Bo¨ttcher, T., Goiny, M., Bering, J., Domhof, S., Nau, R. and Ungerstedt, U. (2003) Regional differences in glutamine synthetase inhibition by L-methionine sulfoximine: a microdialysis study in the rabbit brain. Exp. Brain Res., 150: 194–200. Bradford, H.F., Young, A.M. and Crowder, J.M. (1987) Continuous glutamate leakage from brain cells is balanced by compensatory high-afﬁnity reuptake transport. Neurosci. Lett., 81: 296–302. Breukel, A.I., Besselsen, E., Lopes da Silva, F.H. and Ghijsen, W.E. (1998) A presynaptic N-methyl-D-aspartate autoreceptor in rat hippocampus modulating amino acid release from a cytoplasmic pool. Eur. J. Neurosci., 10: 106–114. Bridges, C.C., Kekuda, R., Wang, H., Prasad, P.D., Mehta, P., Huang, W., Smith, S.B. and Ganapathy, V. (2001) Structure, function, and regulation of human cystine/glutamate transporter in retinal pigment epithelial cells. Invest. Ophthalmol. Vis. Sci., 42: 47–54. Bruno, V., Wroblewska, B., Wroblewski, J.T., Fiore, L. and Nicoletti, F. (1998) Neuroprotective activity of N-acetylaspartylglutamate in cultured cortical cells. Neuroscience, 85: 751–757. Cartmell, J. and Schoepp, D.D. (2000) Regulation of neurotransmitter release by metabotropic glutamate receptors. J. Neurochem., 75: 889–907. Cavelier, P. and Attwell, D. (2005) Tonic release of glutamate by a DIDS-sensitive mechanism in rat hippocampal slices. J. Physiol., 564: 397–410. Cho, Y. and Bannai, S. (1990) Uptake of glutamate and cysteine in C-6 glioma cells and in cultured astrocytes. J. Neurochem., 55: 2091–2097. Cochilla, A.J. and Alford, S. (1998) Metabotropic glutamate receptor-mediated control of neurotransmitter release. Neuron, 20: 1007–1016. Conn, P.J. and Pin, J.P. (1997) Pharmacology and functions of metabotropic glutamate receptors. Annu. Rev. Pharmacol. Toxicol., 37: 205–237. Danbolt, N.C. (2001) Glutamate uptake. Prog. Neurobiol., 65: 1–105. Dawson, L.A., Stow, J.M., Dourish, C.T. and Routledge, C. (1995) Analysis of glutamate in striatal microdialysates using capillary electrophoresis and laser-induced ﬂuorescence detection. J. Chromatogr. A, 700: 81–87. Dehnes, Y., Chaudhry, F.A., Ullensvang, K., Lehre, K.P., Storm-Mathisen, J. and Danbolt, N.C. (1998) The glutamate transporter EAAT4 in rat cerebellar Purkinje cells: a glutamate-gated chloride channel concentrated near the synapse in parts of the dendritic membrane facing astroglia. J. Neurosci., 18: 3606–3619.

44 Drew, K.L., Pehek, E.A., Rasley, B.T., Ma, Y.L. and Green, T.K. (2004) Sampling glutamate and GABA with microdialysis: suggestions on how to get the dialysis membrane closer to the synapse. J. Neurosci. Methods, 140: 127–131. Fillenz, M. (2005) In vivo neurochemical monitoring and the study of behaviour. Neurosci. Biobehav. Rev., 29: 949–962. Fonnum, F. (1993) Regulation of the synthesis of the transmitter glutamate pool. Prog. Biophys. Mol. Biol., 60: 47–57. Garcia-Espinosa, M.A., Rodrigues, T.B., Sierra, A., Benito, M., Fonseca, C., Gray, H.L., Bartnik, B.L., Garcia-Martin, M.L., Ballesteros, P. and Cerdan, S. (2004) Cerebral glucose metabolism and the glutamine cycle as detected by in vivo and in vitro 13C NMR spectroscopy. Neurochem. Int., 45: 297–303. Glavinovic´, M.I. (1999) Monte Carlo simulation of vesicular release, spatiotemporal distribution of glutamate in synaptic cleft and generation of postsynaptic currents. Pﬂugers Arch., 437: 462–470. Gomes, A.R., Correia, S.S., Carvalho, A.L. and Duarte, C.B. (2003) Regulation of AMPA receptor activity, synaptic targeting and recycling: role in synaptic plasticity. Neurochem. Res., 28: 1459–1473. Gorji, A. (2001) Spreading depression: a review of the clinical relevance. Brain Res. Brain Res. Rev., 38: 33–60. Gottesman, J. and Miller, R.F. (2003) N-methyl-D-aspartate receptors contribute to the baseline noise of retinal ganglion cells. Vis. Neurosci., 20: 329–333. Had-Aissouni, L., Re, D.B., Nieoullon, A. and Kerkerian-Le Goff, L. (2002) Importance of astrocytic inactivation of synaptically released glutamate for cell survival in the central nervous system – are astrocytes vulnerable to low intracellular glutamate concentrations? J. Physiol. Paris, 96: 317–322. Haydon, P.G. (2001) GLIA: listening and talking to the synapse. Nat. Rev. Neurosci., 2: 185–193. Hertz, L., Dringen, R., Schousboe, A. and Robinson, S.R. (1999) Astrocytes: glutamate producers for neurons. J. Neurosci. Res., 57: 417–428. Hofer, A. and Dermietzel, R. (1998) Visualization and functional blocking of gap junction hemichannels (connexons) with antibodies against external loop domains in astrocytes. Glia, 24: 141–154. Hu, G., Duffy, P., Swanson, C., Ghasemzadeh, M.B. and Kalivas, P.W. (1999) The regulation of dopamine transmission by metabotropic glutamate receptors. J. Pharmacol. Exp. Ther., 289: 412–416. Hutchinson, P.J., O’Connell, M.T., Kirkpatrick, P.J. and Pickard, J.D. (2002) How can we measure substrate, metabolite and neurotransmitter concentrations in the human brain? Physiol. Meas., 23: R75–R109. Jabaudon, D., Shimamoto, K., Yasuda-Kamatani, Y., Scanziani, M., Gahwiler, B.H. and Gerber, U. (1999) Inhibition of uptake unmasks rapid extracellular turnover of glutamate of nonvesicular origin. Proc. Natl. Acad. Sci. U.S.A., 96: 8733–8738. Jackson, P.S. and Strange, K. (1993) Volume-sensitive anion channels mediate swelling-activated inositol and taurine efﬂux. Am. J. Physiol., 265: C1489–C1500.

Kamermans, M., Fahrenfort, I., Schultz, K., Janssen-Bienhold, U., Sjoerdsma, T. and Weiler, R. (2001) Hemichannel-mediated inhibition in the outer retina. Science, 292: 1178–1180. Kato, T., Otsu, Y., Furune, Y. and Yamamoto, T. (1992) Different effects of L-, N- and T-type calcium channel blockers on striatal dopamine release measured by microdialysis in freely moving rats. Neurochem. Int., 21: 99–107. Kharazia, V.N. and Weinberg, R.J. (1999) Immunogold localization of AMPA and NMDA receptors in somatic sensory cortex of albino rat. J. Comp. Neurol., 412: 292–302. Kimelberg, H.K., Goderie, S.K., Higman, S., Pang, S. and Waniewski, R.A. (1990) Swelling-induced release of glutamate, aspartate, and taurine from astrocyte cultures. J. Neurosci., 10: 1583–1591. Krebs, M.O., Trovero, F., Desban, M., Gauchy, C., Glowinski, J. and Kemel, M.L. (1991) Distinct presynaptic regulation of dopamine release through NMDA receptors in striosomeand matrix-enriched areas of the rat striatum. J. Neurosci., 11: 1256–1262. Kreft, M., Stenovec, M., Rupnik, M., Grilc, S., Krzan, M., Potokar, M., Pangrsic, T., Haydon, P.G. and Zorec, R. (2004) Properties of Ca(2+)-dependent exocytosis in cultured astrocytes. Glia, 46: 437–445. Kulagina, N.V., Shankar, L. and Michael, A.C. (1999) Monitoring glutamate and ascorbate in the extracellular space of brain tissue with electrochemical microsensors. Anal. Chem., 71: 5093–5100. Kvamme, E., Roberg, B., Johansen, L. and Torgner, I.A. (1988) Interrelated effects of calcium and sulfhydryl reagents on renal phosphate-activated glutaminase. Contrib. Nephrol., 63: 156–160. Lada, M.W., Vickroy, T.W. and Kennedy, R.T. (1998) Evidence for neuronal origin and metabotropic receptormediated regulation of extracellular glutamate and aspartate in rat striatum in vivo following electrical stimulation of the prefrontal cortex. J. Neurochem., 70: 617–625. Larowe, S.D., Mardikian, P., Malcolm, R., Myrick, H., Kalivas, P., McFarland, K., Saladin, M., McRae, A. and Brady, K. (2006) Safety and tolerability of N-acetylcysteine in cocaine-dependent individuals. Am. J. Addict., 15: 105–110. Lehre, K.P. and Danbolt, N.C. (1998) The number of glutamate transporter subtype molecules at glutamatergic synapses: chemical and stereological quantiﬁcation in young adult rat brain. J. Neurosci., 18: 8751–8757. Levi, G. and Raiteri, M. (1993) Carrier-mediated release of neurotransmitters. Trends Neurosci., 16: 415–419. Liu, T.F., Li, H.Y., Atkinson, M.M. and Johnson, R.G. (1995) Intracellular lucifer yellow leakage from Novikoff cells in the presence of ATP or low extracellular Ca: evidence for hemigap junction channels. Methods Find Exp. Clin. Pharmacol., 17: 23–28. Malva, J.O., Carvalho, A.P. and Carvalho, C.M. (1994) Modulation of dopamine and noradrenaline release and of intracellular Ca2+ concentration by presynaptic glutamate receptors in hippocampus. Br. J. Pharmacol., 113: 1439–1447.

45 Matsui, K. and Jahr, C.E. (2003) Ectopic release of synaptic vesicles. Neuron, 40: 1173–1183. Matsui, K., Jahr, C.E. and Rubio, M.E. (2005) High-concentration rapid transients of glutamate mediate neuralglial communication via ectopic release. J. Neurosci., 25: 7538–7547. McCall, A., Glaeser, B.S., Millington, W. and Wurtman, R.J. (1979) Monosodium glutamate neurotoxicity, hyperosmolarity, and blood-brain barrier dysfunction. Neurobehav. Toxicol., 1: 279–283. McFarland, K., Lapish, C.C. and Kalivas, P.W. (2003) Prefrontal glutamate release into the core of the nucleus accumbens mediates cocaine-induced reinstatement of drug-seeking behavior. J. Neurosci., 23: 3531–3537. Melendez, R.I., Vuthiganon, J. and Kalivas, P.W. (2005) Regulation of extracellular glutamate in the prefrontal cortex: focus on the cystine glutamate exchanger and group I metabotropic glutamate receptors. J. Pharmacol. Exp. Ther., 314: 139–147. Mena, F.V., Baab, P.J., Zielke, C.L., Huang, Y. and Zielke, H.R. (2005) Formation of extracellular glutamate from glutamine: exclusion of pyroglutamate as an intermediate. Brain Res., 1052: 88–96. Miele, M., Boutelle, M.G. and Fillenz, M. (1996) The source of physiologically stimulated glutamate efﬂux from the striatum of conscious rats. J. Physiol., 497(Pt 3): 745–751. Moran, M.M., McFarland, K., Melendez, R.I., Kalivas, P.W. and Seamans, J.K. (2005) Cystine/glutamate exchange regulates metabotropic glutamate receptor presynaptic inhibition of excitatory transmission and vulnerability to cocaine seeking. J. Neurosci., 25: 6389–6393. Morari, M., O’Connor, W.T., Ungerstedt, U. and Fuxe, K. (1993) N-methyl-D-aspartic acid differentially regulates extracellular dopamine, GABA, and glutamate levels in the dorsolateral neostriatum of the halothane-anesthetized rat: an in vivo microdialysis study. J. Neurochem., 60: 1884–1893. Murphy, T.H., Schnaar, R.L. and Coyle, J.T. (1990) Immature cortical neurons are uniquely sensitive to glutamate toxicity by inhibition of cystine uptake. FASEB J., 4: 1624–1633. Newman, E.A. (2003) New roles for astrocytes: regulation of synaptic transmission. Trends Neurosci., 26: 536–542. Osborne, P.G., O’Connor, W.T. and Ungerstedt, U. (1991) Effect of varying the ionic concentration of a microdialysis perfusate on basal striatal dopamine levels in awake rats. J. Neurochem., 56: 452–456. Ottersen, O.P., Takumi, Y., Matsubara, A., Landsend, A.S., Laake, J.H. and Usami, S. (1998) Molecular organization of a type of peripheral glutamate synapse: the afferent synapses of hair cells in the inner ear. Prog. Neurobiol., 54: 127–148. Ozawa, S., Kamiya, H. and Tsuzuki, K. (1998) Glutamate receptors in the mammalian central nervous system. Prog. Neurobiol., 54: 581–618. Parpura, V., Basarsky, T.A., Liu, F., Jeftinija, K., Jeftinija, S. and Haydon, P.G. (1994) Glutamate-mediated astrocyteneuron signalling. Nature, 369: 744–747.

Parpura, V. and Haydon, P.G. (2000) Physiological astrocytic calcium levels stimulate glutamate release to modulate adjacent neurons. Proc. Natl. Acad. Sci. U.S.A., 97: 8629–8634. Patneau, D.K. and Mayer, M.L. (1990) Structure–activity relationships for amino acid transmitter candidates acting at Nmethyl-D-aspartate and quisqualate receptors. J. Neurosci., 10: 2385–2399. Paulsen, R.E. and Fonnum, F. (1989) Role of glial cells for the basal and Ca2+-dependent K+-evoked release of transmitter amino acids investigated by microdialysis. J. Neurochem., 52: 1823–1829. Pierce, R.C., Bell, K., Duffy, P. and Kalivas, P.W. (1996) Repeated cocaine augments excitatory amino acid transmission in the nucleus accumbens only in rats having developed behavioral sensitization. J. Neurosci., 16: 1550–1560. Pittaluga, A. and Raiteri, M. (1990) Release-enhancing glycinedependent presynaptic NMDA receptors exist on noradrenergic terminals of hippocampus. Eur. J. Pharmacol., 191: 231–234. Reid, M.S. and Berger, S.P. (1996) Evidence for sensitization of cocaine-induced nucleus accumbens glutamate release. Neuroreport, 7: 1325–1329. Reid, M.S., Hsu, K. Jr. and Berger, S.P. (1997) Cocaine and amphetamine preferentially stimulate glutamate release in the limbic system: studies on the involvement of dopamine. Synapse, 27: 95–105. Ronne-Engstrom, E., Carlson, H., Liu, Y., Ungerstedt, U. and Hillered, L. (1995) Inﬂuence of perfusate glucose concentration on dialysate lactate, pyruvate, aspartate, and glutamate levels under basal and hypoxic conditions: a microdialysis study in rat brain. J. Neurochem., 65: 257–262. Rossell, S., Gonzalez, L.E. and Hernandez, L. (2003) One-second time resolution brain microdialysis in fully awake rats Protocol for the collection, separation and sorting of nanoliter dialysate volumes. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 784: 385–393. Rothman, D.L., Behar, K.L., Hyder, F. and Shulman, R.G. (2003) In vivo NMR studies of the glutamate neurotransmitter ﬂux and neuroenergetics: implications for brain function. Annu. Rev. Physiol., 65: 401–427. Rumbaugh, G. and Vicini, S. (1999) Distinct synaptic and extrasynaptic NMDA receptors in developing cerebellar granule neurons. J. Neurosci., 19: 10603–10610. Rusakov, D.A. and Kullmann, D.M. (1998) Extrasynaptic glutamate diffusion in the hippocampus: ultrastructural constraints, uptake, and receptor activation. J. Neurosci., 18: 3158–3170. Sato, H., Tamba, M., Ishii, T. and Bannai, S. (1999) Cloning and expression of a plasma membrane cystine/glutamate exchange transporter composed of two distinct proteins. J. Biol. Chem., 274: 11455–11458. Saulskaya, N.B. and Soloviova, N.A. (2004) Tetrodotoxindependent glutamate release in the rat nucleus accumbens during concurrent presentation of appetitive and conditioned aversive stimuli. J. Neurosci. Methods, 140: 15–21.

46 Schousboe, A., Westergaard, N., Waagepetersen, H.S., Larsson, O.M., Bakken, I.J. and Sonnewald, U. (1997) Trafﬁcking between glia and neurons of TCA cycle intermediates and related metabolites. Glia, 21: 99–105. Slusher, B.S., Vornov, J.J., Thomas, A.G., Hurn, P.D., Harukuni, I., Bhardwaj, A., Traystman, R.J., Robinson, M.B., Britton, P., Lu, X.C., Tortella, F.C., Wozniak, K.M., Yudkoff, M., Potter, B.M. and Jackson, P.F. (1999) Selective inhibition of NAALADase, which converts NAAG to glutamate, reduces ischemic brain injury. Nat. Med., 5: 1396–1402. Smith, Q.R. (2000) Transport of glutamate and other amino acids at the blood-brain barrier. J. Nutr., 130: 1016S–1022S. Stevens, C.F. and Tsujimoto, T. (1995) Estimates for the pool size of releasable quanta at a single central synapse and for the time required to reﬁll the pool. Proc. Natl. Acad. Sci. U.S.A., 92: 846–849. Stocca, G. and Vicini, S. (1998) Increased contribution of NR2A subunit to synaptic NMDA receptors in developing rat cortical neurons. J. Physiol., 507(Pt 1): 13–24. Strange, K., Emma, F. and Jackson, P.S. (1996) Cellular and molecular physiology of volume-sensitive anion channels. Am. J. Physiol., 270: C711–C730. Timmerman, W. and Westerink, B.H. (1997) Brain microdialysis of GABA and glutamate: what does it signify? Synapse, 27: 242–261. Tovar, K.R. and Westbrook, G.L. (1999) The incorporation of NMDA receptors with a distinct subunit composition at nascent hippocampal synapses in vitro. J. Neurosci., 19: 4180–4188. Ventura, R. and Harris, K.M. (1999) Three-dimensional relationships between hippocampal synapses and astrocytes. J. Neurosci., 19: 6897–6906. Verkhratsky, A., Orkand, R.K. and Kettenmann, H. (1998) Glial calcium: homeostasis and signaling function. Physiol. Rev., 78: 99–141. Verma, A. and Moghaddam, B. (1998) Regulation of striatal dopamine release by metabotropic glutamate receptors. Synapse, 28: 220–226. Volterra, A. and Meldolesi, J. (2005) Astrocytes, from brain glue to communication elements: the revolution continues. Nat. Rev. Neurosci., 6: 626–640. Waagepetersen, H.S., Bakken, I.J., Larsson, O.M., Sonnewald, U. and Schousboe, A. (1998) Comparison of lactate and

glucose metabolism in cultured neocortical neurons and astrocytes using 13C-NMR spectroscopy. Dev. Neurosci., 20: 310–320. Wang, J.K., Andrews, H. and Thukral, V. (1992) Presynaptic glutamate receptors regulate noradrenaline release from isolated nerve terminals. J. Neurochem., 58: 204–211. Westbrook, G.L., Mayer, M.L., Namboodiri, M.A. and Neale, J.H. (1986) High concentrations of N-acetylaspartylglutamate (NAAG) selectively activate NMDA receptors on mouse spinal cord neurons in cell culture. J. Neurosci., 6: 3385–3392. Westerink, B.H., Damsma, G. and de Vries, J.B. (1989) Effect of ouabain applied by intrastriatal microdialysis on the in vivo release of dopamine, acetylcholine, and amino acids in the brain of conscious rats. J. Neurochem., 52: 705–712. Westerink, B.H., Tuntler, J., Damsma, G., Rollema, H. and de Vries, J.B. (1987) The use of tetrodotoxin for the characterization of drug-enhanced dopamine release in conscious rats studied by brain dialysis. Naunyn Schmiedebergs Arch. Pharmacol., 336: 502–507. Wroblewska, B., Wroblewski, J.T., Pshenichkin, S., Surin, A., Sullivan, S.E. and Neale, J.H. (1997) N-acetylaspartylglutamate selectively activates mGluR3 receptors in transfected cells. J. Neurochem., 69: 174–181. Xi, Z.X., Baker, D.A., Shen, H., Carson, D.S. and Kalivas, P.W. (2002) Group II metabotropic glutamate receptors modulate extracellular glutamate in the nucleus accumbens. J. Pharmacol. Exp. Ther., 300: 162–171. Ye, Z.C., Wyeth, M.S., Baltan-Tekkok, S. and Ransom, B.R. (2003) Functional hemichannels in astrocytes: a novel mechanism of glutamate release. J. Neurosci., 23: 3588–3596. You, Z.B., Herrera-Marschitz, M., Nylander, I., Goiny, M., O’Connor, W.T., Ungerstedt, U. and Terenius, L. (1994) The striatonigral dynorphin pathway of the rat studied with in vivo microdialysis – II. Effects of dopamine D1 and D2 receptor agonists. Neuroscience, 63: 427–434. Zhang, Q., Pangrsic, T., Kreft, M., Krzan, M., Li, N., Sul, J.Y., Halassa, M., Van Bockstaele, E., Zorec, R. and Haydon, P.G. (2004) Fusion-related release of glutamate from astrocytes. J. Biol. Chem., 279: 12724–12733. Zhou, J., Neale, J.H., Pomper, M.G. and Kozikowski, A.P. (2005) NAAG peptidase inhibitors and their potential for diagnosis and therapy. Nat. Rev. Drug Discov., 4: 1015–1026.

CHAPTER 1.4

The validity of intracerebral microdialysis Kevin C. Chen Department of Chemical and Biomedical Engineering, Florida State University, Tallahassee, FL, USA

Abstract: Notwithstanding its invasiveness, intracerebral microdialysis has become a popular tool in monitoring and interpreting neurochemical processes in the brain in conjunctions with various stimuli or pharmacological/pathological alterations in behavioral or pharmacological studies. The most widely used quantitative methods are the zero-net-ﬂux (ZNF) method and the varied ﬂow rate method, both of which have been questioned about their validity to determine the true ‘‘unperturbed’’ basal concentration of extracellular neurotransmitters. This inability is attributed to the inevitable disruption of both neurotransmitter release and uptake sites during microdialysis probe implantation. Model simulations show that if microdialysis impairs the neurotransmitter release site more than the corresponding uptake sites, underestimation of the unperturbed basal concentration by these quantitative methods occurs and the trend of the in vivo recovery of the neurotransmitter with the uptake inhibition would reverse as compared with that of microdialysis extraction. Two modiﬁed DA kinetic models accounting for the presence of DA baseline were presented and compared to assess the basal extracellular DA, [DA]e, in rat striatum. Examining the data from the microdialysis ZNF measurements and fast-scan cyclic voltammetry experiments on striatal DA systems shows reasonable doubt that the current ZNF measurement on striatal DA (5–10 nM) did underestimate the basal level [DA]e However, the underestimation is not as serious as previously thought by Michael and coworkers. Although the striatal [DA]e could be much higher than the single-digit nanomolar range determined by the microdialysis ZNF method, the basal [DA]e is probably still conﬁned within nanomolar range, perhaps fewer than 100 nM.

neuroscientists to observe real-time chemical events and intercellular processes in the brains. This widespread use is reﬂected on the fact that as of February, 2007, PubMed listed some 9000+scientiﬁc and clinical papers describing applications of this technique; and the number is still growing. Depending on different research needs, the microdialysis probe also has different designs. This chapter focuses on intracerebral microdialysis of the concentric probe, as this is the one which is mostly used in neuroscientiﬁc research. The methodology and principles of microdialysis operation have been introduced in other chapters and complete reviews in this regard can be found in the literature (e.g., Benveniste and Huttemeier, 1990; Justice, 1993; Kehr, 1993; Stenken, 1999; Bourne, 2003; Plock

I. Introduction Originally developed for neuroscientiﬁc research, microdialysis has now become a popular tool applied to monitor extracellular analyte concentrations in many different tissues and organs in many disciplines, including pharmacophysiology (De Lange et al., 2000; Chefer et al., 2003), neurobiology/neurochemistry (Smith and Justice, 1994; Cosford et al., 1996; Bourne, 2003), sports medicine (MacLean et al., 1999), and tumor oncology (Dabrosin, 2005). Albeit invasive, intracerebral microdialysis becomes popular by allowing Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

47

DOI: 10.1016/S1569-7339(06)16004-X Copyright 2007 Elsevier B.V. All rights reserved

48

and Kloft, 2005). Therefore, this chapter will mainly focus on the validity of microdialysis data associated with tissue trauma and current status in assessing such knowledge on quantitative microdialysis. Concern over the possibility of microdialysis data compromised by probe-induced tissue trauma is non-trivial if one is to establish microdialysis as a credible tool in neuroscience research.

I.A. Microdialysis – an invasive technique Two major criticisms about microdialysis are its inability to monitor real-time responses of in vivo neurochemical events and its invasive nature. What is lacking in the time resolution of microdialysis sampling is more or less compensated by its capability to analyze diverse compounds when coupled to other high-precision analytical instrument, such as high-performance liquid chromatography or scintillation counter. Nonetheless, the tissue damage caused by probe implantation is a serious issue. Because this method must insert its probe into the desired brain region to collect extracellular ﬂuid for compositional analysis, microdialysis sampling inevitably disrupts the morphological structure of connective tissue and processes at the insertion site. As one learns more from and about microdialysis, concerns over its invasive nature, and hence the credibility of resulting data, also increase (Di Chiara et al., 1996; Westerink and Timmerman, 1999; De Lange et al., 2000). Many works have already subjected the tissue trauma exerted by microdialysis implantation to morphological and neurochemical examination. Generally, insertion of the microdialysis probe causes edema and hemorrhage, as well as gliosis and ﬁbrosis, in the surrounding tissue (Benveniste and Diemer, 1987; Benveniste and Huttemeier, 1990; Grabb et al., 1998). Probe insertion also disrupts the otherwise intact blood-brain barrier (Morgan et al., 1996; Groothuis et al., 1998). Whereas consensus regarding the microdialysis-induced morphological disruption and subsequent inﬂammatory responses is ﬁrmly established, the inﬂuences of these damages on the functional neurochemical data obtained by microdialysis are often unclear and still under debates. It is not

uncommon to ﬁnd controversial experimental results and conclusions regarding how traumatized tissues affect microdialysis data. For example, destructive and irrecoverable effects of the microdialysis-induced trauma on the nearby neuronal processes and synaptic number density have been reported (e.g., Shuaib et al., 1990; ClappLilly et al., 1999). In particular, Grabb et al. (1998) compared several amino acids with purine metabolites in the dialysate contents in the rat brains implanted with microdialysis probes chronically (24 h) or acutely (2 h), and reported that rat brains implanted with chronic microdialysis showed abnormal responses of these dialysates, both in magnitudes and response time, to ischemic insults. In contrast, more positive conclusions on the intrusive nature of microdialysis sampling can also be found (e.g., Georgieva et al., 1993; Martin-Fardon et al., 1997; West et al., 2002), both from observing a fairly constant dialysate content of various neurotransmitters and peptides over repeated probe insertion (assuming the use of a guide cannula implanted permanently) or from morphological examination. It appears that in chronic microdialysis, the dialysate content of many neurotransmitters can maintain quite a constant level, while their metabolites increased initially (a few hours after probe insertion), followed by a gradual decline over a long-time period (2–3 days later). This trend has been observed in the microdialysis studies for dopamine (DA) and its metabolite L-DOPA in rat striatum (Robinson and Camp, 1991; Marburger et al., 2000). The question has been raised as to whether the neurotransmitter metabolites, not the neurotransmitters themselves, should be used as a better indicator for tissue integrity and recovery from microdialysis probe implantation. Whereas many pharmacological and neurochemical studies have ﬁrmly established that microdialysis implantation inevitably disrupted tissue structure and possibly damaged neuronal processes, no quantitative suggestion was made as to how and to what extent these tissue damages affect microdialysis sampling accuracy and interpretation of the corresponding data. Note that the concern brought in this chapter is not whether the dialysate neurotransmitters genuinely reﬂect functional release from nearby neurons, as discussed in

49

Di Chiara et al. (1996) and Westerink and Timmerman (1999). On the contrary, the concern herein is whether the microdialysis-induced tissue trauma would be so signiﬁcant as to completely compromise the validity of microdialysis, and hence askew the interpretation of quantitative microdialysis data. These quantitative methods include the zeronet-ﬂuxmethod and the varied perfusion method, and will be introduced later. The same issue is also discussed in Chapter 1.6 by A.C. Michael.

I.B. The main issues I.B.1. Passive diffusion To apply microdialysis to quantify the endogenous concentration of compounds of interests in the brain extracellular space (ECS), one needs to deal with two issues. One is the fact that the diffusion resistance encountered in the tissue is different from that of the quiescent medium in vitro. This difference in the diffusion resistances between in vitro and in vivo experiments originates partly from: (i) the much reduced diffusible space in the brain [20% of the total brain volume belongs to ECS (Nicholson and Sykova, 1998)], which increases the interstitial diffusional resistance, and partly from (ii) the active processes in the tissue. According to the existing microdialysis theory (e.g., Benveniste and Huttemeier, 1990; Bungay et al., 1990; Chen et al., 2002), the reduced ECS (characterized by the ECS volume fraction fe) and the increased interstitial diffusion path (characterized by a tortuosity factor l) make the tissue diffusion resistance much larger than that of the medium in vitro (which usually is a quiescent bulk liquid, with a volume percentage of diffusible space equal to 100%), thus reducing the microdialysis recovery in vivo. But the active processes of the endogenous analyte (such as release, uptake, active transporters and exchangers, or metabolism) act in the opposite way to enhance the in vivo recovery of these neurotransmitters in the brains (Parsons and Justice, 1992, 1994). The theoretical basis of how active processes can affect microdialysis characteristics (such as microdialysis recovery and extraction) was ﬁrst proposed by Bungay et al. (1990), and later experimentally veriﬁed by

Justice and coworkers (Parsons et al., 1991; Parsons and Justice, 1992; Smith and Justice, 1994; Cosford et al., 1996). Because of the interplays between these unknown factors in the brains, it was soon realized that in vitro calibration cannot be applied in vivo due to the unknown tissue diffusion resistance, which can be either larger or smaller than that of the in vitro resistance1. I.B.2. Active processes A second issue related to practicing quantitative microdialysis to sample endogenous neurotransmitters is whether the microdialysis probe signiﬁcantly damages the neurotransmitter release and/or uptake sites. The role of active processes in deﬁning the basal level of extracellular neurotransmitters is different from their roles in shaping the microdialysis extraction and recovery ratio. Relative changes in the neurotransmitter release and uptake directly modiﬁes the basal level of extracellular neurotransmitter, a consequence of the mass balance between neurotransmitter release and uptake rates. As microdialysis only samples the extracellular neurotransmitters in its immediate surroundings, quantitative methods, like the zero-net-ﬂux (ZNF) method or the varied ﬂow rate method, can only reﬂect such changed basal level near the probe but not the intact one at further distances away. Hence, disruption in the neurotransmitter release and uptake brought about by probe insertion can seriously compromise the ability of microdialysis to quantify the basal level of extracellular neurotransmitters in the undisturbed state. There have been reports indicating that microdialysis probe insertion overwhelmingly destroyed DA release sites in the vicinity of the probe (Lu et al., 1998; Peters and Michael, 1998; Yang et al., 1998; Borland et al., 2005). A consequence of this preferential destruction on the neurotransmitter release sites (including the closed synapses at axonal terminals as well as the open synapses on somatic dendrites) will be a lowered level of the extracellular neurotransmitters surrounding the microdialysis probe. Both issues discussed above are all related to the microdialysis-induced tissue trauma. The passive 1 For the in vivo microdialysis, only the quiescent case is discussed and compared.

50

II. Theory of single-probe microdialysis II.A. Quantitative microdialysis (blind shoot method) Before introducing some more complete mathematical microdialysis theory, it is necessary to introduce a few microdialysis methods (speciﬁcally, the ZNF method and the slow perfusion method), which are widely employed to quantify extracellular concentrations of compounds of interest. Because both methods are based on the fundamental mass balance principle, it was once believed that these two methods can adequately sample the extracellular level of the substances of interest without any prior knowledge of the tissue diffusive properties and neurochemical processes. II.A.1. Zero-net-flux method This method, originally proposed by Lonnroth et al. (1987) as the ‘‘concentration difference method’’, is based on the idea that the extracellular concentration of analyte in the surrounding tissue should equal the concentration in the infused dialysate ﬂuid when a net-zero exchange of the analyte occurs between the dialysate within the probe and the extracellular ﬂuid outside the probe. The difference between the infusion (‘in’) and outﬂow (‘out’) analyte concentrations in the dialysates

3

Cin Cout (nM)

diffusion is closely dependent on the diffusion properties (i.e., the ECS volume fraction and the effective diffusion coefﬁcient) in the surrounding tissue, which can be altered due to edema or ﬁbroblast proliferation resulting from probe insertion. The active release or uptake processes, which set the basal level of extracellular analyte to be sampled by microdialysis, can be physically destroyed on sites during probe insertion or later on when neuronal degeneration occurs. Thus, a quantitative microdialysis model should try to address both issues and provide theoretical predictions on how these factors inﬂuence microdialysis performance. This chapter will brieﬂy review the history of microdialysis theory development; then discuss current development on assessing how tissue trauma affects the practice of quantitative microdialysis.

control uptake inhibition

2 1 0 -1 -2 0

2.5

5.0

7.5

10.0

12.5

15.0

Cin(nM) Fig. 1. The concentration difference plot showing the hypothetic situation that after neurotransmitter extrasynaptic uptake is inhibited, the plot slope (E) decreases and the ZNF concentration (CZNF) increases; but the dialysate Cout (when Cin ¼ 0) decreases. Thus using the Cout solely to judge the increase or decrease of extracellular neurotransmitter could be problematic, as in this ﬁgure conclusion regarding increase or decrease of C* after pharmacological application based on the dialysate Cout (when Cin ¼ 0) is opposite to that based on CZNF.

(denoted as Cin and Cout, respectively) is plotted against the infusion concentration Cin (Fig. 1). The dialysate concentration at the ZNF point (CZNF) is taken to be the infusion concentration Cin when a ZNF condition is achieved, that is, Cin Cout ¼ 0. This value CZNF at the ZNF condition was initially taken as the mean ECS concentration C* in the unperturbed tissue. As the issue of tissue trauma was later recognized, the equivalency between CZNF and C* was challenged. To distinguish between the speciﬁc infusing analyte concentration at the ZNF point and the true tissue ECS concentration C*, this chapter designates the former value the CZNF, and will not assume the equivalency of CZNF and C*. It should be noted that even if CZNF is not equal to the C* in the distant unperturbed tissue, the CZNF so measured is still the mean ECS concentration in the damaged tissue nearby the inserted probe. From the microdialysis theory (Bungay et al., 1990), the slope of the concentration difference curve is termed the ‘‘in vivo extraction efﬁciency’’, E, of the probe (Bungay et al., 1990). Another commonly used factor characterizing microdialysis performance is the relative recovery, deﬁned as C out R ¼ when C in is zero; (1) C

51

Traditional qualitative approach in studying the effect of speciﬁc pharmacological agents on neurochemistry is to perfuse the probe with the blank Ringer (Cin ¼ 0) and observe the relative changes of Cout in the dialysate in comparison with the control. The relative change in Cout is used to infer the increase/decrease of the analyte in the ECS due to the action of the pharmacological agent, assuming that relative changes in Cout monotonically reﬂects relative changes in C*. It is possible, thought not yet demonstrated in experiments, that a decrease in Cout after pharmacological applications may not necessarily imply a decrease in the CZNF or C* (see Fig. 1). However, a quantitative microdialysis method like the ZNF method can clarify if the relative increase/decrease of the dialysate Cout can truly reﬂect the same trend of the CZNF, and hence of the basal C*. Of course, it is based on the implicit assumption that a monotonic relationship exists between CZNF and C*. Initially, the relative recovery R was implicitly viewed to be the same as the extraction efﬁciency E. This equality was demonstrated in the theoretical works of Bungay et al. (1990), where the tissue was treated as a single-phase uniform medium. For the ZNF method, Bungay et al. (1990) elegantly showed that the following equation, 0

1

C in C out ¼ E @C in |{z} C A

(2)

C ZNF

can be derived from the mass balance equation across microdialysis membrane and between dialysate inlet and outlet. From Eq. (2), it is obvious that when Cin Cout ¼ 0, Cin ¼ CZNF ¼ C*. Furthermore, Eq. (2) also shows that when Cin ¼ 0, R ¼ E. Initially, Eq. (2) was believed to be adequate since the concentration difference plot constructed from experimental data always yields a straight line, implying that reversal operation from recovery (when Cin is lower than CZNF) to delivery (when Cin is higher than CZNF) around the ZNF point is symmetric (Parsons and Justice, 1994). As recent works from Michael and coworkers (Peters and Michael, 1998; Yang et al., 1998; Peters et al., 2000; Yang et al., 2000) started to question the equality between R and E due to the

probe-induced damages to neurotransmitter release (and possibly uptake), a more general relations was proposed, 0

1

B R C C C in C out ¼ E B @C in E C A, |ﬄ{zﬄ}

(3)

C ZNF

where it is seen that the measured ZNF concentration CZNF will not be the true ECS concentration C* if R is not equal to E, and vice versa. Furthermore, any difference between R and E should be reﬂected on the difference between CZNF and C*. When Eq. (3) is used to interpret the concentration difference plot, such plots can only determine E and CZNF, but not R (since C* is unknown and cannot be assumed to be equal to CZNF). II.A.2. Varied flow rate method This method is based on the idea that if a nonperfusing microdialysis probe is placed in the tissue, long enough, eventually diffusion equilibrium is achieved and the neurochemical contents inside and outside of the probe should be the same. Analyzing the dialysate content (Cout) of various analytes in equilibrium gives a direct indication of their mean extracellular concentrations in the surrounding tissue, which, hypothetically, should be the same as the CZNF determined by the ZNF method2. Since a true zero perfusion ﬂow rate cannot be achieved for the purpose of collecting dialysate ﬂuid in sufﬁcient quantity, a non-zero ﬂow rate, notwithstanding its smallness, disturbs concentration equilibrium and creates diffusion gradients for all kinds of analytes across the probe membrane. Hence, an alternative to circumvent this dilemma is to perform microdialysis sampling at various low perfusion ﬂow rates. Then the curve of the dialysate concentration versus the ﬂow rate is extrapolated to the zero ﬂow condition to obtain the ideal extracellular concentration. Whereas extrapolation may create quite signiﬁcant errors, theory had been developed to help 2 Originally, it was believed that the concentration determined at the no-ﬂow condition will reﬂect the true C* in the undisturbed tissue. Now it is understood that it is in fact CZNF.

52 10 best fit with theory extrapolations

Cout (nM)

8 6 4 2 0 0

0.25

0.5

0.75 1.0 Q (µL/min)

1.25

1.5

Fig. 2. Plot of the varied perfusion ﬂow method illustrating the possible errors in the extrapolation process. As can be seen, extrapolation without theoretical ﬁtting could return erroneous results at the zero ﬂow rate. Hence, prediction of CZNF can be greatly facilitated by ﬁtting the data ﬁrst.

reduce possible extrapolation errors (Jacobson et al., 1985; Bungay et al., 1990), C out ¼ 1 expðK m A=QÞ, C ZNF

(4)

where Km is the mean mass transfer coefﬁcient, A the active membrane surface area, and Q the probe perfusion ﬂow rate. In Fig. 2, one can see that the theoretical curve of Eq. (4) actually levels off slightly as the ﬂow rate Q approaches zero. Thus extrapolation to the no-ﬂow condition without the help of theoretical ﬁtting is dangerous. II.B. More theoretical models The development of microdialysis models can be roughly divided into two stages, depending on whether the tissue trauma was speciﬁcally considered. In the early stage (the ﬁrst stage), the microdialysis model was developed without explicitly considering the traumatized tissue (in terms of altered tissue ECS structure and impaired neurotransmitter release and uptake). It is noticed, however, that if the traumatized tissue extends sufﬁciently wide, it becomes the sole medium interacting with the microdialysis probe. In this case, the presence of the healthy tissue can be ignored and the single-medium model is still applicable as long as the parameters for the single-phase tissue

can properly reﬂect those of the traumatized tissue. In the second (also the latest) stage, microdialysis models explicitly incorporated a layer of traumatized tissue between the microdialysis probe and the outermost healthy tissue. Apparently, model of this kind, although more complex by nature due to incorporating more unknown parameters, is better equipped to study how preferential damages in the neurotransmitter release and uptake affect the accuracy of quantitative microdialysis. II.B.1. Theory without specifically considering tissue trauma Benveniste et al. (1989) were the ﬁrst to recognize that the in vitro calibration method (i.e., using the in vitro recovery ratio Rin vitro to estimate C* by CoutXRin vitro) could be problematic in predicting the extracellular concentrations of some analyte (especially true for K+, Ca2+, and mannitol), mainly because of the uncertainty of the diffusional resistances of these analytes in the surrounding tissue. To correct the in vitro recovery to estimate C* in vivo, they proposed a semi-empirical relation to account for the increased diffusion resistance in the tissue, l2 C out C ¼ K (5) fe Rin vitro where l is the diffusional tortuosity in the tissue ECS (Nicholson and Sykova, 1998), fe the ECS volume fraction. The constant K represents a compensating factor for the concentration gradient across the probe membrane, but can be used to account for other factors as well. A value of K ¼ 0.7 was estimated based on microdialysis data of K+, Ca2+, and mannitol (Benveniste et al., 1989). It should be noted that derivation of Eq. (5) was based on pure diffusion and did not consider inﬂuence of active processes. Another notable microdialysis theory was at the same time developed by Amberg and Lindefors (1989), who considered the diffusion resistances in the tissue as well as in the probe membrane and the perfusing dialysate. Their main result can be summarized by the following equation, R¼

4pLfe De hðtÞ, Q

(6)

53

C ¼

C out Q C out Q C out ¼ . R 4pLfe De hðtÞ 0:8pLfe De

(7)

Similarly, Amberg and Lindefors (1989) did not consider the active processes in their model. Because of this, models of both groups cannot be applied to compounds governed by highly active processes such as neurotransmitters. Their theory can be applied to quantify extracellular neurotransmitter concentrations only when the active processes are obliterated (e.g., by 6-OHDA lesions to destroy dopaminergic innervation) or inhibited (by various pharmacological blockers). II.B.2. Theory considering tissue trauma Theoretical modeling to study the impact of possible tissue trauma on microdialysis sampling characteristics started with Peters and Michael (1998), who constructed a simple, discrete release model to explain their experimental observations (Lu et al., 1998; Yang et al., 1998) in rat striatum. In their experiments incorporating both microdialysis sampling and fast-scan cyclic voltammetry (FSCV), they found that evoked response of extracellular DA as recorded by a carbon-ﬁber microelectrode placed immediately adjacent to the microdialysis probe was absent, but became detectable after DA uptake inhibitor nomifensine was systemically applied. Whereas their model can capture qualitative behaviors in agreement with their experimental data and support their claims, their model is not without drawback. For example, only one single DA release site was considered in the initial model of Peters and Michael (1998). Although later on, a more realistic model with multi-discrete release sites was introduced (Yang et al., 2000), it appears that the more systematic investigation on how microdialysis-induced tissue trauma affect microdialysis sampling can be obtained from more rigorous and well-constructed models. Since the works of Michael and

coworkers, Bungay et al. (2003) and Chen (2005a, 2005b, 2006) also attempted to build mathematical models to examine the effects of tissue trauma on the sampling composition and accuracy of quantitative microdialysis methods, especially the ZNF method. In these models, the probe-induced ‘‘tissue trauma’’ is categorized as (i) tissue morphological alterations in terms of abnormal tissue ECS and extracellular diffusivity and (ii) physiological impairment in neurotransmitter release and uptake. The model in Bungay et al. (2003) will be further introduced in another chapter of the book. Hence only the model from Chen (2006) is described here. It should be noted that both models from Bungay et al. (2003) and Chen (2006) are extremely similar and were built on the same framework. Fig. 3 depict the schematics of Chen’s microdialysis model. A glossary of the symbols used in the model is given in Table 1. Along the radial direction, the diffusible space was divided into three regions, the probe membrane (region 1: [ri, ro]), the traumatized tissue layer (region 2: [ro, rt]), and the external normal tissue region (region 3: [rt, N]). The parameters for ECS diffusion and

membrane

traumatized tissue

normal tissue

Di =

Dm

Dt

De

φi =

φm

φt

φe

κi =

0

κt = εκ κe

κe

σi =

0

σt = εσσe

σe

r

~ ~

where h(t) is a scaled time-dependent factor, De the effective diffusivity of analyte in the brain ECS, and L the probe membrane length. The function h(t) declines monotonically to 0.2 after a reasonably long time (1 h). Hence,

ri

ro

rt

Fig. 3. Schematic representation of the microdialysis model showing the three consecutive diffusion regions in the radial diffusion direction, respectively, the membrane [ri, ro], the trauma tissue layer [ro, rt], and the normal tissue [rt, N]. Each region is characterized by its distinct diffusion properties (D and f), neurotransmitter release rate density (s) and linear uptake rate constant (k), with different subscripts indicating the different regions. Figure reproduced from Chen (2005a) with permission.

54 Table 1. Glossary of the model variables and parameters Symbols ri ro rt Q L Cin Cout CZNF C* fm ft fe Dm Dt De kt ke kt se rm rt re Gt Ge w

Inner radius of the microdialysis probe membrane Outer radius of the microdialysis probe membrane Outer radius of the traumatized tissue layer Microdialysis infusion ﬂow rate Probe membrane length Dialysate analyte concentration at inlet Dialysate analyte concentration at outlet Dialysate analyte concentration at zero-net-ﬂux point Basal ECS concentration in normal tissue Volume fraction of diffusible space in probe membrane Volume fraction of the ECS in the trauma tissue Volume fraction of the ECS in the normal tissue Effective diffusivity of the analyte in the membrane Effective diffusivity of the analyte in the trauma tissue Effective diffusivity of the analyte in the normal tissue First-order uptake rate constant in the trauma tissue First-order uptake rate constant in the normal tissue Constant release rate of the analyte in the trauma tissue Constant release rate of the analyte in the normal tissue Diffusion resistance in the membrane region Diffusion resistance in the trauma tissue Diffusion resistance in the normal tissue Diffusion penetration distance in the trauma tissue pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ð¼ Dt =kt Þ Diffusion penetration distance in the normal tissue pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ð¼ De =ke Þ Concentration weighting function, deﬁned in Eq. (10)

neurotransmitter release/uptake processes in each region are indicated in the ﬁgure. By solving the coupled mass balance equation in each region simultaneously, the analytical concentration proﬁle within each region was obtained and the theoretical expression for E was found to be 1 E ¼ 1 exp , (8) Qðrm þ rt þ re Þ where rm, rt, and re are the diffusion resistances in the probe membrane, the traumatized tissue layer, and the normal tissue, respectively, and are

speciﬁcally given as ln rroi rm ¼ 2pLfm Dm Gt F t Grot ; Grtt ro and rt ¼ 2pLft Dt F ro ; rt ; rt G t G t Ge Ge K 0 Grte F e Grot ; Grtt rt re ¼ 2pLfe De K 1 rt F ro ; rt ; rt Ge Gt Gt Ge where K0( ) and K1( ) are the modiﬁed Bessel functions of the second kind of order zero and one, respectively, Fe, Ft, and F* are some geometric functions associated with the dimensions and properties of the microdialysis probe and tissues. These functions exist partly due to the arbitrary distinction of the tissue trauma layer from the normal tissue3, and are deﬁned as ro rt ro rt rt ro Ft ; ¼ K0 I0 K0 I0 Gt Gt Gt Gt Gt Gt

rt ro rt Fe K0 ¼ I1 ro Gt Gt rt ro þK 1 I0 Gt Gt ro rt ro rt rt F ; ; ¼ K1 I0 Gt Gt Ge Gt Gt rt ro þ K0 I1 Gt Gt ro K 0 Ge G e f t Dt þ Gt fe De K 1 rt Ge ro rt K0 I1 Gt Gt rt ro þ K1 I0 Gt Gt ro rt ; Gt Gt

3

It should be understood that microdialysis-induced tissue trauma is not distinctly conﬁned within close proximity of the probe but is progressively distributed. Hence, the rt in the model represents a cut-off distance similar in concept to the molecular size cut-off for the microdialysis membrane permeability. The Dt, ft, es, and ek represent the averaged properties within the conceptualized distinct trauma region.

55

where I0( ) and I1( ) are the modiﬁed Bessel functions of the ﬁrst kindﬃ of order zero and ﬃone, pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ respectively, Gt ð¼ Dt =kt Þand Ge ð¼ De =ke Þare the diffusion penetration distances in the traumatized and normal tissue, respectively. Another important theoretical result from the model is that the CZNF is related to C* by s C ZNF ¼ ð1 wÞ þ w C , k |ﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄ{zﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄ}

(9)

R=E

where es and ek are the remaining fractions of the release and uptake strengths, respectively, in the traumatized tissue (see Fig. 3 for symbol clariﬁcation). The w is a weighting function between the normal baseline C* and the altered baseline (Ct ¼ es C*/ek) in the traumatized layer, and is deﬁned as w¼

G =r t o

F Grot ; Grtt ; Grte

.

(10)

Since the factor between CZNF and C* is the ratio R/E, the in vivo recovery R is theoretically expressed as R¼

1 1 exp Qðrm þ rt þ re Þ s ð1 wÞ þ w , k

ð11Þ

in which the ﬁrst term in the brackets represents the extraction efﬁcient E deﬁned in Eq. (8). As the model was speciﬁcally built for studying the microdialysis ZNF method sampling endogenous neurotransmitter, several conclusions regarding the ZNF operation were made from the model simulations. First, the model shows that changes in the ECS structure (reﬂected by the modiﬁed Dt and ft) of the surrounding tissue do not affect the accuracy of the ZNF method. This prediction is reasonable since microdialysis ZNF method is supposed to operate under quasi-steady states. Second, microdialysis-induced damage on neurotransmitter release and uptakes can cause either underestimate or overestimate of the basal

ECS concentration of the affected neurotransmitters as determined by the microdialysis ZNF method, depending on which process is damaged more. Underestimation (i.e., CZNFoC*) occurs when the probe impairs more of the neurotransmitter release sites than the uptake site in the tissue surrounding the probe, and the reverse is also true. Third, the consequence of a serious ZNF underestimation (due to the overwhelming damage to the neurotransmitter release sites) is that the in vivo recovery R of the endogenous neurotransmitter would ‘‘increase’’ after systemic application of the speciﬁc neurotransmitter uptake blockers, exactly opposite of the known trend of the microdialysis extraction efﬁciency E with neurotransmitter uptake inhibition4. The three predictions all agree with the experimental and theoretical works of Michael and coworkers. However, one more in-depth observation from the model simulation is that the in vivo recovery of neurotransmitter R starts to increase with systemic uptake inhibition only when the neurotransmitter release site are damaged more than the corresponding uptake sites in the surrounding tissue. How much more of the neurotransmitter release damage is required for the R to exhibit a rising trend with systemic uptake inhibition? Model simulations (Chen, 2006) indicate that for this rising trend (i.e., an increasing R of [DA]e following DA uptake inhibition) to occur is when es/eko0.4 (see the Fig. 11 in Chen, 2006). This criterion is based on a relative scale. That is, if the uptake sites surrounding the microdialysis probe were also damaged such that ek is, say, 0.5, then the maximum es to exhibit a rising R upon uptake inhibition is 0.4 0.5 ¼ 0.2. 4 Experiments on the microdialysis ZNF method have demonstrated that the slope of the concentration difference plot (i.e., the extraction efﬁciency E) of several neurotransmitters decreased upon application of neurotransmitter uptake inhibitors. The tested neurotransmitters include dopamine (Smith and Justice, 1994), serotonin and norepinephrine (Cosford et al., 1996). In the opposite direction of uptake inhibition, Zapata and Shippenberg (2002) showed that dopamine D3 receptor agonist signiﬁcantly ‘‘increased’’ the microdialysis extraction efﬁciency of dopamine and ‘‘decreased’’ dopamine ZNF concentration, indicating involvement of D3 receptors on DA transporter activity.

56

Although the mathematical models of Bungay et al. (2003) and Chen (2005b, 2006) allow the effects of tissue trauma on microdialysis performance to be studied, the models also introduce more unknown parameters, most of which cannot be obtained experimentally. For example, the models characterize the tissue trauma by a distinct tissue layer of different diffusive and release/uptake properties immediately outside of the probe. It should be understood that microdialysis-induced tissue trauma is not distinctly conﬁned within close proximity of the probe but is progressively distributed. Therefore, the trauma layer distance rt in the model represents a cut-off distance similar in concept to the molecular size cut-off for the microdialysis membrane permeability. Furthermore, the es and ek represent the averaged properties within the conceptualized distinct trauma region. What the models offer allow us to analyze, from a theoretical perspective, how the microdialysis characteristics would be changed if tissue trauma did occur. III. Studying basal DA to infer errors in quantitative microdialysis III.A. Introduction The microdialysis-induced tissue damage on neurotransmitter release and uptake processes has particularly important implications in the microdialysis data studying extrasynaptic transmission and tonic regulation of neuronal activity by extracellular neurotransmitters. The experimental and theoretical works of Michael and coworkers (Lu et al., 1998; Peters and Michael, 1998; Yang et al., 1998; Borland et al., 2005) on extracellular DA in rat striatum had led them to the following conclusions: (1) Microdialysis probe implantation would traumatize the tissue immediately surrounding the probe, within which an overwhelmingly large number of the DA release sites are damaged compared with uptake sites. (2) Because of the preferential damage in DA release, the probe relative recovery is much smaller than the corresponding extraction efﬁciency of DA.

(3) For DA, not only is R smaller than E, R also exhibits the opposite trend of E in response to pharmacological DA uptake inhibitors. (4) The measured ZNF concentration CZNF signiﬁcantly underestimates the basal ECS concentration of DA in striatal tissues. From their experimental data (Lu et al., 1998; Yang et al., 1998; Borland et al., 2005), the damaged DA release induced by acute probe implantation was estimated to yield a CZNF as low as only 2% of the unperturbed ECS DA concentration C* in the striatum (Peters and Michael, 1998; Peters et al., 2000). Since the microdialysis ZNF concentration of ECS DA in the general region of rat striatum has been reported to be within the 6–12 nM range; in rat striatum (Sam and Justice, 1996; Jones et al., 1998), in nucleus accumbens (Parsons et al., 1991; Parsons and Justice, 1992; Smith and Justice, 1994; Yim and Gonzales, 2000), Michael and coworkers concluded that the resting level of striatal [DA]e should be at least 500 nM (Kulagina et al., 2001), and could be as high as 2 mM (Borland and Michael, 2004; Borland et al., 2005). Obviously, the conclusions and criticisms drawn from Michael and coworkers regarding the errors in microdialysis ZNF method have a profound impact on the use of microdialysis. Theoretical simulations modeling microdialysis characteristics under damaged neurotransmitter release and uptake (Peters and Michael, 1998; Bungay et al., 2003; Chen 2005a, 2005b, 2006) have shown that the above predictions are true if probe implantation indeed caused preferential damage on the neurotransmitter release sites. What is not shown independently is whether striatal [DA]e is at the micromolar level. To investigate if the reported ZNF concentrations for ECS DA seriously underestimate the basal [DA]e in rat striatum, one must examine current evidence of the true resting level of striatal [DA]e obtained by means other than microdialysis. As the tonic [DA]e sets the background level of DA signaling as well as the transmission of other excitatory neurotransmitters, investigation of the tonic level of ECS DA is itself highly important aside from the issue of microdialysis validity.

57

III.B. Basal level of extracellular DA in rat striatum Since the accuracy of the ZNF method in measuring the true, unperturbed C* is being questioned, the data obtained by real-time FSCV are particularly relevant because the smaller carbon-ﬁber sensor is believed to induce less tissue disruption (Khan and Michael, 2003). Unfortunately, most FSCV studies are designed to measure only the dynamic changes of extracellular monoamines during electrical stimulation because stable signals from the basal [DA]e in vivo cannot be reliably obtained. Besides the fact that the voltammetric currents are prone to shifts in the ionic environment of the brain, the active DA uptake and metabolism are thought to keep the [DA]e so low that other electroactive substances that oxidize at similar voltage potentials, such as DOPAC (5 mM) and ascorbate acid (400 mM), seriously mask the true DA signals (Rice and Nicholson, 1995). If striatal [DA]e is indeed 2 mM, there seems to be no reason why such a high basal [DA]e cannot be detected by FSCV. The hypothesis of a low basal ECS DA and high interferences from DA metabolites in striatum is plausible because the basal [DA]e in the retina, where we believe DA uptake is minimal and extracellular DOPAC is low, could be voltammetrically measured without difﬁculty (300 nM; Witkovsky et al., 1993). The same reasoning also explains why in vitro a much lower DA concentration (at the nanomalar range) can always be measured in a beaker (perhaps because no signiﬁcant interferences from other ions or metabolites).

III.B.1. Evidence for a nanomolar DA Within the few voltammetry studies that used different techniques to measure the basal DA in the striatum of anesthetized rats, the reported values for [DA]e are quite random but consistently at the nanomolar range. Crespi and Mobius (1992) reported a basal DA level of 1.5 nM with differential pulse voltammetry (DPV), whereas Gonon and Buda (1985) and Suaud-Chagny et al. (1992) obtained higher values (26 and 8 nM, respectively) with differential normal pulse voltammetry. Particular attention should be paid to the work of

Blaha (1996), where he applied chronoamperometry with stearate–graphite paste electrode (SGE) and estimated a striatal [DA]e 85 nM. However, this estimation may be erroneous because Blaha used the total delivered DA concentration (Cin – Cout) to be the ECS concentration immediately outside of the microdialysis probe in his calculation. To make corrections on his data, it is necessary to ﬁrst summarize his experimental procedure. Blaha (1996) placed the SGE (o.d. 0.15 mm) next to a concentric microdialysis probe (o.d. 0.34 mm) implanted in rat striatum. After maximally inhibiting DA uptake by nomifensine (a DA transporter blocker) followed by perfusing the probe with Ca2+-free Ringer, Blaha performed retrodialysis with exogenous DA ranging between 0.5 and 2.5 mM and recorded the DA signals immediately outside the probe with the SGE. The combined use of a Ca2+-free Ringer and repeated nomifensine applications ensured that the surrounding tissue was free of spontaneous DA release and active uptake, and hence could be viewed as a passive medium. Under this circumstance, the DA signals recorded next to the microdialysis probe may truly reﬂect the infused concentration of exogenous DA. From the retrodialysis calibration, he found that the DA signals corresponding to Cin ¼ 2 mM was the same as the pre-nomifensine baseline (Fig. 4). Since Cin ¼ 2 mM also resulted in Cin Cout, ¼ 85 nM, he concluded that the basal [DA]e is about 85 nM, which obviously may not be the ECS DA concentration surrounding the probe. To recalculate the actual [DA]e outside of the probe, we can assume that the perfusion of Ca2+free Ringer completely inhibits synaptic DA release in the surrounding tissue so that the CZNF outside of the probe is zero and the microdialysis extraction efﬁciency can be calculated as (Cin Cout)/Cin ¼ 85 nM/2000 nM ¼ 0.043. According to the microdialysis theory developed in Chen et al. (2002), the steady-state extracellular concentration distribution under passive diffusion (valid when a very small k was used) is CðrÞ ¼ C in UðrÞ; with the normalized distribution U(r) deﬁned as r K 0 Ge r U (12) ¼ ro ln ro ro fe De K 1 ro þ K 0 ro ri

Ge f m D m

Ge

Ge

58

Fig. 4. In vivo calibration by retrodialysis. Nomifensine was ﬁrst applied (2.5 mg/kg) with a bolus i.v. injection, followed by the same dose for every 30 min for the ﬁrst hour and every hour thereafter. At the time nomifensine induced maxima uptake inhibition (observed by the increase of SGE signals), dialysis probe was perfused with a Ca2+-free Ringer at 5 mL/min. One hour after the start of Ca2+-free perfusion, which was when SGE signals started to decline, DA was added to the Ca2+-free Ringer in concentrations of 0.5, 1.0, 1.5, 2.0, and 2.5 mM. HPLC analysis indicated that Cin Cout ¼ 1976, 4573, 67710, 8676, and 11079 nM for each DA concentration infusion, respectively. The horizontal dashed line represents the pre-nomifensine level. Figure reproduced from Blaha (1996) with permission.

In the work of Blaha (1996), the microdialysis probe had an outer diameter 0.34 mm, an active membrane length 0.4 mm, and was perfused at the ﬂow rate 5 mL/min. Using ro ¼ 0.17 mm, L ¼ 0.4 mm, Q ¼ 5 mL/min, and other assumed parameters [fe ¼ 0.4 (according to Bungay et al., 2003), De ¼ 1.5 (2.4 10–6) cm2/s (control value increased by 1.5-fold due to increased fe), k ¼ 0.1 1/s (due to nomifensine application)], the following parameters are adjusted: ri ( ¼ 0.12 mm), fm ( ¼ 0.2), and Dm ( ¼ De), until the extraction efﬁciency, ! 2pro fe De L dU E ¼ 1 exp , (13) Q dr r¼ro agrees with the experimental value 0.043. To obtain the averaged [DA]e outside of the probe as measured by SGE, the spatial distribution of the theoretical ECS DA was averaged from r ¼ ro to r ¼ 5ro, which corresponds to the regions possibly interacted with the SGE, and yielded an averaged C¯ ¼ 20 nM. Hence, the baseline signals obtained in Blaha (1996) correspond to [DA]e ¼ 20 nM instead of the higher 85 nM. Nonetheless, it is cautioned that this nanomolar estimate of [DA]e in

the vicinity of the microdialysis probe can also be used to support the preferential damages on DA release sites surrounding the probe, hence favoring the claim of a much higher basal [DA]e in the distant, undisrupted tissue. In view of the available voltammetry data for baseline DA, the randomness in the reported data from different groups probably reﬂects the aforementioned hypothesis that voltammetry is not suitable for measuring the baseline level of electroactive substances that are in scarce quantity, especially in vivo where other interferences may be present. Hence the randomness is more or less expected. Nonetheless, all reported data fall within the nanomolar range, which is consistent with the conventional hypothesis that the basal [DA]e in the striatum is low. In addition to directly measuring ECS DA by voltammetry, several studies also attempted to estimate the basal ECS level of DA using voltammetrically derived kinetic data. Kawagoe et al. (1992) measured the DA uptake kinetic parameters and used them in a further analysis, estimating the time-averaged ECS DA 2 mm away from a DA synaptic cleft to be 6 nM. In a complex ﬁnitedifference model to simulate the tonic and phasic

59

DA releases in the caudate-putamen, Venton et al. (2003) predicted a tonic background [DA]e 30 nM using voltammetrically derived parameters. Although a consensus among all simulation works is that a low-basal ECS DA is consistent with the current voltammetrically measured DA kinetics, these simulation results may not be very conclusive because some other unknown mechanisms may not be accounted for. III.B.2. Evidence for a micromolar DA Nonetheless, evidence supporting a higher level of ECS DA is also available, mainly from the laboratory of Michael and coworkers. For example, Kulagina et al. (2001) recently found that intrastriatal infusion of the ionotropic glutamate receptor antagonist kynurenate caused a steep drop in the DA-like electrochemical signals equivalent to 500 nM of the ECS DA, implying that the ECS DA level must be at least above 500 nM to afford such concentration drops. Borland and Michael (2004) found even larger drops in the FSCV signals equivalent to 2 mM of extracellular DA. Some other indirect evidence also comes from the voltammetry recording of evoked DA release immediately adjacent to and distant from an acutely implanted microdialysis in the striatum while electrically stimulating the medial forebrain bundle (Lu et al., 1998; Yang et al., 1998; Borland et al., 2005). They found that the evoked DA response immediately adjacent to the probe was lacking while the DA response 1 mm away from the probe was robust. After systemic administration of nomifensine, evoked [DA]e response adjacent to the probe became detectable. They concluded that the lack of evoked DA response immediately adjacent to the microdialysis probe before nomifensine is due to the probe-induced destruction of essentially all the DA release sites in the same tissue region, while most uptake sites remained intact. After applying nomifensine to inhibit uptake, noticeable DA response adjacent to the probe was due to the prolonged diffusion of extracellular DA released from distant viable DAergic synapses. If this argument is true, one would expect a time delay in the DA signals at the probe after nomifensine because most of the DA signals recorded here should come from diffusion of the evoked DA

released at certain distances away from the probe. However, their recorded traces did not show such time delays. It is uncertain whether the recorded signals adjacent to the probe after nomifensine resulted from (i) a local DA release or (ii) a distant DA release plus diffusion. Additionally, Borland et al. (2005) found no increase in the non-evoked base line for [DA]e after nomifensine application. This is in contrast to the common understanding that nomifensine increases the local motor activity (Nakachi et al., 1995; Stanford et al., 2002; Garris et al., 2003), which can also be triggered by increases in [DA]e (Costall et al., 1983; Steinpreis and Salamone, 1993; Ando et al., 1994; Fan et al., 2000). The lack of noticeable changes in DA baseline in response to nomifensine may be attributed to the inability of FSCV to measure the non-evoked basal [DA]e, even after DA uptake process was blocked. However, a similar study performed by Blaha et al. (1996) also measured the basal [DA]e level in striatum from a SGE (stearate–graphite past electrode) standing alone or placed next to a microdialysis probe. Contrary to the ﬁndings of Borland et al. (2005), Balha et al. (1996) found increased basal [DA]e recorded from the stand-alone SGE after nomifensine application. Although their electrode is quite large in size (0.15 mm o.d.), they argued that the active recording area, conﬁned to the tip surface of the graphite paste, is quite small and not in direct contact with the highly traumatized tissue in the shaft region of the electrode. It appears that even with voltammetry technique, many results obtained from different groups regarding the basal [DA]e itself and the effect of DA uptake inhibitors on basal [DA]e can still be inconsistent. Currently, there appear to be no ready explanations to satisfactorily resolve these conﬂicting ﬁndings for the basal level of striatal [DA]e as measured by different voltammetry methods. Could the differences be attributed to acute and chronic microdialysis? Whereas the lack of evoked DA response in the tissue adjacent to an acutely implanted microdialysis probe has been positively veriﬁed by Michael and coworkers, it is uncertain whether this unresponsiveness of neurotransmitter release to impulses still persists over longer periods of time, as it has been hypothesized that synaptic

60

connection and synaptic transmission could be self-repaired to certain extent after inﬂammatory response subsided (Georgieva et al., 1993; MartinFardon et al., 1997; Hasbani et al., 2000). An ad hoc hypothesis here is that the traumatized tissue, given time, could partially recover from the implantation of the foreign object at about the same time the inﬂammatory response subside (which is what the recovery period is for in chronic microdialysis). Thus, it would be desirable if the microdialysis/voltammetry experiments of Michael and coworkers could be extended to chronic situations for further veriﬁcation.

III.C. Extracting basal extracellular DA level from DA kinetic study

assumed to be zero. In its current form, Eq. (14) is insufﬁcient to describe the DA kinetics immediately following electrical stimulation, for it implies that after stimulation the absolute [DA]e eventually declines to zero. An appropriate DA kinetic model must also recognize the fact that during the evoked DA overﬂow experiments, the carbon ﬁber electrode in fact registers the relative changes in [DA]e, DC, rather than the absolute [DA]e level, C ¼ C*+DC (Wightman and Zimmerman, 1990; Jones et al., 1998). Since the Michaelis–Menten uptake mechanism is commonly known to apply to the absolute concentration C, not DC (as demonstrated by many receptor binding experiments on synaptosomes), Eq. (14) cannot be used with the symbol C simply replaced by DC. Thus, this DA kinetic model needs to be modiﬁed.

Indirect estimates for a low ECS concentration of DA can also be made from DA kinetic studies using FSCV combined with carbon ﬁber microelectrodes (e.g., Wightman and Zimmerman, 1990; Kawagoe et al., 1992; Garris et al., 1994; Wu et al., 2001). Before making such estimates, it is necessary to address the baseline issue often ignored in the existing DA kinetic analyses.

III.C.2. Modifications of the DA kinetic model Assuming that [DA]e is non-zero, a basal level for [DA]e (denoted as C*) under non-stimulated condition should be sustained by the basal release rate sand can be described by Eq. (14) under steady ¯ state as

III.C.1. Existing DA kinetic model from voltammetry study In the experiments of cyclic voltammetry coupled with electrical stimulations, the non-linear DA uptake parameters, the maximum uptake velocity Vmax and the Michaelis–Menten constant Km, were obtained by ﬁtting the Michaelis–Menten uptake model (Wightman and Zimmerman, 1990) against the transient proﬁle of the evoked DA,

in which the basal level C* is similarly deﬁned as a mean value averaged over time and space. As noted, the FSCV technique is best suited for monitoring concentration changes (DC) of catecholamine over a short period of time, during which the baseline C* can be assumed constant. Hence, Eq. (14) during electrical stimulation (in which the basal release term is replaced by the synchronized release term, [DA]p f ) can be rewritten as

dC V max C ¼s , dt Km þ C

(14)

where s (nM/s) is the mean DA release rate averaged over space and time as deﬁned in Chen (2005b), and C represents the ECS DA concentration. During the electric stimulation, s is represented as a product of the unit [DA]e increase per stimulus pulse, [DA]p, and the stimulation frequency f (Hz). After stimulation ceases, s is

s¯ ¼

sK V max C ¯ m or C ¼ K m þ C V max s¯

(15)

dðC þ DCÞ dDC V max ðC þ DCÞ ¼ ¼ ½DAp f dt dt K m þ C þ DC (16) To obtain an equation for the measured variable, DC, Eq. (16) is subtracted from the governing equation at steady state, that is, 0¼

dDC V max C ¼ s¯ , dt K m þ C

61

yielding dDC V app max DC ¼ ½DAp f s¯ app , dt |ﬄﬄﬄﬄﬄﬄﬄﬄ{zﬄﬄﬄﬄﬄﬄﬄﬄ} K m þ DC

(17)

½DAapp p f

during electrical stimulation, where ½DAapp ¯ ; V app p ¼ ½DAp s=f max ¼ V max K m = ðK m þ C Þ ¼ V max s; ¯ and K app m ¼ K m þ C are the apparent parameters of the current DA kinetic model used in experimental ﬁtting for DC, as in contrast to the intrinsic parameters [DA]p, Vmax, and Km in Eq. (14) for C. The reason to arrange the original Eq. (16) for DC to Eq. (17), which retains the same form as Eq. (14), is that the existing ﬁtting algorithm used in many studies need not be changed then. After the cessation of electrical stimulation, the original DA kinetic model assumed that the tonic ﬁring of DAergic neurons is temporarily inhibited (Wightman and Zimmerman, 1990), thus leaving only DA uptake. However, for a more general form, we can designate the tonic DA release rate after electrical stimulation to be s(t), so that dDC V app max DC ¼ ½sðtÞ s ¯ app dt K m þ DC after electrical stimulation:

ð18aÞ

Note that in theory, one must assume 0 sðtÞ s¯ to obtain sensible results. If the tonic DA release after electrical stimulation is not inhibited, sðtÞ ¼ s; ¯ yielding another version of Eq. (18a) as dDC V app max DC ¼ app ; after electrical stimulation: dt K m þ DC (18b) The reason for particularly mentioning this special case of no post-stimulation inhibition in tonic DA release is that Eq. (18b) is the current equation used to ﬁt the uptake data of evoked DA following electrical stimulation. This implies that the current ﬁtting methodology, commonly applied by many researchers, in fact assumed no post-stimulation inhibition on DA release. On the other hand, since

the work of Kuhr et al. (1987) indicated that DA tonic release after electrical stimulation was indeed temporarily suppressed, one would expect to see the recorded DA signals in the uptake phase fall below the pre-stimulation level. In such case, the extent of the signal undershoot can provide an estimate of the basal [DA]e. Since all voltammetry studies on kinetic DA overﬂow and uptake did not show such undershoots, it might imply that either the post-stimulation inhibition on DA release is absent or short-lived, or the tonic DA release is largely impulse-independent and hence not affected by electrical stimulation. Alternative explanations for the lack of obvious undershoots in post-stimulation DA kinetics might be that undershoot did occur, but the relative scale of undershoot compared with the evoked DA overﬂow was too small to be noticed, the baseline level of the ECS DA was too low, or some combination of both. III.C.3. Interpretation of the fitted parameters Since the current DA kinetic model ﬁtted the kinetic data of the evoked DA changes, DC, against Eqs. (17) and (18b), it is obvious that the ﬁtted parameters are the apparent ones when the whole DC curve was ﬁtted (Wu et al., 2001). This means that for most kinetic data measured in rat striatum, including caudate-putamen and nucleus accumbens, the experimental V app max ¼ V max K m = ðK m þ C Þ ¼ 2:526:0mM=s and K app m ¼ Km þ C ¼ 0:1520:35mM (Kawagoe et al., 1992; Wu et al., 2001; Garris et al., 2003). The ﬁtted unit increase of ECS DA concentration is also the apparent one, ½DAapp ¼ ½DAp s=f ¯ ; which is not a constant p but depends on the stimulation frequency f. On average, ½DAapp has a value 0.27 mM/s when f is p 10 Hz (Kawagoe et al., 1992). Assuming s¯ ¼ 0:25mM=s 5Hz ¼ 1:25mM=s (the frequency 5 Hz is the assumed resting ﬁring frequency of DAergic app neurons), the ﬁtted V app max ¼ 3:8mM=s and K m ¼ 0:2mM; one can determine the intrinsic Vmax, Km, and the basal C* from the aforementioned relations by recursive iterations and self-corrections, as demonstrated in Chen (2005b). It was found that the intrinsic Vmax and Km are 5.05 mM/s and 0.15 mM, respectively, resulting from a basal C*49.5 nM.

62

In the literature, there is another type of data analysis: the individual parameters were ﬁtted separately from different regions of the response curve (Kawagoe et al., 1992; Wu et al., 2001). For example, the maximum velocity V app max (or Vmax) was ﬁtted from the largest descending slope of the signal curve immediately following the stimulation, assuming that C at this stage was much higher than Km. The apparent V app max was obtained from the slope if Eq. (18b) was used (i.e., sðtÞ ¼ s). ¯ This experimental V app ð¼ V sÞ is a lower ¯ max max bound of the intrinsic Vmax. In contrast, if sðtÞ ¼ 0; s¯ þ V app max ¼ V max is the slope obtained. Thus the measured slope can range from ðV max sÞ ¯ to V max : Similarly, [DA]p (or ½DAapp ) was deterp mined from the difference of the slopes on both sides of the curve at the end of stimulation: ½DAp ð or

1 dDC dDC ½DAapp , p Þ¼ f dt before dt after (19)

which could yield either [DA]p or ½DAapp p ; depending on if sðtÞis equal to its minimum (zero) or maximum (s) ¯ immediately after electrical stimulation, respectively. As to the Michaelis–Menten constant, Wu et al. (2001) proposed that at lowfrequency stimulations, a new elevated steady state, DCSS, can be reached, such that K app m

¼

! V app max 1 DC SS , ½DAapp p f

(20)

which was derived from Eq. (17) assuming steady states. Regardless of whether the experimental paapp rameters in Eq. (20) are V app or Vmax max and ½DAp and [DA]p, the Michaelis–Menten constant so determined by Eq. (20) appears to be the apparent one, that is, Km+C*. Recently the DA kinetic model of Chen (2005b) was further modiﬁed by Michael et al. (2005) by separating the tonic release rate, s(t), from the forced DA release during electrical stimulation: dDC V max ðC þ DCÞ ¼ ½DAp f þ sðtÞ . dt K m þ C þ DC

(21)

The advantage of this slightly modiﬁed model is that both the evoked DA release and uptake during and after electric stimulation can be described by one single equation, Eq. (21), by adjusting the time course of s(t), which must similarly be bound between [0, s], ¯ with s¯ being the steady-state resting level. Note that unlike the model of Michael et al. (2005), the s(t) in Eq. (18a) exists only after the cessation of electrical stimulation; during electrical stimulation the s(t) in this chapter is viewed as zero. In contrast, the s(t) in Eq. (21) during electrical stimulations could be non-zero. If the tonic release s(t) during electric stimulations is nonzero, which is the case in the theoretical work of Michael et al. (2005), the DA release during pulse stimulation is composed of two modes, a synchronized release enforced by the stimulation frequency f and a non-synchronized release s(t). Michael and coworkers also contended that the experimentally determined Michaelis–Menten constant K app m is related to its intrinsic counterpart byK app ¼ K m C ;not K app m m ¼ K m þ C : If the C* for striatal DA is indeed 2 mM as claimed (Borland and Michael, 2004; Michael et al., 2005), their model would imply that the intrinsic Km for the striatal DA system would be 2+0.16 ¼ 2.16 mM. This is very different from the many Km measurements found in striatal synaptosomes (e.g., Morel et al., 1998), which, however, are consistent with the FSCV measurements in the lower range (0.15–2.0 mM). A comparison of both models is summarized in Table 2. It should be noted that the relation K app m ¼ K m C exists only when K app was ﬁtted from the m DA concentration corresponding to one-half of the maximal uptake rate, d(–DC)/dt|max. If ﬁtting the whole curve by non-linear regression, K app m ¼ K m þ C should be the relation implied in the ﬁtted parameters. Furthermore, Michael et al. (2005) suspected that the model of Chen (2005b) yields ‘‘unexpected’’ results when the product value ½DAp f is smaller than the basal release rate s: ¯ This ‘‘unexpected’’ result should be attributed to the use of unreasonable parameter values rather than the model’s defect. During electrical stimulations, if the voltammetry signals yield a positive response (i.e., dDC/dt>0), it is expected that the total DA release rate during this stage be larger

63 Table 2. Comparisons of the DA kinetic models incorporating non-zero DA baseline Models

Chen (2005b)

Michael et al. (2005)

Comments

Governing equations

Eqs. (17) and (18a, 18b)

Eq. (21)

Essentially the same

Tonic release time course s(t)

Zero during electrical stimulation

Could be non-zero during electrical stimulation

Baseline C*

Low (nanomolar level); bound by experimental K app m value

High (2 mM); not bound by experimental K app m value

Fundamentally different

Apparent maximal uptake velocity, V app max

V app max ¼ V max K m =ðK m þ C Þ

V app max ¼ V max sðtÞ

Essentially the same. Depending on the poststimulation s(t), the experimental V app max can range from (Vmaxs) ¯ to Vmax

Apparent Michaelis–Menten constant, K app m

K app m ¼ Km þ C

K app m ¼ Km C

Fundamentally different

Intrinsic Michaelis–Menten constant, Km

Low; bound by experimental K app m value

High; must be higher than the baseline C*

Unit DA increase, ½DAapp p

½DAapp ¼ ½DAp s=f ¯ p

No speciﬁc relation

Uptake behaviors

Rapid uptake after electrical stimulations is due to a high Vmax and a low s; ¯ release inhibition after electrical is not critical

high. Both Vmax and sare ¯ Rapid uptake after electrical stimulation is due to temporary inhibition of poststimulation DA release

than the basal release rate s: ¯ Irrespective whether the experimentally measured quantity is the intrinsic ½DAp or the apparent ½DAapp p ; it is necessary to impose that ½DAp f 4s¯ in Chen’s model to obtain reasonable simulations. This restriction on reasonable parameters is necessary and universally applicable to all models with no exceptions. For example, if the value

assigned for the total release rate ½DAp f þ sðtÞ in Eq. (21) during electrical stimulation is smaller than the constant basal release rate s; ¯ the model of Michael et al. (2005) similarly produces the ‘‘unexpected’’ negative response. Nonetheless, the most salient point in the model of Michael et al. (2005) is that they offer another possible explanation for the experimentally observed DA clearance kinetics. Since they argued for the high basal [DA]e (2 mM), the tonic DA release rate s¯ must be comparable with the maximal uptake rate constant Vmax, hence resulting a

Experimental ½DAapp p could be the intrinsic or the apparent one

slower than expected DA clearance kinetics. In this case, the rapid DA clearance seen in many voltammetry experiments must be attributed to the temporary inhibition of the on-going tonic release s(t) after electrical stimulation. Without this temporary inhibition in the tonic DA release, the overall DA clearance would be slow, as is the case for the clearance of exogenous DA (Michael et al., 2005), wherein dopaminergic neurons were not electrically stimulated. Whereas this hypothesis is very appealing and can explain the case of exogenous DA clearance, one should bear in mind that to apply this theory to interpret the DA kinetics experiments involving electrical stimulation, the release function s(t) must be manipulated in a rather precise manner: to reduce s(t) greatly immediately after the cessation of electrical stimulation so that a rapid clearance response can be seen, and to recover s(t) back to its basal level s¯ at the right moment so that no undershoot of the signals

64

below the pre-stimulation level could occur. The latter imposes an intriguing question. Could the recovery period of the DA tonic release be delayed by any means (e.g., by applying a stronger or faster stimulation amplitude/frequency, or using various pharmacological agents) to reveal the true signal undershoot in the voltammetry experiments to support the contention of a high basal [DA]e? Answer to this question should be the direction of future studies in the DA kinetics to resolve the current controversies in baseline [DA]e. III.C.4. Behavior of DA recovery with DA uptake inhibition To assess ‘‘indirectly’’ whether microdialysis implantation preferentially destroys the neurotransmitter uptake sites and hence causes the microdialysis ZNF method to underestimate the extracellular neurotransmitter concentrations, one can examine whether the relative recovery of the collected neurotransmitters increases or decreases after blocking the neurotransmitter uptake processes. If the neurotransmitter recovery ‘‘increases’’ upon inhibiting neurotransmitter uptake, this will imply that the claim by Michael and coworkers regarding the preferential damage of neurotransmitter release and the higher DA concentration is true. As mentioned earlier, the difﬁculty in investigating the trend of microdialysis relative recovery R ( ¼ Cout/C*) with neurotransmitter uptake inhibition lies in the uncertainty of determining C*, as the dialysate content (Cout) can be easily analyzed by HPLC-ECD. Since the validity of the microdialysis ZNF method in determining the basal C* for neurotransmitters is under question, it is necessary to use independent technique such as voltammetry to measure the steady-state C*. In this aspect, the recent data from Budygin et al. (1999, 2000) prove useful in evaluating the behavior of DA recovery with DA uptake inhibition, as both microdialysis and voltammetry techniques were used to evaluate the DA dialysate concentration (Cout) and DA uptake/release parameters, separately. The main ﬁndings of their works are that administration of GBR12909 (a selective DA uptake inhibitor) to freely moving rats increased the dialysate DA recovered from the striatum to 300% of the control level, while the K app m value, as ﬁtted from the evoked

DA kinetic data, increased from 160 to 2,500 nM, a 15-fold increase. On the contrary, no signiﬁcant app were observed with changes in V app max and ½DAp respect to GBR12909. Note that their voltammetry and microdialysis measurements were performed in different groups of rats; so the measured increase app in K app and constant V app by the m max and ½DAp smaller-size carbon-ﬁber microelectrodes should be free of the concern in tissue disruption as is commonly criticized with microdialysis experiments. If the basal level [DA]e is solely determined by rate balance between DA release and uptake, and indeed remains low at the nanomolar level, a 15-fold app increase in K app m and a constant V max ð¼ 3:8 mM=sÞ would imply that the basal [DA]e also increased 15 fold from the estimated control level (49 nM; Chen, 2005b). Even after considering the action of auto-inhibition exerted by DA autoreceptors, 10fold increase in [DA]e after GBR12909 should still exist. This interpretation is based on the model of Chen (2005b). In contrast, if adopting the rela tionship that K app m ¼ K m C and C* ¼ 2 mM as assumed in Michael et al. (2005), increase of K app m from 160 to 2,500 nM only modiﬁes the intrinsic Km from 2.2 to 4.5 mM. This small increase in Km will probably elevate the basal [DA]e by about twofold, making the DA recovery after GBR12909 larger than the control. In view of this discrepancy in the model predictions, it is impossible at present to draw credible conclusions regarding how DA recovery varies with DA uptake inhibition. It is necessary to clarify which model depicts the actual DA uptake system in the brains before accurate predictions could be drawn. III.C.5. Physiological perspectives It is worth noting the range of the binding afﬁnities for DA receptors, as their physiological levels may also give hints to the ‘‘expected’’ basal [DA]e from the perspective of DA transmission. Numerous binding competition studies have shown that both DA D1 and D2 receptors can exist in two afﬁnity states (e.g., George et al., 1985; Mireylees et al., 1986; Seeman and Grigoriadis, 1987; Richﬁeld et al., 1989; Neve and Neve, 1997; Seeman and Tallerico, 2003). Whereas the low-afﬁnity binding constants of the DA D1 and D2 receptors are at the micromolar level, their high-afﬁnity

65

binding constants fall within the nanomolar ranges. Binding competition between DA and [3H]DA has revealed an equilibrium dissociation constant of the high-afﬁnity D1 receptor with exogenous DA in the range of 0.8–50 nM [0.8 nM – Seeman and Grigoriadis (1987); 10–32 nM – Neve and Neve (1997); 49720 nM – Richﬁeld et al. (1989)]. For the high-afﬁnity D2 receptors, the equilibrium dissociation constant has been reported to be 1.75–50 nM [1.75 nM – Seeman (2002); 7.5 nM – Seeman and Grigoriadis (1987); 4–50 nM – Neve and Neve (1997); 4374 nM – Richﬁeld et al. (1989)]. In view of the high-afﬁnity D1 and D2 receptors and their extrasynaptic distributions, a temporally and spatially averaged ECS DA concentration at the nanomolar level does not exclude the possibility of tonic participation of DA D1 and D2 receptors in DAergic transmission. In contrast, a micromolar-level [DA]e would completely saturate these high-afﬁnity receptors. Considering the functions of DA D1 and D2 receptors, it is likely that the high-afﬁnity receptors regulate the tonic DAergic transmission, as they can be easily activated by the nanomolar [DA]e. During phasic discharge, in which local [DA]e can easily shoot up to micromolar level, the low-afﬁnity receptors are activated as a second gating mechanism to regulate the excitatory synaptic activity. From a neurochemical point of view, too high background DA risks potentiating the DA high-afﬁnity receptors, overwhelming the DAT uptake system during phasic discharges, and consequently, overcoming any concomitant presynaptic inhibition of DA release. According to the current Km values measured from in vivo FSCV and in vitro synaptosomes experiments, a 500 nM ECS DA will keep 500/(200+500) ¼ 71% of the extrasynaptic DAT population occupied, leaving little capacity to further clear the more massive DA overﬂow that occurs during intense phasic ﬁring. A [DA]e ¼ 2 mM would essentially saturate the complete DAT uptake system. Of course, this estimate is based on that the intrinsic Km is 0.2 mM. If the Km is 2.2 mM instead of 0.2 mM, and [DA]e ¼ 2 mM as claimed (Michael et al., 2005), 50% of the DAT population is still occupied under the resting condition. Whether 50% saturation of the DAT population

at rest constitutes an effective extrasynaptic DAT uptake system deserves further investigation. Excessive [DA]e may induce DA excitoxicity (Alagarsamy et al., 1997; McLaughlin et al., 1998). However, noticeable DA toxicity can only be demonstrated at [DA]e concentrations higher than 50 mM, which hence does not help resolve the issue of baseline [DA]e here. Too much DA in the brain has been implicated in the DAergic attention-deﬁcit hyperactivity disorder (Carboni et al., 2003) and the positive symptoms of schizophrenia (AbiDargham et al., 2000). Many antipsychotic drugs, therefore, act to relieve positive schizophrenic symptoms by reducing the brain DA level (the dopamine hypothesis). In summary, from the aspects of a robust DA homeostatic mechanism and a high signal-to-noise ratio (a clear differentiation between backgrounds or baseline ﬁring signals and those that are evoked by afferent stimulation), a low ECS DA level may be more desirable.

IV. Data from other neurotransmitters IV.A. Glutamate Numerous microdialysis studies of extracellular glutamate (Di Chiara et al., 1996; Herrera-Marschitz et al., 1996; Westerink and Timmerman, 1999) have established that the glutamate content in the dialysate is insensitive to the application of TTX, a fast Na+ channel blocker. For glutamate, this TTX-insensitivity was attributed to the substantial contribution from glia-released glutamate to the ECS pool (Di Chiara et al., 1996; Timmerman and Westerink, 1997; Westerink and Timmerman, 1999). A recent study further supports the hypothesis that the primary origin of extrasynaptic glutamate in the striatum arises from non-vesicular glutamate release by the cystine–glutamate antiporters that are preferentially located on glial elements (Baker et al., 2002). Recently, Khan and Michael (2003) offered an alternative explanation that the TTX-insensitivity of the microdialysis-recovered glutamate is due to the probe-induced tissue trauma, which similarly suppressed glutamate release to such an extent that no further suppression of glutamate release could be seen upon TTX

66

application. Whether the probe-induced tissue trauma is the cause of the TTX-insensitivity in microdialysis glutamate is uncertain; but if it was, one still needs to reconcile why probe-induced tissue trauma did not similarly eradicate the dialysate TTX sensitivity for other neurotransmitters, such as DA, serotonin, noradrenaline, and acetylcholine (Di Chiara et al., 1996; Westerink and Timmerman, 1999). The extracellular glutamate level in rat striatum has been measured by microdialysis ZNF method to be 3 mM (Miele et al., 1996). Using a novel glutamate microsensor, Kulagina et al. (1999) recently estimated the extracellular glutamate level, [GLU]e, to be 29 mM, about ten times as high as the ZNF measurement. Kulagina et al. (1999) interpreted these disparate results between microdialysis and their microsensor studies as an indication that the microdialysis ZNF method similarly underestimates the basal ECS glutamate concentration due to the probe-induced tissue damage to glutamatergic nerve terminals. Interestingly, in a recent work of Kennedy et al. (2002) a silica capillary was used to directly withdraw ECS ﬂuid at ultralow rates (1–50 nL/min) to measure ECS glutamate in rat striatum. Although the size of the silica capillary (o.d. 90 mm) is still larger than the typical carbon-ﬁber microelectrode, it is considerably smaller than microdialysis probes. Most importantly, the ECS ﬂuid was collected from the tip instead of the shaft region of the capillary; hence ﬂuid drainage is more of an issue than the insertion-induced tissue disruption. Supposedly, these authors induced less tissue disruption in the implanted site than Miele et al. (1996) did with microdialysis; however, Kennedy et al. (2002) obtained the same low ECS glutamate concentration (1.7670.15 mM). A major concern in the method of Kennedy et al. (2002) is whether directly withdrawing ECS ﬂuid would cause local ﬂuid drainage around the sampling capillary. If drainage was not serious, the sampled glutamate content represented the ECS level in the locally damaged tissue only (more or less similar to the CZNF). However, if drainage did occur, this would mean that the sampled ECS ﬂuid was not only from the locally damaged tissue area but also from some more distant regions, where tissue trauma was less

serious. In this case, the glutamate content obtained from the ﬁrst few withdrawals (before tissue response to ECS ﬂuid drainage kicks in) would still be able to reﬂect better the undisturbed C*. When considering the basal level of extracellular glutamate, it is also important to recognize that glutamate, though the major excitatory neurotransmitter in the central nervous system, can also become a toxin leading to neuronal excitotoxicity when overly accumulated in the ECS. What is the concentration range for extracellular glutamate to become toxic? In vitro studies (Ankarcrona et al., 1995; Yu et al., 2003) showed that incubation of cerebellar and cortical neurons with as low as 30 mM of glutamate was sufﬁcient to induce cell death (both necrosis and apoptosis). In view of this fact, it is uncertain if the resting [GLU]e in intact brains can be as high as 29 mM.

IV.B. Serotonin DA is not the only neurotransmitter that has been extensively studied by either the microdialysis or the voltammetry technique. Another neurotransmitter, serotonin (5-HT) has also been studied using either methods, hence allowing assessment of the possible errors associated with the measurements of the extracellular 5-HT by the microdialysis ZNF method. In the rat dorsal raphe (a brain region rich with serotonergic neurons), the CZNF of extracellular 5-HT has been reported to be 1.3 nM (Tao et al., 2000). Using the DPV method, however, Crespi et al. (1988) measured the extracellular 5-HT in the dorsal raphe to be 10 nM, about eight times as large as the ZNF estimation. Although the accuracy of the DPV technique in determining the basal level of 5-HT is unclear, kinetic data from evoked release and uptake of 5-HT in the same brain region indicate that this estimate might be very close to the actual value based on the following calculations. Using FSCV, Bunin et al. (1998) measured the 5-HT release and uptake parameters in the dorsal raphe of app rat brains as V app max ¼ 1:3mM=s; K m ¼ 170nM; and [5-HT]p ¼ 100 nM/s. If the basal ﬁring rate of serotonergic neurons can be assumed to be 1.4 Hz (Robichaud and Debonnel, 2005), adopting the

67

same approach in Chen (2005b) for calculation yields a basal [5-HT]e 16 nM, which is very close to the DPV estimate (10 nM). The difference in the results obtained by both methods lends support to that tissue trauma affected the accuracy of the ZNF measurement for 5-HT, which underestimates the undisrupted basal [5-HT]e in dorsal raphe by 10-fold.

future similar experiments could be conducted under chronic situations: one may wish to observe whether the lack of evoked DA overﬂow next to the microdialysis probe still persists after a sufﬁciently long recovery time. Only then the validity of the quantitative microdialysis methods sampling brain neurotransmitters could be better assessed.

V. Conclusions

References

Microdialysis probe inevitably damages tissue integrity and neurochemistry, in terms of morphological alterations of tissue ECS structure and physiological impairment of neurotransmitter release/uptake. The damage on the latter two active processes can fundamentally sway the ability of the two commonly used microdialysis methods in sampling the ‘‘unperturbed’’ extracellular neurotransmitters, the ZNF method and the varied ﬂow rate method. Examining the data from the microdialysis ZNF measurements and FSCV experiments on striatal DA systems shows reasonable doubt that the current ZNF measurement on striatal DA (5–10 nM) did underestimate the basal level [DA]e. However, the underestimation is not as serious as previously thought by Michael and coworkers. Namely, although the striatal [DA]e could be much higher than the single-digit nanomolar range determined by the ZNF method, the basal [DA]e is probably still conﬁned within nanomolar level, perhaps fewer than 100 nM. Two modiﬁed DA kinetic models accounting for the presence of DA baseline were presented and compared. Considering the various theoretical interpretations of the model together with other evidence such as the existence of high afﬁnity DA receptors and the low Michaelis–Menten uptake constant obtained from synaptosomes, a nanomolar [DA]e appears to be the more consistent conclusion at present. The most convincing evidence for the serious underestimation of the microdialysis ZNF method comes from the experimental data that compared the evoked DA response at sites close to and distant from an acutely implanted microdialysis probe. It would be greatly desirable if in the

Abi-Dargham, A., Rodenhiser, J., Printz, D., Zea-Ponce, Y., Gil, R., Kegeles, L.S., Weiss, R., Cooper, T.B., Mann, J.J., Van Heertum, R.L., Gorman, J.M. and Laruelle, M. (2000) Increased baseline occupancy of D2 receptors by dopamine in schizophrenia. Proc. Natl. Acad. Sci. U.S.A., 97: 8104–8109. Alagarsamy, S., Phillips, M., Pappas, T. and Johnson, K.M. (1997) Dopamine neurotoxicity in cortical neurons. Drug Alcohol Depend., 48: 105–111. Amberg, G. and Lindefors, N. (1989) Intracerebral microdialysis: II Mathematical studies of diffusion kinetics. J. Pharmacol. Methods, 22: 157–183. Ando, K., Miyata, H. and Yanagita, T. (1994) Effects of methamphetamine, dopamine and noradrenaline administered into the nucleus accumbens of rats discriminating subcutaneous methamphetamine. Jpn. J. Pharmacol., 64: 35–40. Ankarcrona, M., Dypbukt, J.M., Bonfoco, E., Zhivotovsky, B., Orrenius, S., Lipton, S.A. and Nicotera, P. (1995) Glutamate-induced neuronal death: a succession of necrosis or apoptosis depending on mitochondrial function. Neuron, 15: 961–973. Baker, D.A., Xi, Z.-X., Shen, H., Swanson, C.J. and Kalivas, P.W. (2002) The origin and neuronal function of in vivo nonsynaptic glutamate. J. Neurosci., 22: 9134–9141. Benveniste, H. and Diemer, N.H. (1987) Cellular reactions to implantation of a microdialysis tube in the rat hippocampus. Acta Neuropathol. (Berl.), 74: 234–238. Benveniste, H., Hansen, A.J. and Ottosen, N.S. (1989) Determination of brain interstitial concentrations by microdialysis. J. Neurochem., 52: 1741–1750. Benveniste, H. and Huttemeier, P.C. (1990) Microdialysis – theory and application. Prog. Neurobiol., 35: 195–212. Blaha, C.D. (1996) Evaluation of stearate-graphite paste electrodes for chronic measurement of extracellular dopamine concentrations in the mammalian brain. Pharmacol. Bioechm. Behav., 55: 351–364. Blaha, C.D., Coury, A. and Phillips, A.G. (1996) Does monoamine oxidase inhibition by pargyline increase extracellular dopamine concentrations in the striatum? Neuroscience, 75: 543–550. Borland, L.M. and Michael, A.C. (2004) Voltammetric study of the control of striatal dopamine release by glutamate. J. Neurochem., 91: 220–229.

68 Borland, L.M., Shi, G., Yang, H. and Michael, A.C. (2005) Voltammetric study of extracellular dopamine near microdialysis probes acutely implanted in the striatum of the anesthetized rat. J. Neurosci. Methods, 146: 149–158. Bourne, J.A. (2003) Intracerebral microdialysis: 30 years as a tool for the neuroscientist. Clin. Exp. Pharmacol. Physiol., 30: 16–24. Budygin, E.A., Gainetdinov, R.R., Kilpatrick, M.R., Rayevsky, K.S., Mannisto, P.T. and Wightman, R.M. (1999) Effect of tolcapone, a catechol-O-methyltransferase inhibitor, on striatal dopaminergic transmission during blockade of dopamine uptake. Eur. J. Pharmacol., 370: 125–131. Budygin, E.A., Kilpatrick, M.R., Gainetdinov, R.R. and Wightman, R.M. (2000) Correlation between behavior and extracellular dopamine levels in rat striatum: comparison of microdialysis and fast scan cyclic voltammetry. Neurosci. Lett., 281: 9–12. Bungay, P.M., Morrison, P.F. and Dedrick, R.L. (1990) Steady-state theory for quantitative microdialysis of solutes and water in vivo and in vitro. Life Sci., 46: 105–119. Bungay, P.M., Newton-Vinson, P., Isele, W., Garris, P.A. and Justice, J.B. Jr. (2003) Microdialysis of dopamine interpreted with quantitative model incorporating probe implantation trauma. J. Neurochem., 86: 932–946. Bunin, M.A., Prioleau, C., Mailman, R.B. and Wightman, R.M. (1998) Release and uptake rates of 5-hydroxytryptamine in the dorsal raphe and substantia nigra reticulata of the rat brain. J. Neurochem., 70: 1077–1087. Carboni, E., Silvagni, A., Valentini, V. and Di Chiara, G. (2003) Effect of amphetamine, cocaine and depolarization by high potassium on extracellular dopamine in the nucleus accumbens shell of SHR rats: an in vivo microdialysis study. Neurosci. Biobehav. Rev., 27: 653–659. Chefer, V.I., Kieffer, B.L. and Shippenberg, T.S. (2003) Basal and morphine-evoked dopaminergic neurotransmission in the nucleus accumbens of MOR- and DOR-knockout mice. Eur. J. Neurosci., 18: 1915–1922. Chen, K.C. (2005a) Preferentially impaired neurotransmitter release sites not their discreteness compromise the validity of microdialysis zero-net-ﬂux method. J. Neurochem., 92: 29–45. Chen, K.C. (2005b) Evidence on extracellular dopamine level in rat striatum: implications to validity of quantitative microdialysis. J. Neurochem., 92: 46–58. Chen, K.C. (2006) Effects of tissue trauma on characteristics of microdialysis zero-net-ﬂux method sampling neurotransmitters. J. Theor. Biol., 238: 863–888. Chen, K.C., Hoistad, M., Kehr, J., Fuxe, K. and Nicholson, C. (2002) Theory relating in vitro and in vivo microdialysis with one or two probes. J. Neurochem., 81: 108–121. Clapp-Lilly, K.L., Roberts, R.C., Duffy, L.K., Irons, K.P., Hu, YT. and Drew, K.L. (1999) An ultrastructural analysis of tissue surrounding a microdialysis probe. J. Neurosci. Methods, 90: 129–142. Cosford, R.J.O., Vinson, A.P., Kukoyi, S. and Justice, J.B. (1996) Quantitative microdialysis of serotonin and norepinephrine: pharmacological inﬂuences on in vivo extraction fraction. J. Neurosci. Methods, 68: 39–47.

Costall, B., Domeney, A.M. and Naylor, R.J. (1983) A comparison of the behavioural consequences of chronic stimulation of dopamine receptors in the nucleus accumbens of rat brain effected by a continuous infusion or by single daily injections. Naunyn Schmiedebergs Arch. Pharmacol., 324: 27–33. Crespi, F., Martin, K.F. and Marsden, C.A. (1988) Measurement of extracellular basal levels of serotonin in vivo using naﬁon-coated carbon ﬁber electrodes combined with differential pulse voltammetry. Neurosci., 27: 885–896. Crespi, F. and Mobius, C. (1992) In vivo selective monitoring of basal levels of cerebral dopamine using voltammetry with naﬁon modiﬁed (NA-CRO) carbon ﬁber microelectrodes. J. Neurosci. Methods, 42: 149–161. Dabrosin, C. (2005) Microdialysis – an in vivo technique for studies of growth factors in breast cancer. Front. Biosci., 10: 1329–1335. De Lange, E.C.M., De Boer, A.G. and Breimer, D.D. (2000) Methodological issues in microdialysis sampling for pharmacokinetic studies. Adv. Drug Deliv. Rev., 45: 125–148. Di Chiara, G., Tanda, G. and Carboni, E. (1996) Estimation of in vivo neurotransmitter release by brain microdialysis: the issue of validity. Behav. Pharmacol., 7: 640–657. Fan, X.-D., Li, X.-M. and Juorio, A.V. (2000) Substantia nigra pars reticulata lesion induces preconvulsive behavior and changes in glutamate receptor gene expression in the rat brain. Brain Res., 867: 40–51. Garris, P.A., Budygin, E.A., Phillips, P.E.M., Venton, B.J., Robinson, D.L., Bergstrom, B.P., Rebec, G.V. and Wightman, R.M. (2003) A role for presynaptic mechanisms in the actions of nomifensine and haloperidol. Neuroscience, 118: 819–829. Garris, P.A., Ciolkowski, E.L., Pastore, P. and Wightman, R.M. (1994) Efﬂux of dopamine from the synaptic cleft in the nucleus accumbens of the rat brain. J. Neurosci., 14: 6084–6093. George, S.R., Watanabe, M., Di Paolo, T., Falardeau, P., Labrie, F. and Seeman, P. (1985) The functional state of the dopamine receptor in the anterior pituitary is in the high afﬁnity form. Endocrinol., 117: 690–697. Georgieva, J., Luthman, J., Mohringe, B. and Magnusson, O. (1993) Tissue and microdialysis changes after repeated and permanent probe implantation in the striatum of freely moving rats. Brain Res. Bull., 31: 463–470. Gonon, F.G. and Buda, M.J. (1985) Regulation of dopamine release by impulse ﬂow and by autoreceptors as studied by in vivo voltammetry in the rat striatum. Neurosci., 14: 765–774. Grabb, M.C., Sciotti, V.M., Gidday, J.M., Cohen, S.A. and Van Wylen, D.G.L. (1998) Neurochemical and morphological responses to acutely and chronically implanted brain microdialysis probes. J. Neurosci. Methods, 82: 25–34. Groothuis, D.R., Ward, S., Schlageter, K.E., Itskovich, A.C., Schwerin, S.C., Allen, C.V., Dills, C. and Levy, R.M. (1998) Changes in blood-brain barrier permeability associated with insertion of brain cannulas and microdialysis probes. Brain Res., 803: 218–230.

69 Hasbani, M.J., Underhil, S.M., De Eruasquin, G. and Goldberg, M.P. (2000) Synaptic loss and regeneration: a mechanism for functional decline and recovery after cerebral ischemia? Neuroscientist, 6: 110–119. Herrera-Marschitz, M., You, Z.B., Goiny, M., Meana, J.J., Silveira, R., Godukhin, O.V., Chen, Y., Espinoza, S., Pettersson, E., Loidl, C.F., Lubec, G., Andersson, K., Nylander, I., Terenius, L. and Ungerstedt, U. (1996) On the origin of extracellular glutamate levels monitored in the basal ganglia of the rat by in vivo microdialysis. J. Neurochem., 66: 1726–1735. Jacobson, I., Sandberg, M. and Hamberger, A. (1985) Mass transfer in brain dialysis devices – a new method for the estimation of extracellular amino acids concentration. J. Neurosci. Methods, 15: 263–268. Jones, S.R., Gainetdinov, R.R., Jaber, M., Giros, B., Wightman, R.M. and Carton, M.G. (1998) Profound neuronal plasticity in response to inactivation of the dopamine transporter. Proc. Natl. Acad. Sci. U.S.A., 95: 4029–4034. Justice, J.B. Jr. (1993) Quantitative microdialysis of neurotransmitters. J. Neurosci. Methods, 48: 263–276. Kawagoe, K.T., Garris, P.A., Wiedemann, D.J. and Wightman, R.M. (1992) Regulation of transient dopamine concentration gradients in the microenvironment surrounding nerve terminals in the rat striatum. Neuroscience, 51: 55–64. Kehr, J. (1993) A survey on quantitative microdialysis: theoretical models and practical implications. J. Neurosci. Methods, 48: 251–261. Kennedy, R.T., Thompson, J.E. and Vickroy, T.W. (2002) In vivo monitoring of amino acids by direct sampling of brain extracellular ﬂuid at ultralow ﬂow rates and capillary electrophoresis. J. Neurosci. Methods, 114: 39–49. Khan, A.S. and Michael, A.C. (2003) Invasive consequences of using micro-electrodes and microdialysis probes in the brain. Trends Anal. Chem., 22: 503–508. Kuhr, W.G., Wightman, R.M. and Rebec, G.V. (1987) Dopaminergic neurons: simultaneous measurements of dopamine release and single-unit activity during stimulation of the medial forebrain bundle. Brain Res., 418: 122–128. Kulagina, N.V., Shankar, L. and Michael, A.C. (1999) Monitoring glutamate and ascorbate in the extracellular space of brain tissue with electrochemical microsensors. Anal. Chem., 71: 5093–5100. Kulagina, N.V., Zigmond, M.J. and Michael, A.C. (2001) Glutamate regulates the spontaneous and evoked release of dopamine in the rat striatum. Neuroscience, 102: 121–128. Lonnroth, P., Jansson, P.A. and Smith, U. (1987) A microdialysis method allowing characterization of intercellular water space in humans. Am. J. Physiol., 253: E228–E231. Lu, Y., Peters, J.L. and Michael, A.C. (1998) Direct comparison of the response of voltammetry and microdialysis to electrically evoked release of striatal dopamine. J. Neurochem., 70: 584–593. MacLean, D.A., Bangsbo, J. and Saltin, B. (1999) Muscle interstitial glucose and lactate levels during dynamic exercise in humans determined by microdialysis. J. Appl. Physiol., 87: 1483–1490.

Marburger, A., Sohr, R., Reum, T. and Morgenstern, R. (2000) Comparison by microdialysis of striatal L-DOPA after its systemic administration in rats with probes implanted acutely or through a guide cannula. J. Neurosci. Methods, 102: 127–132. Martin-Fardon, R., Sandillon, F., Thibault, J., Privat, A. and Vigon, J. (1997) Long-term monitoring of extracellular dopamine concentration in the rat striatum by a repeated microdialysis procedure. J. Neurosci. Methods, 72: 123–135. McLaughlin, B.A., Nelson, D., Erecinska, M. and Chesselet, M.F. (1998) Toxicity of dopamine to striatal neurons in vitro and potentiation of cell death by a mitochondrial inhibitor. J. Neurochem., 70: 2406–2415. Michael, A.C., Borland, L.M., Mitala, J.J., Willoughby, B.M. and Motzko, C.M. (2005) Theory for the impact of basal turnover on dopamine clearance kinetics in the rat striatum after medial forebrain bundle stimulation and pressure ejection. J. Neurochem., 94: 1202–1211. Miele, M., Berners, M., Boutelle, M.G., Kusakabe, M. and Fillenz, M. (1996) The determination of the extracellular concentration of brain glutamate using quantitative microdialysis. Brain Res., 707: 131–133. Mireylees, S.E., Brammer, N.T. and Buckley, G.A. (1986) A kinetic study of the in vitro uptake of [3H]dopamine over a wide range of concentrations by rat striatal preparations. Biochem. Pharmacol., 35: 4065–4071. Morel, P., Fauconneau, B., Page, G., Mirbeau, T. and Huguet, F. (1998) Inhibitory effects of ascorbic acid on dopamine uptake by rat striatal synaptosomes: relationship to lipid peroxidation and oxidation of protein sulfhydryl groups. Neurosci. Res., 32: 171–179. Morgan, M.E., Singhal, D. and Anderson, B.D. (1996) Quantitative assessment of blood-brain barrier damage during microdialysis. J. Pharmacol. Exp. Ther., 277: 1167–1176. Nakachi, N., Kiuchi, Y., Inagaki, M., Inazu, M., Yamazaki, Y. and Oguchi, K. (1995) Effects of various dopamine uptake inhibitors on striatal extracellular dopamine levels and behaviours in rats. Eur. J. Pharmacol., 281: 195–203. Neve, K.A. and Neve, R.L. (1997) Molecular biology of dopamine receptors. In: Neve, K.A. and Neve, R.L. (Eds.), The Dopamine Receptors. Humana, Totowa, NJ, pp. 27–76. Nicholson, C. and Sykova, E. (1998) Extracellular space structure revealed by diffusion analysis. Trends Neurosci., 21: 207–215. Parsons, L.H. and Justice, J.B. Jr. (1992) Extracellular concentration and in vivo recovery of dopamine in the nucleus accumbens using microdialysis. J. Neurochem., 58: 212–218. Parsons, L.H. and Justice, J.B. Jr. (1994) Quantitative approaches to in vivo brain microdialysis. Crit. Rev. Neurobiol., 8: 189–220. Parsons, L.H., Smith, A.D. and Justice, J.B. Jr. (1991) The in vivo microdialysis recovery of dopamine is altered independently of basal level by 6-hydroxydopamine lesions to the nucleus accumbens. J. Neurosci. Methods, 40: 139–147. Peters, J.L. and Michael, A.C. (1998) Modeling voltammetry and microdialysis of striatal extracellular dopamine: the

70 impact of dopamine uptake on extraction and recovery ratios. J. Neurochem., 70: 594–603. Peters, J.L., Yang, H. and Michael, A.C. (2000) Quantitative aspects of brain microdialysis. Anal. Chim. Acta, 412: 1–12. Plock, N. and Kloft, C. (2005) Microdialysis – theoretical background and recent implementation in applied life sciences. Eur. J. Pharm. Sci., 25: 1–24. Rice, M.E. and Nicholson, C., 1995. Diffusion and ion shifts in the brain extracellular microenvironment and their relevance for voltammetric measurements. In: Boulton, A.A., Baker, G.B., Adams, R.N., (Eds.), Neuromethods. Vol. 27. Voltammetric Methods in Brain Systems. Human, NJ, pp. 27–79. Richﬁeld, E.K., Penney, J.B. and Young, A.B. (1989) Anatomical and afﬁnity state comparisons between dopamine D1 and D2 receptors in the rat central nervous system. Neurosci., 30: 767–777. Robichaud, M. and Debonnel, G. (2005) Oestrogen and testosterone modulate the ﬁring activity of dorsal raphe nucleus serotonergic neurons in both male and female rats. J. Neuroendocrinol., 17: 179–185. Robinson, T.E. and Camp, D.M. (1991) The feasibility of repeated microdialysis for within-subjects design experiments: studies on the mesostriatal dopamine system. In: Robinson, T.E. and Justice, J.B. (Eds.) Microdialysis in the Neuroscience, Vol. 7. Elsevier, Amsterdam, pp. 189–234. Sam, P.M. and Justice, J.B. Jr. (1996) Effect of general microdialysis-induced depletion on extracellular dopamine. Anal. Chem., 68: 724–728. Seeman, P. (2002) Atypical antipsychotics: mechanism of action. Can. J. Psychiatry, 47: 27–38. Seeman, P. and Grigoriadis, D. (1987) Dopamine receptors in brain and periphery. Neurochem. Int., 10: 1–25. Seeman, P. and Tallerico, T. (2003) Link between dopamine D1 and D2 receptors in rat and human striatal tissues. Synapse, 47: 250–254. Shuaib, A., Xu, K., Crain, B., Siren, A.-L., Feuerstein, G., Hallenbeck, J. and Davis, J.N. (1990) Assessment of damage from implantation of microdialysis probes in the rat hippocampus with silver degeneration staining. Neurosci. Lett., 112: 149–154. Smith, A.D. and Justice, J.B. Jr. (1994) The effect of inhibition of synthesis, release, metabolism and uptake on the microdialysis extraction fraction of dopamine. J. Neurochem., 54: 75–82. Stanford, J.A., Currier, T.D. and Gerhardt, G.A. (2002) Acute locomotor effects of ﬂuoxetine, sertraline, and nomifensine in young versus aged Fischer 344 rats. Pharmacol. Biochem. Behav., 71: 325–332. Steinpreis, R.E. and Salamone, J.D. (1993) The role of nucleus accumbens dopamine in the neurochemical and behavioral effects of phencyclidine: a microdialysis and behavioral study. Brain Res., 612: 263–270. Stenken, J.A. (1999) Methods and issues in microdialysis calibration. Anal. Chim. Acta, 379: 337–358. Suaud-Chagny, M.F., Chergui, K., Chouvet, G. and Gonon, F. (1992) Relationship between dopamine release in the rat

nucleus accumbens and the discharge activity of dopaminergic neurons during local in vivo application of amino acids in the ventral tegmental area. Neuroscience, 49: 63–72. Tao, R., Ma, Z. and Auerbach, S.B. (2000) Differential effect of local infusion of serotonin reuptake inhibitors in the raphe versus forebrain and the role of depolarization-induced release in increased extracellular serotonin. J. Pharmacol. Exp. Ther., 294: 571–579. Timmerman, W. and Westerink, B.H.C. (1997) Brain microdialysis of GABA and glutamate: what does it signify? Synapse, 27: 242–261. Venton, B.J., Zhang, H., Garris, P.A., Phillips, P.E.M., Sulzer, D. and Wightman, R.M. (2003) Real-time decoding of dopamine concentration changes in the caudate-putamen during tonic and phasic ﬁring. J. Neurochem., 87: 1284–1295. West, A.R., Moore, H. and Grace, A.A. (2002) Direct examination of local regulation of membrane activity in striatal and prefrontal neurons in vivo using simultaneous intracellular recording and reverse microdialysis. J. Pharmacol. Exp. Ther., 301: 867–877. Westerink, B.H.C. and Timmerman, W. (1999) Do neurotransmitters sampled by brain microdialysis reﬂect functional release? Anal. Chim. Acta, 379: 263–274. Wightman, R.M. and Zimmerman, J.B. (1990) Control of dopamine extracellular concentration in rat striatum by impulse ﬂow and uptake. Brain Res. Rev., 15: 135–144. Witkovsky, P., Nicholson, C., Rice, M.E., Bohmaker, K. and Meller, E. (1993) Extracellular dopamine concentration in the retina of the clawed frog Xenopus laevis. Proc. Natl. Acad. Sci. U.S.A., 90: 5667–5671. Wu, Q., Reith, M.E.A., Wightman, R.M., Kawagoe, K.T. and Garris, P.A. (2001) Determination of release and uptake parameters from electrically evoked dopamine dynamics measured by real-time voltammetry. J. Neurosci. Methods, 112: 119–133. Yang, H., Peters, J.L., Allen, C., Chern, S.-S., Coalson, R.D. and Michael, A.C. (2000) A theoretical description of microdialysis with mass transport coupled to chemical events. Anal. Chem., 72: 2042–2049. Yang, H., Peters, J.L. and Michael, A.C. (1998) Coupled effects of mass transfer and uptake kinetics on in vivo microdialysis of dopamine. J. Neurochem., 71: 684–692. Yim, H.J. and Gonzales, R.A. (2000) Ethanol-induced increases in dopamine extracellular concentration in rat nucleus accumbens are accounted for by increased release and not uptake inhibition. Alcohol, 22: 107–115. Yu, X., Sun, L., Luo, X., Xu, Z. and An, L. (2003) Investigation of the death model induced by glutamate treatment in serum-, antioxidant-free primary cultured cortical neurons. Dev. Brain Res., 145: 263–268. Zapata, A. and Shippenberg, T.S. (2002) D3 receptor ligands modulate extracellular dopamine clearance in the nucleus accumbens. J. Neurochem., 81: 1035–1042.

CHAPTER 1.5

Microdialysis in the brain of anesthetized vs. freely moving animals M.A. de Souza Silva, C.P. Mu¨ller and J.P. Huston Institute of Physiological Psychology, Heinrich-Heine University of Du¨sseldorf, Du¨sseldorf, Germany

Abstract: Microdialysis in the brain is performed in anesthetized or freely moving animals of various species. Both methods have provided valuable results, although they address different aspects of the brain’s functions. They each have advantages and disadvantages that need to be carefully considered when planning experiments or interpreting the data. The freely moving approach offers the opportunity to monitor behavior concurrently with neurochemical changes and is the preferred method, although for certain purposes it may be desirable to use both approaches. Advantages of the anesthetized preparation are that it is easier to use and that it may be of interest to eliminate sensory or behavioral inﬂuences on the neurotransmitter systems of interest. Here we provide a detailed overview of the advantages and disadvantages of each method, discuss how anesthetics affect neuronal function, and how these interactions can alter general brain functions. Side effects of the anesthetics and their consequences for experimentation are summarized. Results show that anesthetics can have profound effects on the basal neurotransmitter levels in the brain, which depend on the speciﬁc neurotransmitter and brain area dialyzed. Anesthetics and/or sensory-motor interactions may also interact with the neurochemical response to a pharmacological challenge, and, thereby inﬂuence dialysis results. We outline for the major neurotransmitter systems the kinds of differences that have been observed between the anesthetized and freely moving preparation and how they might be interpreted.

the profound effects anesthetics can have on the brain’s neurochemical functions. In the freely moving ‘‘conscious’’ animal central neurochemical activity can be monitored in close relation to sensory, motor, or cognitive processes. In the anesthetized ‘‘unconscious’’ animal, which is devoid of motor output and most sensory input, the brain is not only a rather isolated system, but is under the inﬂuence of the anesthetic. Investigating neurochemical changes in such a reduced system, of course, creates problems when making inferences to a functional, sensory responsive and behavior generating system, not being inﬂuenced by an anesthetic drug. In this chapter we discuss the advantages and problems associated with each method, with

I. Introduction Microdialysis (MD) in the neurosciences is performed in either anesthetized or freely moving animals. Each approach has advantages and drawbacks that need to be carefully considered when planning experiments or interpreting the data. This is of particular importance, since, for a given treatment, the anesthetized preparation may yield very different results compared with the freely moving animal. Contradictory ﬁndings in anesthetized and freely moving animals are not surprising, given Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

71

DOI: 10.1016/S1569-7339(06)16005-1 Copyright 2007 Elsevier B.V. All rights reserved

72

particular emphasis on the interaction of anesthetic agents with the brain’s neurochemical systems and interaction with pharmacological challenges, comparing ﬁndings obtained in both preparations.

II. Advantages and drawbacks of microdialysis in anesthetized and freely moving animals The freely moving preparation allows monitoring the neurochemical activity of the brain in an organism that interacts with its environment. Neurochemical measures can then be related to sensory stimulation, to spontaneous or induced behavioral activity, or to the effects of a drug treatment. However, the activity of several neurotransmitter systems of the brain is closely linked to sensory stimulation and/or the occurrence of behaviors. Sensory-motor stimulation and novelty are found, for example, to increase extracellular acetylcholine (ACh) levels (Thiel et al., 1998, 2000; Fournier et al., 2004). ACh, but not dopamine (DA) or serotonin (5-HT) is closely related to spontaneous behavioral activity (Westerink et al., 1990; Mizuno et al., 1991). In the case of sensory/motor-sensitive systems, arousal and attention or behavioral activity may ‘‘confound’’ the effects of a drug treatment. In most cases the ‘‘causal chain’’ is not clear, that is, whether the drug tested inﬂuences behavior by modulating the neurotransmitter activity, or the altered behavior or the feedback therefrom induces the change in neurotransmitter activity. Furthermore, several neurotransmitter systems that do not appear to be related to spontaneous behavioral changes are, nonetheless, involved in the control of drug-induced behaviors (Westerink et al., 1990; Waterhouse et al., 2004). When a drug is applied, which systematically inﬂuences arousal, attention, sensory processing, or motor behavior, the anesthetized approach may be useful to control for such potential confounds. Furthermore, pharmacological studies in freely moving animals often require a short restraint of the animals, which can be a problem when the neurochemical system under investigation is sensitive to the restraint procedure. DA and ACh were shown to increase as a consequence of restraint or injection procedure (Feenstra et al., 1995; De Souza Silva et al., 2000).

Although the magnitude of this effect may be determined in a vehicle control group, and possibly subtracted from the treatment effect, the additional use of an anesthetized approach may also be useful. General anesthesia can be of advantage since it eliminates most sensory inputs, arousal, attention, and locomotor activity, which can interact with the neurochemical systems under investigation and, thereby, create ‘‘noise’’ in the data. Under anesthesia, the brain is kept in a relatively constant physiological state during the whole measurement period, and stable baselines are more easily obtained. However, the anesthetized preparation has several disadvantages, some of which have been summarized by Boix et al. (1993) and will be outlined in more detail here: (1) anesthetics affect general brain functions; for example, they decrease metabolism, glucose and oxygen use, particularly in the gray matter of the brain (Sokoloff et al., 1977; Grome and McCulloch, 1981), and cause hypothermia (Crane et al., 1978; Lafferty et al., 1978). (2) Anesthetics can have side effects on vital functions; volatile anesthetics and barbiturates depress respiratory functions (Ever and Crowder, 2001). (3) Anesthetics interact with ligand-gated ion channels, and possibly with other receptor (R) proteins (Yamakura et al., 2001; Hemmings et al., 2005). The functions of these Rs and, thus, neuronal activity are altered by anesthesia. As a consequence, anesthetics may inﬂuence the basal release of neurotransmitters. An increased or decreased basal release may cause ceiling or ﬂoor effects after pharmacological treatments, respectively. However, although neuronal activity is generally decreased (Warenycia and McKenzie, 1988), the basic electrophysiological properties of the neurochemical systems are preserved under anesthesia (Tepper et al., 1991). (4) During anesthesia the impact of sensory stimulation on brain activity is restricted, although not completely eliminated (Kurosawa et al., 1992). It should be noted that, although the anesthetized animal may not exhibit overt behavior, manipulating the animal, for example, by moving its head, can nevertheless induce a neurochemical response in the brain (Carboni and Rossetti, 2001). (5) An anesthetic may affect the availability of drugs to the brain by interfering

73

with the brain’s blood supply or blood-brain barrier functions (Gumbleton et al., 1989). A decreased availability of the drug in the brain may be responsible for a diminished neurochemical effect.

III. The use of anesthesia in microdialysis Anesthesia is a state classically characterized by immobility, amnesia, hypnosis (sleep or unconsciousness), analgesia, and muscle relaxation (Beattie, 2001). Not all anesthetics have the same effects, but immobility, unconsciousness/hypnosis, and amnesia are produced by most of them (Evers and Crowder, 2001). Anesthesia can be achieved using a variety of drugs and ways of administration. For MD studies, liquid (i.p. injections), volatile, and gaseous (inhalation) applications are used. For inhaled anesthetics, the dose required to induce anesthesia is based on the minimum alveolar concentration (MAC), deﬁned as the concentration that produces immobility and lack of response to noxious stimulation in 50% of the subjects. The dose of i.p.-administered anesthetic required to induce a certain endpoint of anesthesia is difﬁcult to deﬁne due to lack of information on the pharmacokinetic proﬁle and difﬁculties in determining steady-state drug concentrations in the brain (Franks and Lieb, 1994). However, approximate doses can be estimated. Krasowski and Harrison (1999) provide a table with the EC50 values (concentration of the drug that produces the maximal effect in 50% of the subjects) for producing immobility in various animal species for many general anesthetics. Immobility and lack of response to noxious stimulation are very helpful parameters, since, for most general anesthetics, administration of concentrations two- to fourfold above the EC50 for inducing immobility produce deleterious side effects and are lethal (Franks and Lieb, 1994). General anesthetics decrease brain metabolism, blood ﬂow, and temperature (Lonjon et al., 2001; Kiyatkin and Brown, 2005), especially barbiturates (Crane et al., 1978; Lafferty et al., 1978). Pentobarbital reduces brain temperature to 4.0– 4.51C below baseline levels (from 36.5–37.51C to

32–331C) within 10 min of anesthesia (Kiyatkin and Brown, 2005). Similar temperature reductions were found after urethane anesthesia (Moser and Mathiesen, 1996). The decrease in brain metabolism and the derived decrease in heat production are major factors determining the drop in brain temperature. Although a decrease in body metabolism and, consequently, in temperature also occurs, this is not the major factor behind brain hypothermia (Kiyatkin and Brown, 2005). The drop in brain metabolism is caused by a diminished cerebral blood ﬂow induced by anesthetics (Lafferty et al., 1978). Body warming effectively counteracts decrease in body core temperature during anesthesia. Although it significantly attenuates temperature decrease in the brain, it does not completely compensate for it. During halothane and pentobarbital anesthesia in body-warmed cats, cortical tissue temperature is 1.0 and 1.81C lower than body core temperature, respectively, and in pentobarbital anesthetized body-warmed rats brain hypothermia is threefold less than in non-body warmed controls, but still 0.61C lower than body core temperature (Erickson and Lanier, 2003; Kiyatkin and Brown, 2005). During alpha-chloralose anesthesia in body-warmed rats, the difference between the cortex and body core temperature reached 4.31C, while with chloral hydrate anesthesia the difference was of 2.71C (Zhu et al., 2004). Such a decrease in brain temperature interferes with processes governing neuronal excitability, and even small temperature differences can modulate the release and uptake of neurotransmitters, their interaction with Rs and ion channel activity (Rosen, 1996; Xie et al., 2000; Volgushev et al., 2004). General anesthetics can also have side effects on vital functions. Inhalation anesthetics and barbiturates have effects on respiratory function. Depending on the drug, it can be a depression of alveolar ventilation or increase of respiratory frequency associated with a decrease in alveolar ventilation, resulting in elevation of arterial carbon dioxide tension (Evers and Crowder, 2001). Isoflurane, particularly, depresses the response to hypoxia and hypercapnia (Hirshman et al., 1977). While barbiturates decrease blood pressure, ketamine increases blood pressure, heart

74

rate, and cardiac output (Evers and Crowder, 2001). The most commonly used anesthetics for MD experiments are urethane, halothane, isoflurane, chloral hydrate, and pentobarbital (Table 1). Urethane (ethyl carbamate) is a water-soluble compound

that has the advantage of producing a long-lasting steady level of anesthesia with minimal effects on cardiovascular and respiratory systems and maintenance of spinal reﬂexes (Maggi and Meli, 1986). At concentrations close to the anesthetic EC50, urethane produces modest effects on inhibitory and

Table 1. Dose of anesthetics used in experiments with in vivo microdialysis Substance Injectable Chloral hydrate

Pentobarbital

Urethane

Ketamine Propofol Volatile Halothane

Isoﬂurane

Amount

Species

References

350 mg/kg i.p.; suppl.: 100 mg/kg 360–400 mg/kg i.p.; suppl.: 80–100 mg/kg 400 mg/kg i.p. plus suppl.

Rat Rat Rat

Suppl.: 60 mg/kg i.p. Suppl.: 80 mg/kg i.p. Suppl.: 80–100 mg/kg/h i.p. Suppl.: 0.1 ml/0.5 h i.v. 400–500 mg/kg i.p. plus suppl. 50 mg 25–30 mg/kg 10–40 mg/kg 25–50 mg/kg 1.2 g/kg i.p.

Rat Rat Rat Rat Rat Rat Primate Rat Rat Rat

1.25 g/kg i.p.

Rat

1.5 g/kg 40% w/v, 0.65 ml/100 g 25–100 mg/kg 25–50 mg/kg 25–50 mg/kg 100 mg/kg

Rat Chick Rat Rat Rat Rat

Kreiss and Lucki (1994) Hjorth et al. (1997) Church et al. (1987), Bradberry et al. (1993), Luoh et al. (1994), Hjorth and Auerbach (1996), Di Matteo et al. (1999) Done and Sharp (1994) Thomas et al. (1994) Millan et al. (1997) Benloucif et al. (1993) Assie and Koek (1996) Chen et al. (1991) Moghaddam et al. (1993) Kikuchi et al. (1997) Wang et al. (2000) Hong et al. (2005), Okakura-Mochizuki et al. (1996) Kart et al. (2004), Gronier et al. (2000), De Souza Silva et al. (2000, 2002) Boix et al. (1993) Gruss et al. (1999) Kikuchi et al. (1997) Sato et al. (1996) Kikuchi et al. (1998), Wang et al. (2000) Pain et al. (2002)

0.9% in O2 Induction: 2% Maintenance: 0.75–1% in O2 1.0% in O2 1.0% in O2/N2O 1:2 Induction: 2–3% Maintenance: 1% in O2/CO2 95:5 Induction: 2% Maintenance: 1% in O2 1.5% in O2 1% in O2/N2O 1:2 2% in O2/N2O 1:1 Induction: 3–4% in O2 Maintenance: 1.4–1.5% in O2 With O2 1.5–2.0% in O2/N2O 3:7 1.5–2.5% in air

Rat Rat

Wilkinson et al. (1993) Rutter et al. (1998)

Rat Rat Rat

Inada et al. (1992) Bonhomme et al. (1995) Richter et al. (1995)

Rat

Mammoto et al. (1997)

Rat Rat Rat Cat

Hurd et al. (1988) De Deurwaerdere and Spampinato (1999) Kidd et al. (1990) Vazquez and Baghdoyan (2004)

Rat Primate Rat

Hume et al. (2001) Tsukada et al. (1999) Jansson et al. (2004)

75

excitatory neurotransmitter-gated ion channels. However, urethane has marked effects on ion channels above the concentration required for surgical anesthesia, and may significantly alter several neurotransmitter systems in the central nervous system (CNS; Hara and Harris, 2002). Urethane has cytotoxic actions and should be used only in terminal studies, not when recovery is expected (Soma, 1983). Halothane (2-bromo-2-chloro1,1,1-triﬂuoroethane) is a volatile liquid at room temperature. Although it is a non-irritant, welltolerated compound, it depresses respiratory and cardiac function. Spontaneous respiration becomes rapid and shallow and arterial pressure is reduced (Evers and Crowder, 2001). Halothane is normally applied through a calibrated vaporizer, alone, or in a mixture with nitrous oxide, at concentrations of 3% for initial anesthesia and 1–1.5% for anesthesia maintenance (Table 1). Isoflurane (1-chloro-2,2,2triﬂuoroethyl diﬂuoromethyl ether) is increasingly being applied for MD studies. It is also a volatile liquid at room temperature. Unlike with halothane, cardiac function is well maintained, but it decreases arterial blood pressure and depresses respiratory functions (Evers and Crowder, 2001). Owing to the relatively rapid effects in induction and recovery from anesthesia and the fact that more than 99% of the inhaled isoflurane is excreted unchanged through the lungs (Kharasch and Thummel, 1993), it is suitable for short anesthesia for the placement of MD probe in the freely moving preparation. Pentobarbital (5-ethyl-5-(1-methylbutyl)-2,4,6(1 H,3 H,5 H)-pyrimidinetrione) is a short acting barbiturate available as either the free acid or sodium salt. However, pentobarbital causes respiratory depression, impairs myocardial function, and decreases body core temperature (Evers and Crowder, 2001). Chloral hydrate (2,2,2-trichloroethane-1,1-diol) is used as sedative, hypnotic, and, for animals, general anesthetic. At doses that produces surgical plane of anesthesia, it causes marked respiratory and cardiac depression (Soma, 1983). For MD studies level of anesthesia should be assessed by presence or absence of movement, corneal reﬂex, and withdraw reﬂexes (e.g., strong pinching stimulation of the hind paw). A bodywarming device for the control and maintenance of temperature within a small variance (between 36.5

and 37.51C) is necessary. Most of the body-warming devices commercially available control body temperature through a rectal probe. If the anesthetic of choice interferes with respiratory functions, an artiﬁcial ventilator must be employed. A catheter for supply of physiological solution (i.p. or i.v.) is also recommended, since MD experiments are long lasting and ﬂuid balance must be kept within physiological ranges.

IV. The effects of anesthetics on neurotransmitter systems in the brain Although anesthetics have been used since 1846 (Beattie, 2001), only in the last decade has there been considerable progress in understanding their mechanisms of action. For many years the idea prevailed that anesthetics do not have speciﬁc molecular targets, but that they non-specifically interact with neuronal-membrane lipids (Miller and Pang, 1976). An increasing body of evidence now supports the hypothesis that anesthetics interact primarily directly with R proteins. The ligandgated ion channels are considered to be their main molecular targets, although voltage-gated ion channels and G-protein-coupled Rs are also targeted (for reviews see Yamakura et al., 2001; Dilger, 2002; Hemmings et al., 2005). The ligand-gated ion channels include the nicotinic ACh-Rs, the 5-HT3-Rs, the ionotropic glutamate-Rs alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA)-, kainite- and n-methyl-D-aspartate (NMDA)-sensitive R subtypes and the gamma aminobutyric acid type A- (GABAA), and glycine-Rs. Except for the ionotropic glutamate-Rs, they are all part of an evolutionarily related ligand-gated ion channel superfamily of Rs (Ortells and Lunt, 1995). Nicotinic ACh-Rs are found in skeletal muscle, throughout the CNS and at autonomic ganglia (Gotti and Clementi, 2004). In the CNS, presynaptic nicotinic ACh-Rs modulate the release of GABA, glutamate, monoamines such as 5-HT, DA, and noradrenaline (NA), and ACh (McGehee and Role, 1996). Postsynaptic ACh-Rs have been identiﬁed in rat hippocampal interneurons (Frazier et al., 1998). The ligand-gated cation channels of 5-HT3-Rs are widely distributed in the peripheral

76

and CNSs, including the hippocampus, brain-stem, dorsal root ganglia, and vagal axons (Tecott et al., 1993). 5-HT3-Rs of the nucleus tractus solitarius and area postrema are involved in the vomiting reﬂex (Wilde and Markham, 1996). The ones localized in nociceptive primary afferents may play a role in nociception by facilitating the response of some dorsal horn neurons to noxious stimuli (Reeves and Lummis, 2002). The ionotropic glutamate-Rs are excitatory neurotransmitter-Rs localized at most synapses in the CNS (Kew and Kemp, 2005). NMDA-Rs are proposed to be involved in learning and memory (Balderas et al., 2004), but have also been implicated in various neurodegenerative and psychiatric conditions (Kemp and McKernan, 2002). Less is known about AMPAand kainate-Rs in comparison with NMDA-Rs. AMPA-Rs seem to function as the major fast excitatory neurotransmitter-Rs at almost all the synapses in the CNS (Gomes et al., 2003). Kainate-Rs are distributed throughout the CNS and spinal cord, and are implicated in the mediation of nociception (Ruscheweyh and Sandkuhler, 2002). GABAA- and glycine-Rs are chloride-selective ion channels, and, therefore, inhibitory neurotransmitter-Rs, since in most cells opening of chloride channels leads to membrane hyperpolarization and inhibition of action potentials. Glycine-Rs in the Renshaw cells are known to mediate the inhibition of motor neurons in the spinal cord (Betz et al., 1999). Yamakura et al. (2001) and Dilger (2002) review the effects of anesthetics on ligand-gated ion channels. While GABAA- and glycine-Rs are potentiated by most anesthetics, the responses of 5-HT3-Rs are varied, and nicotinic ACh-Rs and NMDA-Rs are inhibited by many anesthetics, which do not potentiate GABAA- or glycine-Rs. Many inhaled anesthetics enhance GABAA-R function by increasing channel opening at both synaptic and extrasynaptic-Rs. Additionally, they depress excitatory synaptic transmission via post- and presynaptic action, apparently by reducing the release of glutamate. Most intravenous anesthetics increase gating of the GABAA-Rs by GABA, or by slowing their desensitization as in the case of propofol (see also Hemmings et al., 2005).

Given the complexity of the anesthetic state, it is likely that different R systems in several brain areas are involved in the diverse actions of anesthesia. Evidence on the differential effect of anesthetics on distinct aspects of anesthesia has been published in recent reviews (Campagna et al., 2003; Sonner et al., 2003; Rudolph and Antkowiak, 2004). While hippocampus, amygdala, and entorhinal/perirhinal cortices seem to be implicated in the amnestic effects, the circuitry underlying the sedative effects appears to involve the hypothalamic tuberomamillary nuclei and cortical areas. For hypnosis, subcortical structures such as the thalamus and midbrain reticular formation have been implicated (Nelson et al., 2002; Veselis et al., 2005). Immobility is likely primarily caused by an action of anesthetics on the spinal cord (Zhang et al., 2003). Anesthetics affect the release of various neurotransmitters in the brain. Although there is evidence that enhancement of inhibitory and/or inhibition of excitatory neurotransmission is the main mechanism of action of anesthetics, other neurotransmitter systems in the brain are involved in some aspects of anesthesia. Table 2 summarizes the effects of several anesthetics on different neurotransmitter systems as evaluated by means of MD.

IV.A. Acetylcholine Acetylcholinergic neurons have been implicated in several behavioral functions that are also affected by anesthetics, such as memory and learning, antinociception, locomotion, and arousal (Lee et al., 2000; Gold, 2003; Decker et al., 2004). Most anesthetics have an inhibitory effect on basal release of ACh from striatal interneurons (Bertorelli et al., 1990; Damsma and Fibiger, 1991; Shichino et al., 1997) and from basal forebrain projection neurons (Sato et al., 1996; Kikuchi et al., 1997, 1998; Jansson et al., 2004), except for ketamine and nitrous oxide, which increased basal levels of ACh in the frontal cortex and hippocampus (Sato et al., 1996; Kikuchi et al., 1997; Shichino et al., 1998). Ketamine had no effect on the activity of striatal interneurons (Sato et al., 1996). ACh levels in the

77 Table 2. Effects of anesthetics on extracellular transmitter levels in the brain determined by microdialysis (m/k significant increase/ decrease; – no significant effect vs. freely moving condition) Anesthetic

Neurotransmitter

Brain area

Effects

References

Chloral hydrate

ACh

Striatum

DA

Frontal cortex Striatum Frontal cortex

k k – k – – m m – m k k – –/m k – – – m m m m – k – k k k k – – – – – – – – k k k –/m m m m – m m m – m m k

Damsma and Fibiger (1991) Bertorelli et al. (1990) Pan and Lai (1995) Zhang et al. (1989) Pan and Lai (1995) Van Gaalen et al. (1997) Tao and Auerbach (1994) Baseski et al. (2005) Baseski et al. (2005) Taguchi et al (1991) Damsma and Fibiger (1991) Keifer et al. (1994) Erb et al. (2001) Adachi et al. (2001) Adachi et al. (2000) Adachi et al. (2005a) Fink-Jensen et al. (1994) Shiraishi et al. (1997) Spampinato et al. (1986) Miyano et al. (1993) Stahle et al. (1990) Osborne et al. (1990) Chave et al. (1996) Kalen et al. (1988a) Page and Abercrombie (1999) Kalen et al. (1988b) Mammoto et al. (1997) Shiraishi et al. (1997) Morales-Villagran and Tapia (1996) Rozza et al. (2000) Morales-Villagran and Tapia (1996) Rozza et al. (2000) Morales-Villagran and Tapia (1996) Rozza et al. (2000) Morales-Villagran and Tapia (1996) Stengard et al. (1993) Campbell et al (1993) Jansson et al. (2004) Shichino et al. (1997, 1998) Shichino et al. (1997) Adachi et al. (2005b) Opacka-Juffry et al. (1991) Anzawa et al. (2001) Kushikata et al. (2005) Kushikata et al. (2005) Kikuchi et al. (1997) Wang et al. (2000) Sato et al. (1996) Sato et al. (1996) Taguchi et al. (1991) Kubota et al. (1999a, b) Rozza et al (2000)

NA

Halothane

5-HT M-ENK L-ENK ACh

DA

NA

Hypothalamus (post.) Hippocampus

5-HT HIS GLU

Striatum Hypothalamus (ant.) Striatum

ACh

Cerebral cortex Striatum Cerebral cortex Striatum Cerebral cortex Striatum Cerebellum Striatum Frontal cortex

DA

Striatum Striatum

NA

POA/hypothalamus

ASP GLY TAU GABA Isoﬂurane

Nucleus accumbens Striatum Striatum IPN Striatum mPRF Hippocampus Striatum

Hypothalamus (post.) Ketamine

ACh

NA GLU

Hippocampus Frontal cortex Striatum IPN Frontal cortex Cerebral cortex

78 Table 2 (continued ) Anesthetic

Neurotransmitter

Brain area

Effects

References

Nitrous oxide Pentobarbital

ASP GLY ACh ACh

Cerebral cortex Cerebral cortex Frontal cortex IPN Frontal cortex Hippocampus Striatum

DA

Frontal cortex Striatum PVN/hypothalamus POA/hypothalamus Frontal cortex

k k m – k k k k – – k k – – k – – – – k k k – m – k – m

Rozza et al. (2000) Rozza et al. (2000) Shichino et al. (1998) Taguchi et al. (1991) Kikuchi et al. (1997) Sato et al. (1996) Damsma and Fibiger (1991) Sato et al. (1996) Pan and Lai (1995) Semba et al. (2005) Shimokawa et al. (1998) Mizuno et al (1994) Pan and Lai (1995) Kubota et al. (1999a) Tao and Auerbach (1994) Semba et al. (2005) Rozza et al. (2000) Rozza et al. (2000) Rozza et al. (2000) Kikuchi et al. (1998) Kikuchi et al. (1998) Wang et al. (2000) Kikuchi et al. (1998) Pain et al. (2002) Semba et al. (2005) Kubota et al. (1999b) Hirose et al. (1998) Shimokawa et al. (1998)

NA

5-HT

Propofol

GLU ASP GLY ACh

DA

Urethane

NA GABA NA

Nucleus accumbens Striatum Cerebral cortex Cerebral cortex Cerebral cortex Frontal cortex Hippocampus Striatum Nucleus accumbens Striatum Frontal cortex Nucleus accumbens PVN/hypothalamus

- - -. Brain structures: IPN, interpeduncular nucleus; mPRF, medial pontine reticular formation; POA, preoptic area; PVN, periventricular nucleus. Neurotransmitters: Ach, acetylcholine; ASP, aspartate; DA, dopamine; GABA, gamma aminobutyric acid; GLU, glutamate; GLY, glycin; HIS, histamine; 5-HT, serotonin; L-Enk, Leu-enkephalin; M-Enk, Met-enkephaline; NA, noradrenaline; TAU, taurine.

medial pontine reticular formation (mPRF), an indicator of arousal state, were decreased by halothane anesthesia (Keifer et al., 1994). In the interpeduncular nucleus (IPN) of the midbrain, one of the few brain structures in which the rate of glucose utilization is increased in contrast to the decrease observed in almost all structures following anesthetics, extracellular ACh level is increased by ketamine and halothane anesthesia (Taguchi et al., 1991).

IV.B. Dopamine Dopaminergic neurons are also believed to participate in the effect of anesthetics. There are contradictory ﬁndings on the effects of halothane on

the levels of DA in the striatum, perhaps due to differences in time between sample collection and analysis (Adachi et al., 2000). Halothane anesthesia was proposed to increase DA release from the axon terminal and simultaneously accelerate DA reuptake via a proportional increase in the activity of DA transporters, maintaining a steady extracellular concentration of DA (Fink-Jensen et al., 1994). This suggestion is in agreement with the ﬁnding that the DA metabolites dihydroxyphenylacetic acid (DOPAC) and homovanillic acid (HVA) increased during halothane anesthesia (Stahle et al., 1990; Fink-Jensen et al., 1994), although others found no such increase (Miyano et al., 1993). However, halothane anesthesia seems to increase the extracellular DA in the striatum via activation of the nigro-striatal pathway rather

79

than via modulation of DA release or reuptake at nerve terminals (Adachi et al., 2001), and seems to inﬂuence the spontaneous and evoked DA release in different ways (Keita et al., 1999; Adachi et al., 2003). Halothane might also act by enhancing the metabolism of presynaptically stored or synthesized DA released from the axon terminal, rather than by altering the metabolism of DA that is returned from the synaptic cleft to the terminal (Adachi et al., 2005a). It has also been suggested that the increase in DA concentration is a consequence of the hypoxia induced by halothane (Miyano et al., 1993). Another volatile anesthetic, isoflurane, increased the extracellular concentration of DA and metabolites (Adachi et al., 2005b). Propofol has reinforcing effects at subanesthetic and anesthetic doses (Pain et al., 1996). At a small dose devoid of locomotor impairment, propofol exhibited anxiolytic-like effects (Pain et al., 1999). Concordantly, at subanesthetic and anesthetic doses, propofol increased levels of DA in the nucleus accumbens (Pain et al., 2002), indicating a marked inﬂuence on the functions of the mesolimbic dopaminergic system. Given the potent agonistic effects of propofol on GABAergic neurotransmission, its effects on DAergic neurons may be mediated by an action on GABAergic neurons, since GABA-releasing neurons exert an inhibitory control on dopaminergic neurons originating in the ventral tegmental area and projecting to the nucleus accumbens (Murai et al., 1994; Krasowski et al., 1998). Chloral hydrate and pentobarbital did not inﬂuence basal DA concentration in the frontal cortex (Pan and Lai, 1995). In the striatum, pentobarbital and propofol had no effect on DA concentration, but propofol increased and pentobarbital decreased DOPAC and HVA levels (Semba et al., 2005).

2003). Volatile anesthetics such as isoflurane, sevoflurane, and halothane had no effect on NA levels in the posterior hypothalamus during anesthesia, but an increase was observed during emergence from anesthesia (Chave et al., 1996; Kushikata et al., 2005). Given the role of the posterior hypothalamus in the control of cardiovascular regulation (Bealer, 1999; Kaehler et al., 1999), and temperature regulation (Monda et al., 2000), the increase in NA in the posterior hypothalamus during emergence from anesthesia may be related to sympathetic mediated control of cardiovascular functions and thermogenesis (Kushikata et al., 2005). In the anterior hypothalamus, particularly in the preoptic area (POA), which is also involved in thermoregulation (Metcalf and Myers, 1978), sevoflurane and isoflurane increased NA concentration during anesthesia (Anzawa et al., 2001; Kushikata et al., 2005). Conversely, pentobarbital decreased NA concentration in the medial POA (Mizuno et al., 1994). In the paraventricular nucleus (PVN) of the hypothalamus, pentobarbital had no effect on NA concentration, while urethane increased it (Shimokawa et al., 1998). Such ﬁndings suggest that the inﬂuence of anesthesia on NA activity in the hypothalamus is dependent on the anesthetic and is subregion speciﬁc. In the frontal cortex, ketamine anesthesia markedly increased basal NA, whereas propofol decreased it, while pentobarbital and chloral hydrate had no effects (Pan and Lai, 1995; Kubota et al., 1999a, b). However, pentobarbital and chloral hydrate decreased the in vivo relative recovery (dialysate extraction fraction) of NA in the frontal cortex, suggesting that pentobarbital and chloral hydrate selectively suppress NA release, reuptake, and metabolism (Pan and Lai, 1995).

IV.C. Noradrenaline

IV.D. Serotonin

The noradrenergic neurons in the brain are also regarded as an important target for general anesthetics (Angel, 1993). The noradrenergic activity in the hypothalamus is involved in the control of sleep, arousal, body temperature, and stress response (Feenstra, 2000; Berridge and Waterhouse,

In comparison to the other monoamines, less information is available on the effects of anesthetics on the serotonergic system. Anesthetic doses of chloral hydrate induced a transient increase, and ketamine a sustained increase, in 5-HT levels in the nucleus accumbens, while pentobarbital led to a

80

sustained decrease (Tao and Auerbach, 1994). Although the mechanisms by which chloral hydrate, ketamine, and pentobarbital inﬂuence the serotonergic neurons are presumably distinct, they seem to have in common the potentiation or inhibition of GABAergic-Rs localized on the cell bodies of serotonergic neurons (Nishikawa and Scatton, 1983). In the striatum pentobarbital had no effect on 5-HT, but decreased 5-hydroxy-3-indolacetic acid (5-HIAA) concentration, while propofol increased 5-HT and 5-HIAA (Semba et al., 2005). Such ﬁndings indicate that both propofol and pentobarbital can inﬂuence the activity of the serotonergic system in the brain in opposite directions.

concentrations, while pentobarbital had no effect (Rozza et al., 2000). Although most anesthetics act on GABAA-Rs, there is no evidence for modulation of extracellular GABA concentration by anesthetics in MD studies. Halothane had no effects on GABA concentration in the cerebellum (Stengard et al., 1993), and in the striatum only a tendency to decrease was observed (Campbell et al., 1993). In spite of the propofol-induced increase in DA in the nucleus accumbens (Pain et al. 2002), anesthetic doses of propofol had no effect on accumbal GABA concentration (Hirose et al., 1998). V. Comparison of neurotransmitter dynamics in the freely moving and anesthetized preparation

IV.E. Histamine Histamine is a neurotransmitter or a neuromodulator involved in various brain activities, such as regulation of neuroendocrine and cardiovascular functions, thermoregulation, circadian rhythms, antinociception, anesthesia and behaviors such as learning and memory (Haas and Panula, 2003; Haseno¨hrl and Huston, 2004). Although there is little information on interactions between the activity of histaminergic neurons and anesthetics, there is indication that halothane anesthesia inhibits the release of neuronal histamine in the hypothalamus (Mammoto et al., 1997). IV.F. Amino acids Volatile anesthetics affect glutamate-mediated transmission through postsynaptic action and presynaptic inhibition. Halothane anesthesia decreased extracellular glutamate concentration in the striatum (Morales-Villagran and Tapia, 1996; Shiraishi et al., 1997), but did not affect striatal levels of aspartate, glycine, and taurine (Morales-Villagran and Tapia, 1996). This effect of halothane was not observed on levels of glutamate, aspartate, and glycine in the fronto-parietal cortex (Rozza et al., 2000). However, in the fronto-parietal cortex, ketamine, an anesthetic that effectively inhibits the function of NMDA-Rs (Yamakura et al., 1993, 2000), decreased glutamate, aspartate, and glycine

When comparing the effects of pharmacological treatments on the activity of a particular transmitter between freely moving and anesthetized animals, in most cases similar directions of effects have been observed. However, several studies have shown either differences in magnitude of the response, complete absence of response, or opposite directions of effects. All ﬁndings reported here are based on MD studies in rats, unless indicated otherwise. V.A. Acetylcholine The activity of the cholinergic system highly depends on sensory inputs and parallels spontaneous changes in motor activity (Westerink et al., 1990; Mizuno et al., 1991). Accordingly, pharmacological treatments which interact with stimulus processing, or which have gross effects on behavioral activity, are likely to yield different neurochemical responses in the anesthetized versus the freely moving animal. De Souza Silva et al. (2000) compared the effects of the neurokinin substance P, administered into the nucleus basalis magnocellularis region, on ACh levels in the frontal cortex of urethane-anesthetized and freely moving animals. In freely moving animals vehicle administration increased frontal cortex ACh and general behavioral activity. The latter may have been caused by the restraint procedure

81

or by the vehicle. Application of substance P, however, reduced both the increases in frontal cortex ACh and behavioral activity caused by vehicle administration. In urethane-anesthetized animals, vehicle administration no longer affected ACh levels, while substance P increased frontal cortex ACh levels. Given the interaction of the pharmacological effects with the injection procedure and the resulting hyperactivity, the anesthetized preparation appeared to be the more appropriate approach for further experimentation (De Souza Silva et al., 2000). A similar dissociation in the ACh response was observed in the striatum after local application of the neuropeptide galanin. Whereas galanin decreased ACh levels in freely moving animals, it increased ACh levels under enflurane anesthesia. The s.c. application of the muscarinergic antagonist, scopolamine, led to a comparable increase in ACh levels in the striatum of both preparations (Antoniou et al., 1997). A study by Dringenberg et al. (1998) reported comparable effects of the histamine1-R antagonist, chlorpheniramine (i.p.), on ACh levels in the frontal cortex and hippocampus of freely moving and urethane-anesthetized animals. Halothane anesthesia also did not inﬂuence the ACh response to the GABAA-R antagonist, bicuculline, in the pontine reticular formation of cats (Vazques and Baghdoyan, 2004). Thus, pharmacological treatment effects on ACh levels in the brain can be comparable or different in anesthetized vs. freely moving animals. For ACh, the high sensitivity to the handling, restraint, and injection procedure in the freely moving preparation may be a major source of this difference.

V.B. Dopamine Although the dopaminergic system can be activated by handling, novelty, and stressors (e.g., Feenstra et al., 1995; Brudzynski and Gibson, 1997), it is generally assumed to be less sensitive to spontaneously occurring behavioral activity than, for example, the cholinergic system. However, significant stressors or aversive stimulation may also lead to increased DA activity (e.g., Kawahara et al., 1999).

Stahle et al. (1990) compared the effects of apomorphine and a-methyl-p-tyrosin (a-MPT), which reduce dopaminergic transmission, on DA activity in the striatum of freely moving and halothaneanesthetized rats. Under halothane-anesthesia the a-MPT-induced decrease in DA was potentiated (Stahle et al., 1990), while no effect (Spampinato et al., 1986) or a delayed decrease in DA was found after apomorphine (Stahle et al., 1990). These ﬁndings are in line with recordings from substantia nigra and other mesolimbic DA neurons, which showed that the suppressive effects of the DA agonists, quinpirole, and apomorphine, on neuronal ﬁring frequency were enhanced by chloral hydrate, urethane, and pentobarbital (Kelland et al., 1989, 1990). Interestingly, halothane can potentiate not only a drug-induced drop in DA levels, but also a drug-induced increase. The increase of DA in the striatum induced by the DA-reuptake blockers, vanoxerine and nomifensine, was potentiated in halothane- or isoflurane-anesthetized rats compared with the freely moving animals (Opacka-Juffry et al., 1991; Fink-Jensen et al., 1994; Adachi et al., 2001). The release of DA by methamphetamine was also potentiated by halothane anesthesia (Adachi et al., 2001). In addition to the systemic treatment effects, the inﬂuence of local drug application on DA levels may be potentiated by halothane. Perfusion of NMDA into the striatum locally increased DA levels in freely moving animals, which was abolished by atropine and potentiated by bicuculline (Whitehead et al., 2001). Under halothane anesthesia, the NMDA effects were potentiated compared with freely moving animals, although basal DA levels were comparable. Atropine abolished the NMDA-induced increase of DA in the striatum, while bicuculline lost its potentiating effect under halothane anesthesia (Whitehead et al., 2004). These studies suggest that interactions between different transmitter systems can also be altered by anesthetics. In rhesus monkeys, Tsukada et al. (1999) showed that isoflurane anesthesia enhanced the increase in DA levels in the striatum induced by cocaine or by the selective DA-reuptake blocker, GBR 12909. However, anesthetics do not always modify treatment effects on DA levels. The effects of the DA releaser, d-amphetamine, were not affected in

82

halothane-, isoflurane-, or barbiturate-anesthetized rats compared with freely moving animals (Glick et al., 1988; Fink-Jensen et al. 1994), which is in line with the effects of d-amphetamine on DA cell activity, which was not altered by halothane or chloral hydrate anesthesia (Bunney et al., 1973). Also, the effect of cocaine on DA levels in the striatum was not altered by halothane anesthesia in spontaneously hypertensive and Wistar-Kyoto rats (Inada et al., 1992). Furthermore, the local application of a pharmacological treatment may yield comparable effects on extracellular DA levels in freely moving and anesthetized animals: NMDA caused a comparable increase in striatal DA levels in freely moving (Marek et al., 1992) and halothane-anesthetized animals (Morari et al., 1993). Some studies have reported a disappearance or reversal of treatment effects under anesthesia. While the systemic administration of substance P caused a sustained increase in DA levels in the neostriatum and nucleus accumbens of freely moving animals (Boix et al., 1992), it led to a long-lasting decrease in the neostriatum, and had no effect in the nucleus accumbens in urethane-anesthetized animals (Boix et al., 1993). Also, the haloperidolinduced increase of DA in the striatum of conscious animals was no longer observed under halothane- (Spampinato et al., 1986) or chloral hydrate anesthesia (Zhang et al., 1989). Interestingly, halothane increased, while chloral hydrate decreased basal DA levels in the striatum (Spampinato et al., 1986; Zhang et al., 1989), which may suggest different mechanisms by which the two anesthetics reduce the treatment effect. An alternative explanation for the disappearance of a drug effect in the anesthetized preparation can be a shift in the dose response curve to the right, either by the anesthetic or the elimination of the sensory-motor interaction. For example, Spanagel et al. (1991) compared the effects of b-endorphin (i.c.v.) on DA in the nucleus accumbens between freely moving and halothaneanesthetized animals and found that b-endorphin increased DA in both preparations, paralleled by an increase in locomotor activity in the freely moving animals. The dose-response function for the DA level was shifted to the right under halothane anesthesia (Spanagel et al., 1991).

Another study compared the effects of chloral hydrate and pentobarbital on the cocaine-induced increase in DA in the medial prefrontal cortex. While chloral hydrate anesthesia had no inﬂuence, pentobarbital led to a diminished response compared with freely moving animals (Pan et al., 1995). Interestingly, a study by Moghaddam and Bunney (1993) reported that chloral hydrate did not affect the DA response in the striatum after chronic treatment with haloperidol and an acute challenge. Although it was suggested that a diminished or absent DA response to stimulating drugs under anesthesia may be due to the elimination of the (mostly drug-increased) motor output under anesthesia (Spanagel et al., 1991; Boix et al., 1993), the effects may depend on the anesthetic used. Studies that apply the same drug treatment in freely moving and anesthetized animals can also yield comparable results. Gronier et al. (2000) reported a strong behavioral activation after intra-ventral tegmental area application of the muscarinergic agonist, oxotremorine. The oxotremorine-induced increase in DA in the ventral tegmental area and frontal cortex was comparable in urethane-anesthetized and freely moving animals. This may suggest that the drug-induced changes of DA activity in some parts of the DA system are more susceptible to anesthesia and/or behavioral changes than others. In conclusion, the DA response to a pharmacological treatment can be potentiated, attenuated, or unchanged by anesthesia. Especially treatments that cause a strong behavioral activation, and, thus, a strong feedback to the DA system, seem to yield stronger effects in freely moving animals.

V.C. Noradrenaline Noradrenergic activity appears to be less sensitive to spontaneous behavioral activity. However, handling and other stressors, or sensory stimulation may also increase its activity (Kalen et al., 1988a; Kawahara et al., 1999). While chloral hydrate did not alter the cocaineinduced NA increase in the medial prefrontal cortex, pentobarbital led to a diminished response

83

compared with freely moving animals (Pan et al., 1995). Chloral hydrate also did not inﬂuence the increase in NA levels in the hippocampus after s.c. administration of nicotine (Mitchell, 1993). The local application of corticotrophin-releasing factor (CRF) into the locus coeruleus increased NA in the hippocampus of freely moving animals. This effect was not inﬂuenced by halothane anesthesia (Page and Abercrombie, 1999). Also, for NA, some studies suggest that neurochemical responses to pharmacological treatments may interact with the behavioral effects of the treatment. The local application of the a2-adrenoceptor agonist, clonidine, and the cholinergic R agonist, carbachol, into the locus coeruleus decreased and increased NA levels in the prefrontal cortex, respectively, in both freely moving and chloral hydrate-anesthetized rats. Neither treatment inﬂuenced behavior (Van Gaalen et al., 1997). Conversely, the local application of NMDA and kainate caused behavioral activation (chewing, grooming, and locomotion) in freely moving animals. The concomitant NMDA- and kainateinduced increase in NA in the frontal cortex was larger in freely moving than in chloral hydrateanesthetized animals (Van Gaalen et al., 1997). In freely moving animals the systemic application of either the 5-HT1A-R agonists, 8-OH-DPAT, MDL 73005EF, and buspirone, or the 5-HT2-R antagonist, ritanserin, increased NA in the ventral hippocampus. Under chloral hydrate, the buspirone-induced increase was attenuated, while 8-OHDPAT, MDL 73005EF, and ritanserin were no longer effective (Done and Sharp, 1994). In conclusion, the majority of the comparative studies that measured NA levels in anesthetized and freely moving animals showed that under anesthesia an attenuated treatment response may occur.

V.D. Serotonin In the brain 5-HT levels may parallel drug- and novelty-induced increases of locomotor activity (Bickerdike et al., 1993; Mu¨ller et al., 2002). The muscarinergic agonist, oxotremorine, increased 5-HT in the ventral tegmental area of freely moving animals, parallel to a profound behavioral

activation, and a comparable effect on 5-HT activity in this area was found under urethaneanesthesia (Gronier et al., 2000). However, pharmacological treatment effects on 5-HT levels were found to be modulated by other general anesthetics. The increase in 5-HT in freely moving animals, induced by either s.c. or intra dorsal raphe nucleus morphine application, was absent under chloral hydrate anesthesia, and pentobarbital anesthesia eliminated the increase in 5-HT induced by s.c. application of morphine (Tao and Auerbach, 1994). Thus, neurochemical treatment effects can be inﬂuenced by anesthetics in the 5-HT system. As is the case with DA, the behavioral effects of a pharmacological treatment may be an important source for the inﬂuence of anesthesia on 5-HT response. V.E. Histamine The responsiveness of the histaminergic system of the brain can also differ between freely moving and anesthetized animals. The histamine3-R antagonist, thioperamide (i.v.), increased the extracellular histamine in the anterior hypothalamus of freely moving animals, but not under halothaneanesthesia (Mammoto et al., 1997). However, histaminergic responsiveness is not always altered by anesthesia: The effects of the adenosine2A-R agonist, CGS21680, applied into the subarachnoid space on histamine levels in the frontal cortex were investigated in freely moving and urethane-anesthetized rats. Although the agonist induced sleep in freely moving animals, the attenuating effects on frontal cortex histamine was not affected by anesthesia (Hong et al., 2005). The few studies on histaminergic responses show that in some cases differences between the anesthetized and the freely moving animal may be expected. V.F. Amino acids Gruss et al. (1999) compared the effects of NMDA applied into the forebrain of chicks on the local extracellular levels of glutamate and taurine between freely moving and urethane-anesthetized

84

animals. While NMDA administration did not inﬂuence glutamate, it increased taurine levels. This effect was abolished under urethane anesthesia. In freely moving rats Young and Bradford (1991) found an increase in extracellular glutamate in the striatum after intra-striatal application of NMDA. This effect was absent under halothane anesthesia (Morari et al., 1993). Morales-Villagran and Tapia (1996) investigated the effect of intra-striatal administration of the potassium channel blocker, 4-aminopyridine (4-AP), on the concentration of amino acids in the striatum. 4-AP led to hyperactivity, rearing and jumping, and an increase in levels of glutamate, aspartate, taurine, and glycine. Halothane anesthesia prevented the strong impact of 4-AP on behavior, but not its neurochemical effects, with the exception of the taurine response, which was potentiated (Morales-Villagran and Tapia, 1996). Accordingly, also the amino acid neurotransmitter response to a pharmacological treatment can differ between anesthetized and freely moving animals. However, the inﬂuence of a potential locomotor response feedback may be weaker than for DA or 5-HT.

V.G. Interpreting outcome differences In general, interpreting these data requires some caution, since there are at least two major factors that cannot be easily segregated. On the one hand the cause can be a different neurotransmitter dynamic in the anesthetized preparation. This may especially account for those neurotransmitter systems and brain regions, which show an effect of the anesthetic on basal release. A shift in the basal neurotransmitter release in the one or other direction can alter a pharmacological treatment effect and lead to ceiling or ﬂoor effects, and, thereby, mask the ‘‘true’’ magnitude of the treatment effect. It may even result in a zero effect when the treatment is a competitive ligand to one of the anesthetic binding sites in the brain. An alternative explanation for a difference between the freely moving and anesthetized approach can be the impact of the missing sensory input and motor output in the anesthetized animals. In some cases it was the declared goal of the study to

determine the extent of this difference (e.g., De Souza Silva et al., 2000; Gronier et al., 2000). Such information would seem to be especially useful for transmitters that are easily inﬂuenced by spontaneous and drug-induced locomotor activity. However, there is no anesthetic available at the moment that does not inﬂuence multiple brain functions. Therefore, an attribution of outcome differences between freely moving and anesthetized animals solely to reduced sensory input and/or motor output remains limited.

VI. Conclusions When planning experiments with anesthetized animals one should be aware of the diverse effects of the anesthetic on the brain’s functions. Considering that the Rs involved in anesthesia are ubiquitous in the brain, particularly the GABAA- and glutamate-Rs, it can be expected that potentially all neurotransmitter systems can be affected by anesthesia. Moreover, by affecting in distinct ways different neurotransmitter systems, anesthetics can change the inhibitory, excitatory, or modulatory control from one set of neurons over others. The anesthetized approach may be useful in particular for neurotransmitters whose activity is highly correlated with spontaneous behavioral alterations, such as ACh, or neurotransmitters that parallel drug-induced behavioral changes, for example, DA in studies involving administration of stimulant drugs. The interpretation of results from anesthetized preparations must be considered in view of possibility that the anesthetic can increase or decrease baseline levels, and affect the neurochemical responses being measured. In case of an additional pharmacological treatment, interactions between anesthesia and the pharmacological treatment must be considered. Given possible effects of anesthetics on vital functions, in addition to their direct and indirect actions on the brain’s neurotransmitter systems, it would seem that the freely moving preparation should be employed whenever possible. Conversely, in freely moving animals, the inﬂuence of behavioral changes on the effects of a pharmacological treatment, or even the inﬂuence of the behavioral changes induced by

85

the treatment procedure (e.g., i.p. injection), needs to be carefully considered. Ideally, we should have the information from both the freely moving and anesthetized preparation, with a dose-response analysis of the anesthetic.

Acknowledgments This work was supported by the grants DE-792/2 and Hu 306/ 23-3 from the Deutsche Forschungsgemeinschaft.

References Adachi, Y.U., Aramaki, Y., Satomoto, M., Higuchi, H. and Watanabe, K. (2003) Halothane attenuated haloperidol and enhanced clozapine-induced dopamine release in the rat striatum. Neurochem. Int., 43: 113–119. Adachi, Y.U., Satomoto, M., Higuchi, H., Watanabe, K., Yamada, S. and Kazama, T. (2005a) Halothane enhances dopamine metabolism at presynaptic sites in a calciumindependent manner in rat striatum. Br. J. Anaesth., 95: 485–494. Adachi, Y.U., Uchihashi, Y., Watanabe, K. and Satoh, T. (2000) Halothane anesthesia decreases the extracellular level of dopamine in rat striatum: a microdialysis study in vivo. J. Anaesth., 14: 82–90. Adachi, Y.U., Watanabe, K., Satoh, T. and Vizi, E.S. (2001) Halothane potentiates the effect of methamphetamine and nomifensine on extracellular dopamine levels in rat striatum: a microdialysis study. Br. J. Anaesth., 86: 837–845. Adachi, Y.U., Yamada, S., Satomoto, M., Higuchi, H., Watanabe, K. and Kazama, T. (2005b) Isoﬂurane anesthesia induces biphasic effect on dopamine release in the rat striatum. Brain Res. Bull., 67: 176–181. Angel, A. (1993) Central neuronal pathways and the process of anaesthesia. Br. J. Anaesth., 71: 148–163. Antoniou, K., Kehr, J., Snitt, K. and Ogren, S.O. (1997) Differential effects of the neuropeptide galanin on striatal acetylcholine release in anaesthetized and awake rats. Br. J. Pharmacol., 121: 1180–1186. Anzawa, N., Kushikata, T., Ohkawa, H., Yoshida, H., Kubota, T. and Matsuki, A. (2001) Increased noradrenaline release from rat preoptic area during and after sevoﬂurane and isoﬂurane anesthesia. Can. J. Anaesth., 48: 462–465. Assie, M.B. and Koek, W. (1996) Effects of 5-HT1A receptor antagonists on hippocampal 5-hydroxytryptamine levels: (S)-WAY100135, but not WAY100635, has partial agonist properties. Eur. J. Pharmacol., 304: 15–21. Balderas, I., Ramirez-Amaya, V. and Bermudez-Rattoni, F. (2004) Memory-linked morphological changes. Rev. Neurol., 38: 944–948.

Baseski, H.M., Watson, C.J., Cellar, N.A., Shackman, J.G. and Kennedy, R.T. (2005) Capillary liquid chromatography with MS3 for the determination of enkephalins in microdialysis samples from the striatum of anesthetized and freely-moving rats. J. Mass Spectrom., 40: 146–153. Bealer, S.L. (1999) Preoptic recess noradrenergic receptors control modiﬁcation of baroreﬂex sensitivity by hypertonicity. Am. J. Physiol., 276: R44–R51. Beattie, C. (2001) History and principles of anesthesiology. In: Hardman, J.G., Limbird, L.E. and Gilman, A.G. (Eds.), The Pharmacological Basis of Therapeutics. McGraw-Hill, New York, pp. 321–335. Benloucif, S., Keegan, M.J. and Galloway, M.P. (1993) Serotonin-facilitated dopamine release in vivo: pharmacological characterization. J. Pharmacol. Exp. Ther., 265: 373–377. Berridge, C.W. and Waterhouse, B.D. (2003) The locus coeruleusnoradrenergic system: modulation of behavioral state and statedependent cognitive processes. Brain Res. Rev., 42: 33–84. Bertorelli, R., Hallstrom, A., Hurd, Y.L., Karlsson, A., Consolo, S. and Ungerstedt, U. (1990) Anesthesia effects on in vivo acetylcholine transmission – comparisons of radioenzymatic and HPLC assays. Eur. J. Pharmacol., 175: 79–83. Betz, H., Kuhse, J., Schmieden, V., Laube, B., Kirsch, J. and Harvey, R.J. (1999) Structure and functions of inhibitory and excitatory glycine receptors. Ann. N. Y. Acad. Sci., 868: 667–676. Bickerdike, M.J., Wright, I.K. and Marsden, C.A. (1993) Social-isolation attenuates rat forebrain 5-HT release induced by KCl stimulation and exposure to a novel environment. Behav. Pharmacol., 4: 231–236. Boix, F., Mattioli, R., Adams, F., Huston, J.P. and Schwarting, R.K.W. (1992) Effects of Substance-P on extracellular dopamine in neostriatum and nucleus-accumbens. Eur. J. Pharmacol., 216: 103–107. Boix, F., Mattioli, R., Adams, F., Huston, J.P. and Schwarting, R.K.W. (1993) Peripherally administered Substance-P affects extracellular dopamine concentrations in the neostriatum but not in the nucleus-accumbens under anesthesia. Brain Res. Bull., 31: 655–660. Bonhomme, N., De Deurwaerdere, P., Le Moal, M. and Spampinato, U. (1995) Evidence for 5-HT4 receptor subtype involvement in the enhancement of striatal dopamine release induced by serotonin – a microdialysis study in the halothane-anesthetized rat. Neuropharmacology, 34: 269–279. Bradberry, C.W., Nobiletti, J.B., Elsworth, J.D., Murphy, B., Jatlow, P. and Roth, R.H. (1993) Cocaine and cocaethylene – microdialysis comparison of brain drug levels and effects on dopamine and serotonin. J. Neurochem., 60: 1429–1435. Brudzynski, S.M. and Gibson, C.J. (1997) Release of dopamine in the nucleus accumbens caused by stimulation of the subiculum in freely moving rats. Brain Res. Bull., 42: 303–308. Bunney, B.S., Walters, J.R., Roth, R.H. and Aghajanian, G.K. (1973) Dopaminergic neurons – Effect of antipsychotic drugs and amphetamine on single cell activity. J. Pharmacol. Exp. Ther., 185: 560–571. Campagna, J.A., Miller, K.W. and Forman, S.A. (2003) Mechanisms of actions of inhaled anesthetics. New Engl. J. Med., 348: 2110–2124.

86 Campbell, K., Kalen, P., Lundberg, C., Wictorin, K., Rosengren, E. and Bjorklund, A. (1993) Extracellular gamma-aminobutyric-acid levels in the rat caudate-putamen – monitoring the neuronal and glial contribution by intracerebral microdialysis. Brain Res., 614: 241–250. Carboni, S. and Rossetti, Z.L. (2001) Increased extracellular glutamate in rat prefrontal cortex: movement artifact contribution. Brain Res., 890: 189–191. Chave, S., Kushikata, T., Ohkawa, H., Ishiara, H., Grimaud, D. and Matsuki, A. (1996) Effects of two volatile anesthetics (sevoﬂurane and halothane) on the hypothalamic noradrenaline release in rat brain. Brain Res., 706: 293–296. Chen, J.P., van Praag, H.M. and Gardner, E.L. (1991) Activation of 5-HT3 receptor by 1-phenylbiguanide increases dopamine release in the rat nucleus accumbens. Brain Res., 543: 354–357. Church, W.H., Justice, J.B. and Byrd, L.D. (1987) Extracellular dopamine in rat striatum following uptake inhibition by cocaine, nomifensine and benztropine. Eur. J. Pharmacol., 139: 345–348. Crane, P.D., Braun, L.D., Cornford, E.M., Cremer, J.E., Glass, J.M. and Oldendorf, W.H. (1978) Dose dependent reduction of glucose utilization by pentobarbital in rat brain. Stroke, 9: 12–18. Damsma, G. and Fibiger, H.C. (1991) The effects of anesthesia and hypothermia on interstitial concentrations of acetylcholine and choline in rat striatum. Life Sci., 48: 2469–2474. Decker, M.W., Rueter, L.E. and Bitner, R.S. (2004) Nicotinic acetylcholine receptor agonists: a potential new class of analgesics. Curr. Top. Med. Chem., 4: 369–384. De Deurwaerdere, P. and Spampinato, U. (1999) Role of serotonin(2A) and serotonin(2B/2C) receptor subtypes in the control of accumbal and striatal dopamine release elicited in vivo by dorsal raphe nucleus electrical stimulation. J. Neurochem., 73: 1033–1042. De Souza Silva, M.A., Haseno¨hrl, R.U., Tomaz, C., Schwarting, R.K.W. and Huston, J.P. (2000) Differential modulation of frontal cortex acetylcholine by injection of substance P into the nucleus basalis magnocellularis region in the freely-moving vs. the anesthetized preparation. Synapse, 38: 243–253. De Souza Silva, M.A., Jezek, K., Weth, K., Mu¨ller, H.W., Huston, J.P., Brandao, M.L. and Haseno¨hrl, R.U. (2002) Facilitation of learning and modulation of frontal cortex acetylcholine by ventral pallidal injection of heparin glucosaminoglycan. Neuroscience, 113: 529–535. Dilger, J.P. (2002) The effects of general anaesthetics on ligandgated ion channels. Br. J. Anaesth., 89: 41–51. Di Matteo, V., Di Giovanni, G., Di Mascio, M. and Esposito, E. (1999) SB 242084, a selective serotonin2C receptor antagonist, increases dopaminergic transmission in the mesolimbic system. Neuropharmacology, 38: 1195–1205. Done, C.J.G. and Sharp, T. (1994) Biochemical evidence for the regulation of central noradrenergic activity by 5-HT1A and 5-HT2 receptors – microdialysis studies in the awake and anesthetized rat. Neuropharmacology, 33: 411–421.

Dringenberg, H.C., de Souza Silva, M.A., Rossmu¨ller, J., Huston, J.P. and Schwarting, R.K.W. (1998) Histamine H-1 receptor antagonists produce increases in extracellular acetylcholine in rat frontal cortex and hippocampus. J. Neurochem., 70: 1750–1758. Erb, C., Troost, J., Kopf, S., Schmitt, U., Loffelholz, K., Soreq, H. and Klein, J. (2001) Compensatory mechanisms enhance hippocampal acetylcholine release in transgenic mice expressing human acetylcholinesterase. J. Neurochem., 77: 638–646. Erickson, K.M. and Lanier, W.L. (2003) Anesthetic technique inﬂuences brain temperature, independently of core temperature, during craniotomy in cats. Anesth. Analg., 96: 1460–1466. Evers, A.S. and Crowder, C.M. (2001) General anesthetics. In: Hardman, J.G., Limbird, L.E. and Gilman, A.G. (Eds.), The Pharmacological Basis of Therapeutics. McGraw-Hill, New York, pp. 321–335. Feenstra, M.G. (2000) Dopamine and noradrenaline release in the prefrontal cortex in relation to unconditioned and conditioned stress and reward. Prog. Brain Res., 126: 133–163. Feenstra, M.G.P., Botterblom, M.H.A. and Vanuum, J.F.M. (1995) Novelty-induced increase in dopamine release in the rat prefrontal cortex in-vivo – inhibition by diazepam. Neurosci. Lett., 189: 81–84. Fink-Jensen, A., Ingwersen, S.H., Nielsen, P.G., Hansen, L., Nielsen, E.B. and Hansen, A.J. (1994) Halothane anesthesia enhances the effect of dopamine uptake inhibition on interstitial levels of striatal dopamine. Naunyn Schmiedebergs Arch. Pharmacol., 350: 239–244. Fournier, G.N., Semba, K. and Rasmusson, D.D. (2004) Modality- and region-speciﬁc acetylcholine release in the rat neocortex. Neuroscience, 126: 257–262. Franks, N.P. and Lieb, W.R. (1994) Molecular and cellular mechanisms of general anesthesia. Nature, 367: 607–614. Frazier, C.J., Buhler, A.V., Weiner, J.L. and Dunwiddie, T.V. (1998) Synaptic potentials mediated via alpha-bungarotoxinsensitive nicotinic acetylcholine receptors in rat hippocampal interneurons. J. Neurosci., 18: 8228–8235. Glick, S.D., Carlson, J.N., Baird, J.L., Maisonneuve, I.M. and Bullock, A.E. (1988) Basal and amphetamine-induced asymmetries in striatal dopamine release and metabolism – bilateral invivo microdialysis in normal rats. Brain Res., 473: 161–164. Gold, P.E. (2003) Acetylcholine modulation of neural systems involved in learning and memory. Neurobiol. Learn. Mem., 80: 194–210. Gomes, A.R., Correia, S.S., Carvalho, A.L. and Duarte, C.B. (2003) Regulation of AMPA receptor activity, synaptic targeting and recycling: role in synaptic plasticity. Neurochem. Res., 28: 1459–1473. Gotti, C. and Clementi, F. (2004) Neuronal nicotinic receptors: from structure to pathology. Prog. Neurobiol., 74: 363–396. Grome, J.J. and McCulloch, J. (1981) The effects of chloral hydrate anesthesia on the metabolic response in the substantia nigra to apomorphine. Brain Res., 214: 223–228.

87 Gronier, B., Perry, K.W. and Rasmussen, K. (2000) Activation of the mesocorticolimbic dopaminergic system by stimulation of muscarinic cholinergic receptors in the ventral tegmental area. Psychopharmacology, 147: 347–355. Gruss, M., Bredenkotter, M. and Braun, K. (1999) N-methyl-Daspartate receptor-mediated modulation of monoaminergic metabolites and amino acids in the chick forebrain: an in vivo microdialysis and electrophysiology study. J. Neurobiol., 40: 116–135. Gumbleton, M., Taylor, G., Sewell, R.D.E. and Nicholls, P.J. (1989) Anesthetic inﬂuences on brain hemodynamics in the rat and their significance to biochemical, neuro-pharmacological and drug disposition studies. Biochem. Pharmacol., 38: 2745–2748. Haas, H. and Panula, P. (2003) The role of histamine and the tuberomamillary nucleus in the nervous system. Nat. Rev. Neurosci., 4: 121–130. Hara, K. and Harris, R.A. (2002) The anesthetic mechanism of urethane: the effects on neurotransmitter-gated ion channels. Anesth. Analg., 94: 313–318. Haseno¨hrl, R.U. and Huston, J.P. (2004) Histamine. In: Riedel, G. and Platt, B. (Eds.), Memories are Made of These: From Messengers to Molecules. Landes Bioscience, Georgetown. Hemmings, H.C., Akabas, M.H., Goldstein, P.A., Trudell, J.R., Orser, B.A. and Harrison, N.L. (2005) Emerging molecular mechanisms of general anesthetic action. Trends Pharmacol. Sci., 26: 503–510. Hirose, N., Yoshida, Y., Koide, S., Takada, K., Saigusa, T. and Koshikawa, N. (1998) Effects of propofol and fentanyl on extracellular levels of gamma-aminobutyric acid in the rat nucleus accumbens: an in vivo microdialysis study. J. Oral. Sci., 40: 165–170. Hirshman, C.A., McCullough, R.E., Cohen, P.J. and Weil, J.V. (1977) Depression of hypoxic ventilatory response by halothane, enﬂurane and isoﬂurane in dogs. Br. J. Anaesth., 49: 957–963. Hjorth, S. and Auerbach, S.B. (1996) 5-HT1A autoreceptors and the mode of action of selective serotonin reuptake inhibitors (SSRI). Behav. Brain Res., 73: 281–283. Hjorth, S., Westlin, D. and Bengtsson, H.J. (1997) WAY100635-induced augmentation of the 5-HT-elevating action of citalopram: relative importance of the dose of the 5-HT1A (auto)receptor blocker versus that of the 5-HT reuptake inhibitor. Neuropharmacology, 36: 461–465. Hong, Z.Y., Huang, Z.L., Qu, W.M., Eguchi, N., Urade, Y. and Hayaishi, O. (2005) An adenosine A(2A) receptor agonist induces sleep by increasing GABA release in the tuberomammillary nucleus to inhibit histaminergic systems in rats. J. Neurochem., 92: 1542–1549. Hume, S., Hirani, E., Opacka-Juffry, J., Myers, R., Townsend, C., Pike, V. and Grasby, P. (2001) Effect of 5-HT on binding of [(11)C] WAY 100635 to 5-HT(IA) receptors in rat brain, assessed using in vivo microdialysis nd PET after fenﬂuramine. Synapse, 41: 150–159. Hurd, Y.L., Kehr, J. and Ungerstedt, U. (1988) In vivo microdialysis as a technique to monitor drug transport – correlation of extracellular cocaine levels and dopamine overﬂow in the rat-brain. J. Neurochem., 51: 1314–1316.

Inada, T., Polk, K., Jin, C., Purser, C., Hume, A., Hoskins, B., Ho, I.K. and Rockhold, R.W. (1992) Cocaine elevates striatal dopamine efﬂux in spontaneously hypertensive and Wistar-Kyoto rats. Brain Res. Bull., 28: 227–231. Jansson, A., Olin, K., Yoshitake, T., Hagman, B., Herrington, M.K., Kehr, J. and Permert, J. (2004) Effects of isoﬂurane on prefrontal acetylcholine release and hypothalamic Fos response in young adult and aged rats. Exp. Neurol., 190: 535–543. Kaehler, S.T., Singewald, N. and Philippu, A. (1999) The release of catecholamines in hypothalamus and locus coeruleus is modulated by peripheral chemoreceptors. Naunyn Schmiedebergs Arch. Pharmacol., 360: 428–434. Kalen, P., Kokaia, M., Lindvall, O. and Bjorklund, A. (1988a) Basic characteristics of noradrenaline release in the hippocampus of intact and 6-hydroxdopamine-lesioned rats as studied by in vivo microdialysis. Brain Res., 474: 374–379. Kalen, P., Strecker, R.E., Rosengren, E. and Bjorklund, A. (1988b) Endogenous release of neuronal serotonin and 5-Hydroxyindoleacetic acid in the caudate putamen of the rat as revealed by intracerebral dialysis coupled to high-performance liquid-chromatography with ﬂuorimetric detection. J. Neurochem., 51: 1422–1435. Kart, E., Jocham, G., Mu¨ller, C.P., Schlo¨mer, C., Brandao, M.L., Huston, J.P. and de Souza Silva, M.A. (2004) Neurokinin-1 receptor antagonism by SR140333: enhanced in vivo ACh in the hippocampus and promnestic post-trial effects. Peptides, 25: 1959–1969. Kawahara, Y., Kawahara, H. and Westerink, B.H.C. (1999) Comparison of effects of hypotension and handling stress on the release of noradrenaline and dopamine in the locus coeruleus and medial prefrontal cortex of the rat. Naunyn Schmiedebergs Arch. Pharmacol., 360: 42–49. Keifer, J.C., Baghdoyan, H.A., Becker, L. and Lydic, R. (1994) Halothane decreases pontine acetylcholine-release and increases EEG spindles. Neuroreport, 5: 577–580. Keita, H., Henzel-Rouelle, D., Dupont, H., Desmonts, J.M. and Mantz, J. (1999) Halothane and isoﬂurane increase spontaneous but reduce the N-methyl-D-aspartate-evoked dopamine release in rat striatal slices – evidence for direct presynaptic effects. Anesthesiology, 91: 1788–1797. Kelland, M.D., Chiodo, L.A. and Freeman, A.S. (1990) Anesthetic inﬂuences on the basal activity and pharmacological responsiveness of nigrostriatal dopamine neurons. Synapse, 6: 207–209. Kelland, M.D., Freeman, A.S. and Chiodo, L.A. (1989) Chloral hydrate anesthesia alters the responsiveness of identiﬁed midbrain dopamine neurons to dopamine agonist administration. Synapse, 3: 30–37. Kemp, J.A. and McKernan, R.M. (2002) NMDA receptor pathways as drug targets. Nat. Neurosci., 5: 1039–1042. Kew, J.N. and Kemp, J.A. (2005) Ionotropic and metabotropic glutamate receptor structure and pharmacology. Psychopharmacology, 179: 4–29. Kharasch, E.D. and Thummel, K.E. (1993) Identiﬁcation of cytochrome P450 2E1 as the predominant enzyme catalyzing human liver microsomal deﬂuorination of sevoﬂurane, isoﬂurane, and methoxyﬂurane. Anesthesiology, 79: 795–807.

88 Kidd, E.J., Leysen, J.E. and Marsden, C.A. (1990) Chronic 5-HT2 receptor antagonist treatment alters 5-HT1A autoregulatory control of 5-HT release in rat-brain in vivo. J. Neurosci. Methods, 34: 91–98. Kikuchi, T., Wang, Y., Sato, K. and Okumura, F. (1998) In vivo effects of propofol on acetylcholine release from the frontal cortex, hippocampus and striatum studied by intracerebral microdialysis in freely moving rats. Br. J. Anaesth., 80: 644–648. Kikuchi, T., Wang, Y., Shinbori, H., Sato, K. and Okumura, F. (1997) Effects of ketamine and pentobarbitone on acetylcholine release from the rat frontal cortex in vivo. Br. J. Anaesth., 79: 128–130. Kiyatkin, E.A. and Brown, P.L. (2005) Brain and body temperature homeostasis during sodium pentobarbital anesthesia with and without body warming in rats. Physiol. Behav., 84: 563–570. Krasowski, M.D. and Harrison, N.L. (1999) General anaesthetic actions on ligand-gated ion channels. Cell Mol. Life Sci., 55: 1278–1303. Krasowski, M.D., Koltchine, V.V., Rick, C.E., Ye, Q., Finn, S.E. and Harrison, N.L. (1998) Propofol and other intravenous anesthetics have sites of action on the gamma-aminobutyric acid type a receptor distinct from that for isoﬂurane. Mol. Pharmacol., 53: 530–538. Kreiss, D.S. and Lucki, I. (1994) Differential regulation of serotonin (5-HT) release in the striatum and hippocampus by 5-HT1A autoreceptors of the dorsal and median raphe nuclei. J. Pharmacol. Exp. Ther., 269: 1268–1279. Kubota, T., Anzawa, N., Hirota, K., Yoshida, H., Kushikata, T. and Matsuki, A. (1999a) Effects of ketamine and pentobarbital on noradrenaline release from the medial prefrontal cortex in rats. Can. J. Anaesth., 46: 388–392. Kubota, T., Hirota, K., Yoshida, H., Takahashi, S., Anzawa, N., Ohkawa, H., Kushikata, T. and Matsuki, A. (1999b) Effects of sedatives on noradrenaline release from the medial prefrontal cortex in rats. Psychopharmacology, 146: 335–338. Kurosawa, M., Sato, A. and Sato, Y. (1992) Cutaneous mechanical sensory stimulation increases extracellular acetylcholine-release in cerebral-cortex in anesthetized rats. Neurochem. Int., 21: 423–427. Kushikata, T., Hirota, K., Kotani, N., Yoshida, H., Kudo, M. and Matsuki, A. (2005) Isoﬂurane increases norepinephrine release in the rat preoptic area and the posterior hypothalamus in vivo and in vitro: relevance to thermoregulation during anesthesia. Neuroscience, 131: 79–86. Lafferty, J.J., Keykhah, M.M., Shapiro, H.M., Van Horn, K. and Behar, M.G. (1978) Cerebral hypometabolism obtained with deep pentobarbital anesthesia and hypothermia (301C). Anesthesiology, 49: 159–164. Lee, M.S., Rinne, J.O. and Marsden, C.D. (2000) The pedunculopontine nucleus: its role in the genesis of movement disorders. Yonsei Med. J., 41: 167–184. Lonjon, M., Risso, J.J., Palmier, B., Negrin, J. and Darbin, O. (2001) Effects of hypothermic deep-anaesthesia on energy metabolism at brain and peripheral levels: a multi-probe microdialysis study in free-moving rat. Neurosci. Lett., 304: 21–24.

Luoh, H.F., Kuo, T.B.J., Chan, S.H.H. and Pan, W.H.T. (1994) Power spectral-analysis of electroencephalographic desynchronization induced by cocaine in rats – correlation with microdialysis evaluation of dopaminergic neurotransmission at the medial prefrontal cortex. Synapse, 16: 29–35. Maggi, C.A. and Meli, A. (1986) Suitability of urethane anesthesia for physiopharmacological investigations in various systems. Part 1: general considerations. Experientia, 42: 109–114. Mammoto, T., Yamamoto, Y., Kagawa, K., Hayashi, Y., Mashimo, T., Yoshiya, I. and Yamatodani, A. (1997) Interactions between neuronal histamine and halothane anesthesia in rats. J. Neurochem., 69: 406–411. Marek, P., Chapman, C.D. and Howard, S. (1992) The effect of phencyclidine and di-2-amino-5-phosphonovaleric acid on Nmethyl-D-aspartic acid-induced changes in extracellular concentration of dopamine and DOPAC in the rat neostriatum. Neuropharmacology, 31: 123–127. McGehee, D.S. and Role, L.W. (1996) Presynaptic ionotropic receptors. Curr. Opin. Neurobiol., 6: 342–349. Metcalf, G. and Myers, R.D. (1978) Precise location within the preoptic area where noradrenaline produces hypothermia. Eur. J. Pharmacol., 51: 47–53. Millan, M.J., Hjorth, S., Samanin, R., Schreiber, R., Jaffard, R., DeLadonchamps, B., Veiga, S., Goument, B., Peglion, J.L., Spedding, M. and Brocco, M. (1997) S 15535, a novel benzodioxopiperazine ligand of serotonin (5-HT)(1A) receptors. 2. Modulation of hippocampal serotonin release in relation to potential anxiolytic properties. J. Pharmacol. Exp. Ther., 282: 148–161. Miller, K.W. and Pang, K.Y. (1976) General anaesthetics can selectively perturb lipid bilayer membranes. Nature, 263: 253–255. Mitchell, S.N. (1993) Role of the locus ceruleus in the noradrenergic response to a systemic administration of nicotine. Neuropharmacology, 32: 937–949. Miyano, K., Tanifuji, Y. and Eger, E.I. (1993) The effect of halothane dose on striatal dopamine – an in vivo microdialysis study. Brain Res., 605: 342–344. Mizuno, T., Endo, Y., Arita, J. and Kimura, F. (1991) Acetylcholine-release in the rat hippocampus as measured by the microdialysis method correlates with motor-activity and exhibits a diurnal variation. Neuroscience, 44: 607–612. Mizuno, T., Ito, E. and Kimura, F. (1994) Pentobarbital sodium inhibits the release of noradrenaline in the medial preoptic area in the rat. Neurosci. Lett., 170: 111–113. Moghaddam, B., Berridge, C.W., Goldman-Rakic, P.S., Bunney, B.S. and Roth, R.H. (1993) In vivo assessment of basal and drug-induced dopamine release in cortical and subcortical regions of the anesthetized primate. Synapse, 13: 215–222. Moghaddam, B. and Bunney, B.S. (1993) Depolarization inactivation of dopamine neurons – terminal release characteristics. Synapse, 14: 195–200. Monda, M., Viggiano, A. and De Luca, V. (2000) Intracerebroventricular injection of prostaglandin E(1) changes concentrations of biogenic amines in the posterior hypothalamus of the rat. Brain Res., 873: 197–202.

89 Morales-Villagran, A. and Tapia, R. (1996) Preferential stimulation of glutamate release by 4-aminopyridine in rat striatum in vivo. Neurochem. Int., 28: 35–40. Morari, M., Oconnor, W.T., Ungerstedt, U. and Fuxe, K. (1993) N-methyl-D-aspartic acid differentially regulates extracellular dopamine, GABA, and glutamate levels in the dorsolateral neostriatum of the halothane-anesthetized rat – an in vivo microdialysis study. J. Neurochem., 60: 1884–1893. Moser, E.I. and Mathiesen, L.I. (1996) Relationship between neuronal activity and brain temperature in rats. Neuroreport, 7: 1876–1880. Mu¨ller, C.P., Carey, R.J., de Souza Silva, M.A., Jocham, G. and Huston, J.P. (2002) Cocaine increases serotonergic activity in the hippocampus and nucleus accumbens in vivo: 5-HT1A-receptor antagonism blocks behavioral but potentiates serotonergic activation. Synapse, 45: 67–77. Murai, T., Koshikawa, N., Kanayama, T., Takada, K., Tomiyama, K. and Kobayashi, M. (1994) Opposite effects of midazolam and beta-carboline-3-carboxylate ethyl ester on the release of dopamine from rat nucleus accumbens measured by in vivo microdialysis. Eur. J. Pharmacol., 261: 65–71. Nelson, L.E., Guo, T.Z., Lu, J., Saper, C.B., Franks, N.P. and Maze, M. (2002) The sedative component of anesthesia is mediated by GABA(A) receptors in an endogenous sleep pathway. Nat. Neurosci., 5: 979–984. Nishikawa, T. and Scatton, B. (1983) Evidence for a GABAergic inhibitory inﬂuence on serotonergic neurons originating from the dorsal raphe. Brain Res., 279: 325–329. Okakura-Mochizuki, K., Mochizuki, T., Yamamoto, Y., Horii, A. and Yamatodani, A. (1996) Endogenous GABA modulates histamine release from the anterior hypothalamus of the rat. J. Neurochem., 67: 171–176. Opacka-Juffry, J., Ahier, R.G. and Cremer, J.E. (1991) Nomifensine-induced increase in extracellular striatal dopamine is enhanced by isoﬂurane anesthesia. Synapse, 7: 169–171. Ortells, M.O. and Lunt, G.G. (1995) Evolutionary history of the ligand-gated ion-channel superfamily of receptors. Trends Neurosci., 18: 121–127. Osborne, P.G., O’Connor, W.T., Drew, K.L. and Ungerstedt, U. (1990) An in vivo microdialysis characterization of extracellular dopamine and GABA in dorsolateral striatum of awake freely moving and halothane anesthetized rats. J. Neurosci. Methods, 34: 99–105. Page, M.E. and Abercrombie, E.D. (1999) Discrete local application of corticotropin-releasing factor increases locus coeruleus discharge and extracellular norepinephrine in rat hippocampus. Synapse, 33: 304–313. Pain, L., Gobaille, S., Schleef, C., Aunis, D. and Oberling, P. (2002) In vivo dopamine measurements in the nucleus accumbens after nonanesthetic and anesthetic doses of propofol in rats. Anesth. Analg., 95: 915–919. Pain, L., Oberling, P., Launoy, A. and Di Scala, G. (1999) Effect of nonsedative doses of propofol on an innate anxiogenic situation in rats. Anesthesiology, 90: 191–196. Pain, L., Oberling, P., Sandner, G. and Di Scala, G. (1996) Effect of propofol on affective state as assessed by place conditioning paradigm in rats. Anesthesiology, 85: 121–128.

Pan, W.H.T. and Lai, Y.J. (1995) Anesthetics decreased the microdialysis extraction fraction of norepinephrine but not dopamine in the medial prefrontal cortex. Synapse, 21: 85–92. Pan, W.H.T., Lai, Y.J. and Chen, N.H. (1995) Differential effects of chloral hydrate and pentobarbital sodium on a cocaine level and its catecholamine response in the medial prefrontal cortex – a comparison with conscious rats. J. Neurochem., 64: 2653–2659. Reeves, D.C. and Lummis, S.C. (2002) The molecular basis of the structure and function of the 5-HT3 receptor: a model ligand-gated ion channel. Mol. Membr. Biol., 19: 11–26. Richter, R.M., Pich, E.M., Koob, G.F. and Weiss, F. (1995) Sensitization of cocaine-stimulated increase in extracellular levels of corticotropin-releasing factor from the rat amygdala after repeated administration as determined by intracranial microdialysis. Neurosci. Lett., 187: 169–172. Rosen, A.D. (1996) Temperature modulation of calcium channel function in GH3 cells. Am. J. Physiol., 271: C863–C868. Rozza, A., Masoero, E., Favalli, L., Lanza, E., Govoni, S., Rizzo, V. and Montalbetti, L. (2000) Inﬂuence of different anaesthetics on extracellular aminoacids in rat brain. J. Neurosci. Meth., 101: 165–169. Rudolph, U. and Antkowiak, B. (2004) Molecular and neuronal substrates for general anaesthetics. Nat. Rev. Neurosci., 5: 709–720. Ruscheweyh, R. and Sandkuhler, J. (2002) Role of kainate receptors in nociception. Brain Res. Rev., 40: 215–222. Rutter, J.J., Baumann, M.H. and Waterhouse, B.D. (1998) Systemically administered cocaine alters stimulus-evoked responses of thalamic somatosensory neurons to perithreshold vibrissae stimulation. Brain Res., 798: 7–17. Sato, K., Wu, J., Kikuchi, T., Wang, Y., Watanabe, I. and Okumura, F. (1996) Differential effects of ketamine and pentobarbitone on acetylcholine release from the rat hippocampus and striatum. Br. J. Anaesth., 77: 381–384. Semba, K., Adachi, N. and Arai, T. (2005) Facilitation of serotonergic activity and amnesia in rats caused by intravenous anesthetics. Anesthesiology, 102: 616–623. Shichino, T., Murakawa, M., Adachi, T., Arai, T., Miyazaki, Y. and Mori, K. (1998) Effects of inhalation anaesthetics on the release of acetylcholine in the rat cerebral cortex in vivo. Br. J. Anaesth., 80: 365–370. Shichino, T., Murakawa, M., Adachi, T., Nakao, S., Shinomura, T., Kurata, J. and Mori, K. (1997) Effects of isoﬂurane on in vivo release of acetylcholine in the rat cerebral cortex and striatum. Acta Anaesthesiol. Scand., 41: 1335–1340. Shimokawa, A., Jin, Q.H., Ishizuka, Y., Kunitake, T., Takasaki, M. and Kannan, H. (1998) Effects of anesthetics on norepinephrine release in the hypothalamic paraventricular nucleus region of awake rats. Neurosci. Lett., 244: 21–24. Shiraishi, M., Kamiyama, Y., Huttemeier, P.C. and Benveniste, H. (1997) Extracellular glutamate and dopamine measured by microdialysis in the rat striatum during blockade of synaptic transmission in anesthetized and awake rats. Brain Res., 759: 221–227.

90 Sokoloff, L., Reivich, M., Kennedy, C., Desrosiers, M.H., Patlak, C.S., Pettigrew, K.D., Sakurada, O. and Shinohara, M. (1977) Deoxyglucose-C-14 method for measurement of local cerebral glucose-utilization – theory, procedure, and normal values in conscious and anesthetized albino rat. J. Neurochem., 28: 897–916. Soma, L.R. (1983) Anesthetic and analgesic considerations in the experimental animal. Ann. N. Y. Acad. Sci., 406: 32–47. Sonner, J.M., Antognini, J.F., Dutton, R.C., Flood, P., Gray, A.T., Harris, R.A., Homanics, G.E., Kendig, J., Orser, B., Raines, D.E., Rampil, I.J., Trudell, J., Vissel, B. and Eger, E.I. (2003) Inhaled anesthetics and immobility: mechanisms, mysteries, and minimum alveolar anesthetic concentration. Anesth. Analg., 97: 718–740. Spampinato, U., Girault, J.A., Danguir, J., Savaki, H.E., Glowinski, J. and Besson, M.J. (1986) Apomorphine and haloperidol effects on striatal H-3 dopamine release in anesthetized, awake restrained and freely moving rats. Brain Res. Bull., 16: 161–166. Spanagel, R., Herz, A., Balskubik, R. and Shippenberg, T.S. (1991) Beta-endorphin-induced locomotor stimulation and reinforcement are associated with an increase in dopamine release in the nucleus accumbens. Psychopharmacology, 104: 51–56. Stahle, L., Collin, A.K. and Ungerstedt, U. (1990) Effects of halothane anesthesia on extracellular levels of dopamine, dihydroxyphenylacetic acid, homovanillic-acid and 5-hydroxyindolacetic acid in rat striatum – a microdialysis study. Naunyn Schmiedebergs Arch. Pharmacol., 342: 136–140. Stengard, K., Tham, R., O’connor, W.T., Hoglund, G. and Ungerstedt, U. (1993) Acute toluene exposure increases extracellular GABA in the cerebellum of rat – a microdialysis study. Pharmacol. Toxicol., 73: 315–318. Taguchi, K., Andresen, M.J. and Hentall, I.D. (1991) Acetylcholine release from the midbrain interpeduncular nucleus during anesthesia. Neuroreport, 2: 789–792. Tao, R. and Auerbach, S.B. (1994) Anesthetics block morphine-induced increases in serotonin release in rat CNS. Synapse, 18: 307–314. Tecott, L.H., Maricq, A.V. and Julius, D. (1993) Nervous system distribution of the serotonin 5-HT3 receptor mRNA. Proc. Natl. Acad. Sci. U.S.A., 90: 1430–1434. Tepper, J.M., Creese, I. and Schwartz, D.H. (1991) Stimulusevoked changes in neostriatal dopamine levels in awake and anesthetized rats as measured by microdialysis. Brain Res., 559: 283–292. Thiel, C.M., Huston, J.P. and Schwarting, R.K.W. (1998) Hippocampal acetylcholine and habituation learning. Neuroscience, 85: 1253–1262. Thiel, C.M., Mu¨ller, C.P., Huston, J.P. and Schwarting, R.K.W. (2000) Auditory noise can prevent increased extracellular acetylcholine levels in the hippocampus in response to aversive stimulation. Brain Res., 882: 112–119. Thomas, D.N., Post, R.M. and Pert, A. (1994) Focal and systemic cocaine differentially affect extracellular norepinephrine in the locus coeruleus, frontal cortex and hippocampus of the anesthetized rat. Brain Res., 645: 135–142.

Tsukada, H., Nishiyama, S., Kakiuchi, T., Ohba, H., Sato, K., Harada, N. and Nakanishi, S. (1999) Isoﬂurane anesthesia enhances the inhibitory effects of cocaine and GBR12909 on dopamine transporter: PET studies in combination with microdialysis in the monkey brain. Brain Res., 849: 85–96. Van Gaalen, M., Kawahara, H., Kawahara, Y. and Westerink, B.H.C. (1997) The locus coeruleus noradrenergic system in the rat brain studied by dual-probe microdialysis. Brain Res., 763: 56–62. Vazquez, J. and Baghdoyan, H.A. (2004) GABA(A) receptors inhibit acetylcholine release in cat pontine reticular formation: Implications for REM sleep regulation. J. Neurophysiol., 92: 2198–2206. Veselis, R.A., Feshchenko, V.A., Reinsel, R.A., Beattie, B. and Akhurst, T.J. (2005) Propofol and thiopental do not interfere with regional cerebral blood ﬂow response at sedative concentrations. Anesthesiology, 102: 26–34. Volgushev, M., Kudryashov, I., Chistiakova, M., Mukovski, M., Niesmann, J. and Eysel, U.T. (2004) Probability of transmitter release at neocortical synapses at different temperatures. J. Neurophysiol., 92: 212–220. Wang, Y., Kikuchi, T., Sakai, M., Wu, J.L., Sato, K. and Okumura, F. (2000) Age-related modiﬁcations of effects of ketamine and propofol on rat hippocampus acetylcholine release studied by in vivo brain microdialysis. Acta Anesthesiol. Scand., 44: 112–117. Warenycia, M.W. and McKenzie, G.M. (1988) Excitation of striatal neurons by dexamphetamine is not abolished by either chloral hydrate or urethane anesthesia. Neuropharmacology, 27: 1309–1312. Waterhouse, B.D., Devilbiss, D., Seiple, S. and Markowitz, R. (2004) Sensorimotor-related discharge of simultaneously recorded, single neurons in the dorsal raphe nucleus of the awake, unrestrained rat. Brain Res., 1000: 183–191. Westerink, B.H.C., Deboer, P. and Damsma, G. (1990) Dopamine–acetylcholine interaction in the striatum studied by microdialysis in the awake rat – some methodological aspects. J. Neurosci. Methods, 34: 117–124. Whitehead, K.J., Rose, S. and Jenner, P. (2001) Involvement of intrinsic cholinergic and GABAergic innervation in the effect of NMDA on striatal dopamine efﬂux and metabolism as assessed by microdialysis studies in freely moving rats. Eur. J. Neurosci., 14: 851–860. Whitehead, K.J., Rose, S. and Jenner, P. (2004) Halothane anesthesia affects NMDA-stimulated cholinergic and GABAergic modulation of striatal dopamine efﬂux and metabolism in the rat in vivo. Neurochem. Res., 29: 835–842. Wilde, M.I. and Markham, A. (1996) Ondansetron. A review of its pharmacology and preliminary clinical ﬁndings in novel applications. Drugs, 52: 773–794. Wilkinson, L.S., Mittleman, G., Torres, E., Humby, T., Hall, F.S. and Robbins, T.W. (1993) Enhancement of amphetamine-induced locomotor activity and dopamine release in nucleus accumbens following excitotoxic lesions of the hippocampus. Behav. Brain Res., 55: 143–150.

91 Xie, T., McCann, U.D., Kim, S., Yuan, J. and Ricaurte, G.A. (2000) Effect of temperature on dopamine transporter function and intracellular accumulation of methamphetamine: implications for methamphetamine-induced dopaminergic neurotoxicity. J. Neurosci., 20: 7838–7845. Yamakura, T., Bertaccini, E., Trudell, J.R. and Harris, R.A. (2001) Anesthetics and ion channels: molecular models and sites of action. Ann. Rev. Pharmacol. Toxicol., 41: 23–51. Yamakura, T., Chavez-Noriega, L.E. and Harris, R.A. (2000) Subunit-dependent inhibition of human neuronal nicotinic acetylcholine receptors and other ligand-gated ion channels by dissociative anesthetics ketamine and dizocilpine. Anesthesiology, 92: 1144–1153. Yamakura, T., Mori, H., Masaki, H., Shimoji, K. and Mishina, M. (1993) Different sensitivities of NMDA receptor channel

subtypes to non-competitive antagonists. Neuroreport, 4: 687–690. Young, A.M.J. and Bradford, H.F. (1991) N-methyl-D-aspartate releases excitatory amino-acids in rat corpus striatum in vivo. J. Neurochem., 56: 1677–1683. Zhang, Y., Laster, M.J., Hara, K., Harris, R.A., Eger, E.I., Stabernack, C.R. and Sonner, J.M. (2003) Glycine receptors mediate part of the immobility produced by inhaled anesthetics. Anesth. Analg., 96: 97–101. Zhang, W., Tilson, H., Stachowiak, M.K. and Hong, J.S. (1989) Repeated haloperidol administration changes basal release of striatal dopamine and subsequent response to haloperidol challenge. Brain Res., 484: 389–392. Zhu, M., Nehra, D., Ackerman, J. and Yablonskiy, D. (2004) On the role of anesthesia on the body/brain temperature differential in rats. J. Therm. Biol., 29: 599–603.

CHAPTER 1.6

Quantitative aspects of brain microdialysis: insights from voltammetric measurements of dopamine next to microdialysis probes Adrian C. Michael Department of Chemistry, University of Pittsburgh, Pittsburgh, PA, USA

Abstract: Microdialysis offers many attributes that make it extremely popular as a technique for sampling the extracellular space in the brain of living animals. Correlating the chemical composition of the extracellular space with behavioral events, pharmacological responses, and brain pathophysiology is a central objective in the ﬁeld of neurochemistry. Microdialysis has evolved to a position of prominence in efforts to achieve this objective. Nonetheless, questions remain unanswered about the relationship between concentrations as measured by microdialysis and the actual in vivo concentration of target substances in the extracellular space of brain tissue. These questions remain open because it is the case that methods for the in vivo calibration of microdialysis probes are not available and are difﬁcult to envision. Although a technique called quantitative microdialysis, based on the determination of no-net-ﬂux concentrations, has been developed, it actually only provides quantitative information when two key parameters, the extraction fraction and relative recovery, of the target substance are the same during the microdialysis procedure. Although quantitative microdialysis has been applied to dopamine, voltammetric recordings of extracellular dopamine in the tissue adjacent to microdialysis probes caste doubt on the idea that the dopamine extraction fraction and relative recovery are equal to each other. This chapter reviews the quantitative basis of in vivo microdialysis, describes the evidence that suggests a difference between the in vivo extraction fraction and relative recovery of dopamine, and considers the consequences of that difference in terms of the nature of the relationship between dopamine as measured by microdialysis and the actual in vivo concentration of dopamine in the brain extracellular space.

maturity, so there is no shortage of literature and commercial resources to support new users and novel applications. Perhaps the most important resource is the commercial availability of probes and the other apparatus (syringe pumps, ﬂuid swivels, and test chambers) customized for microdialysis procedures. Probe implantation, especially in rats and mice, involves conventional and straightforward stereotaxic procedures and unanesthetized rodents are very tolerant of the attendant guide cannulas and ﬂow lines. Although microdialysiscompatible analytical systems, especially those

I. Introduction The widespread interest in microdialysis as a tool for monitoring the chemical composition of the extracellular space of the living brain is well documented. This interest is spurred on by the numerous advantages and beneﬁts that microdialysis offers to neurochemistry, neuropharmacology, and related ﬁelds. The technique has reached advanced Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

93

DOI: 10.1016/S1569-7339(06)16006-3 Copyright 2007 Elsevier B.V. All rights reserved

94

based on capillary electrophoresis, continue to emerge from academic research laboratories, commercial HPLC systems provide essentially turn-key performance for brain microdialysis applications. Microdialysis probes provide very clean samples that can be transferred directly to chromatography or capillary electrophoresis systems. This has enabled the development of on-line perfusion methods in which the probe outlet line is plumbed directly to an interface, sometimes just a valve, to the analytical system. This enables automated continuous monitoring with minimal interference of the animal’s behavior by the experimenter. Modern analytical systems provide very high performance where sensitivity, selectivity, stability, and reproducibility are concerned. Consequently, there is essentially no doubt about the reported amounts or identities of target substances in brain microdialysate samples.

II. The issue of in vivo calibration Despite the many and well-known advantages of microdialysis, the absence of methodologies for calibrating microdialysis probes once they are implanted in brain tissue is a potential drawback of the technique. The quantitative analysis of microdialysate samples provides information regarding the concentrations or amounts of substances in the microdialysate samples themselves. However, due to the lack of methods for in vivo probe calibration, the relationship between the concentrations of substances in microdialysate samples and the actual in vivo concentrations of those substances is unknown. Because it is widely recognized that microdialysis probes are used without in vivo calibration, microdialysis results are usually normalized with respect to baseline data, a practice that acknowledges but does not solve the calibration problem. The underlying assumption is that relative changes in the normalized data are equivalent to relative changes in the actual in vivo concentrations of target substances in the brain extracellular space. But, in the absence of methods for in vivo probe calibration it has not been possible to verify this equivalency. Overall,

the relationship between absolute and relative microdialysis results and the actual chemical composition of the brain extracellular space is not as clear as one might expect or desire. The in vitro calibration of microdialysis probes in standard solutions of known composition is straightforward. However, in vitro calibration is not applicable to results obtained during in vivo microdialysis because the probe calibration factors are affected by the medium in which the probe is immersed. Hence, the quantitative performance of the probe changes from one medium to the next and so is not determined solely by the probe itself. The diffusion of small molecules between the ﬂuids on either side of the dialysis membrane, the internal and external media, determines the quantitative performance of the probe. Diffusion is a mode of mass transport driven by the continuous random motion of molecules in ﬂuids (Crank, 1975; Berg, 1993). In the presence of a concentration gradient, random motion carries more molecules from regions of higher concentration to regions of lower concentration than vice versa, resulting in net mass transport down the concentration gradient. The concentration gradient that drives mass transport from the external medium to the internal medium during microdialysis extends into the external medium itself: if this were not the case, the dialysate could not provide information about the composition of the external medium (this statement would not be strictly true if the velocity of mass transport in the external medium were to greatly exceed that in the membrane or in the internal medium, e.g., if the external medium were vigorously stirred (Bungay et al., 2003), but such a condition does not apply to brain microdialysis). Since the concentration gradient extends into the external medium, the physicochemical properties of the external medium play an important role in determining the quantitative performance of the probe. Any process that affects the concentration gradient in the external medium will also affect the diffusion process that delivers substances into the dialysate sample. During in vivo microdialysis, such processes might include the release or uptake of substances by neurons, glia, or terminals as well as metabolism, binding to receptors, or transport

95

between the extracellular space and the local vasculature (Morrison et al., 1991). Clearly, it is impossible to emulate these processes in vitro calibration procedures, so calibration would have to be performed in vivo.

II.A. The impact of calibration uncertainty A classical application of microdialysis in the brain is to monitor how drugs affect extracellular neurotransmitter levels. If a drug affects the concentration gradient of the neurotransmitter in the extracellular space near the microdialysis probe, then that drug will change the diffusion of the neurotransmitter into the probe. Any change in the dialysate concentration of the neurotransmitter therefore reﬂects, ﬁrst, how the drug affects the extracellular concentration of the neurotransmitter and, second, how the drug affects the diffusion of the neurotransmitter into the probe. In vivo calibration of the probe would be required to dissect apart these two independent contributions to the change in concentration of the neurotransmitter in the dialysate samples. Note that to fully dissect these two contributions it would be necessary to calibrate the probe both before and after delivery of the drug to account for the drug-induced change in probe performance. Anything that affects the tissue surrounding the probe might affect the quantitative performance of the probe. This includes the probe itself. Implantation of objects into the brain, may they be microdialysis probes, stimulating electrodes, recording electrodes, or voltammetric electrodes, leads to disruption of the surrounding tissue (Clapp-Lilly et al., 1999; Szarowski et al., 2003; Peters et al., 2004; Retterer et al., 2004). This is simply unavoidable as the implantation is an invasive procedure. Tissue disruption might affect the concentration gradients of substances near the probes either by changing the physical geometry and composition of the extracellular space or by changing the activity of biochemical pathways acting on the target substances. Again, in vivo calibration of the probe would be necessary to shed light on these contributions.

II.B. The obstacle to in vivo calibration Calibration of any analytical determination requires a primary standard. For the majority of substances of interest in brain microdialysis studies, it is not possible to independently know or control the extracellular concentration of the analyte; indeed, if this were possible, microdialysis would probably be unnecessary. Thus, primary standards for in vivo calibration are generally not available. It should be emphasized that ‘model compounds’ cannot be used as primary standards, because each target substance likely has speciﬁc interactions with brain tissue that affect its diffusion properties. For example, the diffusion of dopamine in the extracellular space is affected by dopamine uptake (Gulley and Zahniser, 2003); a model compound that is not equally recognized by the dopamine transporter (DAT) cannot, therefore, serve as a primary standard for dopamine. The absence of methods for in vivo probe calibration stems from the absence of primary standards. Unfortunately perhaps, it is not the objective of this chapter to solve the mystery of how to calibrate microdialysis probes in the brain. Rather, this chapter will explore at least one approach to identifying whether the absence of in vivo calibration is a ‘big deal’ or just an unavoidable annoyance that goes with the territory of in vivo analysis. The answer might very well depend on which particular substance is under consideration.

II.C. The quantitative basis of microdialysis The lack of methods for the in vivo calibration of microdialysis probes might be surprising to those who have encountered the literature on ‘quantitative microdialysis’ (Morrison et al., 1991; Parsons and Justice, 1994; Watson et al., 2006 and citations therein). So a discussion of the quantitative aspects of microdialysis is in order. The objective of microdialysis is to deduce information regarding the concentration of a target substance in the external medium, usually denoted Cext,N, by measuring the concentration of the substance in the dialysate that exits the probe via

96

the outlet line, Cout. Sometimes, the target substance is added to the perfusion ﬂuid that is pumped into the probe inlet, in which case the inlet concentration is denoted Cin. The external concentration is labeled with an inﬁnity symbol, Cext,N, to denote locations sufﬁciently far from the probe that the external concentration is not affected by the probe itself. The labeling of these various concentration terms is by now conventional in the microdialysis literature. The quantitative aspects of microdialysis are embodied in two parameters, the extraction fraction, E and relative recovery, R. The relative recovery is the ratio of Cout to Cext,N obtained when Cin is zero (sometimes referred to as conventional microdialysis): R¼

C out C ext;1

or C out ¼ R C ext;1

(1)

According to Eq. (1), R is the probe calibration factor for conventional microdialysis. To measure R it is necessary to measure Cout in the presence of a primary standard for Cext,N. While this is simple during in vitro calibration, the same is not true for in vivo microdialysis. The in vivo values of R for substances of interest in brain microdialysis are unknown. When the target substance is added to the perfusion ﬂuid some of it is lost (extracted) from the perfusion ﬂuid while the remainder exits via the probe outlet. By measuring the concentration that reaches the probe outlet it is possible to deduce what fraction was extracted. When Cext,N is zero, this concentration is: C out ¼ ð1 EÞ C in

(2)

Eq. (2) expresses that the fraction of Cin reaching the probe outlet is the fraction that was not extracted from the probe, which is given by 1 E. Then, the fraction that was extracted is E Cin. For the sake of completeness, it is perhaps worth mentioning that the extraction fraction, E, in Eq. (2) is not relevant to the technique known as retrodialysis, in which substance is added to the perfusion ﬂuid for the purposes of intracranial delivery. For intracranial delivery, it is not appropriate to write C ext ¼ E C in ; where E is obtained from Eq. (2). This is because the volume of the

external medium is vast in comparison to the volume of the internal volume of the microdialysis probe, so substances extracted from the probe are vastly diluted. The quantity E C in is just the fraction of Cin that does not arrive at the probe outlet during the concentration differences measurement. In general, Cout is the sum of two contributions: the amount of substance recovered from the external medium, R C ext;1 (from Eq. 1), and the amount of substance not extracted from the perfusion medium ð1 EÞ C in (from Eq. 2): C out ¼ R C ext;1 þ ð1 EÞ C in

(3)

Eq. (3) can be rearranged to obtain the mathematical form of the concentration differences plot, which is a plot of Cin Cout as a function of Cin (Lonnroth et al., 1987): C in C out ¼ E C in R C ext;1

(4)

According to Eq. (4), E can be obtained empirically from the slope of the concentration differences plot. Hence, unlike R, E can be measured during in vivo experiments since one does not need a primary standard for Cext,N to construct the concentration differences plot and determine its slope. Equation (4), which is of central importance in the quantitative aspects of microdialysis, is derived by simply summing two contributions to Cout, the concentration recovered from the external medium (Eq. 1) and the concentration not extracted from the perfusion ﬂuid (Eq. 2). By simply adding them together, we are invoking the idea that the two contributions are independent of one another, that is, the concentration recovered from the external medium is independent of Cin and that the concentration extracted from the internal medium is independent of Cext,N. The validity of this idea is derived from the concept that diffusion involves the stochastic motion of non-interacting particles (Crank, 1975; Berg, 1993). For this reason, E and R are normally concentration-independent quantities. It has been suggested that the concentration differences plot (Eq. 4) enables quantitative in vivo microdialysis measurements without actually requiring in vivo probe calibration (Parsons and Justice, 1994; Watson et al., 2006). There is a

97

special point on the concentration differences plot where Cin and Cout are mutually equal to the concentration of no-net-ﬂux, CNNF. According to Eq. (4), if E and R are equal to each other, then CNNF equals Cext,N; let Cout ¼ Cin ¼ CNNF and let E ¼ R ¼ F and rewrite Eq. (4) as: C NNF C NNF ¼ F C NNF F C ext;1 and therefore; C NNF ¼ C ext;1

ð5Þ

Eq. (5) is the basis for quantitative microdialysis, which employs CNNF as a measure of Cext,N. The derivation of Eq. (5) makes it very clear that a key assumption underlying quantitative microdialysis is that the extraction fraction and relative recovery are equal. While E and R may be equal under certain circumstances (Song and Lunte, 1999) they are not always equal. When they are not equal, CNNF is not a direct measure of Cext,N; let Cout ¼ Cin ¼ CNNF and rewrite Eq. (4) as: R C ext;1 (6) E Hence, the probe calibration factor for no-net-ﬂux microdialysis is R=E; which means that quantitative microdialysis unavoidably demands knowledge of the R value. The value of R must be measured to use Eq. (6) to relate CNNF to Cext,N or to verify that it is appropriate to use Eq. (5). Thus, quantitative microdialysis is not possible without knowledge of the in vivo R value of the substance of interest.

C NNF ¼

III. An example of in vitro microdialysis where E and R are different When the quantitative aspects of microdialysis are studied under the well-controlled conditions of an in vitro calibration procedure, it is likely that measured values of E and R will be the same. However, this is not always the case. Julie Stenken’s group developed specialized perfusion media designed to enhance the microdialysis recovery of certain drugs (Khramov and Stenken, 1999). Employing the principles of molecular recognition, Stenken and coworkers included bulky ligands in

the perfusion ﬂuid capable of binding, and thereby trapping, the target drugs. Once trapped in the probe, the drugs are unable to diffuse back across the membrane, so their recovery is enhanced to the point that the dialysate concentration of the drug actually exceeded the concentration in the surrounding solution. According to Eq. (1), this means the microdialysis recovery value of the drug exceeds 1. However, under the usual conditions of a microdialysis experiment, where the volume of the external medium is greater than the internal volume of the probe, it is not possible for the extraction fraction to be greater than 1 as this would lead to negative values of Cout (Eq. 2). Hence, Stenken’s work is an example of in vitro microdialysis where E and R are not equal.

III.A. Does E equal R in vivo? The absence of evidence for the equality of E and R values under in vivo conditions means that the validity of the concentration differences method for quantitative microdialysis has not been conﬁrmed. However, the literature contains arguments in favor of the validity of quantitative microdialysis, that is, E and R are equal in vivo. Two of these arguments are discussed next. One argument is derived from the observation that concentration difference plots obtained under in vivo conditions are straight and exhibit the same slope when Cin is greater than CNNF and when it is less than CNNF (Parsons and Justice, 1994; Watson et al., 2006). The equality of the slope above and below CNNF has been described as evidence for the equality of E and R. The argument is that when Cin is greater than Cout, the slope of the concentration differences plot measures E because there is net ﬂux from the probe to the external medium. In contrast, when Cin is less than Cout, the slope of the concentration differences plot measures R because the net ﬂux is from the external medium into the probe. This description of the concentration differences plot, however, is not consistent with Eq. (4), which explains that the slope only contains information about E; the slope contains no information about R, which only contributes to the intercept value. Negative

98

values of Cin Cout simply reﬂect the fact that the quantity E Cin is smaller than the quantity R Cext,N, that is, less substance is being extracted from the probe than is being recovered from the external medium, causing Cout to be greater than Cin. Positive values of Cin Cout simply reﬂect the opposite case that E Cin>R Cext,N. The observation that the concentration differences plot has the same slope above and below CNNF says nothing about the equality of E and R because the slope only measures E and contains no information about R. Another argument offered in support of quantitative microdialysis is based on the method of extrapolating Cout to zero perfusion ﬂow (Parsons and Justice, 1994; Watson et al., 2006). In this procedure, Cout is measured at a series of different perfusion rates and the results are extrapolated to estimate the value of Cout at zero, or very slow, perfusion velocity. If one assumes that R approaches 1 as the perfusion velocity approaches zero, then Cout at zero perfusion velocity appears to be a quantitative measure of Cext,N. The few extrapolated Cout values that have been determined agree well with the corresponding CNNF values, which has led to claims that CNNF is a valid quantitative measure of Cext,N. However, it is quite obvious that there is no assurance during in vivo microdialysis that R will indeed approach 1 at zero perfusion velocity; this is another untested assumption in the ﬁeld of quantitative microdialysis, much like the assumption that E and R are always the same. In fact, the zero perfusion velocity produces a state of no-net-ﬂux in the microdialysis probe. It is obvious that this is the case, because at zero perfusion velocity the values of Cin and Cout must be equal, so that there is no-net-ﬂux between the internal and external media. Moreover, CNNF is independent of the perfusion velocity (Peters and Michael, 1998), so Cout at zero ﬂow has to be equal to CNNF measured at any other ﬂow rate. The agreement between Cout at zero ﬂow and CNNF is not evidence for the validity of quantitative microdialysis; these quantities agree because they are in fact the same quantity. The concentration differences procedure and the extrapolation to zero perfusion velocity are two procedures for measuring the same quantity, CNNF, so

nothing is learned from the fact that the two procedures yield the same result.

IV. A comment about the concentrationindependence of E and R As mentioned above, Eq. (4), which is a key relationship in the quantitative aspects of microdialysis, is derived by assuming that E and R are independent of the concentration of the target substance. However, there is a possible exception to this case that arises when the target substance is subject to non-linear kinetics, which may very well be encountered during in vivo experiments. An example of a non-linear kinetic process is dopamine uptake that is mediated by the DAT, which exhibits Michaelis–Menton-like kinetics (Wu et al., 2001). Because the transporter removes dopamine molecules that might contribute to the concentration gradient that controls dopamine extraction and recovery, changes in the kinetics of the transporter are expected to change the value of E and R. Indeed, Smith and Justice (1994) reported sometime ago that DAT inhibition changes the slope of the dopamine concentration differences plot, consistent with the idea that the dopamine extraction fraction is affected by DAT kinetics. The fact that DAT kinetics are affected by the concentration of dopamine itself has led to the theoretical prediction of non-linear concentration differences plots (Peters and Michael, 1998). This would occur if the dopamine Cin value were so high that the dopamine extracted from the probe would saturate the transporter, effectively inhibiting dopamine uptake. To date, non-linear dopamine concentration differences plots have not been observed even when very high Cin values were tested; however, this is not surprising. Although the value of Cin might be very large, it is also true that the internal volume of the probe is very small and the volumetric perfusion ﬂow through the probes is also small (typically 2 mL/min or less). Hence, the total amount of dopamine delivered to the brain while a concentration differences plot is measured is very small. This small amount of dopamine is rapidly diluted in the large volume of the external medium and is further diluted by the

99

uptake process itself. For these reasons, dopamine concentration differences plots retain their linear behavior despite the non-linearity of DAT kinetics. Thus, at present it appears to be satisfactory to regard both E and R for dopamine as concentration-independent quantities.

V. Voltammetry as a tool to investigate in vivo recovery of dopamine As mentioned above, the quantitative aspects of microdialysis are embodied in two parameters, E and R. The value of E can be obtained in vivo from the slope of concentration differences plots even in the absence of a primary standard. The same is not true, however, of the value of R, measurement of which requires a primary standard for probe calibration. Sometime ago, we became interested in the idea that it might be possible to study the in vivo recovery of dopamine from the extracellular space of brain tissue by the means of voltammetric measurements in the tissue surrounding microdialysis probes (Lu et al., 1998). Originally, our motivation for this work was to investigate how the temporal dynamics of concentration events in the extracellular space relate to the time course in changes in dialysate dopamine concentrations. We were interested to know the transfer function for in vivo microdialysis (see Engstrom et al., 1988). However, the measurements have provided more insight into the dopamine recovery process. At the outset, however, it is necessary to clearly state that these measurements do not solve the in vivo calibration problem because voltammetric results do not constitute a primary standard against which the microdialysis probe can be calibrated. The value of voltammetric measurements next to microdialysis probes lies in the identiﬁcation of the mechanism that underpins the inequality between E and R. The inequality between E and R of dopamine during in vivo microdialysis is a consequence of disruption of the dopamine system in the tissue surrounding the microdialysis probe. It appears that dopamine release from terminals near the probe is suppressed even though dopamine uptake remains relatively intact (the data in support of

this assessment are discussed below). Because of this differential disruption of dopamine release and uptake, full dopamine release activity only occurs in tissue that is remote from the probe, that is, tissue separated from the probe by an intervening layer of disrupted tissue. In order for dopamine released in this remote normal tissue to reach a microdialysis probe, it has to diffuse considerable distances (100 s of mm) through the extracellular space of the disrupted tissue. While diffusing in the disrupted tissue, however, dopamine molecules are subject to removal from the extracellular ﬂuid compartment by the DAT. Thus, dopamine uptake acts to decrease the likelihood of dopamine released in the remote normal tissue from contributing to Cout. The consequence is an inherently low in vivo dopamine recovery value. Recall that the recovery value (Eq. 1) is deﬁned relative to the quantity Cext,N, which is the concentration in the external medium sufﬁciently far from the probe that the external medium is unaffected by the probe. In the context of an in vivo microdialysis experiment, this means tissue beyond the disrupted zone as well as beyond the region of diffusional depletion. The results that support our description of the impairment of the dopamine system in the vicinity of microdialysis probes were obtained by means of voltammetric measurements next to the probe. The rationale for this approach is based on the difference in the size of microdialysis probes and voltammetric microelectrodes. The probes have a diameter of least 225 mm and create a disrupted zone of tissue that in one histological study extended as far as 1.4 mm from the implantation site (Clapp-Lilly et al., 1999). Carbon ﬁber microelectrodes come in a variety of sizes but the one we use for dopamine experiments have a diameter of 7 mm. Electron microscopy of the track created by these carbon ﬁbers reveals that the tissue damage they cause is mainly conﬁned to a distance of 3 mm (Peters et al., 2004). Thus, the rationale of our voltammetric measurements near microdialysis probes is based on the idea that the probe is responsible for the majority of the tissue disruption and that the microelectrode can be used to query the consequence of that disruption on the dopamine system.

100

The conclusion that the in vivo recovery of dopamine is inherently low is based on observations from the striatum of chloral hydrate anesthetized rats. Voltammetry in conjunction with carbon ﬁber microelectrodes has been widely employed to monitor extracellular dopamine in the striatum during electrical stimulation of dopaminergic axons in the medial forebrain bundle (MFB). In one study, we ﬁrst placed a microelectrode in the striatum and recorded evoked dopamine release (Borland et al., 2005). Then, we implanted a microdialysis probe so that the gap between the probe and the microelectrode was just over 200 mm. (To make the electrode and the probe ﬁt so close to one another it was necessary to tilt the microelectrode at a slight angle with respect to the microdialysis probe.) Two hours after implantation of the probe, an identical electrical stimulus was delivered to the MFB. The response evoked by this stimulus was 10-fold smaller in amplitude than the response evoked prior to implanting the microdialysis probe (Fig. 1). This result shows that implantation of the probe acutely disrupts evoked dopamine release for a distance of at least 200 mm from the outer surface of the probe itself. We have reported other versions of this experiment (Lu et al., 1998; Yang et al., 1998). In some cases, we have used microdialysis probes equipped with a carbon ﬁber microelectrode that is actually mounted onto the outer surface of the dialysis membrane. These electrodes provide information about electrically evoked dopamine release immediately adjacent to the probe itself. Two hours after implantation of the probe–electrode combination, MFB stimulation evokes no response at the electrode mounted directly onto the probe (Yang et al., 1998). In contrast, we have also used microelectrodes positioned 1 mm from the probe. Probe implantation does not affect the stimulus response at these microelectrodes (Yang et al., 1998; Borland et al., 2005). Thus, probe implantation leads to the formation of a gradient in disruption of evoked DA release, with a total abolition of the response immediately adjacent to the probe, a 10-fold decrease at 200 mm from the probe, and no significant effect 1 mm from the probe. Although no evoked release was observed at the microelectrode mounted on the outer surface of

A

20 nA

C

B

10

20

30

40

time, sec Fig. 1. Voltammetric responses recorded during electrical stimulation of the medial forebrain bundle (MFB) with a carbon ﬁber microelectrode positioned about 200 mm from a microdialysis probe. (A) The response recorded prior to implanting the microdialysis probe. (B) The response recorded 2 h after implanting the microdialysis probe. (C) The response recorded 20 min after the administration of the dopamine uptake inhibitor, nomifensine (20 mg/kg i.p.). Implantation of the microdialysis probe approximately 200 mm from the voltammetric microelectrode decreased the response amplitude by 90% (A vs. B). Nomifensine more than doubled the response amplitude after probe implantation (B vs. C), which is much larger to the proportional effect of nomifensine in the absence of a microdialysis probe (see Borland et al., 2005 for full details).

the probe, evoked release was observed after inhibition of the DAT with nomifensine. Furthermore, a systematic study of the impact of nomifensine on the amplitude of the evoked response clearly shows that uptake inhibition has a larger relative effect on the response at electrodes nearer to the probe (Borland et al., 2005; Fig. 1). Thus, not only is there a gradient of evoked dopamine release near the probe, there is also a gradient in the activity of the dopamine uptake mechanism. These data support the idea that probe implantation has a differential effect on the activity of evoked dopamine release and dopamine uptake. The results we obtained with the microelectrodes mounted on the outer surface of the microdialysis probes correlate very closely with results observed at microelectrodes mounted into the distal end of the outlet line of the probe

101

(Lu et al., 1998; Yang et al., 1998). When electrical stimulation was performed 2 h after probe implantation, no evoked dopamine release was observed by voltammetry in the probe outlet line. However, evoked dopamine release was readily observed after uptake inhibition with nomifensine. These results also correlate with analyses of 30-s duration microdialysis samples by capillary electrophoresis with electrochemical detection (Qian et al., 1999). Prior to uptake inhibition, electrical stimulation increased the dialysate dopamine content by a few nanomolar (the capillary electrophoresis with electrochemical detection (CE-EC) system offers lower detection limits than on-line voltammetry). However, after uptake inhibition, electrical stimulation dramatically increased the dopamine content of the dialysate samples. These observations conﬁrm that inhibition of dopamine uptake promotes the ability of dopamine to diffuse into the probe and contribute to Cout, that is, dopamine uptake increases the in vivo recovery of dopamine. All the results discussed to this point were obtained during acute experiments performed starting 2 h after implantation of the probe. Conventional microdialysis procedures for experiments involving dopamine include a wait time after probe implantation of approximately 24 h. To date, however, we have only performed simultaneous microdialysis and voltammetry in anesthetized animals for technical reasons. The longest time after probe implantation we have studied in the anesthetized animals is 16 h (Yang and Michael, unpublished observations). Again, 16 h after probe implantation, MFB stimulation produced no detectable response at microelectrodes mounted onto the outer surface of microdialysis probes while the response 1 mm from the probe was unaffected. Thus, evoked dopamine release remains inhibited for at least 16 h. Furthermore, two studies have examined the effect of MFB stimulation of dialysate dopamine levels in unanesthetized animals at 24 h after probe implantation (Tepper et al., 1991; Manley et al., 1992). Qualitatively, the results are similar to what we have reported at shorter times. Electrical stimulation produces relatively small (nanomolar) changes in dialysate dopamine levels, with a dramatic increase after uptake inhibition.

VI. Comparison of the in vivo extraction and recovery of dopamine Collectively, the results described above support the idea that once dopamine is released into the extracellular space, it is far more likely to be taken back into dopamine terminals via the DAT than to diffuse to and be recovered by a microdialysis probe (Yang et al., 2000). After inhibition of the DAT, the likelihood that dopamine molecules will diffuse to and be recovered by the probe increases; uptake inhibition increases in vivo dopamine recovery. Thus, we conclude that the effect of dopamine uptake on the in vivo microdialysis recovery and extraction of dopamine are opposite, because Smith and Justice (1994) demonstrated that inhibition of dopamine uptake decreases the slope of the dopamine concentration differences plot, that is uptake inhibition decreases the in vivo dopamine extraction fraction. The opposite effect of dopamine uptake inhibition on the in vivo dopamine extraction and recovery proves that the relationship between CNNF and Cext,N is given by Eq. (6) rather than Eq. (5). At ﬁrst glance, it might seem strange that dopamine uptake should have opposite effects on the in vivo extraction and recovery of dopamine, especially since some prior descriptions of microdialysis have assumed that extraction and recovery are symmetrical processes (Morrison et al., 1991). As it turns out, the idea that uptake has opposite effects on extraction and recovery is quite easy to understand and reﬂects a symmetrical effect on the dopamine Cout value. We have already explained that uptake decreases dopamine recovery by preventing dopamine from diffusing to the probe and contributing to Cout. So, uptake decreases recovery by decreasing Cout. Uptake has the same effect on the exogenous dopamine that is delivered via the probe inlet line during a measurement of the extraction fraction. Uptake removes this dopamine from the extracellular space and prevents it from diffusing back to the probe to contribute to Cout. So, uptake increases extraction by decreasing Cout. Thus, the same mechanism underpins the opposite effect on uptake recovery and extraction. It is the effect of uptake on Cout that is symmetric, rather than its effect on E and R.

102

VII. A comment regarding in vivo dopamine recovery at zero perfusion velocity Earlier, we mentioned that the value of Cout obtained by extrapolation to zero perfusion velocity agrees well the CNNF obtained from x-intercept of the concentration differences plot (see Watson et al., 2006). This agreement has been used to support the validity of quantitative microdialysis on the grounds of the assumptions that R approaches a value of 1 at zero perfusion velocity, so that Cout should then equal Cext,N. However, the assumption that R approaches one is not valid in the case of the in vivo microdialysis of dopamine. The ability of the dopamine uptake mechanism to prevent recovery of dopamine by the probe does not suddenly disappear at zero perfusion velocity. When the probe is operated at zero or even very slow perfusion velocity, the ability of dopamine to diffuse to and be recovered by the probe continues to be limited by the ability of the transporter to remove dopamine from the extracellular space before it reaches the probe. For this reason, recovery cannot approach one just because the perfusion velocity is slow. Although recovery approaches one at slow perfusion rates during in vitro microdialysis in a beaker, this same behavior cannot be assumed during an in vivo experiment. A bit of mathematical work is helpful to reinforce the idea that uptake prevents 100% recovery at zero perfusion velocity. Mathematical models of the diffusion processes involved in microdialysis usually divide the space near the probe into zones that describe the probe, the space adjacent to the probe that is disrupted, and the non-disrupted surroundings (e.g., Bungay et al., 2003; Chen, 2005). At zero perfusion velocity, once the system reaches steady state, there are no gradients in the direction axial to the probe, so it is only necessary here to consider the radial direction. At zero perfusion velocity, diffusion is the only process that affects the concentration of substances in the probe (there is no convection because the perfusion ﬂuid is stationary). Diffusion in the probe can be described by Fick’s second law (Crank, 1975): @C @2 C ¼D 2 @t @x

(7)

where D is the diffusion coefﬁcient, C is concentration, t is time, and x is distance. In the disrupted layer, the concentration gradient of the diffusing substance is also affected by a clearance reaction, which is described by Fick’s law with an additional term to describe uptake: @C @2 C ¼ D 2 kC @t @x

(8)

where k is a ﬁrst-order rate constant. In the surrounding normal layer, the concentration gradient of the diffusing substance is affected by both a clearance reaction and a generation term: @C @2 C ¼ D 2 kC þ G @t @x

(9)

where G is a steady state generation rate. Eqs. (7–9) are not intended as a rigorous model of microdialysis but rather a simpliﬁed model that illustrates the effect of clearance and generation terms on diffusion gradients. Fig. 2 shows steady-state concentration proﬁles calculated by solving Eqs. (7–9) with several values of the rate constant, k. In the top panel of Fig. 2, the zones corresponding to the probe (a) and disrupted layer (b) are arbitrarily given the same dimension, l. The four concentration proﬁles were calculated with different values of the dimensionless quantity l 2 k=D; with the values listed in the ﬁgure itself. According to this calculation, the Cout value is very nearly zero with the largest clearance rate constant used and increases as the rate constant decreases. This behavior of the steady-state concentration proﬁle is consistent with our earlier discussion pointing out that a clearance process in the external medium can prevent the attainment of 100% recovery at zero perfusion velocity. Note that the concentration proﬁle exhibits no-net-ﬂux behavior (zero slope) only within the probe zone (a), where diffusion is the only process that affects the concentration proﬁle. This reinforces the idea that CNNF is not equal to Cext,N under the conditions of this model. The bottom panel of Fig. 2 repeats the same calculation after doubling the thickness of the disrupted zone to 2 l. The effect of increasing the thickness of the disrupted zone is to decrease the predicted value of CNNF at each value of the clearance rate constant, which is consistent with the idea that the ability of

103 ∂C ∂ 2C =D 2 ∂t ∂x

VIII. How important is the difference between E and R?

∂C ∂ 2C = D 2 _ kC ∂t ∂x ∂ C ∂C = D 2 _ kC + G ∂t ∂x 2

concentration, C/Cext,inf

1.0

0.00001 0.8 0.6

0.0001

0.4

a

0.2

b

c

0.001 0.01

0.0 0

2

4

6

8

6

8

distance, x/l concentration, C/Cext,inf

1.0

a 0.8

b 0.00001

0.6

0.0001

0.4

c

0.2

0.001 0.01

0.0 0

2

4

distance, x/l Fig. 2. Steady-state solutions to Eqs. (7–9) of the text. Each plot is the solution obtained with a different value for the parameter l 2 k=D; where l is the thickness of the probe zone (a), k is the clearance rate constant, and D is the diffusion coefﬁcient. In the top panel, the probe zone (a) and disrupted zone (b) have the same thickness. In the bottom panel, the disrupted zone is twice as thick as the probe zone. In both cases, the normal zone (c) is inﬁnitely large but only a portion is included in the plots. Note that the concentration proﬁles are ﬂat in zone (a), which is where the no-net-ﬂux condition exists. Even though these are steady-state proﬁles, concentration gradients exist in zones (b) and (c) due the impact of the generation and clearance processes, which act as a source and sink, respectively, of the diffusing substance.

the substance to reach the probe is limited by clearance in the disrupted zone even when the perfusion velocity is zero: the thicker the disrupted zone the greater the limitation on recovery. The point of the comparison between the two panels is that the geometry of the disrupted zone must be known to employ diffusion models in any effort to quantitatively predict values of E and R.

According to Eqs. (5) and (6), microdialysis would underestimate the extracellular concentration of dopamine if the value of R is smaller than the value of E. The magnitude of this underestimation depends on the relative values of E and R. Several authors have used diffusion models in attempts to quantitatively predict in vivo values of E and R (Morrison et al., 1991; Peters and Michael, 1998; Bungay et al., 2003; Chen, 2005). Chen used such a model to suggest that dopamine uptake should not have opposite effects on the in vivo dopamine E and R values, as we have suggested here and in several publications. However, Chen used a rather small value of 20 mm as the thickness of the disrupted zone in his model, whereas our results from voltammetry in the tissue surrounding microdialysis probes suggest the actual value is at least 10-fold larger. Moreover, Chen’s model did not incorporate the idea that a gradient in dopamine release and uptake activity might exist in the disrupted tissue layer, the existence of which is also revealed by our voltammetry ﬁndings. Thus, whereas Chen concludes that CNNF values of dopamine only slightly underestimate the actual in vivo Cext,N value, we suspect that CNNF greatly underestimates Cext,N. Answering the question of exactly by how much CNNF, or Cout of conventional microdialysis, underestimates the actual in vivo value of Cext,N for dopamine demands a primary standard for probe calibration, which does not exist. However, we do have the ability to compare responses recorded at voltammetric electrodes next to microdialysis probes and base estimates on the relative change in the amplitude of the stimulus responses caused by probe implantation. According to postcalibration of the microelectrodes, the amplitude of a typical stimulus response, with the stimulation parameters we have routinely employed (35 s, 45 Hz, biphasic, constant current pulses with a width of 2 ms and amplitude of 250 mA), corresponds to dopamine concentrations in the vicinity of 10 mM. After probe implantation that amplitude falls below our detection limit, which in the worstcase is 0.5 mM and more typically 0.1 mM, at

104

microelectrodes mounted onto the outer surface of microdialysis probes. Hence, conservatively, there is a 50-fold loss in evoked dopamine release in the immediate vicinity of the probe. With a 50-fold loss in dopamine release adjacent to the probe, the in vivo recovery value for dopamine could not exceed 0.02, or 2%. However, there is reason to suspect that the recovery value may be even lower, considering that CE-EC analysis of dialysate samples collected during MFB stimulation revealed that the evoked change in dialysate dopamine concentration, without inhibition of dopamine uptake, was about 5 nM (Qian et al., 1999). This value suggests a dopamine recovery value in the vicinity of 5 nM/10 mM ¼ 0.0005, or 0.05%, in which case microdialysis results significantly underestimate extracellular dopamine values. The majority of our prior work with voltammetry has involved the recording of dopamine changes evoked by electrical stimulation of MFB, which raises the question of whether the ﬁndings are relevant to measurements of dopamine in the absence of electrical stimulation. One reason for thinking that the stimulation results are relevant to non-stimulated results derives from the short lifetime of dopamine in the extracellular space, especially in the striatum where DAT expression is high. Estimates place the half-life of dopamine in the striatal extracellular space at about 70 ms (Wu et al., 2003), which is far shorter than stimulus duration of 25 s we have employed during voltammetric measurements near microdialysis probes. The recovery value is controlled by the ability of dopamine to diffuse to the probe after it is released into the extracellular space. Since dopamine’s halflife is only 70 ms, dopamine normally only diffuses for about 210 ms, that is, three half-lives, which is the time required for 95% of released dopamine molecules to be cleared by the transporter. Because the stimulus we use lasts so much longer than the lifetime of dopamine in the extracellular space, we believe that the recovery of dopamine observed during electrical stimulation is a good indicator of the recovery of non-evoked dopamine. Some voltammetric data on non-evoked dopamine is now available in the literature. The ﬁrst example provides evidence in support of our idea that inhibition of dopamine uptake increases the in

vivo recovery of dopamine. When we performed voltammetric recording with a microelectrode placed about 200 mm from a microdialysis probe, we observed a large amplitude, long lasting increase in extracellular dopamine levels in response to the systemic administration of the dopamine uptake inhibitor, nomifensine (Borland et al., 2005). In the absence of a microdialysis probe we normally see no increase in dopamine after nomifensine administration, only a small voltammetric response that is attributed to a change in the pH of the extracellular space. We conclude from this ﬁnding that the long-known increase in dialysate dopamine levels after inhibition of dopamine uptake is a consequence of a change in the dopamine concentration in the disrupted tissue near the probe, which exhibits a different response to uptake inhibition than the surrounding normal tissue. We conclude that the increase in dialysate dopamine levels after uptake inhibition is due to an increase in the dopamine recovery that results from the atypical drug response of the disrupted tissue close to the probe. Overall, we conclude that in the in vivo recovery of dopamine is very small due to the preferential disruption of dopamine release in the vicinity of the microdialysis probe. The small recovery implies that dialysate dopamine levels, both Cout from conventional microdialysis and CNNF from the concentration differences plot, underestimates the actual in vivo dopamine concentration. To conﬁrm whether or not this is the case that requires independent knowledge of the extracellular concentration of dopamine. At present, voltammetry is the main alternative to microdialysis for measuring extracellular dopamine concentrations but voltammetry has the reputation for being best suited to monitoring rapid changes in extracellular dopamine concentrations and not so well suited to monitoring the spatiotemporal average concentration, which would be the most appropriate for comparison to microdialysis basal levels. The available voltammetry literature does not reach a clear consensus on the average extracellular dopamine concentration. Gonon and coworkers reported a value of 25 nM using a differential pulse voltammetry protocol in pargyline-pretreated rats (Gonon and Buda, 1985). This value is ﬁvefold

105

higher than typical values of CNNF, which might seem like a reasonable agreement considering the differences in experimental design. However, using fast scan cyclic voltammetry, we reported that the intrastriatal infusion of a glutamate antagonist decreased extracellular dopamine concentration by ca. 2 mM from its resting value (Borland and Michael, 2004). At present, neither voltammetric estimate of the basal dopamine concentration can be validated by in vivo calibration, so ﬁnal resolution of this issue is not possible. However, a resting concentration of dopamine as high as 2 mM would imply an in vivo dopamine recovery value of 5 nM/2 mM ¼ 0.0025, which is in accordance with the value deduced from measurements of stimulated dopamine release. We are not the ﬁrst to suggest that microdialysis leads to underestimation of extracellular dopamine concentrations. Early papers considering the impact of uptake on recovery reached the very same conclusion. By accounting for the impact of uptake on recovery, those authors concluded that extracellular dopamine levels were about 1.6 mM (Lindefors et al., 1989; Amberg and Lindefors, 1989), even though the measured dialysate dopamine concentrations were in the nanomolar range, which is in excellent agreement with the 2 mM value we have reported.

IX. What do microdialysis results tell us? In general, microdialysis is adopted as a tool to sample the extracellular space of the brain to gain insight into the neurochemical underpinnings of behaviors, responses to drugs, and disease pathology. However, when the probe is implanted, the tissue adjacent to the probe is disrupted. The traumatic injury created during probe implantation has been unequivocally revealed by histology (e.g., Clapp-Lilly et al., 1999). The disruption of dopaminergic kinetics attendant to the traumatic injury has been specifically revealed by voltammetric recordings within the implantation injury (Yang et al., 1998; Borland et al., 2005). Thus, a new concept in the quantitative aspects of microdialysis is that the dialysate concentration of target substances reﬂects a weighted sum of the extracellular

concentrations in the injured tissue and in the surrounding normal tissue (Bungay et al., 2003). This weighting strategy is probably unrealistic, however, because it is based on the idea that there are two distinct tissue zones to consider, the disrupted zone and the normal surrounding zone, each with their own distinct properties. Voltammetric measurements near microdialysis probes suggest instead that the disruption to the dopamine system creates a gradient in activity, not two distinct regions. At present, no one has considered how to mathematically account for this gradient. Furthermore, it may not be necessary to account for the gradient because our results suggest that the dialysate concentration of dopamine mainly reﬂect a single point on the gradients, that is, the concentration where the gradient intersects the outer surface of the probe. During MFB stimulation, voltammetric responses recorded in the probe outlet line correlated best with those recorded with electrodes mounted onto the outer surface of the microdialysis probe (Lu et al., 1998). The same is true for dopamine as measured by CE-EC (Qian et al., 1999). Theoretically, we have suggested that CNNF is the concentration at the outer surface of the probe under no-net-ﬂux conditions (Peters et al., 1998), further suggesting that microdialysis predominantly reports dopamine in the disrupted tissue immediately adjacent to the probe. The contributions to dialysate dopamine concentrations from the surrounding normal tissue are not apparent.

X. Concluding remarks Voltammetric measurements have revealed disruption of dopaminergic kinetics in the tissues adjacent to implanted microdialysis probes. The disruption has differential effects on the kinetics of dopamine release and uptake, severely compromising dopamine release but leaving dopamine uptake relatively intact. Thus, microdialysis mainly measures dopamine released into the extracellular space at some distance from the probe. This results in very low dopamine recovery because the vast majority of dopamine is removed from the extracellular space by the dopamine

106

uptake process before ﬁnding its way to the probe. This has two significant consequences for the quantitative relationship between dialysate and extracellular dopamine concentrations. The ﬁrst is that dialysate concentrations inherently underestimate extracellular concentrations of dopamine. The second is that the quantitative performance of the microdialysis sampling changes in the face of anything that might affect the kinetics of the transporter. This certainly includes drugs that target the transporter. However, the transporter is regulated in a multitude of ways (Gulley and Zahniser, 2003), meaning there are many ways in which the quantitative performance of the probe might be altered. For this reason, the quantitative performance of the probe might change during an experiment and this needs to be factored into the interpretation of microdialysis results. Nonetheless, this has never been done. Finally, we point out that our studies have focused on the subject of dopamine microdialysis because of the availability of voltammetry as a second tool for in vivo dopamine measurements. Nonetheless, the issue of how the microdialysis probe affects the kinetics of processes that regulate the concentration gradients of other target substances also deserves consideration and analysis. Acknowledgments Several individuals have contributed significantly to the efforts in the author’s laboratory aimed at understanding the in vivo microdialysis of dopamine. The author is indebted to Yi Lu, Jennifer Peters, Hua Yang, Cassandra Allen, and Laura Borland. Our experiments on voltammetry near microdialysis probes have been supported by grants from National Institutes of Health: NS 31442, MH 63122, DA 13661. References Amberg, G. and Lindefors, N. (1989) Intracranial microdialysis II: mathematical studies of diffusion kinetics. J. Pharmacol. Methods, 22: 157–183. Berg, H.C. (1993) Random Walks in Biology. Princeton University Press, Princeton, NJ.

Borland, L.M. and Michael, A.C. (2004) Voltammetric study of the control of striatal dopamine release by glutamate. J. Neurochem., 91: 220–229. Borland, L.M., Shi, G., Yang, H. and Michael, A.C. (2005) Voltammetric study of extracellular dopamine near microdialysis probes acutely implanted in the striatum of the anesthetized rat. J. Neurosci. Methods, 146: 149–158. Bungay, P.M., Newton-Vinson, P., Isele, W., Garris, P.A. and Justice, J.B. Jr. (2003) Microdialysis of dopamine interpreted with quantitative model incorporating probe implantation trauma. J. Neurochem., 86: 932–946. Chen, K.C. (2005) Evidence on extracellular dopamine level in rat striatum: implication for the validity of quantitative microdialysis. J. Neurochem., 92: 46–58. Clapp-Lilly, K.L., Roberts, R.C., Duffy, L.K., Irons, K.P., Hu, Y. and Drew, K.L. (1999) An ultrastructural analysis of tissue surrounding a microdialysis probe. J. Neurosci. Methods, 90: 129–142. Crank, J. (1975) The Mathematics of Diffusion. 2nd ed. Clarendon Press, Oxford, UK. Engstrom, R.C., Wightman, R.M. and Kristensen, E.W. (1988) Diffusional distortion in the monitoring of dynamic events. Anal. Chem., 60: 652–656. Gonon, F.G. and Buda, M.J. (1985) Regulation of dopamine release by impulse ﬂow and by autoreceptors as studied by in vivo voltammetry in the rat striatum. Neuroscience, 14: 765–774. Gulley, J.M. and Zahniser, N.R. (2003) Rapid regulation of dopamine transporter function by substrates, blockers and presynaptic ligands. Eur. J. Pharmacol., 479: 139–152. Khramov, A.N. and Stenken, J.A. (1999) Enhanced microdialysis extraction efﬁciency of ibuprofen in vitro by facilitated transport with beta-cyclodextrin. Anal. Chem., 71: 1257–1264. Lindefors, N., Amberg, G. and Ungerstedt, U. (1989) Intracerebral microdialysis: I. Experimental studies of diffusion kinetics. J. Pharmacol. Methods, 22: 141–156. Lonnroth, M.T., Jansson, P.A. and Smith, U. (1987) A microdialysis method allowing characterization of intercellular water space in humans. Am. J. Physiol., 253: E228–E231. Lu, Y., Peters, J.L. and Michael, A.C. (1998) Direct comparison of the response of voltammetry and microdialysis to electrically evoked release of striatal dopamine. J. Neurochem., 70: 584–593. Manley, L.D., Kuczenski, R., Segal, D.S., Young, S.J. and Groves, P.M. (1992) Effects of frequency and pattern of medial forebrain bundle stimulation on caudate dialysate dopamine and serotonin. J. Neurochem., 58: 1491–1498. Morrison, P.F., Bungay, P.M., Hsiao, J.K., Mefford, I.N., Dykstra, K.H. and Dedrik, R.L. (1991) Quantitative microdialysis. In: Robinson, T.E. and Justice, J.B. (Eds.), Micodialysis in the Neurosciences. Elsevier, Amsterdam. Qian, J., Wu, Y., Yang, H. and Michael, A.C. (1999) An integrated decoupler for capillary electrophoresis with electrochemical detection: application to analysis of brain microdialysate. Anal. Chem., 71: 4486–4492.

107 Parsons, L.H. and Justice, J.B. Jr. (1994) Quantitative approaches to in vivo microdialysis. Crit. Rev. Neurobiol., 8: 189–220. Peters, J.L. and Michael, A.C. (1998) Modeling voltammetry and microdialysis of striatal extracellular dopamine: the impact of dopamine uptake on extraction and recovery ratios. J. Neurochem., 70: 594–603. Peters, J.L., Miner, L.H., Michael, A.C. and Sesack, S.R. (2004) Ultrastructure at carbon ﬁber microelectrode implantation sites after acute voltammetric measurements in the striatum of anesthetized rats. J. Neurosci. Methods, 137: 9–23. Retterer, S.T., Smith, K.L., Bjornsson, C.S., Neeves, K.B., Spence, A.J.H., Turner, J.N., Shain, W. and Isaacson, M.S. (2004) Model neural prostheses with integrates microﬂuidics: a potential intervention strategy for controlling reactive cell and tissue responses. IEEE Trans. Biomed. Eng., 51: 2063–2073. Smith, A.D. and Justice, J.B. Jr. (1994) The effect of inhibition of synthesis, release, metabolism and uptake on the microdialysis extraction fraction of dopamine. J. Neurosci. Methods, 54: 75–82. Song, Y. and Lunte, C.E. (1999) Comparison of calibration by delivery versus no net ﬂux for quantitative in vivo microdialysis sampling. Anal. Chim. Acta,, 379: 251–262.

Szarowski, D.H., Anderson, M.D., Retterer, S., Spence, A.J., Isaacson, M., Craighead, H.G., Turner, J.N. and Shain, W. (2003) Brain responses to mico-machined silicon devices. Brain Res., 983: 23–35. Tepper, J.M., Creese, I. and Schwatz, D.H. (1991) Stimulusevoked changes in neostriatal dopamine levels in awake and anesthetized rats as measured by microdialysis. Brain Res., 559: 283–292. Watson, C.J., Venton, B.J. and Kennedy, R.T. (2006) In vivo measurement of neurotransmitters by microdialysis sampling. Anal. Chem., 78: 1391–1399. Wu, Q., Reith, M.E.A., Wightman, R.M., Kawagoe, K.T. and Garris, P.A. (2001) Determination of release and uptake parameters from electrically evoked dopamine dynamics measured by real-time voltammetry. J. Neurosci. Methods, 112: 119–133. Yang, H., Peters, J.L., Allen, C., Chern, S., Coalson, R.D. and Michael, A.C. (2000) A theoretical description of microdialysis with mass transport coupled to chemical events. Anal. Chem., 72: 2042–2049. Yang, H., Peters, J.L. and Michael, A.C. (1998) Coupled effects of mass transfer and uptake kinetics on in vivo microdialysis of dopamine. J. Neurochem., 71: 684–692.

CHAPTER 2.1

New methodological aspects of microdialysis Jan Kehr Department of Physiology and Pharmacology, Karolinska Institutet, Stockholm, Sweden

Abstract: This chapter aims to give an overview of the latest advancements of microdialysis with special emphasis on the implications of microsystems and nanotechnology and suggests possible future development. Microdialysis offers a unique opportunity to monitor chemistry of the brain microenvironment and this feature, often, is overwhelming the main limitation of microdialysis being an invasive technique. For further advancement of microdialysis technology, probably the most interesting efforts are those aiming to integrate microdialysis sampling with biosensors, micrototal analysis systems (mTAS), mass spectrometry, and imaging techniques. This progress is characterized by implementing microelectromechanical systems (MEMS) and nanotechnology, including fabrication of biosensors and microﬂuidic devices and applications of functionalized nanoparticles. There is a great potential to apply MEMS and biosensor technologies to automatize and further improve the speed and feasibility of microdialysis for bed-site monitoring. In experimental neuropharmacology, thousands of published papers provide a solid ground to validate microdialysis as a key technique in pharmacological neurochemistry in vivo and as a major tool for studies of neurochemical basis of behavior. The ‘‘peptidomic’’ approach, using capillary liquid chromatography– tandem mass spectrometry and microdialysis, enables identiﬁcation of new peptides and their fragments and search for putative endogenous ligands to orphan receptors. Further development of faster and more sensitive detection techniques conjugated to microdialysis is expected to strengthen the use of microdialysis in functional pharmacology and enable searching for novel biomarkers of CNS diseases. shown a strong promise for clinical use as a neuromonitoring technique in intensive care (for review, see Ungerstedt 1991; Ungerstedt and Rostami, 2004; Hillered et al., 2005; Nordstro¨m, 2005). For pharmaceutical companies active in the CNS therapeutic area, microdialysis undoubtedly provides the crucial data on mechanisms of action and functional validity of candidate drugs. This chapter aims to give an overview of the latest advancements of microdialysis with special emphasis on the implications of microsystems and nanotechnology and suggestions for possible future development.

I. Introduction Microdialysis has become a well-established technique for in vivo studies of the chemistry of the central nervous system (CNS) and, in particular, for monitoring chemical neurotransmission and energy metabolism following various physiological, pathological, and/or pharmacological stimuli. Microdialysis offers a unique opportunity to study the mechanisms of drug actions in experimental neuropharmacology (for review, see Ungerstedt et al., 1982; Ungerstedt, 1984; Hamberger et al., 1985; Robinson and Justice, 1991; Sharp and Zetterstro¨m, 1992; Kehr, 1998, 1999; Kehr and Yoshitake, 2006) and recently, microdialysis has

I.A. Brain microenvironment – an important communication channel The extracellular space (ECS) of the brain plays an important role as a common communication

Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

111

DOI: 10.1016/S1569-7339(06)16007-5 Copyright 2007 Elsevier B.V. All rights reserved

112

channel for chemical signaling between (1) neurons to neurons, (2) neurons and glia cells, as well as for (3) neurovascular communication. The ECS compartment under normal conditions comprises 20% of the total brain volume (for review, see Nicholson and Sykova, 1998; Nicholson, 2005). A physical model describing the diffusion of molecules or ions in the microenvironment of the brain ECS was derived by Nicholson (1979) and Nicholson and Phillips (1981). In a simpliﬁed model, the porosity of the brain microenvironment resembles the liquid phase of soap foam and the movement of molecules in this phase follows the classic Fick’s equations for diffusion (Nicholson, 2005). The diffusion of molecules in the brain extracellular ﬂuid (ECF) is driven by the concentration gradients existing between the delivery (e.g., release and active transport) sites and the clearance (e.g., uptake, metabolism, and transport) sites for a particular molecule and governs the propagation of chemical signals within the brain microenvironment. The diffusion model serves as a basis for further mathematical characterization of quantitative aspects of molecular diffusion and monitoring techniques implementing microdialysis probes (Bungay et al., 1990; Chen et al., 2002) and implantable sensors such as voltammetric electrodes (May et al., 1988), ion-selective microelectrodes (Nicholson, 1993; Nicholson and Sykova, 1998), and optodes (Mayevsky and Chance, 1973). The most recent models for quantitative microdialysis also consider factors associated with tissue trauma caused by the implantation of the microdialysis probe and are detailed by Chen, Michael, and Bungay in other chapters of this book. I.B. Volume transmission – concept verification by microdialysis sampling The concept of volume transmission (VT) was derived on the basis of immunocytochemical data, at both light and electron microscopic levels, which revealed the existence of neuroanatomically substantial distances between the release sites of neurotransmitters and their respective post-synaptic receptors or transporter proteins (for review, see Fuxe and Agnati, 1991; Fuxe et al., 2000). The phenomenon of VT in the brain is characterized by

a long-distance diffusion of neuroactive molecules, extra-synaptic release, and the release of transmitters such as monoamines or acetylcholine, and neuromodulators such as neuropeptides or adenosine from non-junctional varicosities, which was conﬁrmed by several investigators (Fuxe and Agnati, 1991;Vizi and Kiss, 1998; Zoli et al., 1999; Descarries and Mechawar, 2000). A number of morphologic and physiological techniques have shown that many neurotransmitters, traditionally believed to act only within the limited space of synaptic junctions, indeed can diffuse over distances exceeding the volume of the synaptic cleft. In addition, electron microscopic studies have shown that the transporter proteins involved in the reuptake of released neurotransmitters from the ECF are often rather distal from the release sites (Beaudet and Descarries, 1978). Further support for the VT hypothesis was provided by neurotransplantation techniques in lesioned rat brains demonstrating that neurotransmitters such as dopamine (DA) can diffuse over long distances out from the grafted tissue (Stro¨mberg et al., 1984). Temperature gradients caused by neuronal activity (Rivera et al., 2006), as well as a slow convective movement of the ECF toward the perivascular Virchow–Robin space, ventricles, and cortical subarachnoid space due to pulsatile arterial expansions causing the movement of the brain in a piston-like fashion (Greitz et al., 1993), were some additional factors suggested to facilitate VT. Several lines of evidence exist suggesting that even CSF can function as a conduit for transporting signaling molecules involved in the regulation of physiological functions such as circadian rhythms (Silver et al., 1996). A more recent approach to study VT makes use of dual-probe microdialysis, which allows continuous infusion of a molecular label via one probe and sampling the molecules diffusing through the brain microenvironment by a second probe implanted at a short distance (e.g., 1 mm) from the ﬁrst probe (for review, see Kehr et al., 2000). A major advantage of using microdialysis to study VT is the possibility to elucidate the mechanisms regulating the size and the shape of the ECS in health and disease, and the role of extracellular molecular transport for design of novel drug therapies. Thus, dual-probe microdialysis was used to

113

study the diffusion of tritiated DA and its metabolites, as well as of the ECS marker tritiated mannitol (Ho¨istad et al., 2000; Kehr et al., 2000). The corresponding mathematical model for dualprobe microdialysis was derived by Chen (Ho¨istad

Fig. 1. Applications of dual-probe microdialysis for studies of long-distance diffusion. Diffusion proﬁles of 3H-DA and its tritiated metabolites in the brains of (A) control and (B) 6-OHDAlesioned rats following 5 h of continuous infusion of 3H-DA through the microdialysis probe implanted into the rat striata. The total tritium content was visualized in 10-mm brain sections with the phospho-imager after 5 days of exposure. The second microdialysis probe, not visible in the ﬁgures, was implanted at 1mm distance from the infusion probe and used for collection of diffusing tritiated labels. Following HPLC separation of DA and its metabolites, the radioactivity of the corresponding fractions was measured by a liquid scintillation counter. The lower panel shows the time courses of diffusing 3H-DA and labeled metabolites 3H-DOPAC and 3H-HVA formed in situ in (A) control and (B) 6-OHDA-lesioned rats. With permission from Blackwell Publishing, 2006 (Hoistad et al. 2000).

et al., 2002; chapter by Chen in this book). The example in Fig. 1 illustrates the feasibility of using the dual-probe approach to evaluate long-distance (1 mm) diffusion of tritiated DA and its acidic metabolites DOPAC and HVA in the striatum of control and 6-OHDA-lesioned rats. At steady state, the levels of recovered 3H-DA were only slightly higher over the background threshold, whereas the basal levels of endogenous (cold) DA in control animals were easily detectable with a mean value of 6.2 nM. 3H-DA radioactivity did not increase in the lesioned rats. Surprisingly, the overall recoveries of labeled metabolites DOPAC and HVA formed in situ from 3H-DA were signiﬁcantly reduced in 6-OHDA-lesioned rats. These results were supported by histological analysis (using a phospho-imager) showing strong attenuation of diffusion proﬁles of tritiated DA and its metabolites at the infusion site following a 5-h infusion of 3H-DA in lesioned rats. These data suggest that the dual-probe technique can be used to study molecular transport within the brain microenvironment in neuropathological models of brain diseases or following various pharmacological stimuli. Another application of dual-probe microdialysis for diffusion studies of drugs in the same anatomical region was developed by Westerink and De Vries (2001). Here, the drug affecting the release of DA was infused via the ﬁrst probe, whereas released DA was collected by the second probe implanted at 1 mm distance from the infusion probe. Interestingly, no effects were seen at the second probe when high potassium chloride was infused, suggesting that diffusion of both the potassium ions and the DA occurs only at short distances, as shown earlier by in vivo voltammetry monitoring. The study showed that most of the compounds affecting DA release had rather slow infusion rates, indicating that relatively high infusion concentrations (1–10 mM) were required to reach substantial brain concentrations at a distance of 1 mm from the infusion site. In conclusion, the fact that microdialysis technique as such can generate valuable information on release, reuptake, and metabolism of neurotransmitters under both tonic and stimulatory or inhibitory conditions provides further experimental evidence for the existence of VT.

114

I.C. Microdialysis in relation to other in vivo techniques Current technologies for monitoring molecules or signals reﬂecting functional changes in brain chemistry can be divided into two main groups including minimally invasive techniques and neuroimaging techniques, as depicted in Fig. 2. Minimally invasive techniques require stereotaxic implantation of a sensing device directly into the target brain area, thereby causing tissue trauma. Depending on the physical detection principle, the intracorporeal sensors can be divided into potentiometric, amperometric, and optical sensors. Thus, the potentiometric sensors measure directly the concentrations of endogenous ions such as potassium or calcium, voltammetric microelectrodes detect electrochemically active molecules such as DA, and optodes measure interactions of externally introduced light with brain tissue (reﬂectance, absorbance, and ﬂuorescence) or by (bio)chemical means (luminescence). Biosensors are electrodes or optodes utilizing speciﬁc biochemical (enzymatic) or immunochemical (antibodies) reactions occurring on the sensor surface, thereby transforming the chemical energy of these analyte-speciﬁc reactions into physical signals (electrons and photons), which are easily detectable and recorded by external instrumental devices. In this respect, microdialysis can be considered a universal biosensor system where the microdialysis probe represents a central, analyte sampling part of the biosensor assembly, and the allied analytical technique (whether in an off-line or an on-line mode) provides identiﬁcation and quantiﬁcation of the analyte of interest. The response time can be as low as 0.1–1 s for directly detecting biosensors, whereas a relatively laborious and time-consuming analysis of neurotransmitters and other molecules sampled by microdialysis requires the fractions to be collected in 1–60-min intervals. To combine the speed and feasibility of biosensors with microdialysis, which is slower but a more universal sampling technique, several interesting attempts have been made to integrate the biosensor and the microdialysis probe into a single device which will be discussed below. Non-invasive neuroimaging techniques such as magnetic resonance imaging (MRI), single photon

emission computed tomography (SPECT), and positron emission tomography (PET) are well established in clinical radiology for diagnosis and three-dimensional localization of tumors, infarcts, trauma, or similar pathophysiological conditions. There is a growing interest to use functional MRI (fMRI) to measure the changes in blood ﬂow or oxygenation levels as a correlate to neuronal activity in patients with neurological and psychiatric disorders (for review, see Walker and Walker, 2005; Mitterschiffthaler et al., 2006; Norris, 2006). This strategy has become a feasible alternative to radioligand-based PET and SPECT. In general, radiological techniques suffer from high running costs, poor spatial resolution, short lifetime, and complicated preparation of the isotope-labeled tracers. However, the current development of research instruments with a higher spatial resolution makes it possible to also use neuroimaging technologies for studies on animal models of mental and neurodegenerative diseases. Another promising in vivo imaging technology is near-infrared spectroscopy (NIRS). NIRS can operate either in a simple absorption mode, which allows the monitoring of blood oxygenation (oxyhemoglobin) levels (Jobsis, 1977; Ferrari et al., 2004) or in the ﬂuorescence mode. In the latter case, the NIR multiphoton laser-scanning microscopy can be used for detection of systemically administered molecular disease state markers (e.g., speciﬁc antibodies) linked to the NIR ﬂuorescence dies, which emit within the spectral region of 700–900 nm (Sevick-Muraca et al., 2002). So far, this technology has been almost exclusively applied for imaging subcutaneously inoculated tumors in experimental animals. The major limitation for imaging deeper body organs is a rather poor penetration of NIR light through the soft tissues, and in the case of brain tissue, there is an additional strong absorbing barrier represented by the skull bone. A possible solution could be the optode-based approach by implanting an optical ﬁber directly into the brain parenchyma as proposed by Crespi et al. (2005). For further advancement of microdialysis technology, probably the most interesting effort is the one aiming to integrate microdialysis sampling with biosensors, micrototal analysis systems

115

Fig. 2. A scheme illustrating possible implementations of microsystems and nanotechnology in development of in vivo techniques for monitoring molecules in the neuroscience. The monitoring techniques can be divided into two main groups including minimally invasive techniques and neuroimaging techniques. Minimally invasive techniques require stereotaxic implantation of a sensing device directly into the target brain area, thereby causing tissue trauma. Non-invasive neuroimaging techniques include magnetic resonance imaging (MRI), positron emission tomography (PET), single photon emission computed tomography (SPECT), and optical imaging in the near-infrared (NIR) region or using ﬂuorescent (FL) dyes. The advancements of microdialysis sampling are characterized by implementation of sophisticated analytical techniques such as liquid chromatography/mass spectrometry (LC/MS) or nuclear magnetic resonance (NMR) spectroscopy on one hand, and development of simpliﬁed integrated microdialysis systems with biosensors and online read-out devices for bed-site monitoring on the other hand. Both these approaches strongly rely on the progress in miniaturization of liquid sampling and handling protocols, particularly those implementing microelectromechanical systems (MEMS) for fabrication of sensors, microﬂuidic devices, and interfaces.

(mTAS), or imaging techniques. This progress is characterized by implementing nanotechnology, particularly nanoparticles and microelectromechanical systems (MEMS) including sensor and microﬂuidic devices as schematically illustrated in Fig. 2 and discussed in the following paragraphs.

II. Advancements in microdialysis sampling II.A. Pioneering studies – emphasizing the importance of small membrane size and sensitive detection techniques The very ﬁrst attempt to use microdialysis principle for in vivo sampling of molecules diffusing in

the brain microenvironment of the Mongrel dogs was developed by Bito et al. (1966). The dogs were implanted with sterile dialysis sacs (length 8–12 mm and ﬂat width 9 mm) ﬁlled with 6% dextran in saline into the cortex and subcutaneously in the neck. Ten weeks later, the content of the sacs was analyzed for amino acids and ions and compared with the concentrations in plasma and CSF. For most of the amino acids, Bito et al. (1966) have found existing concentration gradients increased in the order: CSFoECFoblood plasma, which led to the conclusion that the brain sac ﬂuid was not a dialysate of either blood or CSF, thereby representing a third ﬂuid compartment. These data supported the idea of a carrier-mediated transport system for amino acids. Further development of

116

microdialysis sampling allowing for a continuous perfusion of dialysis bags and measuring timely changes of recovered compounds was described by Delgado et al. (1972). These authors constructed a brain ‘‘dialytrode’’ for monkeys, which was basically a push–pull cannula with a small (5 1 mm) polysulfone membrane bag glued on its tip. Though the dialytrode principle was tested in some conceptually correctly designed experiments, derived from earlier push–pull ‘‘chemitrode’’ experiments (Roth et al., 1969), it was not successful. For example, dialytrodes failed to detect the presence of radioactive DA in the perfusates following pre-labeling with 14C L-DOPA. It was Ungerstedt and Pycock (1974) who ﬁrst succeeded in providing a functional proof of concept for brain dialysis using a probe constructed with a thin hollow ﬁber dialysis tube and measuring amphetamine-induced release of DA-like radioactivity after pre-labeling the rat striatum with tritiated DA (Fig. 3). This simple experiment, mimicking the protocols developed for radio-labeled pre-loading of brain slices in vitro, opened exciting new possibilities for bioanalytical sampling and monitoring of local chemistry in the organs of the body body. The introduction of highly sensitive analytical techniques for the detection of catecholamines and indolamines by column LC with electrochemical detection (Kissinger et al., 1973) and amino acid derivatives with ﬂuorescence detection (Lindroth

Fig. 3. The ﬁrst successful proof of concept for brain dialysis as a technique capable to follow brain neurochemistry in vivo was provided by Ungerstedt and Pycock (1974). The authors implanted a microdialysis probe constructed with a thin hollow ﬁber dialysis tube into the rat striatum and measured amphetamine-induced release of DA-like radioactivity following prelabeling the striatum with tritiated DA.

and Mopper, 1979) accelerated the development and applications of microdialysis in both pre-clinical and clinical research. The basic principles and applications of microdialysis including the allied analytical techniques have been described in several monographs and book chapters (Ungerstedt et al., 1982; Ungerstedt, 1984; Hamberger et al., 1985; Robinson and Justice, 1991; Sharp and Zetterstro¨m, 1992; Kehr, 1998, 1999; Kehr and Yoshitake, 2006). II.B. Microdialysis probe II.B.1. Microdialysis membrane, implantation protocol, tissue trauma, blood-brain barrier To a large extent, the construction of most of the currently used microdialysis probes follows the original design described by Delgado et al. (1972). Several variants of concentric or linear microdialysis probes of rigid (for intracerebral use) or ﬂexible (intravenous and peripheral organs) design with their respective guide cannulas are commercially available in today’s market. In addition, numerous clinical microdialysis catheters have been developed and commercialized for use in clinical research. All these probe and catheter constructions use a polymeric dialysis membrane, which is a tubular hollow ﬁber with an outside diameter typically in the range of 220–600 mm depending on the polymer material and the supplier. The tubular membranes are widely used in perfusion cartridges of kidney dialyzers. Typical membranes used for the construction of microdialysis probes include the following materials: cuprophane (regenerated cellulose), polycarbonate, polyamide, polysulfone, polyacrylonitrile AN69, or polyarylethersulfone. Along with their geometry, the membranes can differ in ﬁnal electric charge on the surface, pore structure (symmetric vs. asymmetric membranes), and pore size deﬁned as the molecular cut-off (in Daltons). The molecular cut-off is deﬁned as the average point where 80–90% of molecules with a nominal size will be retained by the membrane. In general, all the membranes suitable for microdialysis should have a low ultraﬁltration coefﬁcient (UFRo10 mL/(h mmHg)), which is the minimal transport of the solvent (water) across the membrane. In the ideal case, the microdialysis sampling

117

should be based entirely on the diffusion of molecules across the concentration gradients with no ﬂuid being removed from, or delivered into, the sampling area. Few microdialysis studies have addressed the issues of minimizing brain tissue reactions due to the size, membrane type, and implantation protocol of the microdialysis probes. It is generally accepted that implantation of the microdialysis probe, the guide cannula, or any similar device into the brain parenchyma will cause local tissue disruption and damage to the blood capillary network. The trauma caused by probe implantation leads to abnormal morphology and physiology of the neuropil surrounding the microdialysis membrane. These factors have been studied implementing various experimental protocols and using both light microscopic techniques (Benveniste and Diemer, 1987; Benveniste et al., 1987; Ruggeri et al., 1990; Shuaib et al., 1990; Georgieva et al., 1993; Grabb et al., 1998; Clapp-Lilly et al., 1999) and ultrastructural analysis (Martin-Fardon et al., 1997; Clapp-Lilly et al., 1999). In addition, the functional responsiveness of DA to various pharmacological and electrophysiological stimuli as a function of implantation trauma and duration of the experiment was examined by microdialysis (Westerink and Tuinte, 1986; Robinson and Camp, 1991; Camp and Robinson, 1992; Fumero et al., 1994; Westerink and De Vries, 2001) and combined microdialysis and/or voltammetry techniques (Lu et al., 1998; Yang et al., 1998) and is also detailed by A. Michael in a separate chapter of this book. Using assemblies consisting of a voltammetric electrode in conjunction with a microdialysis probe, Michael and collaborators (Lu et al., 1998; Peters and Michael, 1998; Yang et al., 1998; Borland et al., 2005) revealed that the DA response to electric stimulation of the medial forebrain bundle was severely impaired when the carbon-ﬁber electrode was placed at a distance of 20–40 mm from the surface of the microdialysis membrane, but not at a distance of 1 mm from the probe in the striatum of the anesthetized rat (Yang et al., 1998). Similar conclusions were made based on quantitative morphologic analysis (Clapp-Lilly et al., 1999), which showed signiﬁcant neuronal cell and synaptic density loss at the edge of the

microdialysis probe 40 h following surgery. There is also an upper limit on how many days the microdialysis sampling can provide neurophysiologically relevant data. The implantation injury leads to a progressive generation of cellular barriers, of which the most important is the formation of the glial scar consisting predominantly of reactive astrocytes and proteoglycans (Silver and Miller, 2004). Applying the neurotransmitter criteria for voltage-sensitive and calcium-dependent release, it was demonstrated that chronically implanted intracerebral microdialysis probes could be used for up to 4 days following 24-h recovery after the surgery (Osborne et al., 1991). Similarly, the data of the no-net-ﬂux study on DA levels in the nucleus accumbens showed that there were no differences in the extraction fraction of DA during 2 days of continuous sampling (Tang et al., 2003). However, 1 week after the probe implantation, both basal and stimulated DA release were signiﬁcantly reduced and showed slower kinetics as a consequence of tissue gliosis (Westerink and Tuinte, 1986). Mechanical trauma induced by the implantation of the microdialysis probe increased extracellular levels of interleukin-1 and decreased intereleukin-6 at 24–48 h after the probe insertion, suggesting a microglia stimulation-preceded astroglial (GFAP) activation (Woodroofe et al., 1991). Histochemical techniques revealed that severe gliosis around the implanted device takes place around 4 days after the surgery (Benveniste and Diemer, 1987; Benveniste et al., 1987) and that the gliotic reaction is dependent on the size of the implanted probe (Zini et al., 1990). The most recent study aimed to examine the effects of probe implantation on brain energy metabolism and DA receptors in the rat brain by combining microdialysis and microPET. Here, 18FDG was used as a label of glucose metabolism and the DA, D2 receptor antagonist 11C-raclopride as a label for the binding potential of DA, D2 receptors (Schiffer et al., 2006). The microPET experiments revealed a widespread, prolonged decrease in glucose metabolism, which lasted for the duration of the probe implantation (up to 25 days). In contrast, the studies in the same animals using 11C-raclopride showed that there were no signiﬁcant differences in D2 receptor binding potential in the control and

118

probe-implanted hemispheres, as shown in Fig. 4. The binding potential of 11C-raclopride paralleled the extracellular levels of DA measured by microdialysis. The microPET data did not conﬁrm the original ﬁndings of Benveniste et al. (1987), who have reported the full recovery of energy metabolism 24–48 h after probe insertion. Rather, these ﬁnding support an earlier study using telemetry, demonstrating that a complete recovery of physiological functions such as periodicity of circadian changes in temperature and motor activity occurred at the earliest at 5–7 days after the microdialysis surgery (Drijfhout et al., 1995). An interesting possibility is to use repeated insertions of the microdialysis probes into the same animal via the ﬁxed guide cannula. It was shown that this approach could be applied for repeated measurements of acetylcholine (Moore et al., 1995) and DA release (Robinson and Camp, 1991). Another issue, particularly important for applications of microdialysis for pharmacokinetic studies, is the damage to the blood-brain barrier (BBB) induced by probe implantation. Several studies have reported relatively fast (within hours) restoration of the BBB following the insertion of the microdialysis probe (Benveniste et al., 1989; Dykstra et al., 1992; de Lange et al., 1995), whereas other reports call for caution when using microdialysis to study normal BBB function (Westergren et al., 1995; Morgan et al., 1996; Groothuis et al., 1998). The latter reports suggest that microdialysis may overestimate the rate of drug transfer into and out of the brain. However, up to date, a number of studies demonstrate the feasibility of using microdialysis for estimating the pharmacokinetic or pharmacodynamic proﬁles of CNS drugs by monitoring concentrations of the drug in blood (free unbound fraction) and in the ECF of the brain (for review, see Elmquist and Sawchuk, 1997, 2000; de Lange et al., 2000; Hammarlund-Udenaes, 2000). On the basis of experimental data on the damaged tissue layer surrounding the microdialysis probe, the quantitative microdialysis model was modiﬁed to include the contribution of the traumatized layer to the relative recovery (Bungay et al., 2003). The thickness of the trauma layer was estimated to vary between 4 and 30 mm. A detailed

description of this model can be found in the chapters written by P. Bungay and K. Chen. II.B.2. Minimizing tissue trauma – sterilization, biocompatibility Only a few studies have addressed the effects of applying sterile vs. non-sterile conditions on tissue reactions, altered levels of sampled molecules, or probe recovery in microdialysis experiments on animals. The requirement of sterilized microdialysis catheters and sterile perfusion ﬂuid is a must for the clinical use; however, in animal experiments, the sterility and membrane biocompatibility issues have thus far received little attention. This issue is certainly of growing importance, particularly when considering strengthened ethical guidelines regulating the use of animals in experimental medicine. According to the Guide for the Care and Use of Laboratory Animals (National Research Council, USA, 1996) and the Council of the European Communities (86/809/EEC), all surgical procedures where the animals are expected to recover must be conducted under aseptic conditions. The existing microdialysis studies have suggested that non-sterile conditions may aggravate tissue inﬂammatory response (Benveniste and Diemer, 1987) and even promote bacterial growth inside the probe, affecting the recovery of the neurotransmitter glutamate (Glu) (Zhou et al., 2002). The type of sterilization or disinfection process does not seem to have a critical impact on tissue reaction and probe recovery (Huff et al., 2003). In the latter study, two disinfection methods (70% ethanol and a commercial contact lens solution) and two sterilization procedures (hydrogen peroxide plasma and e-beam radiation) were used on microdialysis probes aseptically implanted in the livers of rats and monitored for 72 h. Surprisingly, histopathological analysis showed only minor neutrophic inﬁltration adjacent to the membrane regardless of the sterilization or disinfection protocol used and there was no difference in tissue response or probe lesion between the control and the sterilized groups. The authors concluded that the clean room environment applied by most commercial probe manufacturers is sufﬁcient to provide aseptic properties of the microdialysis probes, whereas the situation could be quite different for the laboratory-made

119

Fig. 4. The effect of trauma caused by the implantation of the microdialysis probe in the rat striatum on the radioactivity distribution of 18FDG (upper and middle panels) and 11C-raclopride binding potential (BP) and DA levels (lower panel) monitored by microPET and microdialysis, respectively. The graphs show the time courses of regional alterations in 18FDG uptake, expressed as normalized ROI ratios, in the prefrontal cortex and striatum following surgical implantation of a microdialysis cannula (upper panel) and a sham microdialysis surgery (middle panel). The rats received 18FDG scans one day (-24 h) prior to the microdialysis cannulation and again 2, 12, 24, 48, 120, 168, 360 and 500 h after the surgical implant, corresponding to 0.2, 0.5, 1, 2, 5, 7, 15 and 25 days, respectively. Guide cannulae were surgically positioned above the striatum and the microdialysis probes extending 4.0 mm beyond the end of the guide cannula were inserted into the striatum 48 hrs later. For the sham surgery, guide cannula were inserted into the cortex above the striatum and were immediately removed (time 0 on the abscissa), and the remainder of the surgical procedure was performed using the same protocol as for chronically cannulated rats. At each time point, signiﬁcant differences between ipsilateral and contralateral hemispheres are given asterisks and signiﬁcant differences from the initial, pre-surgical scan are indicated with crosses. The graph in the lower panel shows that changes in 11C-raclopride binding did not track the changes in 18FDG. The microPET experiments revealed a widespread, prolonged decrease in glucose metabolism, whereas 11C-raclopride binding in the same animals showed that there were no signiﬁcant differences in D2 receptor binding potential in the control and probe implanted hemispheres. The binding potential of 11Craclopride paralleled the increase in extracellular levels of dopamine measured by microdialysis. With permission from Elsevier, 2006.

120

probes. Further studies focusing on brain microdialysis probe implants are warranted. Another interesting study focusing on membrane biofouling and glucose recovery following 8 days of subcutaneous implantation of ethylene oxide-sterilized microdialysis probes in rats was published by Wisniewski et al. (2001). Three membranes were compared, poly(ether sulfone), polyacrylonitrile, and polycarbonate, all showing signiﬁcant reduction in glucose recovery on day 8. Furthermore, the tissue reaction (resistance to glucose transport) was three to ﬁve times higher than the resistance induced by the membrane biofouling, suggesting that future development should be focused on improved membrane biocompatibility. Similar conclusions could be drawn from another study, which showed markedly different performance of the same type of sterile probes with polyamide membranes when implanted subcutaneously in humans and rats for 7 days (Wisniewski et al., 2002). The hollow ﬁber membranes used for the construction of the microdialysis probes have been developed for renal dialysis and as such, thoroughly tested for their biocompatible properties in contact with blood cells. However, the importance of biocompatibility of different membranes has been questioned, for example, when used for dialysis in acute renal failure (Macleod et al., 2005). In addition, the mechanisms by which the hemodialysis membranes activate tissue (microglia and astroglia) reactions in the brain may differ from those in the blood. In that respect, for further improvement of microdialysis membranes, it might be more relevant to consider adopting the biocompatibility strategies developed for the neural electrodes (for review, see Polikov et al., 2005). Future success of neuroprosthetic devices is critically dependent on the unaltered performance of recording and/or stimulating electrodes chronically implanted in the CNS. Several strategies have focused on coating the electrodes with bioactive molecules, such as cell adhesion-promoting polypeptides or proteins such as collagen and ﬁbronectin (Ignatius et al., 1998), or cell adhesion peptides such as RGD, YIGSR, IKVAV, and KHIFSDDSSE (Kam et al., 2002). These results also indicate a promising direction for microdialysis probe research. Another

approach may be to employ growth factors or chemoattractants to promote neuronal survival and growth toward the membrane surface, although the growth factor-based strategies have generally been disappointing in the adult CNS. Standard wound healing suppression and immunosuppression techniques might be other options for minimizing the initial immune response and perhaps even the glial scar formation. Shain et al. (2003) found that peripheral injections of dexamethasone at the time of neural electrode insertion greatly attenuated glial scar formation at 1 and 6 weeks as shown by GFAP staining. A similar strategy to minimize tissue reactions induced by the implantation of the microdialysis probe was proposed by Drew et al. (2004). The proposed strategy builds on a possibility to use microdialysis for local infusion of ‘‘neuroprotective’’ substances during the acute phase of probe insertion, which could reduce trauma-induced neurodegeneration and thereby improve the relevance of microdialysis monitoring following various physiological or pharmacological stimuli. II.B.3. Advancements in microdialysis – implementing microfabricated devices During the recent years, a tremendous progress has been achieved in the design and fabrication of microdevices used as the ﬂuidic components, chips, or sensor parts of the instruments devoted to bioanalytical screening of molecules in both the laboratory and the clinical settings. In the ideal case, the integrated mTAS is a single chip, which comprises a device for sample acquisition, transport of liquid sample including its processing (e.g., pre-concentration, separation, and conjugation), and ﬁnally sample detection by a built-in sensor element. Microdialysis offers a challenging opportunity to implement the mTAS strategy and construct a miniaturized monitoring device, which could be implanted into the body without any need for the attachment of connecting lines or cables to external instruments. The wireless technology to transmit or accept the radio signals by internally implanted devices is fairly well developed, for example, the telemetry transducers for studies on animals including small rodents or devices for functional electric stimulation in humans. Recently, it was

121

demonstrated that wireless transmission could be used for monitoring DA signals by cyclic voltammetry at carbon-ﬁber microelectrodes implanted in rat brains (Crespi et al., 2004; Garris et al., 2004). Combining microdialysis probes with biosensors would allow detection of a wider spectrum of analytes, and yet use of a simpler mode of amperometric detection at a constant potential, which could be provided by a small battery placed within the implanted device. These techniques are expected to grow quite extensively in the near future. In terms of the construction of an integrated microdialysis or biosensor device, three principal design alternatives have been described: (i) the biosensor is simply ‘‘coated’’ by, or placed into, the membrane in order to prevent the biosensor from contaminating with proteins; no liquid needs to be pumped through the assembly during the measurement; (ii) the biosensor or microdialysis probe assembly needs to be regularly reﬁlled with solution containing the enzymes necessary for conversion of the analyte, for example, to hydrogen peroxide, which is easily oxidized at the platinum electrode; (iii) the biosensor is placed more distally to the membrane of the microdialysis probe or even outside the body, the probe must be continuously perfused with physiological solution. Typical examples representing the ﬁrst group of the dialysis or biosensors could be an electrochemical enzyme-based glucose sensor coated with a protective layer of nanoporous silicon (Piechotta et al., 2005) and the hollow ﬁber sensor for transdermal glucose monitoring described by Ballerstadt and Schultz (2000). The principle of this biosensor is that glucose diffusing through the membrane can competitively displace Alexa 488 dye, which is immobilized inside the Sephadex beads. The displaced Alexa 488 dye induces a strong ﬂuorescence light emission at 520 nm. The dialysis-assisted ﬁber optic spectroscopy can be considered as an interesting alternative to the more common combination of microdialysis with electrochemical biosensors. The second group of integrated microdialysis or sensor devices is best represented by a concept of the dialysis electrode (Walker et al., 1995; Asai et al., 1996). The dialysis electrode is basically an enzyme-based sensor built within the microdialysis probe. A typical system consists of three electrodes

including a silver or silver chloride (reference), a silver (auxiliary) electrode, and a platinum (working) electrode in the form of wires, which are inserted into the microdialysis probe. The probe is ﬁlled with Ringer’s solution containing the enzymes – ascorbate oxidase to remove interfering ascorbic acid and glutamate oxidase when detecting Glu. The enzyme molecules do not pass through the dialysis membrane and their concentrations should be in sufﬁcient excess (0.05 U/mL for glutamate oxidase and 1 U/mL for ascorbate oxidase (Walker et al., 1995)) to provide a constant degree of enzymatic conversion of Glu during the entire experimental period. However, the concept of dialysis electrodes, though quite sound and economically feasible, has not yet gained further attention or wider use in scientiﬁc community. Probably the most thoroughly described microdialysis or biosensor devices are those represented by the third group of above-mentioned conﬁgurations. In this setting, the on-line sensor device is placed extracorporally. Monitoring glucose and lactate levels by such systems also shows the strongest application potential in various clinical settings. In the initial studies, brain extracellular glucose and lactate levels were measured by microdialysis combined with an on-line biosensor system in a form of immobilized enzyme reactor cartridges (Boutelle et al., 1992) or a thin-layer ﬂow-through cell with immobilized enzyme electrodes (Osborne et al., 1998). Further miniaturization of the sensor manifold was achieved by using microlithographically fabricated integrated systems. The BioMEMS devices offer simultaneous monitoring of several analytes by implementing microarray electrodes coated with speciﬁc enzymes, for example, glucose oxidase and lactate oxidase (Dempsey et al., 1997; Perdomo et al., 2000; Rhemrev-Boom et al., 2002; Petrou et al., 2002, 2003; Zahn et al., 2005). A separate group of integrated microdialysis or MEMS devices comprises the microdialysis probe connected on-line to the microﬂuidic interface, which facilitates sample preparation and control sample injections into the analytical instrument such as capillary electrophoresis and MS (Sandlin et al., 2005; Li et al., 2006, for review, see Kennedy et al., 2002; Powell and Ewing, 2005; and also see chapter by Kennedy in this book).

122

30 nm diffusion passage

Permeable Polysilicon

Fig. 5. Microdialysis microneedles for continuous medical monitoring. The SEM picture of a microdialysis microneedle (left) and a closeup of the microdialysis membrane (right). The arrayed dots on the needle represent 2-mm features deﬁning the dialysis membrane. Thin (30 nm) openings can be seen between the polysilicon layers (top right). In addition, when the layers were fractured, they delaminated and fractured at different points indicating the layers are separate. The permeable polysilicon layer can be seen at the bottom of the trenches (bottom right). With permission from Elsevier, 2006.

In summary, one can conclude that a number of exciting new devices have been constructed aiming to miniaturize the microdialysis sampling apparatus and make it more effective, that is, a faster response and a better integration with detection systems. However, these microdevices are typically incorporated at the outlet of the microdialysis probe or at best, within the lumen of existing microdialysis membranes. Until now, only a few attempts have been made aiming to reduce the size of the microdialysis probe by implementing MEMS technology. Thus, Zahn et al. (2005) described the fabrication of microdialysis microneedles from a silicon substrate using silicon on an insulator device layer as a mechanical support.

The diffusion membranes on silicon needles were fabricated using either a layered polysilicon sandwich with etched holes or a permeable polysilicon with pore defects (5–20 nm wide), which allow small molecules to permeate between the polysilicon grains. The SEM picture of a dialysis microneedle is shown in Fig. 5. This microdialysis microneedle is an order of magnitude smaller in almost all characteristics compared with existing probes, while meeting the same performance criteria. It is fabricated as a complete integrated device during the microlithographic processing steps, so no assembly is necessary and the needle can also be integrated with planar electrochemical sensors.

123

In addition, nanoporous silicon surface coatings on neural electrodes have been shown to decrease adhesion of astrocytes and increased extension of neurites in vitro (Moxon et al., 2004). The dramatic miniaturization and reshaping of the proﬁle of microdialysis microneedle probes is expected to reduce the tissue trauma and provide better contact between the probe and the neuropil, thereby, facilitating more rapid diffusion and higher recovery of analytes sampled from the ECF. The ﬁrst indication of the potential future development in this direction was provided by Cellar et al. (2005) who developed a miniaturized push–pull perfusion system in conjunction with a microﬂuidic chip sampling interface and capillary electrophoresis with laser-induced ﬂuorescence for the determination of amino acids in 45-s intervals. In the context of discussing the general effort to design an integrative miniaturized microdialysis system, a ﬁnal comment should be given to the infusion pump and the sample collector. Collecting samples in an off-line mode is necessary for many complex analytes, for which a simple biosensor approach is not available. Fraction collectors developed for microdialysis can handle sample collecting in microliter volumes, often including a refrigeration option to prevent sample evaporation and the degradation of labile molecules. However, these are all the bench-top instruments and as to date, no portable or chip-based collector device is available. In that sense, the only alternative is to load the microdialysis perfusate into a long piece of the outlet capillary tubing, which is then cut into the pieces providing the volumes necessary for analysis. Using this strategy in combination with CE or LIF of ﬂuorescein isothiocyanate-derivatized amines, Rossell et al. (2003) were able to detect Glu at a 1-s-time resolution in fully conscious rats. Each 1-s fraction corresponded to 30 nL sample in a 4-mm long piece of a fused-silica capillary (see also Chapter 3.3). In a vast majority of all applications, the syringe pump is used to deliver constant, pulse-free ﬂow of the physiological solution through the microdialysis probe. Typical ﬂow-rates range between 0.1 and 2 mL/min. The recently introduced batterydriven syringe pumps for clinical applications deliver the physiological solution in a pulse mode,

typically as pulses of 0.3 mL/min (CMA 106 Microdialysis Pump). An alternative to the syringe pump is an osmotic minipump (e.g., Alzet), which can be implanted intraperitoneally and therefore does not require external connections. However, the ﬂow variations of the commercially available pumps are probably too high; the standard deviation for the ﬂow delivery is 10% at the ﬂowrates of 10 mL/h and below. Currently, there is no study available for evaluating the precision of recovery of small molecules when combining the osmotic minipumps and microdialysis. Recently, a novel design of the osmotic minipump was described based on the principle of phloem loading (a diffusion gradient of sugars) in plant leaves (Ehwald et al., 2006). The pump provided pulsefree liquid delivery at a constant ﬂow of 0.6 mL/ min for at least 26 days. However, the prototype pump was just too large to be implanted into the laboratory animal; hence, further development and experiments are required to examine a potential use of this pump in microdialysis. II.B.4. Enhanced microdialysis – perfusing probes with functionalized nanoparticles Today, all known neurotransmitters and neuromodulators can be recovered by microdialysis sampling. Furthermore, a number of large molecular weight compounds, such as peptides and even certain proteins and enzymes, can diffuse through the microdialysis membranes with a higher molecular cut-off. The distribution and pharmacokinetics of drugs can be studied by applying various models of quantitative microdialysis to estimate absolute extracellular concentrations of the drugs and their metabolites (for review, see Elmquist and Sawchuk, 1997, 2000). However, the applicability of microdialysis to monitor certain molecules in vivo is entirely dependent on the recovery of such molecules and the sensitivity of allied analytical techniques. Sampling large molecules such as neuropeptides involved in brain chemical communication has been of eminent interest to neuropharmacologists since the early development of microdialysis technique in experimental brain research (Brodin et al., 1983). However, the applicability of microdialysis for sampling and monitoring brain extracellular levels of peptides and soluble proteins is limited

124

by several factors: (i) diminutive concentrations of freely diffusing large molecular weight compounds in the ECF; (ii) poor recovery of lipophilic compounds; (iii) common tubular membranes used for microdialysis probes possess low recoveries for molecules larger than 5,000 Da; (iv) large pore membranes (molecular weight cut-off> 100,000 Da) are associated with ﬂux of water across the membrane, which necessitates the use of the push–pull perfusion system; (v) high-sensitive methods such as radioimmunoassays or LC–tandem MS have to be applied. Typically, the competitive radioimmunoassays offering detection limits of 0.1 fmol/sample are used for the determination of neuropeptides in the microdialysates. Recently, methods implementing enzyme-linked immunosorbent assays (ELISA), the chip array immunoassays utilizing ﬂorescent or chemiluminescent labels, and ﬂow cytometry techniques implicating ﬂuorescent molecular, or nanoparticle (quantum dots) tags were shown to achieve comparable detection limits to the solidphase RIA format. Rapid advancements in LC/ MS/MS techniques have enabled sensitive detection of exogenous compounds such as drugs and their metabolites recovered by microdialysis sampling (Emmett et al., 1995) and the identiﬁcation of peptides and their fragments (see chapter written by Kennedy). An increasing effort is given to the optimization of microdialysis sampling protocols, which will allow higher recovery of

polypeptide molecules. One alternative is to use the membranes with larger pores, for example, a polyethersulfone membrane with a 100-kDa molecular weight cut-off (Schutte et al., 2004) or a plasmaphoresis membrane (30,000-kDa m.w. cutoff) in a push–pull mode (Winter et al., 2002). Another possibility is to manipulate the perfusion medium by additives, which will facilitate the recovery of lipophilic peptides and proteins. Thus, the addition of bovine serum albumin (BSA), glycerol, or intralipid emulsion to the Ringer’s perfusion medium was evaluated for improving the in vitro recovery of oleic acid (Carneheim and Stahle, 1991) and in a similar manner, the addition of cyclodextrin polymers was shown to increase relative recovery of eicosanoids and other hydrophobic analytes in vitro (Ao and Stenken, 2003; Sun and Stenken, 2003). Recently, it was demonstrated that the addition of hydrophobic microspheres can enhance the in vitro recovery of selected neuropeptides for MS analysis (Pettersson et al., 2004). Likewise, the addition of antibody-coated microspheres was shown to markedly improve recovery of inﬂammatory cytokines when analyzed by ﬂow cytometry (Ao et al., 2004; Ao and Stenken, 2006a, b). The efﬁcacy of using antibody-immobilized microbeads suspended in the perfusion medium to enhance the microdialysis recovery in vitro of selected cytokines is summarized in Table 1 (Ao and Stenken, 2006b). As shown, the ratio of enhanced recovery vs. control recovery was dependent not

Table 1. Cytokine in vitro relative recovery and relevant physicochemical propertiesa Cytokine

MW (kDa)

Conformation

Control RR% 1.0 mL/min

Enhanced RR% 1.0 mL/min

Ratio of enhanced to control

IL-2 IL-4 IL-5 IL-6 IL-10 IL-12p70 IFN-g MCP-1 TFN-a

17.2 13.6 13.1 21.7 18.8 35 and 40 15.9 13.1 17.3

Monomer Monomer Homodimer Monomer Homodimer Heterodimer Homodimer Homodimer Homodimer

3.170.7 8.572.3 0.770.1 6.470.3 2.770.2 ND 1.370.2 21.571.4 4.371.0

17.872.3 78.374.8 5.170.4 24.173.8 10.171.9 7.270.5 17.471.7 51.476.4 90.875.6

5.7 9.2 7.3 3.8 3.7 ND 13.4 2.4 21.1

Note: The efﬁcacy of using antibody-immobilized microbeads suspended in the perfusion medium to enhance the microdialysis recovery in vitro of selected cytokines. As seen, the ratio of enhanced vs. control recovery was not only dependent on the molecular weight of the sampled protein, but also on complex factors including gyration radius of each particular protein, overall electric charge, and binding afﬁnity to the antibody-coated microbeads. With permission from Elsevier, 2006. a

Cytokine standards were 2,500 pg/mL. All solutions were quiescent at room temperature. RR% value denotes a mean of n ¼ 3. All data are reported as mean7SD. ND, not detected.

125

only on the molecular weight of the sampled protein, but also on complex factors including gyration radius of each particular protein, overall electric charge, and the binding afﬁnity to the antibody-coated microbeads. Taken together, these results indicate a strong potential of using microdialysis in combination with mass spectrometric techniques for the quantiﬁcation of known neuropeptides, as well as for the identiﬁcation of new peptidergic sequences and proteins. This allows to investigate the potential role of peptides in chemical signaling or as markers of brain injury. III. Conclusions Microdialysis offers a unique opportunity to monitor the chemistry of the brain microenvironment and this feature, often, is overwhelming the main limitation of microdialysis being an invasive technique. In clinical microdialysis, the major attention is given to the monitoring of changes in energy metabolism (glucose, lactate, and pyruvate) and glycerol levels serving as an index of membrane decomposition. The samples are collected manually and the biomarker molecules are determined off-line by the use of a dedicated analyzer. There is a great potential to apply MEMS and biosensor technologies to automatize and further improve the speed and feasibility of microdialysis bed-site monitoring. In experimental neuropharmacology, thousands of published papers provide a solid ground to validate microdialysis as a key technique in pharmacological neurochemistry in vivo and as a major tool for studies of the neurochemical basis of behavior. Further development of faster and more sensitive detection techniques conjugated to microdialysis is expected to strengthen the use of microdialysis in functional pharmacology and enable the search for novel biomarkers of CNS diseases. References Ao, X. and Stenken, J.A. (2003) Water-soluble cyclodextrin polymers for enhanced relative recovery of hydrophobic analytes during microdialysis sampling. Analyst, 128: 1143–1149.

Ao, X. and Stenken, J.A. (2006a) No delayed temporal response to sample concentration changes during enhanced microdialysis sampling using cyclodextrins and antibody-immobilized microspheres. Analyst, 131: 62–67. Ao, X. and Stenken, J.A. (2006b) Microdialysis sampling of cytokines. Methods, 38: 331–341. Ao, X., Sellati, T.J. and Stenken, J.A. (2004) Enhanced microdialysis relative recovery of inﬂammatory cytokines using antibody-coated microspheres analyzed by ﬂow cytometry. Anal. Chem., 76: 3777–3784. Asai, S., Iribe, Y., Kohno, T. and Ishikawa, K. (1996) Real time monitoring of biphasic glutamate release using dialysis electrode in rat acute brain ischemia. Neuroreport, 7: 1092–1096. Ballerstadt, R. and Schultz, J.S. (2000) A ﬂuorescence afﬁnity hollow ﬁber sensor for continuous transdermal glucose monitoring. Anal. Chem., 72: 4185–4192. Beaudet, A. and Descarries, L. (1978) The monoamine innervation of rat cerebral cortex: synaptic and nonsynaptic axon terminals. Neuroscience, 3: 851–860. Benveniste, H. and Diemer, N.H. (1987) Cellular reactions to implantation of a microdialysis tube in the rat hippocampus. Acta Neuropathol. (Berl.), 74: 234–238. Benveniste, H., Drejer, J., Schousboe, A. and Diemer, N.H. (1987) Regional cerebral glucose phosphorylation and blood ﬂow after insertion of a microdialysis ﬁber through the dorsal hippocampus in the rat. J. Neurochem., 49: 729–734. Benveniste, H., Hansen, A.J. and Ottosen, N.S. (1989) Determination of brain interstitial concentrations by microdialysis. J. Neurochem., 52: 1741–1750. Bito, L., Davson, H., Levin, E., Murray, M. and Snider, N. (1966) The concentrations of free amino acids and other electrolytes in cerebrospinal ﬂuid, in vivo dialysate of brain, and blood plasma of the dog. J. Neurochem., 13: 1057–1067. Borland, L.M., Shi, G., Yang, H. and Michael, A.C. (2005) Voltammetric study of extracellular dopamine near microdialysis probes acutely implanted in the striatum of the anesthetized rat. J. Neurosci. Methods, 146: 149–158. Boutelle, M.G., Fellows, L.K. and Cook, C. (1992) Enzyme packed bed system for the on-line measurement of glucose, glutamate, and lactate in brain microdialysate. Anal. Chem., 64: 1790–1794. Brodin, E., Lindefors, N. and Ungerstedt, U. (1983) Potassium evoked in vivo release of substance P in rat caudate nucleus measured using a new technique of brain dialysis and an improved substance P-radioimmunoassay. Acta Physiol. Scand. Suppl., 515: 17–20. Bungay, P.M., Morrison, P.F. and Dedrick, R.L. (1990) Steady-state theory for quantitative microdialysis of solutes and water in vivo and in vitro. Life Sci., 46: 105–119. Bungay, P.M., Newton-Vinson, P., Isele, W., Garris, P.A. and Justice, J.B. (2003) Microdialysis of dopamine interpreted with quantitative model incorporating probe implantation trauma. J. Neurochem., 86: 932–946. Camp, D.M. and Robinson, T.E. (1992) On the use of multiple probe insertions at the same site for repeated intracerebral microdialysis experiments in the nigrostriatal dopamine system of rats. J. Neurochem., 58: 1706–1715.

126 Carneheim, C. and Stahle, L. (1991) Microdialysis of lipophilic compounds: a methodological study. Pharmacol. Toxicol., 69: 378–380. Cellar, N.A., Burns, S.T., Meiners, J.C., Chen, H. and Kennedy, R.T. (2005) Microﬂuidic chip for low-ﬂow push–pull perfusion sampling in vivo with on-line analysis of amino acids. Anal. Chem., 77: 7067–7073. Chen, K.C., Hoistad, M., Kehr, J., Fuxe, K. and Nicholson, C. (2002) Quantitative dual-probe microdialysis: mathematical model and analysis. J. Neurochem., 81: 94–107. Clapp-Lilly, K.L., Roberts, R.C., Duffy, L.K., Irons, K.P., Hu, Y. and Drew, K.L. (1999) An ultrastructural analysis of tissue surrounding a microdialysis probe. J. Neurosci. Methods, 90: 129–142. Crespi, F., Bandera, A., Donini, M., Heidbreder, C. and Rovati, L. (2005) Non-invasive in vivo infrared laser spectroscopy to analyse endogenous oxy-haemoglobin, deoxyhaemoglobin, and blood volume in the rat CNS. J. Neurosci. Methods, 145: 11–22. Crespi, F., Dalessandro, D., Annovazzi-Lodi, V., Heidbreder, C. and Norgia, M. (2004) In vivo voltammetry: from wire to wireless measurements. J. Neurosci. Methods, 140: 153–161. de Lange, E.C., de Boer, A.G. and Breimer, D.D. (2000) Methodological issues in microdialysis sampling for pharmacokinetic studies. Adv. Drug Deliv. Rev., 45: 125–148. de Lange, E.C., Hesselink, M.B., Danhof, M., de Boer, A.G. and Breimer, D.D. (1995) The use of intracerebral microdialysis to determine changes in blood-brain barrier transport characteristics. Pharm. Res., 12: 129–133. Delgado, J.M., DeFeudis, F.V., Roth, R.H., Ryugo, D.K. and Mitruka, B.M. (1972) Dialytrode for long term intracerebral perfusion in awake monkeys. Arch. Int. Pharmacodyn., 198: 9–21. Dempsey, E., Diamond, D., Smyth, M.R., Malone, M.A., Rabenstein, K., McShane, A., McKenna, M., Keaveny, T.V. and Freaney, R. (1997) In vitro optimisation of a microdialysis system with potential for on-line monitoring of lactate and glucose in biological samples. Analyst, 122: 185–189. Descarries, L. and Mechawar, N. (2000) Ultrastructural evidence for diffuse transmission by monoamine and acetylcholine neurons of the central nervous system. Prog. Brain Res., 125: 27–47. Drew, K.L., Pehek, E.A., Rasley, B.T., Ma, Y.L. and Green, T.K. (2004) Sampling glutamate and GABA with microdialysis: suggestions on how to get the dialysis membrane closer to the synapse. J. Neurosci. Methods, 140: 127–131. Drijfhout, W.J., Kemper, R.H., Meerlo, P., Koolhaas, J.M., Grol, C.J. and Westerink, B.H. (1995) A telemetry study on the chronic effects of microdialysis probe implantation on the activity pattern and temperature rhythm of the rat. J. Neurosci. Methods, 61: 191–196. Dykstra, K.H., Hsiao, J.K., Morrison, P.F., Bungay, P.M., Mefford, I.N., Scully, M.M. and Dedrick, R.L. (1992) Quantitative examination of tissue concentration proﬁles associated with microdialysis. J. Neurochem., 58: 931–940.

Ehwald, M., Adleff, H., Geggier, P. and Ehwald, R. (2006) A long-term stable and adjustable osmotic pump for small volume ﬂow based on principles of phloem loading. Biotechnol. Bioeng., 94: 37–42. Elmquist, W.F. and Sawchuk, R.J. (1997) Application of microdialysis in pharmacokinetic studies. Pharm. Res., 14: 267–288. Elmquist, W.F. and Sawchuk, R.J. (2000) Use of microdialysis in drug delivery studies. Adv. Drug Deliv. Rev., 45: 123–124. Emmett, M.R., Andren, P.E. and Caprioli, R.M. (1995) Speciﬁc molecular mass detection of endogenously released neuropeptides using in vivo microdialysis/mass spectrometry. J. Neurosci. Methods, 62: 141–147. Ferrari, M., Mottola, L. and Quaresima, V. (2004) Principles, techniques, and limitations of near infrared spectroscopy. Can. J. Appl. Physiol., 29: 463–487. Fumero, B., Guadalupe, T., Valladares, F., Mora, F., O’Neill, R.D., Mas, M. and Gonzalez-Mora, J.L. (1994) Fixed versus removable microdialysis probes for in vivo neurochemical analysis: implications for behavioral studies. J. Neurochem., 63: 1407–1415. Fuxe, K. and Agnati, L.M. (1991) Volume Transmission in the Brain. Novel Mechanisms for Neural Transmission. Advances in Neuroscience. Raven Press, New York. Fuxe, K., Agnati, L.M., Nicholson, C. and Sykova´, E. (2000) Volume Transmission Revisited. Progress in Brain research, Vol. 125. Elsevier, Amsterdam. Garris, P.A., Ensman, R., Poehlman, J., Alexander, A., Langley, P.E., Sandberg, S.G., Greco, P.G., Wightman, R.M. and Rebec, G.V. (2004) Wireless transmission of fast-scan cyclic voltammetry at a carbon-ﬁber microelectrode: proof of principle. J. Neurosci. Methods, 140: 103–115. Georgieva, J., Luthman, J., Mohringe, B. and Magnusson, O. (1993) Tissue and microdialysate changes after repeated and permanent probe implantation in the striatum of freely moving rats. Brain Res. Bull., 31: 463–467. Grabb, M.C., Sciotti, V.M., Gidday, J.M., Cohen, S.A. and van Wylen, D.G. (1998) Neurochemical and morphological responses to acutely and chronically implanted brain microdialysis probes. J. Neurosci. Methods, 82: 25–34. Greitz, D., Franck, A. and Nordell, B. (1993) On the pulsatile nature of intracranial and spinal CSF-circulation demonstrated by MR imaging. Acta Radiol., 34: 321–328. Groothuis, D.R., Ward, S., Schlageter, K.E., Itskovich, A.C., Schwerin, S.C., Allen, C.V., Dills, C. and Levy, R.M. (1998) Changes in blood-brain barrier permeability associated with insertion of brain cannulas and microdialysis probes. Brain Res., 803: 218–230. Hamberger, A., Berthold, C.-H., Jacobson, I., Karlsson, B., Lehmann, A., Nystro¨m, B. and Sandberg, M. (1985) In vivo brain dialysis of extracellular neurotransmitter and putative transmitter amino acids. In: Bayon, A. and Drucker-Colin, R. (Eds.), In Vivo Perfusion and Release of Neuroactive Substances. Alan R. Liss, New York, pp. 473–492. Hammarlund-Udenaes, M. (2000) The use of microdialysis in CNS drug delivery studies. Pharmacokinetic perspectives and results with analgesics and antiepileptics. Adv. Drug Deliv. Rev., 45: 283–294.

127 Hillered, L., Vespa, P.M. and Hovda, D.A. (2005) Translational neurochemical research in acute human brain injury: the current status and potential future for cerebral microdialysis. J. Neurotrauma, 22: 3–41. Ho¨istad, M., Chen, K.C., Nicholson, C., Fuxe, K. and Kehr, J. (2002) Quantitative dual-probe microdialysis: evaluation of [3 H]mannitol diffusion in agar and rat striatum. J. Neurochem., 81: 80–93. Ho¨istad, M., Kehr, J., Andbjer, B., Jansson, A. and Fuxe, K. (2000) Intracerebral infusion of H-dopamine and H-mannitol in the striatum of halothane-anaesthetized male rats. A dualprobe microdialysis study of long-distance diffusion. Eur. J. Neurosci., 12: 2505–2514. Huff, J.K., Bresnahan, J.F. and Davies, M.I. (2003) Preliminary evaluation of several disinfection/sterilization techniques for use with microdialysis probes. Life Sci., 73: 275–287. Ignatius, M.J., Sawhney, N., Gupta, A., Thibadeau, B.M., Monteiro, O.R. and Brown, I.G. (1998) Bioactive surface coatings for nanoscale instruments: effects on CNS neurons. J. Biomed. Mater. Res., 40: 264–274. Jobsis, F. (1977) Non-invasive infrared monitoring of cerebral and myocardial oxygen sufﬁciency and circulatory parameters. Science, 198: 1264–1267. Kam, L., Shain, W., Turner, J.N. and Bizios, R. (2002) Selective adhesion of astrocytes to surfaces modiﬁed with immobilized peptides. Biomaterials, 23: 511–515. Kehr, J. (1998) Microdialysis: an in vivo method for the analysis of body ﬂuids. In: Rossomando, E.F. (Ed.), HPLC in Enzymatic Analysis (2nd ed.). Wiley, New York, pp. 115–136. Kehr, J. (1999) Monitoring chemistry of brain microenvironment: biosensors, microdialysis and related techniques. In: Windhorst, U. and Johansson, H. (Eds.), Modern Techniques in Neuroscience Research. Springer-Verlag, Heidelberg, pp. 1149–1198. Kehr, J. and Yoshitake, T. (2006) Monitoring brain chemical signals by microdialysis. In: Grimes, C.A., Dickey, E.C. and Pishko, M.V. (Eds.) Encyclopedia of Sensors, Vol. 6. American Scientiﬁc Publishers, CA, USA, pp. 287–312. Kehr, J., Ho¨istad, M. and Fuxe, K. (2000) Diffusion of radiolabeled dopamine, its metabolites and mannitol in the rat striatum studied by dual-probe microdialysis. Prog. Brain Res., 125: 179–190. Kennedy, R.T., Watson, C.J., Haskins, W.E., Powell, D.H. and Strecker, R.E. (2002) In vivo neurochemical monitoring by microdialysis and capillary separations. Curr. Opin. Chem. Biol., 6: 659–665. Kissinger, P.T., Refshuage, C.J., Dreiling, R., Blank, L., Freeman, R. and Adams, R.N. (1973) An electrochemical detector for liquid chromatography with picogram sensitivity. Anal. Lett., 6: 465–477. Li, M.W., Huynh, B.H., Hulvey, M.K., Lunte, S.M. and Martin, R.S. (2006) Design and characterization of poly(dimethylsiloxane)-based valves for interfacing continuousﬂow sampling to microchip electrophoresis. Anal. Chem., 78: 1042–1051.

Lindroth, P. and Mopper, K. (1979) High performance liquid chromatographic determination of subpicomole amounts of amino acids by pre-column ﬂuorescence derivatisation with o-phthalaldehyde. Anal. Chem., 51: 1667–1674. Lu, Y., Peters, J.L. and Michael, A.C. (1998) Direct comparison of the response of voltammetry and microdialysis to electrically evoked release of striatal dopamine. J. Neurochem., 70: 584–593. Macleod, A.M., Campbell, M., Cody, J.D., Daly, C., Donaldson, C., Grant, A., Khan, I., Rabindranath, K.S., Vale, L. and Wallace, S. (2005) Cellulose, modiﬁed cellulose and synthetic membranes in the haemodialysis of patients with end-stage renal disease. Cochrane Database Syst. Rev., 3: CD003234. Martin-Fardon, R., Sandillon, F., Thibault, J., Privat, A. and Vignon, J. (1997) Long-term monitoring of extracellular dopamine concentration in the rat striatum by a repeated microdialysis procedure. J. Neurosci. Methods, 72: 123–135. May, L.J., Kuhr, W.G. and Wightman, R.M. (1988) Differentiation of dopamine overﬂow and uptake processes in the extracellular ﬂuid of the rat caudate nucleus with fast-scan in vivo voltammetry. J. Neurochem., 51: 1060–1069. Mayevsky, A. and Chance, B. (1973) A new long-term method for the measurement of NADH ﬂuorescence in intact rat brain with chronically implanted cannula. Adv. Exp. Med. Biol., 37A: 239–244. Mitterschiffthaler, M.T., Ettinger, U., Mehta, M.A., MataixCols, D. and Williams, S.C. (2006) Applications of functional magnetic resonance imaging in psychiatry. J. Magn. Reson. Imaging, 23: 851–861. Moore, H., Stuckman, S., Sarter, M. and Bruno, J.P. (1995) Stimulation of cortical acetylcholine efﬂux by FG 7142 measured with repeated microdialysis sampling. Synapse, 21: 324–331. Morgan, M.E., Singhal, D. and Anderson, B.D. (1996) Quantitative assessment of blood-brain barrier damage during microdialysis. J. Pharmacol. Exp. Ther., 277: 1167–1176. Moxon, K.A., Kalkhoran, N.M., Markert, M., Sambito, M.A., McKenzie, J.L. and Webster, J.T. (2004) Nanostructured surface modiﬁcation of ceramic-based microelectrodes to enhance biocompatibility for a direct brain-machine interface. IEEE Trans. Biomed. Eng., 51: 881–889. Nicholson, C. (1979) Brain cell microenvironment as a communication channel. In: Schmitt, F.O. and Worden, F.G. (Eds.), The Neurosciences Fourth Study Program. MIT Press, Cambridge, MA, pp. 457–476. Nicholson, C. (1993) Ion-selective microelectrodes and diffusion measurements as tools to explore the brain cell microenvironment. J. Neurosci. Methods, 48: 199–213. Nicholson, C. (2005) Factors governing diffusing molecular signals in brain extracellular space. J. Neural Transm., 112: 29–44. Nicholson, C. and Phillips, J.M. (1981) Ion diffusion modiﬁed by tortuosity and volume fraction in the extracellular microenvironment of the rat cerebellum. J. Physiol. (Lond.), 321: 225–257.

128 Nicholson, C. and Sykova, E. (1998) Extracellular space structure revealed by diffusion analysis. Trends Neurosci., 21: 207–215. Nordstro¨m, C.H. (2005) Physiological and biochemical principles underlying volume-targeted therapy – the ‘‘Lund concept’’. Neurocrit. Care, 2: 83–95. Norris, D.G. (2006) Principles of magnetic resonance assessment of brain function. J. Magn. Reson. Imaging, 23: 794–807. Osborne, P.G., Niwa, O. and Yamamoto, K. (1998) Plastic ﬁlm carbon electrodes: enzymatic modiﬁcation for on-line, continuous, and simultaneous measurement of lactate and glucose using microdialysis sampling. Anal. Chem., 70: 1701–1706. Osborne, P.G., O’Connor, W.T., Kehr, J. and Ungerstedt, U. (1991) In vivo characterisation of extracellular dopamine, GABA and acetylcholine from the dorsolateral striatum of awake freely moving rats by chronic microdialysis. J. Neurosci. Methods, 37: 93–102. Perdomo, J., Hinkers, H., Sundermeier, C., Seifert, W., Martinez Morell, O. and Knoll, M. (2000) Miniaturized real-time monitoring system for L-lactate and glucose using microfabricated multi-enzyme sensors. Biosens. Bioelectron., 15: 515–522. Peters, J.L. and Michael, A.C. (1998) Modeling voltammetry and microdialysis of striatal extracellular dopamine: the impact of dopamine uptake on extraction and recovery ratios. J. Neurochem., 70: 594–603. Petrou, P.S., Moser, I. and Jobst, G. (2002) BioMEMS device with integrated microdialysis probe and biosensor array. Biosens. Bioelectron., 17: 859–865. Petrou, P.S., Moser, I. and Jobst, G. (2003) Microdevice with integrated dialysis probe and biosensor array for continuous multi-analyte monitoring. Biosens. Bioelectron., 18: 613–619. Pettersson, A., Amirkhani, A., Arvidsson, B., Markides, K. and Bergquist, J. (2004) A feasibility study of solid supported enhanced microdialysis. Anal. Chem., 76: 1678–1682. Piechotta, G., Albers, J. and Hintsche, R. (2005) Novel micromachined silicon sensor for continuous glucose monitoring. Biosens. Bioelectron., 21: 802–808. Polikov, V.S., Tresco, P.A. and Reichert, W.M. (2005) Response of brain tissue to chronically implanted neural electrodes. J. Neurosci. Methods, 148: 1–18. Powell, P.R. and Ewing, A.G. (2005) Recent advances in the application of capillary electrophoresis to neuroscience. Anal. Bioanal. Chem., 382: 581–591. Rhemrev-Boom, R.M., Tiessen, R.G., Jonker, A.A., Venema, K., Vadgama, P. and Korf, J. (2002) A lightweight measuring device for the continuous in vivo monitoring of glucose by means of ultraslow microdialysis in combination with a miniaturised ﬂow-through biosensor. Clin. Chim. Acta, 316: 1–10. Rivera, A., Agnati, L.F., Horvath, T.L., Valderrama, J.J., de La Calle, A. and Fuxe, K. (2006) Uncoupling protein 2/3 immunoreactivity and the ascending dopaminergic and noradrenergic neuronal systems: relevance for volume transmission. Neuroscience, 137: 1447–1461.

Robinson, T.E. and Camp, D.M. (1991) The effects of four days of continuous striatal microdialysis on indices of dopamine and serotonin neurotransmission in rats. J. Neurosci. Methods, 40: 211–222. Robinson, T.E. and Justice, J.B. Jr. (1991). Microdialysis in the Neurosciences Techniques in the Behavioral and Neural Sciences, Vol. 7. Elsevier, Amsterdam. Rossell, S., Gonzalez, L.E. and Hernandez, L. (2003) One-second time resolution brain microdialysis in fully awake rats. Protocol for the collection, separation and sorting of nanoliter dialysate volumes. J. Chromatogr. B, 784: 385–393. Roth, R.H., Allikmets, L. and Delgado, J.M. (1969) Synthesis and release of noradrenaline and dopamine from discrete regions of monkey brain. Arch. Int. Pharmacodyn., 181: 273–282. Ruggeri, M., Merlo Pich, E., Zini, I., Fuxe, K., Ungerstedt, U. and Agnati, L.F. (1990) Indole-pyruvic acid increases 5-hydroxyindoleacetic acid levels in the cerebrospinal ﬂuid and frontoparietal cortex of the rat: a microdialysis study. Acta Physiol. Scand., 138: 97–98. Sandlin, Z.D., Shou, M., Shackman, J.G. and Kennedy, R.T. (2005) Microﬂuidic electrophoresis chip coupled to microdialysis for in vivo monitoring of amino acid neurotransmitters. Anal. Chem., 77: 7702–7708. Schiffer, W.K., Mirrione, M.M., Biegon, A., Alexoff, D.L., Patel, V. and Dewey, S.L. (2006) Serial microPET measures of the metabolic reaction to a microdialysis probe implant. J. Neurosci. Methods, 155: 272–284. Schutte, R.J., Oshodi, S.A. and Reichert, W.M. (2004) In vitro characterization of microdialysis sampling of macromolecules. Anal. Chem., 76: 6058–6063. Sevick-Muraca, E.M., Houston, J.P. and Gurﬁnkel, M. (2002) Fluorescence-enhanced, near infrared diagnostic imaging with contrast agents. Curr. Opin. Chem. Biol., 6: 642–650. Shain, W., Spataro, L., Dilgen, J., Haverstick, K., Retterer, S. and Isaacson, M. et al. (2003) Controlling cellular reactive responses around neural prosthetic devices using peripheral and local intervention strategies. IEEE Trans. Neural Syst. Rehabil. Eng., 11: 186–188. Sharp, T. and Zetterstro¨m, T. (1992) In vivo measurement of monoamine neurotransmitter release using brain microdialysis. In: Stamford, J.A. (Ed.), Monitoring Neuronal Activity a Practical Approach. IRL Press, Oxford, pp. 147–179. Shuaib, A., Xu, K., Crain, B., Siren, A.L., Feuerstein, G., Hallenbeck, J. and Davis, J.N. (1990) Assessment of damage from implantation of microdialysis probes in the rat hippocampus with silver degeneration staining. Neurosci. Lett., 112: 149–154. Silver, R., LeSauter, J., Tresco, P.A. and Lehman, M.N. (1996) A diffusible coupling signal from the transplanted suprachiasmatic nucleus controlling circadian locomotor rhythms. Nature, 382: 810–813. Silver, J. and Miller, J.H. (2004) Regeneration beyond the glial scar. Nat. Rev. Neurosci., 5: 146–156.

129 Stro¨mberg, I., Herrera-Marschitz, M., Hultgren, L., Ungerstedt, U. and Olson, L. (1984) Adrenal medullary implants in the dopamine-denervated rat striatum. I. Acute catecholamine levels in grafts and host caudate as determined by HPLC-electrochemistry and ﬂuorescence histochemical image analysis. Brain Res., 297: 41–51. Sun, L. and Stenken, J.A. (2003) Improving microdialysis extraction efﬁciency of lipophilic eicosanoids. J. Pharm. Biomed. Anal., 33: 1059–1071. Tang, A., Bungay, P.M. and Gonzales, R.A. (2003) Characterization of probe and tissue factors that inﬂuence interpretation of quantitative microdialysis experiments for dopamine. J. Neurosci. Methods, 126: 1–11. Ungerstedt, U. (1984) Measurement of neurotransmitter release by intracranial dialysis. In: Marsden, C.A. (Ed.) Measurement of Neurotransmitter Release In Vivo. Methods in Neurosciences, Vol. 6. Wiley, New York, pp. 81–105. Ungerstedt, U. (1991) Microdialysis – principles and applications for studies in animals and man. J. Intern. Med., 230: 365–733. Ungerstedt, U., Herrera-Marschitz, M., Jungnelius, U., Sta˚hle, L., Tossman, U. and Zetterstro¨m, T. (1982) Dopamine synaptic mechanisms reﬂected in studies combining behavioural recordings and brain dialysis. In: Kotisaka, M., Shomori, T., Tsukada, T. and Woodruff, G.M. (Eds.), Advances in Dopamine Research. Pergamon Press, New York, pp. 219–231. Ungerstedt, U. and Pycock, C. (1974) Functional correlates of dopamine neurotransmission. Bull. Schweiz. Akad. Med. Wis., 30: 44–55. Ungerstedt, U. and Rostami, E. (2004) Microdialysis in neurointensive care. Curr. Pharm. Des., 10: 2145–2152. Vizi, E.S. and Kiss, J.P. (1998) Neurochemistry and pharmacology of the major hippocampal transmitter systems: synaptic and nonsynaptic interactions. Hippocampus, 8: 566–607. Walker, M.C., Galley, P.T., Errington, M.L., Shorvon, S.D. and Jefferys, J.G. (1995) Ascorbate and glutamate release in the rat hippocampus after perforant path stimulation: a ‘‘dialysis electrode’’ study. J. Neurochem., 65: 725–731. Walker, Z. and Walker, R.W. (2005) Imaging in neurodegenerative disorders: recent studies. Curr. Opin. Psychiatry, 18: 640–646. Westergren, I., Nystrom, B., Hamberger, A. and Johansson, B.B. (1995) Intracerebral dialysis and the blood-brain barrier. J. Neurochem., 64: 229–234.

Westerink, B.H. and De Vries, J.B. (2001) A method to evaluate the diffusion rate of drugs from a microdialysis probe through brain tissue. J. Neurosci. Methods, 109: 53–58. Westerink, B.H. and Tuinte, M.H. (1986) Chronic use of intracerebral dialysis for the in vivo measurement of 3,4-dihydroxyphenylethylamine and its metabolite 3,4-dihydroxyphenylacetic acid. J. Neurochem., 46: 181–185. Winter, C.D., Iannotti, F., Pringle, A.K., Trikkas, C., Clough, G.F. and Church, M.K. (2002) A microdialysis method for the recovery of IL-1beta, IL-6 and nerve growth factor from human brain in vivo. J. Neurosci. Methods, 119: 45–50. Wisniewski, N., Klitzman, B., Miller, B. and Reichert, W.M. (2001) Decreased analyte transport through implanted membranes: differentiation of biofouling from tissue effects. J. Biomed. Mater. Res., 57: 513–521. Wisniewski, N., Rajamand, N., Adamsson, U., Lins, P.E., Reichert, W.M., Klitzman, B. and Ungerstedt, U. (2002) Analyte ﬂux through chronically implanted subcutaneous polyamide membranes differs in humans and rats. Am. J. Physiol. Endocrinol. Metab., 282: E1316–E1323. Woodroofe, M.N., Sarna, G.S., Wadhwa, M., Hayes, G.M., Loughlin, A.J., Tinker, A. and Cuzner, M.L. (1991) Detection of interleukin-1 and interleukin-6 in adult rat brain, following mechanical injury, by in vivo microdialysis: evidence of a role for microglia in cytokine production. J. Neuroimmunol., 33: 227–236. Yang, H., Peters, J.L. and Michael, A.C. (1998) Coupled effects of mass transfer and uptake kinetics on in vivo microdialysis of dopamine. J. Neurochem., 71: 684–692. Zahn, J.D., Trebotich, D. and Liepmann, D. (2005) Microdialysis microneedles for continuous medical monitoring. Biomed. Microdevices, 7: 59–69. Zhou, F., Braddock, J.F., Hu, Y., Zhu, X., Castellani, R.J., Smith, M.A. and Drew, K.L. (2002) Microbial origin of glutamate, hibernation and tissue trauma: an in vivo microdialysis study. J. Neurosci. Methods, 119: 121–128. Zini, I., Zoli, M., Grimaldi, R., Pich, E.M., Biagini, G., Fuxe, K. and Agnati, L.F. (1990) Evidence for a role of neosynthetized putrescine in the increase of glial ﬁbrillary acidic protein immunoreactivity induced by a mechanical lesion in the rat brain. Neurosci. Lett., 120: 13–16. Zoli, M., Jansson, A., Sykova, E., Agnati, L.F. and Fuxe, K. (1999) Volume transmission in the CNS and its relevance for neuropsychopharmacology. Trends Pharmacol. Sci., 20: 142–150.

CHAPTER 2.2

Principles of quantitative microdialysis Peter M. Bungay1,, Paul F. Morrison1, Robert L. Dedrick1, Vladimir I. Chefer2 and Agustin Zapata2 2

1 Division of Bioengineering and Physical Science, National Institutes of Health, Bethesda, MD, USA Integrative Neuroscience Section, Behavioral Neuroscience Branch, National Institute on Drug Abuse, National Institutes of Health, Baltimore, MD, USA

Abstract: In vivo microdialysis relies principally on diffusion as the mechanism for solute exchange between the tissue and the probe perfusate. Mathematical models for describing microdialysis have been formulated based on known concepts of diffusion through tissue and the membranes with which probes are currently constructed. The models predict that the rate of exchange is strongly inﬂuenced by processes that remove the analyte from the tissue interstitium. This presents an opportunity to derive additional information about the clearance processes from microdialysis measurements. Experimental and mathematical approaches that have been developed to date for exploiting this possibility are presented in a revised framework suitable for both hydrophilic and lipophilic analytes.

experimental calibration are addressed, the primary focus in this chapter will be on the mathematical description of microdialysis and model parameter estimation. The objectives will be to illustrate the utility of mathematical models as aids in understanding microdialysis behavior to better plan and interpret microdialysis experiments and to elicit additional information about the processes that affect analyte concentrations in vivo. Particular emphasis will be placed on the quantitative characterization of the rates of processes that eliminate analyte from the tissue interstitial space, such as cellular uptake, metabolism, and efﬂux to blood.

I. Introduction The expression ‘‘quantitative microdialysis’’ has been interpreted in various ways by its practitioners. Probably the most frequent connotation is the use of microdialysis to estimate concentrations of analytes that would have existed in the observation medium in the absence of the probe. In the context of this handbook, the observation medium is typically the extracellular space (ECS) in neural tissue, and the analytes of most interest are the neurotransmitters and neuromodulators, metabolites, other essential endogenous solutes and exogenous analytes, and also pharmacological agents and drugs of abuse. Many microdialysis users treat concentration estimation as an experimental problem of probe calibration. The risk in adopting an empirical approach to calibration lies in not appreciating what is required for calibrations to be appropriate. Although the problems of

II. Mathematical framework Previous mathematical descriptions of in vivo microdialysis (Amberg and Lindefors, 1989; Benveniste et al., 1989; Bungay et al., 1990; Morrison et al., 1991a, b; Chen et al., 2002a, b) have been directed at interpretation of microdialysis measurements obtained primarily with hydrophilic

Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

131

DOI: 10.1016/S1569-7339(06)16008-7 Copyright 2007 Elsevier B.V. All rights reserved

132

Nomenclature

G

B

l $ Y

compartment concentration of bound analyte in units of mass/length3 b equilibrium binding ratio deﬁned in Eq. (8) C analyte concentration in units of mass/ length3 C relative concentrations deﬁned in Eq. (68) in units of mass/length3 D diffusion coefﬁcient in units of length2/time D weighted diffusion coefﬁcient deﬁned in Eq. (80) in units of length2/time d distance, as in the lateral displacement of the inner cannula in Fig. 1, in units of length E extraction fraction; extraction efﬁciency G analyte generation rate in units of mass/ (length3 time) K weighted overall elimination rate constant deﬁned in Eq. (80) in units of time1 K equilibrium partition coefﬁcient K modiﬁed Bessel function of the second kind k ﬁrst order rate constant in units of time1 L relative loss deﬁned in Eq. (51) l length Mr molecular weight, Daltons m mass P permeability in units of length/time p microvascular permeability in units of length/time Q volumetric ﬂow rate in units of length3/time or length3/(time mass of tissue) R relative recovery deﬁned in Eq. (50) r radial distance from probe membrane axis in units of length S membrane surface area in units of length2 s microvascular surface area in units of length2/length3 of tissue t time V volume in units of length3 v annulus ﬂuid velocity in units of length/time z axial distance from inlet end of accessible probe membrane in units of length b decay rate constant in Eq. (16) in units of time1 d thickness of implantation trauma layer in units of length f volume fraction

y z

penetration depth deﬁned in Eq. (39) in units of length tortuosity deﬁned in Eq. (95) dimensionless quantity deﬁned in Eq. (70) dimensionless elimination modulus deﬁned in Eq. (42) azimuthal angle quantity deﬁned in Eq. (77) in units of length

Subscripts A arterial b blood c cellular compartment cann inner cannula d annulus ﬂuid; dialysate e extracellular compartment efﬂux unidirectional ECS-to-plasma movement ext external medium f ﬂuid compartment in membrane i inner surface of membrane inﬂux unidirectional plasma-to-ECS movement j index number M Michaelis–Menten m membrane; based on whole membrane volume n normal tissue nnf no-net-ﬂux o outer surface of membrane p plasma Y probe (annulus ﬂuid+membrane) q quiescent medium s solid matrix compartment of membrane t tissue; based on whole tissue volume tr implantation trauma layer V venous Superscripts app apparent value in inlet end of accessible membrane out outlet end of accessible membrane q quiescent conditions

133

r ws x N

reaction well-stirred conditions exchange far from probe

— ﬂow-rate-weighted radial average in annulus deﬁned in Eq. (25) 0 dimensionless quantity Brackets

Diacritical marks [] ^

time rate of change overall (probe+tissue)

solutes. For such solutes, the diffusional movement through tissue is assumed to be dominated by the interstitial ﬂuid pathway because of low cell membrane permeability. Since the interstitial path is tortuous and the available interstitial space is usually a small fraction of the tissue volume, the effective coefﬁcient of diffusion for a hydrophilic solute in tissue is considerably less than that in free solution. Highly lipophilic solutes, in contrast, can rapidly penetrate and partition preferentially into cell membranes. The additional contribution from diffusion within the cellular phase may augment their diffusion through whole tissue relative to hydrophilic solutes of comparable molecular weight. This proposition is supported by microdialysis studies with ethanol in skeletal muscle (Hickner et al., 1995) and brain (Gonzales et al., 1998). Quantitative analyses of these measurements (Wallgren et al., 1995; Gonzales et al., 1998) yielded values for the effective diffusion coefﬁcient of this lipophilic substance that were greater than expected for diffusion through ECS. In this presentation, incorporating diffusion through the cellular, as well as interstitial, phases will lead to a reformulation of a mathematical framework for microdialysis (Bungay et al., 1990; Morrison et al., 1991a, b) that is generalized to both lipophilic and hydrophilic solutes. The reformulation is also simpliﬁed in a manner that should facilitate interpretation of microdialysis measurements. The modeling begins by deﬁning the probe geometry shown schematically in longitudinal and cross-sectional views in Fig. 1. The semipermeable dialysis membrane is taken to be a cylinder of uniform thickness with inner and outer radii, ri and ro, respectively, inside of which is a cannula of outer radius, rcann. The inﬂowing ‘‘perfusate’’ is

follows a function symbol to indicate the variables on which it depends /S axial average from z ¼ 0 to z ¼ ‘m

usually directed through the annular space between the membrane and an inner cannula and leaves as the ‘‘dialysate’’ through the lumen of the inner cannula. A pump generally constrains the perfusate to ﬂow at a constant volumetric rate designated by Qin d . This may differ from the ﬂow rate of the efﬂuent dialysate, Qout d , if ﬂuid gain or loss occurs across the membrane. The length of the membrane segment accessible for solute exchange between the annular ﬂuid and the external medium will be denoted by ‘m. This probe conﬁguration of a single inner tubular cannula is sometimes termed a ‘‘concentric’’ design. However, in many commercial probes and most user constructed probes, the lateral position of the cannula is not constrained and is unlikely to be concentrically positioned. In Fig. 1, eccentric positioning of the inner cannula is speciﬁed by the displacement, dcann, between the axes of the cannula and membrane. Spatial locations will be speciﬁed by cylindrical coordinates r, y, and z, for the radial, azimuthal, and axial positions, respectively, relative to the origin ﬁxed on the probe axis at the inlet end of the accessible membrane. To reduce the complexity of the mathematics, a concentric alignment will be assumed in most of the treatment, which eliminates the y-dependence. The models are constructed from mass balances for the solute of interest (termed the analyte) formulated for each of the media: ﬂuid within the probe annulus (d), probe membrane (m), and external medium (ext). In the context of this book, the external medium in vivo is taken to be neural tissue (t); however, the theoretical framework is quite general and applicable to many other kinds of tissue. The intent in formulating the models is to describe the rate of exchange of analyte between

134

Fig. 1. Longitudinal and transverse schematic views of a microdialysis probe showing a cylindrical hollow ﬁber membrane with a single eccentrically positioned inner cannula. Net analyte movement across the membrane results in a difference between the inﬂowing out perfusate concentration, Cin d , and the exiting dialysate concentration, Cd . Diffusional loss or gain of analyte may be supplemented by out transmembrane convection associated with a difference between the perfusate and dialysate volumetric ﬂow rates, Qin d and Qd . Cylindrical coordinates, r, f and z, indicate location with respect to the origin positioned on the axis of the membrane at the inlet end. The geometry of the membrane is speciﬁed by the inner and outer radii, ri and ro, respectively, and the length, ‘m, of the portion accessible for analyte exchange. The single internal cannula of outer radius, rcann, is assumed to be aligned parallel to the membrane, but may be displaced a distance, dcann.

the annulus ﬂuid and the external medium that results in a net gain or loss of analyte by the annulus ﬂuid. The resulting models express relationships between the measurable analyte concentrations in the perfusate and dialysate, C din and C dout ; respectively, and the concentrations in the external medium, which for situations in vivo are the generally unknown analyte concentrations in ECS of the tissue surrounding the probe. The membranes employed in commercially available probes are permeable to ﬂuid to a greater or lesser extent (Snyder et al., 2001). Consequently, analyte may be carried through the membrane and tissue by ﬂuid movement, that is, by convection. In some circumstances, consideration of convective exchange may be necessary or desirable. Efforts have been made to incorporate a convective contribution in microdialysis analyses in when Qout d differs from Qd (Bungay and Gonzales, 1996; Gonzales et al., 1998). However, for this chapter, it will be assumed that the perfusate and dialysate ﬂow rates are the same and given by Qd, and analyte movement through probe membrane and tissue occurs purely by diffusion, which is usually the dominant mechanism. Higher ﬂuid permeability is typically associated with probes designed to extend microdialysis to larger molecular weight analytes, such as macromolecules. The present mathematical framework is applicable to

such solutes in conjunction with techniques to out minimize the difference between Qin d and Qd . One approach is to add an inert macromolecule to the perfusate to oppose the hydrostatic pressure driving ﬂuid loss by increasing the perfusate colloid osmotic pressure (Rosdahl et al., 2000). Net diffusional movement of analyte occurs because of spatial variation in free concentration. Thus, it is essential in microdialysis to consider that the analyte concentration is spatially non-uniform and that net analyte diffusion is in the direction of decreasing free concentration of analyte.

II.A. Differential mass balance equations for transients in the tissue A differential mass balance accounts for the rate of accumulation of analyte within a differential volume dV ¼ r dr dy dz.

(1)

The differential lengths, dr, r dy, and dz are small with respect to the overall probe membrane dimensions, but comparable with or greater than the length scales characteristic of any ﬁne-scale tissue heterogeneity. This assumption is invoked to model the tissue as a continuous medium. The ECS occupies a fraction, fe, of this continuous

135

volume. The remaining volume fraction, fc ¼ 1 fe ,

(2)

will be associated with a ‘‘cellular’’ compartment. Both the cellular and extracellular compartments are aggregates of subcompartmental spaces. For example, in neural tissue, the ECS is assumed to include synaptic and extrasynaptic spaces that are averaged over both the spatial and temporal ﬁne-scales associated with the local heterogeneity. Hence, the extracellular and cellular analyte-free concentrations, Ce[r,z,t] and Cc[r,z,t], are each uniform within dV, but can vary with time, t, and position within the global space. The tissue may be subdivided into more than one layer. Within a given layer, tissue properties, such as the compartment volume fractions, will be assumed uniform. The outer layer is considered normal tissue extending sufﬁciently far in the radial and axial directions for the extracellular analyte-free concentration to attain the same ﬁne-scale averaged level, CN e [t], as in the absence of the probe. An inner layer with one or more properties differing from normal can be included to simulate effects such as damage from probe implantation (Peters and Michael, 1998; Yang et al., 1998; Bungay et al., 2003; Chen, 2005a, b, 2006), ﬂuid accumulation between the probe and tissue, a fouling layer of protein and extracellular matrix deposition on the probe membrane, or abnormal tissue created as a result of a foreign body reaction to the probe. The amount of analyte in a compartment changes locally because of diffusion and various local processes of supply and removal. The rate of change by diffusion within a compartment is characterized by a diffusion coefﬁcient, D, and the proportion of the tissue volume available to the compartment, f. The rate of analyte generation is assumed to be independent of analyte concentration and, hence, can be represented by zero-order constants, G. In addition to dependence on a compartment volume fraction, the rate of change by other supply or removal processes is assumed proportional to analyte concentration and characterizable by ﬁrst-order rate constants, k. Although non-linear representations can be introduced for supply and removal processes

(Chen, 2003), the assumption of linearity is a key to considerable simpliﬁcation. A superscript r on a rate constant denotes removal by reaction and a superscript x denotes exchange between the extracellular compartment (e) and either blood plasma (p) or the cellular compartment (c). The direction of exchange is indicated by the leftto-right order of subscript pairs, ep, pe, ec, and ce. The time rate of change of the amounts in the two tissue compartments within the differential volume, dV, is equated to the sum of the rates of supply and removal. The resulting unsteady analyte mass balance for the extracellular compartment is then

fe b e

@C e ½r; z; t 1 @ @C e ½r; z; t ¼ fe De r @t r @r @r

accumulation

þ

diffusion x fe ðkpe C pA ½t kxep C e ½r; z; tÞ blood-ECS exchange

þðfc kxce C c ½r; z; t fe kxec C e ½r; z; tÞ ECS-cellular exchange þfe ðG e kre C e ½r; z; tÞ . ð3Þ synthesis and degradation

The unbound analyte concentration in the arterial plasma is denoted by C pA : The rate constant for unidirectional inﬂux from the plasma to the ECS, kpe, is related (Patlak and Fenstermacher, 1975) to the blood ﬂow rate per unit volume of tissue, Qb, and microvascular inﬂux permeability, ppe, through kpe ¼ ðQb K bp =fe ÞE p ,

(4)

where Ep is the plasma extraction fraction

Ep

ppe s C pA C pV ¼ 1 exp . C pA ðpep =ppe ÞC 1 Qb K bp e (5)

In the above, C pV is the analyte unbound concentration in the venous plasma, Kbp is a measure of the effective partitioning between whole blood and plasma, s is the microvascular surface area per unit volume of tissue, and pep is the microvascular efﬂux permeability. The permeabilities and rate constants

136

are interrelated by

concentration,

kxep =pep ¼ kxpe =ppe .

(6)

K te

The corresponding balance for the cellular compartment is f c bc

K te C t =C e ¼ be fe þ bc fc K ce , ð7Þ

Binding within both compartments is assumed to be rapid, reversible, and proportional to compartment-free concentration so that the extent of binding can be incorporated through equilibrium binding parameters, be ðC e þ Be Þ=C e

and

bc ðC c þ Bc Þ=C c , (8)

where B is the compartment concentration of bound analyte. All processes for net diffusional movement of analyte through the cellular compartment are combined into an apparent diffusion coefﬁcient, Dc. This is an operational deﬁnition for Dc. No attempt has been made to establish whether a degree of connectedness is presumed for the cellular compartment. The diffusion term was omitted from the balance in Morrison et al. (1991b). The contribution of diffusion in the axial direction has been neglected in the above balances. An approximate approach for incorporating the effect of axial diffusion is included below (see Eq. (44)). The linearity and reversibility of binding and extracellular-to-intracellular exchange implies that at equilibrium the ratio of the compartmental concentrations is a constant K ce C c =C e .

kt C e ½r; z; t,

ð10Þ

which has been simpliﬁed by deﬁning the following constant quantities. The equilibrium ratio of total tissue to free extracellular concentrations is

@C c ½r; z; t 1 @ @C c ½r; z; t ¼ fc Dc r @t r @r @r ðfc kxce C c ½r; z; t fe kxec C e ½r; z; tÞ þ fc ðG c krc C c ½r; z; tÞ.

@C e ½r; z; t 1 @ @C e ½r; z; t ¼ Dt r @t r @r @r þ Gt þ fe kxpe C pA ½t

(9)

Assuming extracellular-to-intracellular exchange to be sufﬁciently rapid that the two compartments are always in quasi-equilibrium collapses the two balances into a single balance. Adding Eqs. (3) and (7) leads to a balance for the total local concentration expressed in terms of the extracellular-free

(11)

the total generation rate is G t fe G e þ fc G c ,

(12)

the coefﬁcient for analyte diffusion through whole tissue is Dt fe De þ fc Dc K ce ,

(13)

and the analyte elimination rate constant for whole tissue is kt fe kxep þ fe kre þ fc krc K ce .

(14)

Far from the probe the contribution of diffusion to the analyte balance becomes negligible and Eq. (10) reduces to the ordinary differential equation dC 1 e ½t ¼ G t þ fe kxpe C pA ½t kt C 1 (15) e ½t. dt The ‘‘distant’’ or ‘‘far ﬁeld’’ extracellular concentration, CN e , is taken to be spatially uniform on the scale of the membrane length, ‘m. In other words, spatial averaging is assumed to remove the local variation in extracellular concentration associated with discrete sites of analyte supply and removal, such as synapses and blood vessels. The justiﬁcation for this assumption is that the length scale of the separation between these sites is typically on the order of 0.05 mm or smaller, whereas ‘m ¼ 1 mm or greater. It is assumed that the temporal variation in arterial plasma concentration can be described by a sum of exponential terms X C pA ½t ¼ C pj expðbj tÞ. (16) K te

j

Then Eq. (15) may be integrated directly to obtain an expression for CN e [t] (Morrison et al., 1991a, b). The other quantities in Eq. (15) have been assumed constant. Subtracting Eq. (15) from

137

Eq. (10) eliminates the terms for the rates of supply to leave, K te

@ðC e C 1 1 @ @ðC e C 1 e Þ e Þ ¼ Dt r @t r @r @r kt ðC e C 1 e Þ,

The boundary conditions at the membrane– tissue interface are continuity of the free aqueous concentration across the interface C f ½ro ; z; t ¼ C e ½ro ; z; t

ð17Þ

CN e

in which the spatial uniformity of has been utilized to express the new dependent variable as the difference Ce[r,z,t]CN e [t]. Solutions to Eq. (17) require coupling to a mass balance for the analyte in the probe. Two approaches that have been employed utilize different ways of coupling depending on whether tissue concentrations are time-independent (steady-state) or time-dependent (transient). The same probe mass balance shared by both approaches will now be developed. II.B. Quasi-steady-state mass balances for the probe The membrane will be treated as a uniform medium consisting of an aqueous ﬂuid (f) and a solid matrix phase (s). The separate differential mass balances for the two phases can be combined into the following joint equation by assuming rapid equilibration between the phases and no elimination processes other than diffusion, @C f 1 @ @C f r ¼ Dm , (18) r @r @r @t in which Cf[r,z,t] is the concentration in the ﬂuid phase and Kmf is a membrane-to-ﬂuid equilibrium partition coefﬁcient

and continuity of the total diffusive ﬂux @C f @C e ¼ Dt . Dm @r r¼ro @r r¼ro

(19)

Eq. (19) is analogous to Eq. (11), except that binding has been neglected in both phases, but analyte partitioning between the two phases has been included through the solid-to-ﬂuid partition coefﬁcient, Ksf. The possibility that both phases can contribute to the diffusive passage of analyte has been incorporated through a whole membrane diffusion coefﬁcient analogous to Eq. (13), Dm ff Df þ fs Ds K sf .

(20)

As in the tissue balance, the contribution of diffusion in the azimuthal and axial directions has been neglected.

(22)

Continuity of the free aqueous concentrations also applies at the annulus ﬂuid–membrane interface C d ½ri ; z; t ¼ C f ½ri ; z; t.

(23)

Because the annulus ﬂuid is in motion, both convection and diffusion contribute to the ﬂux into this medium. The continuity of ﬂux boundary condition at this interface utilizes an annulus ﬂuid permeability, Pdi ; to represent this mass transport ¯ d ½z; tÞ ¼ Dm @C f . (24) Pdi ðC d ½ri ; z; t C @r r¼ri The subscript ‘‘i’’ for permeabilities indicates that they are based on the membrane inner surface area. The overbar denotes a transverse-average annulus ﬂuid concentration deﬁned by the local volumetric ﬂow-rate-weighted average, Z þp Z ri ¯ C d ½z; t ¼ C d ½r; z; y; tvd ½r; yrrdy =Qd , p

rcann

K mf

K mf C m =C f ¼ ff þ fs K sf .

(21)

(25) where vd is the annulus ﬂuid velocity, which is assumed to be everywhere parallel to the membrane axis and is related to the volumetric ﬂow rate by Z þp Z ri Qd ¼ vd ½r; y rrdy. (26) p

rcann

In general, the permeability, Pdi ; varies with axial position. However, following the suggestion in Bungay et al. (1990), the variation is expected to be small and the following expression can be used to predict an approximate constant value for the permeability for a thin annulus (rircann5rcann), Pdi ¼

35Dd , 13ðri rcann Þ

(27)

with Dd denoting the free solution diffusion coefﬁcient. The adequacy of this approximation

138

has been substantiated in the case of ethanol microdialysis (unpublished observation) by a more detailed model of Wallgren et al. (1995). The approximation in Eq. (27) was obtained by modifying the analysis on pp. 291–297 of Bird et al. (1960) that can be used to predict an alternate approximation for the dialysate permeability for probes without an inner cannula (rcann ¼ 0) P di ¼

24Dd . 11ri

(28)

The left-hand sides of the differential mass balances (17) and (18) represent the rate of accumulation of analyte mass within a differential volume when the concentration is varying over time. At steady-state, the mass is not changing and the time derivative terms are then zero. In developing the mass balances for the analyte in the membrane and the annulus ﬂuid, a further simplifying assumption will be made that the rates of accumulation in these media are always zero ¯ d @C f @C s @C ¼ ¼ ¼ 0. @t @t @t

(29)

This requires that the membrane and annulus ﬂuid always be in quasi-steady-state with respect to the free concentration in the tissue at the interface with the probe. Support for quasi-steady-state in the membrane can be found in the diffusion resistance analysis of Chen et al. (2002a), who noted that the probe membrane resistance (proportional to the inverse of the permeability) achieved steady-state so rapidly that it would be constant most of the time. Under this assumption, the concentrations in the membrane and annulus ﬂuid can vary in time, but the amount of analyte in these regions is so small compared with that in the tissue that it does not inﬂuence the time course. Although the volumes of the membrane and annulus ﬂuid are generally small, this assumption may need to be re-examined in situations where only a small volume of tissue participates in analyte exchange with the probe and the transients are rapid. Applying this quasi-steady-state assumption to Eq. (18), integrating twice with respect to r and substituting boundary conditions

(21)–(24) leads to ¯ d ½z; tÞ PPi ðC e ½ro ; z; t C ro @C e ½r; z; t , ¼ Dt @r ri r¼ro

ð30Þ

in which PPi is deﬁned as the permeability of the probe that combines the permeabilities of the annulus ﬂuid and the membrane, PPi

1 1 þ Pdi Pmi

1 ,

(31)

where the membrane permeability is given by Pmi ¼

Dm . ri lnðro =ri Þ

(32)

For quasi-steady-state conditions, the basis for the permeabilities can be transferred to the membrane outer surface area giving 35ri Dd ; 13ro ðri rcann Þ 1 1 1 ¼ þ Pdo Pmo

P do ¼ PPo

Pmo ¼

Dm ro lnðro =ri Þ

and ð33Þ

and an alternate form for Eq. (30), @C e ½r; z; t ¯ . PPo ðC e ½ro ; z; t C d ½z; tÞ ¼ Dt @r r¼ro (34) Transferring the permeabilities from the inner to the outer membrane surfaces is equivalent to the constraint that ri D m

@C f ½r; z; t @C e ½r; z; t ¼ ro D t . @r @r r¼ri r¼ro (35)

Performing a mass balance on the annulus ﬂuid by relating the gain in analyte mass to the diffusional ﬂux from the tissue into the probe and utilizing Eq. (35) leads to Qd

¯ d ½z; t @C @C e ½r; z; t ¼ 2p ro Dt , @z @r r¼ro

(36)

139

which combined with Eq. (34) yields the probe mass balance shared by both the steady-state and transient models ¯ d ½z; t Q d @C ¼ PPo ðC e ½ro ; z; t C¯ d ½z; tÞ. 2pro @z

(37)

The derivation implicitly assumes that binding and chemical conversion of the analyte does not occur within the dialysate. The dialysate balance would need to be modiﬁed for situations in which reagents are added to the perfusate to bind or react with analytes (Sun and Stenken, 2003; Ao et al., 2004; Chen et al., 2004). II.C. Steady-state microdialysis models Much simpliﬁcation can be achieved in developing models for steady-state conditions. We proceed in the ﬁrst section below with the additional assumption that the tissue can be considered as a single region of uniform properties that is in intimate contact with the probe. The latter assumption implies that there is no hindrance to analyte exchange associated with the probe–tissue interface. Then, we review efforts to incorporate the effect of trauma induced by probe implantation in which the trauma is assumed to be conﬁned to a tissue region of altered properties interposed between the probe and the surrounding normal tissue. II.C.1. In vivo model for tissue with uniform properties The general solution to the tissue differential mass balance, Eq. (17), for steady-state conditions is C e ½r; z C 1 e ¼ A½zK0 ½r=G,

(38)

in which the axial variation is conﬁned to A[z], an as yet unknown function of z, and G is a parameter with units of length deﬁned by G

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ Dt =kt .

(39)

This ‘‘penetration depth’’ represents a characteristic radial distance over which Ce is disturbed by analyte exchange between the tissue and probe. The shape of the radial proﬁle is given by K0, the

modiﬁed Bessel function of the second kind of zero order. As a boundary condition for Eq. (38) at the probe– tissue interface, r ¼ ro, we apply the following deﬁnition for the permeability of a uniform tissue, Pto

Dt @ðC e ½r; z C 1 e Þ , @r C e ½ro ; z C 1 e r¼ro (40)

which results in Pto ¼

Dt K1 ½ro =G , G K0 ½ro =G

(41)

where K1 is the modiﬁed Bessel function of the second kind of ﬁrst order. Values for the Bessel functions can be found in mathematical handbooks (Abramowitz and Stegun, 1964) or numerical software, such as Microsoft Excel. Deﬁning the dimensionless elimination modulus as ro Y ¼ ro G

sﬃﬃﬃﬃﬃﬃ kt , Dt

(42)

yields two compact dimensionless forms for Eq. (41), ro Pto Y K1 ½Y ¼ K0 ½Y Dt

and

ro kt Y K0 ½Y . ¼ K1 ½Y Pto (43)

The right-hand sides of Eqs. (43) are functions of the modulus only, Y. Expressing these quantities as functions of a single-variable facilitates their use in graphing or spreadsheet software for analyzing quantitative microdialysis measurements. Eqs. (43) are plotted in Fig. 2 for a portion of the range of Y. A much wider range of Y may be encountered, as illustrated below in Section V. The inﬂuence of analyte diffusion in the axial direction has thus far been neglected. If axial diffusion were to be included in the tissue mass balance, the tissue permeability would be enhanced. Tong and Yuan (2002) found that the degree of enhancement varied with the membrane length and the penetration depth. These authors determined that this effect could be accurately incorporated through increasing the permeability

140

Eliminating Ce[ro,z] from the three ﬂux expressions leads to Qd @C¯ d ¯ ¼ P^ o ðC 1 e C d ½zÞ. 2pro @z

Fig. 2. Alternative representations for the diffusional permeability, Pto ; of the tissue surrounding a microdialysis probe, as derived from the linear, steady-state model for tissue with uniform properties. Dimensional parameters are the analyte coefﬁcient of diffusion, Dt, overall elimination rate constant, kt, and the membrane outer radius, ro. Both Dt and kt are based on the tissue volume. The non-dimensional ratios, ro Pto =Dt and ro kt =Pto are functions only of the abscissa, Y, the elimination modulus pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ deﬁned as the dimensionless group, ro kt =Dt : The equations indicated for each curve correspond to Eq. (43) of the text.

(46)

Eq. (46) can be directly integrated, since CN e is assumed uniform. The result is a relationship between the measurable perfusate and dialysate concentrations, Cdin and Cdout, and the unknown distant extracellular concentration, CN e , in terms of the properties of the probe and tissue, " # P^ o So E d ¼ 1 exp , (47) Qd in which Ed is deﬁned as the extraction fraction or extraction efﬁciency, Ed

out C in d Cd , 1 C in d Ce

(48)

and So is the membrane outer surface area, S o 2pro ‘m .

by a factor of approximately, 1+0.369G/‘m, which transforms Eqs. (43) to ro Pto ð1 þ 0:369G=‘m Þ Y K1 ½Y ¼ K0 ½Y Dt ro k t Y K0 ½Y . ¼ ð1 þ 0:369G=‘m Þ K1 ½Y P to

and ð44Þ

The simplicity of this correction factor suggests that it can easily be applied by iteration from an analysis in which the correction is initially ignored. The correction factor provides an a posteriori criterion, 0.37G/‘m51, for neglecting the contribution of axial diffusion in the tissue. The next step is to relate annulus ﬂuid concentrations to tissue concentrations by equating the analyte ﬂux out of the tissue, Eq. (40), the ﬂux into the probe, Eq. (37), and the ﬂux into the annulus ^ ¯ ﬂuid, P^ o ðC 1 e C d ½zÞ; in which Po is the overall permeability deﬁned as P^ o

1 1 þ PPo Pto

1 .

(45)

(49)

The dimensionless ratio in Eq. (48) is normally a fraction lying in the range, 0rEdr1. As a consequence of the assumed concentration linearity of the equations from which it was derived, Ed is independent of values the experimenter chooses for Cdin. Hence, this expression predicts that the value of Ed will be same whether analyte is diffusing from the tissue to the annulus ﬂuid in (sampling), CN e >Cd , or diffusing from the annulus ﬂuid to the tissue (delivery), Cdin>CN e . Although the value of Ed is independent of both Cdin and CN e , the limiting forms of Eq. (48) will be designated by speciﬁc names. The pure sampling limit of Cdin ¼ 0 will be denoted by ‘‘relative recovery, R’’, which is a more restricted sense than often implied by this commonly used term 1 Relative Recovery ðRÞ ðE d ÞC in ¼0 ¼ C out d =C e . d

(50) For the pure delivery limit of form is

CN e

¼ 0 the limiting

in Relative Loss ðLÞ ðE d ÞC 1 ¼ 1 ðC out d =C d Þ. e ¼0

(51)

141

The qualiﬁer, ‘‘relative’’ in these deﬁnitions is employed to indicate recovery and loss based on concentration, rather than amounts. In this context, it is important to consider, as well, the rate at which analyte mass exchanged between the tissue and the annulus ﬂuid. The rate of analyte delivery from the perfusate to tissue is out _ d ¼ Qd ðC in m d C d Þ.

(52)

Expressed in normalized terms, the model prediction for this rate is _d Q Ed m ¼ d in 1 ^ Po S o ðC d C e Þ P^ o So Qd ¼ ^ P S o o Po S o 1 exp . ð53Þ Qd Eq. (53) is also applicable for sampling. In that case, both the numerator and denominator in the far left term change sign. In addition, the model predicts the spatial variation in concentration in both radial and axial directions. The radial proﬁles in the membrane ﬂuid and tissue ECS are given by the right-handside expressions in the following equations C f ½r; z C 1 ln½r=ro 1 1 e ^ ¼ Po þ ¯ d ½z C 1 ln½ri =ro Pmo Pto C e (54) and C e ½r; z C 1 e ¼ ¯ d ½z C 1 C e

K0 ½r=G P^ o , K0 ½ro =G Pto

(55)

while the axial variation is contained in the lefthand-side denominator " # C¯ d ½z C 1 P^ o So z e ¼ exp . (56) 1 ‘m Qd C in d Ce In the above, the free ECS concentration far from the probe is obtained from the steady-state form of Eq. (15), x C1 e ¼ ðG t þ fe kpe C pA Þ=k t .

(57)

The function, A, in Eq. (38) was determined in the process of deriving the above proﬁles, as

demonstrated by substituting Eqs. (55) and (56) into Eq. (38). As an aid in displaying the radial proﬁles, the z-dependence can be removed by axial averaging. An axial average will be indicated by the brackets /S, as in this deﬁnition for axial averaging of the dummy variable, C, Z 1 ‘m C½r; z; t dz. (58) hCi½r; t ‘m 0 Applying deﬁnition (58) to proﬁles (54)–(56) yields hC f i½r C 1 Qd E d ln½r=ro 1 1 e ¼ þ , 1 ln½ri =ro Pmo Pto So C in d Ce (59) hC e i½r C 1 Qd E d K0 ½r=G e ¼ , 1 Pto S o K0 ½ro =G C in d Ce

(60)

and ¯ di C1 hC Q Ed e ¼ d . 1 C in C P^ o So d e

(61)

II.C.2. In vivo model for tissue with probe implantation trauma layer Inserting a microdialysis probe into tissue produces a variety of disturbances. Benveniste et al. (1987) observed altered blood ﬂow and glucose consumption in the vicinity of the probe. Dialysate dopamine (DA) concentrations transiently increased to high levels following implantation (Westerink and De Vries, 1988) suggesting that injury to dopaminergic neurons either caused abnormal release of intracellular stores or disruption of the active cellular reuptake systems. Clapp-Lilly et al. (1999) found histological evidence of loss of neurons up to 400 mm from microdialysis probes 40 h after insertion and tissue abnormalities at distances up to 1.4 mm away. Yang et al. (1998) placed a carbon ﬁber electrode within 40 mm of a microdialysis probe implanted acutely in anesthetized rats. They were unable to detect DA by fast scan cyclic voltammetry following stimulation of the medial forebrain bundle. They concluded that DA release was suppressed in the tissue close to the probe as a consequence of the implantation trauma. Based on the observations of Yang et al. (1998), Bungay et al. (2003), they proposed a mathematical model in

142

Because concentration linearity is assumed for both tissue layers, Ed remains independent of the value of Cdin, and is still equal between sampling and delivery modes. However, for the Cdin value at which there is no net transfer of analyte between probe and tissue, Cdin is equal to Ceapp, not CN e . Consequently, Ed may differ from the relative recovery, which remains deﬁned in terms of the normal tissue ECS concentration, as in Eq. (50), 1 R C out d =C e .

Fig. 3. Schematic for models simulating effects of tissue trauma produced by probe implantation. Trauma is assumed to alter the rates of processes that supply or remove analyte in a cylindrical layer of tissue of thickness, d, interposed between the probe and surrounding normal tissue. The properties of each layer are assumed uniform, but different between layers. Position is indicated by radial and axial coordinates, r and z, respectively, with the origin on the probe axis at the inlet end of the exchange region of the membrane. Adapted from Bungay et al. (2003, Fig. 5, p. 937).

which the trauma was conﬁned to a thin, concentric layer of tissue interposed between the probe and surrounding normal tissue as indicated in Fig. 3. Processes for either or both analyte supply and removal could be abnormal within the trauma layer. The model was ﬁrst posed as a generic representation for any analyte obeying linear kinetics and then applied to DA with the speciﬁc constraints that release was abolished and uptake was reduced in the trauma layer. In a subsequent more detailed analysis, Chen extended the model to permit the rates of release and uptake to be independently varied in the trauma layer (Chen, 2005a, b, 2006). The models predict that the extraction fraction still varies exponentially with perfusate ﬂow rate as described by Eq. (47). The tissue permeability, Pt, is now a composite function of the permeabilities of the two tissue regions. An important difference is that, instead of Eq. (48), Ed is now expressed in terms of an apparent extracellular concentration, C app e ; which can differ from the distant normal tissue concentration, CN e , Ed

out C in d Cd . app C in d Ce

(62)

(63)

II.C.3. In vitro models The extraction efﬁciency obtained from a microdialysis probe in vitro is signiﬁcantly different, in general, from the performance observed in vivo. This is to be expected because of the difﬁculty of creating an in vitro medium that adequately mimics in vivo diffusion characteristics and clearance processes. However, there are two limiting in vitro situations that deserve consideration. For quantitative microdialysis, the more important of the two is the case of a probe immersed in a well-stirred aqueous solution. Usually the external solution is of similar composition to the perfusate solution except for the analyte concentration. The intention in employing the term ‘‘wellstirred’’ is to imply that the conditions are such that, within the external solution, there is no signiﬁcant resistance (inﬁnite permeability) to analyte exchange with the probe. This is indicated by assuming the analyte concentration, Cext, is uniform throughout the external solution. The corresponding description of the extraction fraction is out C in PPo So ws d Cd E d in ¼ 1 exp . (64) Qd C d C ext The signiﬁcance of this expression is that it provides the means for determining the probe permeability, PPo ; by measurement of Edws. Combined with Eq. (47) and measurement of the in vivo Ed, this permits isolation of the tissue contribution to Ed and estimation of Pto : The other in vitro situation of note is the case in which the solution is quiescent, so analyte movement only occurs by diffusion. Background

143

vibration and natural convection effects can introduce a poorly characterized degree of stirring in a solution with a viscosity similar to water. Even if quiescence is maintained, the approach to steadystate is slow and, hence, difﬁcult to discern. For these reasons, measurement of Ed for a probe immersed in a solution within a vial is of little value, other than as a qualitative check on probe operation. Quiescence is sometimes simulated by placing the probe in an aqueous gel. No realistic steady-state solution to Eq. (17) exists in cylindrical coordinates for pure diffusion, that is, in the absence of an elimination term. Two approaches have been put forward that can provide approximate predictions for the quiescent Ed. Chen et al. (2002a) proposed that a limiting form of their transient model exhibited a quasi-steady-state for which the diffusional resistance in the quiescent medium could be described by an expression from the theory of Amberg and Lindefors (1989). Alternatively, Bungay et al. (1990) suggested that the extraction fraction under quiescent conditions, Edq, can be approximated from a model in spherical coordinates in which the probe membrane is represented as a sphere of the same outer surface area, So, E qd ¼

out C in d Cd 1 C in d C ext "

# 1 1 1 ¼ 1 exp ðSo =Qd Þ þ . PPo Pqo

ð65Þ

According to this approach, the approximate permeability predicted for a quiescent external medium is pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ Pqo ¼ Dq fq 2=ðro ‘m Þ. (66) In a dilute aqueous gel, the product of diffusion coefﬁcient and ﬂuid volume fraction, Dqfq, is similar to Dq.

microdialysis (Amberg and Lindefors, 1989; Benveniste et al., 1989; Benveniste and Hu¨ttemeier, 1990; Morrison et al., 1991a, b; Chen et al., 2002a, b). The models of Amberg and Lindefors and Benveniste et al. are deﬁcient in neglecting ECS clearance processes. The elegant model of Chen et al. is distinctive in being capable of both singleand dual-probe applications. Furthermore, Chen et al. have proposed a clever experimental and analytical protocol for evaluating extracellular effective volume fraction, diffusion coefﬁcient, and elimination parameters. The following analysis is a modiﬁcation of the treatment of Morrison et al. (1991b), which has the advantage that it accommodates time-dependent plasma and distant ECS analyte concentrations, while Chen et al. assume these to be constant. In deriving the steady-state models in Section II.C above, it was possible to perform a sequential integration of the differential mass balances in the radial and axial directions. Instead for the transient model, the z-dependence of the concentrations will be removed by axial averaging as deﬁned in Eq. (58). The axial averaged form of Eq. (34) serves as the basis for a boundary condition on the tissue mass balance, ¯ d i½tÞ ¼ Dt @hCe i½r; t PPo ðhCe i½ro ; t hC , @r r¼ro (67) ¯ in which /CeS and hCd i are the axial averaged relative concentrations deﬁned by hCe i½r; t hC e i½r; t C 1 e ½t; 1 ¯ d i½t hC¯ d i½t C ½t. hC

and

e

ð68Þ

The quasi-steady-state probe mass balance, Eq. (37), has to be integrated to provide a relation be¯ d i½t and the known tween the unknown function hC in perfusate concentration Cd . Eq. (37) can be expressed in a compact non-dimensional form as @C¯ d ½z; t=@z0 ¼ oðC ¯ e ½ro ; z; t C¯ d ½z; tÞ,

(69)

in which, II.D. Transient microdialysis models Several approaches have been employed in deriving models to describe time-dependent

$ ¼ PPo S o =Qd

(70)

and z0 ¼ z=‘m .

(71)

144

Integrating Eq. (69) with respect to z twice, ﬁrst from z0 ¼ 0 to z0 and then from z0 ¼ 0 to 1, yields Z 1 0 ¯ d i ¼ hC e ir¼r e$ eþ$z ðC e Þr¼ro dz0 hC o 0

$ þ C in Þ=$, d ð1 e

ð72Þ

in which the inlet concentration is represented by C¯ d ¼ C in at z0 ¼ 0. (73) d Replacing the remaining integral in Eq. (72) with the approximation Z 1 Z 1 0 0 eþ$z ðC e Þr¼ro dz0 eþ$z dz0 0

0

Z

1

0

ðC e Þr¼ro dz

0

¼ hC e ir¼ro ðeþ$ 1Þ=$,

ð74Þ

leads to the desired relation ¯ d i½t hCe i½ro ; t hC 1 $ ¼ ðC in Þ=$. d C e ½t hCe i½ro ; tÞð1 e

ð75Þ

Combining Eqs. (67) and (75) yields the boundary condition @hCe i½r; t in 1 hCe i½ro ; t C d þ C e ½t z ¼ 0, @r r¼ro (76) where z¼

Dt So =Qd . 1 exp½PPo So =Qd

(77)

The boundary condition at r ¼ ro, Eq. (76), has the same form as that given in Eq. (A1–12) of Morrison et al. (1991b), although the deﬁnition of z is different. The coefﬁcient of the derivative term in Eq. (A1–12) has a typographical error. It should have also contained a Kte factor. The z in Eq. (77) above replaces that entire coefﬁcient. The boundary condition at r ¼ N is the same as in Eq. (A1–11), hCe i½1; t ¼ 0. (78) The axial averaged form of the tissue mass balance, Eq. (17), is @hCe i½r; t 1 @ @hCe i½r; t ¼D r KhCe i½r; t, (79) @t r @r @r in which, to maintain similarity to the notation of Morrison et al. (1991b), the transient balance has been expressed in terms of the weighted coefﬁcients

deﬁned as D ¼ Dt =K te

and

K ¼ kt =K te .

(80)

The spatial uniformity of the b, D, G, and k parameters and of CN e [t] has been invoked in all of the axial averaging. Balance (79) was solved analytically subject to the boundary conditions (76) and (78) to produce an expression for the time-dependent radial concentration proﬁle, /CeS[r,t]. The details are given in Morrison et al. (1991b). An expression for the time-dependent extraction fraction was found by integrating the quasi-steadystate balance (36) with respect to z and combining the result with Eqs. (76) and (77) to obtain in C out d ½t C d in C1 e ½t C d Dt S o hC e i½ro ; t C in d ¼ in Qd z C1 e ½t C d hC i½r ; t C in e o d ¼ 1 ePPo So =Qd . in C1 ½t C e d

E d ½t

ð81Þ

Thus, Ed[t] can be determined from the solution for the axial-average extracellular concentration evaluated at the probe–tissue interface. The analytic solution in Morrison et al. (1991b) to the tissue Eqs. (76), (78), and (79), requires initial values, Cp[0] and /CeS[r,0] ¼ CN e [0]. The latter condition assigns an initial uniform value to the ECS concentration for all r and z. Substituting this initial condition into Eq. (81), together with Eq. (64), gives E d ½0 ¼ E ws d .

(82)

In other words, the predicted initial extraction fraction is the same as the steady-state value for well-stirred conditions in vitro as a result of imposing the constraint that the concentration proﬁles within the probe are always at quasi-steady-state.

III. Experimental methodology III.A. Calibration in vivo The prediction from the steady-state linear model that Ed is the same for sampling and delivery

145

modes provides the basis for several methods for probe calibration under steady-state conditions in vivo. Because external media differ considerably from tissue in both analyte diffusion and elimination characteristics, calibrations performed in vitro do not reliably predict overall probe performance in vivo. However, as indicated in Section II.C.3, the determination of the extraction fraction under well-stirred conditions in vitro, Edws, is a valuable adjunct procedure in quantitative microdialysis. A brief description of the principal in vivo calibration techniques is given below. More discussion is provided in numerous earlier reviews (Justice, 1993; Kehr, 1993; Parsons and Justice, 1994; Elmquist and Sawchuk, 1997; Chaurasia, 1999; Stenken, 1999; de Lange et al., 2000; Gonzales et al., 2002; Cano-Cebria´n et al., 2005; Plock and Kloft, 2005).

III.A.1. Retrodialysis Retrodialysis is a pure delivery approach in which a calibrator solute is delivered to the tissue from the perfusate. The relative loss, L, is measured for the calibrator and used as an estimate for the analyte relative recovery, R. The constraint, CN e ¼ 0, in retrodialysis may be achieved if the analyte is an exogenous solute, such as a drug, and the calibration using the analyte is performed before or after the experiment. This approach (sometimes called ‘‘retrodialysis by drug’’) presumes that Ed does not change as the calibration and experiment proceed. An alternative approach (‘‘retrodialysis by calibrator’’) uses a surrogate similar to the analyte of interest, such as a labeled form or an analog of the analyte. Ideally, the calibrant Ed value should mimic that of the analyte. Also, the presence of the calibrant should not inﬂuence the Ed or the pharmacokinetic behavior of the analyte (Sta˚hle, 1994) and vice-versa. In addition to the terms ‘‘retrodialysis’’ (Wang et al., 1991) and ‘‘delivery method’’, other early proponents suggested referring to the surrogate calibrant as an ‘‘internal reference’’ (Scheller and Kolb, 1991) or ‘‘internal standard’’ (Larsson, 1991). The prime advantage in using a surrogate calibrant is that the retrodialysis can be performed concurrently with the experiment, so that the calibrant and analyte

concentrations are both measured in the dialysate samples. Thus, changes in steady-state levels of Ed occurring during the experiment may be monitored in the form of the calibrant L. That the properties of the calibrant should be sufﬁciently close to those of the analyte was recognized in the initial studies (Larsson, 1991; Scheller and Kolb, 1991; Wang et al., 1993). According to the models in Section II, this refers to similarity in the diffusion coefﬁcients (Dd, Dm, and Dt) and the overall elimination rate constant (kt) appearing in the deﬁnitions for the permeabilities, Eqs. (33) and (40). Such a stringent requirement suggests care should be exercised in selecting a surrogate calibrant. Calibrants have been obtained by radiolabeling [14C-lactate (Scheller and Kolb, 1991) and 3H-glucose (Lo¨nnroth and Strindberg, 1995)] and stable isotope labeling (13C-quinolinic acid; Beagles et al., 1998). In numerous instances derivatives and chemically similar compounds have been used as calibrants in in vivo studies. Some examples are displayed in Table 1. The assumed equality between the calibrant loss, L, and the analyte recovery, R, has been conﬁrmed experimentally in only a few instances, for example Wang et al. (1993). It has been more usual for investigators to compare the L of the analyte to that for the calibrant as a criterion for judging the appropriateness of the calibrant. Other criteria have been employed. In some cases, similarity but not equality, of the L values has been deemed acceptable, since the calibrant/analyte L ratio was used as a correction factor in estimating the analyte ECS concentration from the analyte dialysate concentration and the calibrant L (Sarre et al., 1995; Cle´ment et al., 1998). In contrast to these efforts to identify surrogates with similar diffusion and elimination properties, there have been suggestions for using internal standards in which the relationship to the analytes of interest may be less well-deﬁned. This category of calibrants has included antipyrine for 3H-water, caffeine, and aminopyrine in the hippocampus (Terasaki et al., 1992); urea for glucose, lactate, and glycerol in muscle and subcutaneous tissue (Strindberg and Lo¨nnroth, 2000); urea for glucose in mammary adenocarcinomas (Ettinger et al., 2001); and inulin for insulin in muscle (Sjo¨strand

146 Table 1. Examples of the use of analogs as retrodialysis calibrators Analyte

Retrodialysis calibrator

Reference

Theophylline Zidovudine (AZT) Carbamazepine

Caffeine AZdU 2-Methyl-5H-dibenz(b,f)azepine-5carboxamide a-Methyldopa

Larsson (1991) Wong et al. (1992) and Wang et al. (1993) Van Belle et al. (1993, 1995)

L-DOPA, dopamine (DA) and their metabolites Fluconazole Codeine Morphine

Bupivacaine Fluorescein Amoxicillin Flurbiprofen

UK-54,373 (a ﬂuorinated derivative) Nalorphine Nalorphine Ropivacaine Diﬂuoroﬂuorescein Cefadroxil Naproxen

et al., 1999). Calibrants such as these may offer practical advantages under the circumscribed situations for which they are validated. Hydrophilic solutes of similar molecular weight in tissues such as muscle may exhibit similar L and R values if they are predominantly cleared from the ECS by the same mechanism of loss to the blood. However, for an analyte, such as glucose, that is cleared by additional mechanisms whose rates vary during the course of the experiment, the change in analyte R may be different than the L for a calibrant such as urea. Even when tested under basal conditions, L values obtained from urea differed signiﬁcantly from R values determined directly for glucose by steady-state in vivo calibration (Brunner et al., 2000). As emphasized in the discussion below, Ed and, hence, R can be strongly inﬂuenced by the intrinsic elimination rate represented by kt. Noting that tissue-speciﬁc differences had been observed for drugs as similar as theophylline and caffeine, Sta˚hle (2000) concluded that any suggestion of a general internal standard must be rejected.

III.A.2. Steady-state no-net-flux The use of the analyte as the calibrator avoids the need to establish suitability. This is the appeal of a calibration approach known as the ‘‘concentration-difference’’, ‘‘zero-net-ﬂux (znf)’’ or ‘‘no-netﬂux (nnf)’’ technique. The technique involves perfusing the microdialysis probe successively with

Sarre et al. (1995) Yang et al. (1996, 1997) Xie and Hammarlund-Udenaes (1998) Bouw and Hammarlund-Udenaes (1998) and Xie et al. (1999) Cle´ment et al. (1998) Sun et al. (2001) Marchand et al. (2005) Mathy et al. (2005)

solutions containing the analyte at known concentrations, Cdin, spanning the anticipated analyte ECS concentration, CeN. Each of the perfusions is performed at the same ﬂow rate, Qd, and continued until steady-state in the dialysate concentration is achieved. The difference, CdinCdout, is plotted against Cdin (as illustrated below in Section V, Fig. 12A). For systems with concentration linearity, the ‘‘concentration difference’’ plot yields a straight line, the slope of which is the desired calibration factor, Ed, according to the rearranged form of Eq. (48), out in 1 C in d C d ¼ E d C d ðE d C e Þ.

(83)

Curvature in such a plot may indicate concentration non-linearity. However, in simulating nnf measurements for cellular uptake and release processes with DA -like Michaelis–Menten kinetics, Chen (2003) found that such non-linearities might be difﬁcult to detect unless Cdin values substantially exceed the uptake saturation concentration. In most instances in which the technique has been employed, the information content represented by the magnitude of Ed has been ignored. Instead, the focus has been on the interpolated value of the concentration at which out C nnf ¼ C in d ¼ Cd .

(84)

This intercept is termed the ‘‘point-of-nnf’’ and the intercept concentration is taken to be the analyte ECS concentration unperturbed by the presence of

147

the probe C nnf ¼ C 1 e .

(85)

The technique was initially applied to the measurement of glucose in subcutaneous tissue (Lo¨nnroth et al., 1987), but its use rapidly expanded to neurotransmitters and other analytes in a variety of tissues. In a series of articles Michael and colleagues have argued that for DA the equality in Eq. (85) does not hold because of tissue damage from probe implantation (Lu et al., 1998; Peters and Michael, 1998; Yang et al., 1998, 1999). These authors proposed that Eq. (83) should be replaced by out in 1 C in d Cd ¼ Ed Cd R Ce ,

(86)

over basal ECS levels of DA is unresolved at this juncture. The inﬂuence of implantation trauma is exacerbated for neurotransmitters with high rates of clearance from the ECS, because only the tissue close to the probe is involved in the solute exchange. Except for the controversy arising from the trauma effects, acceptance of the validity of the nnf technique has led to its use in a large number of studies. There are some additional drawbacks to the nnf approach. Achieving steady-state for each of several successive perfusions tends to necessitate lengthy experiments with numerous dialysate samples. The requirement that Ceapp and CN e be independent of time for the duration of the experiment is a major limitation.

and at the point-of-nnf then C nnf ¼ ðR=E d Þ C 1 e .

(87)

These assertions have been accompanied by estimates for basal extracellular DA concentrations on the order of CN e ¼ 500 nM (Kulagina et al., 1999), whereas published values for DA Cnnf have been in the range of 4–12 nM. The authors concluded that Ed values obtained from the slope of the concentration difference plots greatly overestimate R for DA. The model presented above in Section II.C.2 for incorporating implantation trauma accommodates these possibilities in predicting, based on Eq. (62), that the point of nnf intercept corresponds to the apparent ECS concentration C nnf ¼ C app e .

(88)

Chen (2005b) concurred that Cnnf values underestimate CN e for DA, but to a much lesser degree. Based on evidence from voltammetry measurements, Chen concluded that the average basal ECS concentration for DA is within the range 7–20 nM in rat striatum. Michael et al. (2005) subsequently offered an alternative interpretation of the signals obtained by voltammetry following electrically stimulated DA release. They contend that the existence of basal ECS levels of DA above the Michaelis–Menten constant, KM, of the DA transporter could explain observed differences in elimination rates following evoked release and pressure injection of the transmitter. The controversy

III.A.3. Variation of perfusate flow rate Quantitative microdialysis could be said to have originated with Jacobson et al. (1985). These authors proposed both an experimental technique for estimating extracellular concentration and a mathematical model that the technique requires. In terms of the present notation, their model equation is " # 1 ^ o So P C out C d e ¼ exp . (89) 1 Qd C in d Ce This is equivalent to Eqs. (47) and (48), except that P^ o for Jacobson et al. was an empirical mass transfer coefﬁcient. The present theory instead provides an explicit expression for the overall permeability in Eq. (45) and the component permeabilities in Eqs. (33) and (44). The approach of Jacobson et al. uses successive perfusions at speciﬁc ﬁxed ﬂow rates. Each perfusion is continued until a steady-state dialysate concentration is obtained. Eq. (89) or other variants are employed in a non-linear regression of the paired values of measured Cdout and Qd to obtain estimates for unknowns, CN e , and the overall permeability. The published studies have made little use of the permeability results. As with the other calibration techniques, in the presence of implantation trauma, the concentration estimate is Ceapp, rather app than CN or CN e . The concentrations, Ce e must remain constant for an extended time interval,

148

which may partially explain why the method has not been widely adopted. Jacobson et al. (1985) illustrated the method with estimation of endogenous amino acid concentrations in the rabbit olfactory bulb. Ekblom et al. (1992) evaluated the use of Eq. (89) and alternative relations between Cout and Qd for the determination of morphine d levels in rat striatum. Parsons and Justice (1992) compared the ﬂow rate and nnf methods for estimating DA concentrations in the rat nucleus accumbens (NAc). More recently Tang et al. (2003) used the same two methods to demonstrate that, for DA in the rat nucleus accumbens, the overall permeability is predominately determined by probe properties, presumably because the avidity of uptake makes the tissue highly permeable to DA. Sta˚hle (2000) reported reliable parameter estimates were difﬁcult to obtain because of the nature of the non-linearity in Eq. (89) when Qd is used as the regression variable.

III.A.4. Dynamic no-net-flux In general, the calibrations obtained with the techniques described above are only reliable under steady-state conditions (see Section IV.E). Situations in which CN e varies over time are common, particularly in pharmacokinetic experiments. Consequently, calibration techniques valid under transient conditions would be desirable. Although not yet rigorously tested, dynamic no-net-ﬂux calibrations appear to be suitable. In this approach, the subjects are divided into groups according to the concentration of the analyte in the perfusate, that is, the replicate subjects in each group have probes that are perfused with solutions of the same Cdin value. A concentration difference plot is obtained for each time point of the experiment from the variation in CdinCdout values among the groups. The point-of-nnf intercepts provide a time proﬁle of the apparent ECS concentration, Ceapp[t]. The technique was proposed and elegantly illustrated by Olson and Justice (1993). They demonstrated that intraperitoneal bolus dosing of cocaine and amphetamine produces both transient elevation of the DA Ceapp[t] intercepts and transient depression of the DA Ed[t] slopes in the NAc. As a consequence of the latter, the increase in

dialysate DA relative to baseline underestimates the relative increase in Ceapp. These authors also showed, by contrast, that the relative time proﬁles are similar following administration of an agent, the DA receptor antagonist haloperidol, that alters DA ECS concentration but not DA elimination. III.A.5. Slow perfusion Slowing the probe perfusion raises Ed, according to Eq. (47). Under steady-state conditions, a sufﬁciently slow perfusion will allow the annulus ﬂuid to approach equilibration with the adjacent tissue at the outﬂow end resulting in Ed ¼ 1. While not a calibration technique per se, slow perfusion can be a means to facilitate estimation of ECS concentrations for both steady-state and transient operation. However, the sampling intervals tend to be long, unless the assay technique is suitable for very small sample volumes, which may limit the ability to track transients. This approach has been widely adopted in clinical applications of microdialysis where the use of long membrane lengths also promotes equilibration. The objective in these applications is typically to monitor ECS concentrations that may be changing over time. III.B. Measurement of spatial concentration profiles The value of the dialysate measurement techniques of the previous section can be powerfully augmented by complementary tissue measurements. Since microdialysis imposes a spatial variation in ECS concentration, methods for assessing spatial concentration proﬁles are especially useful. Autoradiography is a sensitive approach that is particularly suitable when no radiolabeled derivatives of the analyte are produced in vivo during the microdialysis. Dykstra et al. (1992) perfused probes in anesthetized rat caudate-putamen with solutions of 14C-sucrose followed by quantitative autoradiography. As an extracellular marker, sucrose is useful for determining the accessible extracellular volume fraction. The measurements suggested that the trauma of probe implantation resulted in edema in the surrounding tissue extending for distances of at least 1.5 mm from the probe.

149

A subsequent study involving perfusions with 14Czidovudine ﬁt radial proﬁle simulations from the transient model of Section II.D to autoradiograms (Dykstra et al., 1993). The ﬁts provided an estimate of the rate constant for zidovudine elimination from the caudate-putamen. Autoradiograms obtained following addition of probenecid to the perfusate, together with intraperitoneal probenecid administration, yielded a fourfold lower value for the elimination rate constant. The results were in agreement with other observations suggesting that zidovudine is a substrate for active efﬂux transporters in the blood-brain barrier. Groothuis et al. (1998) applied quantitative autoradiography to demonstrate alterations in the blood-brain barrier associated with probe or cannula insertion. Enhanced inﬂux of intravenously administered analytes persisted for periods of up to 28 days following implantation. Radiography is of questionable utility for volatile solutes. An alternative is to assay the analyte content of tissue slices cut parallel to the probe tract, as adopted for ethanol by Wozniak et al. (1991) and Gonzales et al. (1998). The disadvantage is that this approach yields rectilinear proﬁles that require additional analysis for comparison with the radial proﬁles generated by the mathematical models. A complementary aspect of autoradiography and excised tissue assays is that they yield information on whole tissue concentrations, whereas microdialysis measurements relate to the extracellular-free concentrations. III.C. Elimination rate constant from analyte mass retained in tissue This technique applies to the special case in which an exogenous solute is only being delivered to the tissue via the probe perfusate. At steady-state for _ d ; is the same as this situation, the delivery rate, m the rate at which the analyte is cleared from the tissue _ d. _t ¼ m m

(90)

The elimination rate is given by Z _ t ¼ kt C e dV , m

(91)

in which the integral is over the volume of tissue containing analyte that has diffused from the probe. The total amount of the exogenous analyte retained in this tissue volume is Z (92) mt ¼ C t dV . Determining this mass typically requires rapid cessation of the clearance processes by stopping blood ﬂow and freezing the tissue to inhibit further chemical conversion of the analyte. Combining Eqs. (90)–(92) together with Eq. (11) yields _ d =mt . kt ¼ K te m

(93)

The delivery rate by Eq. (52) is obtained from perfusate and dialysate assays. A separate equilibration experiment may be necessary to estimate the partition coefﬁcient, Kte. Eq. (93) then permits estimation of kt. This result is independent of the shape of the spatial concentration proﬁles. As pointed out by Gonzales et al. (1998), Eq. (93) should be valid when analyte delivery is inﬂuenced by convection through the ECS as well as diffusion.

III.D. Dual probe techniques These are techniques in which a delivery probe is perfused with an exogenous solute. Instead of excising the tissue surrounding the probe to determine analyte content as in the previous method, a second probe is employed to sample the analyte that diffuses from the delivery probe. The two probes are implanted in parallel orientation. Since the analyte ECS concentration decreases with distance from the delivery probe, the efﬂuent concentration from the sampling probe depends on the interprobe spacing. For a given separation distance, the ratio of concentrations from the sampling efﬂuent to the delivery perfusate is a function of a number of parameters, such as the elimination rate constant and the diffusion coefﬁcients in the tissue and probe membranes. The nature of the functionality can be better elucidated by repeating the measurements at various separation distances. De Lange et al. (1995) exploited this approach in measurements with

150

acetaminophen and atenolol in rat frontal cortex. These authors developed a mathematical model for calculating the ratio of the brain-to-blood transfer coefﬁcient to the effective diffusion coefﬁcient in the ECS. They reported that the ratio was larger for acetaminophen, the more lipophilic of the two analytes. This general approach was elaborated in consecutive articles (Chen et al., 2002a, b; Ho¨istad et al., 2002) that presented a method for separately estimating the elimination rate constant, the effective extracellular volume fraction, and the tissue diffusion coefﬁcient from transient dual-probe measurements. Thus far, this powerful technique has apparently been applied only to data obtained with 3H-mannitol (Ho¨istad et al., 2002).

III.E. Probe characterization in vitro Generally, both the probe and the tissue inﬂuence microdialysis measurements in vivo. To use the measurements to infer the inﬂuence of tissue processes requires isolating the probe contribution. As indicated in Section II.C.3, the probe permeability, PPo ; is a convenient measure of the probe characteristics. Eq. (64) suggests that PPo can be evaluated by measurement of Ews d , the extraction fraction for the probe immersed in a ﬂuid medium in vitro that is sufﬁciently well-stirred that the resistance to analyte transport is rendered negligible. Preferably, this is achieved by measuring Ed under a series of well-deﬁned ﬂow conditions of varying intensity and extrapolating to the Ews d limit. One such approach is described by Bungay et al. (2003), but the concept was demonstrated previously by Stenken et al. (1993). A less convincing, but simpler empirical approach, is to suspend the probe in a stirred beaker and increase the degree of stirring until a plateau value of Ed is obtained. Because of the temperature dependence of diffusion coefﬁcients, the measurements should be made at a known constant and uniform temperature. Comparison of delivery and sampling efﬁciencies in vitro may be useful for investigating technical problems, such as trapping or loss of analyte within the system between the perfusate pump and dialysate collection equipment.

There is the possibility that the probe permeability may not be the same in vivo and in vitro. Protein adsorption or other membrane fouling processes occurring in vivo could alter the membrane permeability. Exposure to a solution of bovine albumin in saline at a concentration similar to that in plasma reduced the permeability of polyacrylonitrile (PAN)–methallylsulfonate copolymer membrane but had no effect on the permeability of a cellulosic Cuprophan (CUP) membrane (Collins and Ramirez, 1979). Snyder et al. (2001) reported no signiﬁcant alteration in the permeability of polycarbonate (PC)–polyether, PAN, and CUP membranes when exposed to 4% bovine serum albumin or 0.3% ﬁbrinogen solutions. Although not a conclusive test of the lack of an in vivo and in vitro difference, similarity in Ews d values obtained before and after in vivo use would suggest that the difference is small. Chefer et al. (2006) found no signiﬁcant difference in mean Ews d values measured for DA in ﬁve probes before and after two separate experiments in mouse NAc conducted on consecutive days. Although experimental measurement of PPo is desirable, values can be estimated using the predictive expressions for the permeabilities of the annulus ﬂuid and membrane, Eq. (33). These expressions require values for the coefﬁcients of analyte diffusion in the annulus ﬂuid and membrane, Dd and Dm, respectively. Aqueous diffusivities may be estimated from correlations (Bird et al., 1960) or data in the literature for diffusion coefﬁcients of marker solutes. The plot of a representative

pdata ﬃﬃﬃﬃﬃﬃﬃ set for Dd in Fig. 4A illustrates a near 1 M r dependence on molecular weight for neutral hydrophilic solutes over the range 50oMro5,000 Da. In microdialysis membranes, hydrophilic solutes are usually conﬁned to the ﬂuid phase within which diffusion is slowed by tortuosity and hydrodynamic interaction with the solid matrix. As a consequence, Dm is signiﬁcantly lower than Dd and exhibits a stronger dependence on Mr, as evidenced by the Dm data in Fig. 4A and the corresponding values for Dm/Dd in Fig. 4B. These Dm values were measured in cuprophan(e) hollow ﬁbers. Similar Dm/Dd values have been obtained by Snyder et al. (2001) from microdialysis probes fabricated from cuprophan

151

IV. Analysis IV.A. Steady-state

Fig. 4. Molecular weight dependence of the diffusion coefﬁcient for neutral hydrophilic solutes: (A) the diffusion coefﬁcients in free solution (Dd) and in Cuprophan hollow ﬁber membranes (Dm), and (B) the ratio, Dm/Dd. Data from Klein and Holland (1977).

membranes. Snyder et al. also reported generally higher values and wider ranges for Dm/Dd for probes with PC–polyether and PAN membranes. The results suggested to these authors that analyte partitioning into the polymer might give rise to a solid phase diffusion contribution as incorporated operationally in Eq. (20). Membrane thickness and analyte charge are among the other factors that inﬂuence probe permeability (Zhao et al., 1995).

For systems that are linear in their concentration dependencies, the steady-state extraction efﬁciency predicted by Eq. (47) is a function of a singledimensionless ratio of parameters, Qd =ð2pro ‘m P^ o Þ; as displayed in Fig. 5A. The overall permeability, P^ o ; is in turn a function of several other important parameters, which complicates interpretation of this expression. However, Fig. 5A shows the direct inﬂuence of two independent parameters: membrane length, ‘m, and the perfusate ﬂow rate, Qd. Commercial probes are currently available in lengths between 1 and 30 mm. The practical range for ﬂow rates is as wide as that for ‘m, so these two parameters have a major inﬂuence on Ed. If ﬂow rate is increased, the efﬁciency goes down or if the probe is longer, the efﬁciency is greater. This provides the experimenter with considerable freedom to choose a preferred efﬁciency. For experiments in small animals, permissible membrane lengths are usually on the order of 1–4 mm. For such short lengths and typical ﬂow rates on the order of 1 mL/min, efﬁciencies tend to be low. In this range, Eq. (47) provides the means to compare or convert efﬁciency measurements obtained at different ﬂow rates or with probes of different length. The probes in clinical use are often much longer. For low-molecular-weight metabolites, such as glucose, lactate, and glycerol, at a ﬂow rate of 0.3 mL/min, the efﬁciency with these probes is close to 100% (EdE1). Maintaining operation in this range may avoid the need for probe calibration when the objective is to estimate the distant ECS analyte concentrations. This is particularly desirable if Ed remains near unity when the concentrations are changing over time. It should be noted, however, that this near equilibration between the dialysate and tissue may only occur near the outﬂow end of the membrane. In normal clinical use, the probes are perfused with an analyte-free solution implying a disequilibrium exists near the inﬂow end of the membrane (Sta˚hle, 2000). Varying Qd produces two somewhat opposing effects. The resulting tradeoff often governs the choice of Qd. As just discussed, Ed increases

152

interval or increasing the analyte mass in each sample to improve assay accuracy. However, the mass ﬂow rate asymptotes to a maximum given by 1 the absolute value of P^ o S o ðC in d C e Þ: With the short small animal probes, raising Qd may only yield more dilute dialysate samples of larger volume without much gain in analyte mass. As already stressed, a major consequence of concentration linearity is that the plots in Fig. 5 and the equations they represent apply for either sampling or delivery of analyte.

IV.B. Diffusion

Fig. 5. Steady-state in vivo microdialysis for analytes exhibiting concentration linearity. Predictions apply to either delivery N from the perfusate to tissue (Cin d bCe ) or sampling from tissue in N to annulus ﬂuid (Cd 5Ce ), where CN e represents the spatialﬁne-scale-averaged analyte extracellular concentration sufﬁciently distant to be undisturbed by the presence of the probe. (A). The prediction from Eq. (47) for the extraction efﬁciency, Ed, expressed as a fraction displaying the dependence on the perfusate volumetric ﬂow rate, Qd, and the overall permeability, P^ o ; based on the membrane outer surface area, So ¼ pro ‘m. Values of Ed51 are typical of many microdialysis measurements for short membrane lengths, ‘m, in small laboratory animals with Qd on the order of 1 mL/min for which the abscissa, Xb1. The upper limit of Ed1 is typical of clinical applications with low-molecular weight metabolites, long ‘m and Qd of 0.3 mL/min corresponding to X51. (B). Analyte perfusate-to_ d ; in units of mass per unit time and tissue net exchange rate, m normalized in Eq. (53) to be expressed as a fraction.

toward its maximum as Qd is reduced. In contrast, Eq. (53) predicts that lowering Qd lowers the rate at which analyte mass transfers between tissue and annulus ﬂuid, as shown in Fig. 5B. The higher _ d ; at higher Qd may be analyte mass ﬂow rate, m desirable for either decreasing the sampling

The experimenter has little control over the overall permeability, which can vary considerably among analytes and implantation sites. The contribution of the probe to the overall permeability is discussed in Section III.E on probe characterization. Understanding the contribution of the tissue is a primary focus of quantitative microdialysis. According to Eq. (43), diffusion and elimination are the two processes that, through the parameters Dt and kt, determine the permeability of the tissue to analyte. Each of these two parameters is a composite of several factors. Common factors for both, according to their deﬁning Eqs. (13) and (14), are the volume fractions of the extracellular and cellular compartments, fe and fc ¼ 1fe. A nominal value for brain is often quoted as fe ¼ 0.2, but measured values vary among the different brain structures under normal conditions (Nicholson, 2001). Considerable deviations in these values occur in pathological situations, such as ischemia and other kinds of injury (Nicholson, 2001). Insertion of microdialysis probes into tissue can produce adverse effects. Edema is apparently one of the effects of probe insertion trauma in the brain. Using autoradiography with 14C-sucrose as an extracellular marker, Dykstra et al. (1992) obtained estimates of fe ¼ 0.35–0.40 in the tissue surrounding single probes implanted in the striatum of the anesthetized rat. The edema appeared to extend over distances of at least 1.5 mm from the probes. The estimates were for times on the order of 1.5–2.5 h after implantation. Using a different technique (Section III.D) that involved two parallel probes implanted

153

a distance of 1 mm apart, Ho¨istad et al. (2002) estimated an elevated value of fe ¼ 0.30 for 3H-mannitol in the striatum of anesthetized rats from measurements made during a 2-h interval begun 1 h after the probes were implanted. For hydrophilic solutes with negligible cellular contribution to diffusion, Eq. (13) reduces to Dt De f e .

(94)

Coefﬁcients of diffusion in the ECS can be estimated from the relationship, De ¼ Dd =l2 ,

(95)

in which l, the tortuosity as deﬁned by Eq. (95), is a measure of factors such as the relative increase in path length and hindrance for diffusion in the conﬁnes of the ECS compared with diffusion in free, unbounded solution. As indicated pﬃﬃﬃﬃﬃﬃﬃ in Fig. 4A, Dd is inversely proportional to M r for solutes in the molecular weight range 60oMr o5,200, which covers the range for most analytes that have been studied by microdialysis. In the studies by KumeKick et al. (2002) of diffusion in brain slices with osmotically induced edema, tortuosity did not vary with fe at levels above normal, and was the same for two solutes of Mr ¼ 74 (tetramethylammonium, TMA+) and Mr ¼ 3,000 Da (Dextran). By contrast, l was a function of both Mr and fe for fe values below normal, as illustrated in Fig. 6. Thus, if De is unaffected by enlargement of fe, implantation trauma is likely to produce an increase in Dt proportional to the increase in fe, and, consequently, an increase as well in Pto according to Eq. (41). When the tissue predominates over the probe in determining Ed, the latter can be signiﬁcantly augmented by the trauma-induced edema. Figs. 4 and 6 permit estimation of a nominal range of Dt values anticipated for microdialysis analytes in the brain. For the molecular weights of 60 o Mr o 5,200, free solution diffusivities for non-binding hydrophilic non-electrolytes in Fig. 4A vary by about one order of magnitude, 2o Dd (cm2/s 10+6) o 18. For the range of l values in Fig. 6, then the corresponding range of ECS diffusivities from Eq. (95) would be about 0.36oDe (cm2/s 10+6) o6.5, assuming that the l values for TMA+ and dextran are representative for

Fig. 6. Tortuosity (l) and extracellular volume fraction (a ¼ fe) for 74 Da tetramethylammonium (TMA+) and 3,000 Da dextran in rat cortical slices measured by the realtime iontophoretic method. The changes in l and fe were brought about by varying the osmolality of the medium in which the slices were immersed. The vertical dotted line corresponds to normal brain tissue osmolality. From Kume-Kick et al. (2002, Fig. 4, p. 521). Reprinted with permission from Blackwell Publishing.

analytes at the low and high end of the Mr range. From Fig. 6, the range for effective extracellular volume fraction is 0.12 ofe o 0.42 leading to a prediction for hydrophilic analytes in this Mr range of 0.04 o Dt ¼ De fe (cm2/s 10+6) o 2.7. Low-molecular weight analytes that readily permeate cell membranes exhibit diffusivities in brain that signiﬁcantly exceed this upper limit. For example, Dt estimation from microdialysis experiments has yielded values for 3H–H2O of 7.5 10–6 cm2/s (Bungay et al., 1990) and for ethanol of 12 106 cm2/s (Gonzales et al., 1998). Lower diffusion coefﬁcients can be expected, as well, for two reasons. Microdialysis and the related technology of ‘‘open-ﬂow microperfusion’’ have been accomplished with solutes of higher molecular weight (Bodenlenz et al., 2005; Clough, 2005) and consequently lower diffusivity. Second, under transient conditions, binding in the ECS and partitioning into the cellular phase can reduce the weighted diffusion coefﬁcient, D. This effect has been incorporated for reversible and rapid linear

154

binding and partitioning through the deﬁnitions in Eqs. (11) and (80).

IV.C. Clearance processes A crucial feature of microdialysis is that it necessarily produces spatial variation in the analyte concentration in the perfusate, membrane, and the tissue. The nature of the spatial proﬁles is illustrated in Fig. 7, which displays radial proﬁles calculated from the steady-state model Eqs. (59)–(61). To simplify the presentation of the proﬁles, the zdependence has been removed by axial averaging. A family of proﬁles has been generated to indicate the major inﬂuence of the avidity of clearance processes on several aspects of the proﬁles and consequently on the microdialysis behavior. The constant values of all of the other parameters used for the calculations are listed in Table 2. For clarity, the proﬁles are of normalized free analyte concentration ratios that vary between 0 and 1. Fig. 7A represents delivery of analyte from the perfusate to the tissue in which the perfusate concentration, Cin d is greater than the distant ECS concentration, CN e , and Fig. 7B shows the obverse in case in sampling when CN e >Cd . Because of the assumption of concentration linearity, the shape of the corresponding proﬁles are the same, but inverted, between (A) and (B). Both sets of proﬁles were generated from the same Eqs. (59) to (61), except for the use in Fig. 7B of the substitutions ¯ d i C in ¯ di C1 hC hC d e ¼ 1 in 1 C1 C in e Cd d Ce hC e i C in hC e i C 1 d e ¼ 1 in . 1 in Ce Cd Cd C1 e

and ð96Þ

Varying kt produces three notable changes as exhibited by these spatial proﬁles. First, the smaller the value of kt, the farther the proﬁle extends into the tissue. In other words, the penetration depth in Eq. (39) is greater for a slowly cleared solute than for one that is rapidly cleared. Second, the fraction of the overall concentration drop that occurs across the membrane increases as the elimination rate constant increases. This means that the relative contribution of the probe and the tissue to the

Fig. 7. Steady-state axially averaged radial concentration proﬁles for an analyte within the membrane and tissue surrounding a microdialysis probe. The proﬁles were calculated in Microsoft ¯ d i; Excel from model Eqs. (59) to (61) with /CS representinghC /CfS, and /CeS in the annulus ﬂuid, membrane, and extracellular space (ECS), respectively. The parameter values employed in the simulations are given in Table 2. The small difference between the annulus ﬂuid, hC¯ d i; and the interface membrane ﬂuid, hC f ir¼ri ; concentrations has been neglected. The value for the elimination rate constant, kt, was varied to illustrate the important role of the extracellular clearance processes in microdialysis performance. The kt value strongly affects the inﬂuence of the tissue relative to that of the membrane and the volume of tissue participating in analyte exchange with the N probe. (A) Normalized proﬁles for analyte delivery, Cin d >Ce . in (B) Normalized proﬁles for analyte sampling, CN e >Cd , that have same shape as in A, but inverted. For very rapidly cleared analytes, indicated by the curve labeled neurotransmitter, only a thin volume of tissue participates and Ed is primarily determined by the membrane properties.

efﬁciency depends on the elimination rate. Third, the extraction efﬁciency values, shown to the left of the ﬁgures, increase as elimination rate increases. These three features are underscored most dramatically by the additional curve in Fig. 7B

155 Table 2. Parameter values used in illustrative simulations displayed in Figs. 7–11 for in vivo microdialysis in neural tissue of a 300 Da non-binding analyte Parameter

Values

Units

Source

rcann ri ro ‘m Qd Kte Dd Dm D ¼ Dt/Kte K ¼ kt/Kte

0.012 0.020 0.025 1 1 1 7.5 106 1.0 106 6.0 107 0.1, 1, 10, 10,000

cm cm cm cm mL/min cm2/s cm2/s cm2/s mL/(h g tissue)

Correlation in Fig. 4A Assumed Dm/Dd ¼ 0.13 (Dt/l2) fe with l ¼ 1.6, fe ¼ 0.2

Table 3. Selected analyte elimination rate constants in brain Analyte

kt

Mechanism

Site

Sucrose Zidovudine (AZT) Quinidine Ethanol DA

mL/(h g) 0.02 0.8 6 120 6,000

Passive efﬂux Active efﬂux Active efﬂux Passive efﬂux Active uptake

Rat Rat Rat Rat Rat

labeled ‘‘neurotransmitter’’, which approximates the behavior expected for an analyte with characteristics similar to DA: (1) The clearance is so rapid that almost all of the neurotransmitter sampled by the probe comes from just a thin layer of tissue adjacent to the probe. (2) Most of the concentration drop from CN e to ¯ d i occurs across the membrane indicating hC that the overall permeability is dominated by the membrane contribution, because the tissue permeability to the neurotransmitter is very high. (3) Since the probe permeability predominates, Ed ¼ 0.75 is close to the upper limit given by the well-stirred in vitro value, Ews d ¼ 0.79. The increase in Ed with increasing kt is intuitively reasonable in the delivery situation, since the more rapid elimination of analyte from the ECS equates to a higher rate of extraction from the annulus ﬂuid. In contrast, the direct relationship between Ed and kt is perhaps counter-intuitive in the sampling case, since the clearance processes in the ECS

Reference

striatum frontal cortex striatum striatum

Fenstermacher and Rapoport (1984) Dykstra et al. (1993) Sun et al. (2001) Gonzales et al. (1998) Bungay et al. (2003)

would seem to remove analyte that would otherwise be available to diffuse to the probe. An explanation for the observed result relies on the inherent spatial non-uniformity in the ECS concentration. Since the rate of elimination from the ECS is proportional to the local ECS concentration, the lower Ce implies a lower rate of elimination and hence more analyte in the ECS that is available to diffuse to the probe. The rate constant, kt, is an ampliﬁcation factor for this effect. The higher the kt value, the more elimination is suppressed. Thus, for linear behavior, Ed is the same for sampling and delivery. This corresponds to the symmetry between Figs. 7A and B and parallels the observation that the slope of the nnf plot is continuous about the Cnnf intercept. Arbitrary values of kt were chosen in constructing Fig. 7. The values vary over ﬁve orders of magnitude. As indicated in Table 3, kt values encountered in brain microdialysis span at least as wide a range. The low end of the range represents removal from the ECS of inert hydrophilic solutes by passive diffusion to blood across the bloodbrain barrier. The average value for sucrose of

156

0.02 mL/(h g tissue) from Table 4 of Fenstermacher and Rapoport (1984) was found to be suitable in simulating sucrose microdialysis measurements in rat striatum (Dykstra et al., 1992). A 6,000-fold higher estimate was obtained in the same organ for passive transport of the lipophilic agent, ethanol. The efﬂux of drugs, zidovudine (AZT) and quinidine, that are substrates of active transporters in the blood-brain barrier was characterized by intermediate rate constant values. At the upper end of the range, quantitative microdialysis of DA yielded a kt estimate equivalent to 6,000 mL/(h g tissue). Analysis suggested that this value was reduced by the effects of probe implantation trauma compared with a value of 19,000 mL/(h g tissue) obtainable from cyclic voltammetry measurements (Bungay et al., 2003). Considering the major inﬂuence of clearance processes in microdialysis, one might anticipate that concentration non-linearities in analyte clearance ought to affect quantitative microdialysis measurements. Many clearance mechanisms, such as blood-brain barrier efﬂux pumps, cellular uptake transporters, and enzymatic catabolism exhibit saturable non-linear kinetics. Consequently, in such situations kt is typically dependent on analyte ECS concentration and one would expect to observe concomitant concentration dependence in Ed values. However, there have been few reported instances in which changes in Ed were directly related to non-linearity in clearance. Chen (2003) has shown that nnf measurements can be markedly insensitive to non-linear uptake and release processes. One report illustrating a non-linear effect involved microdialysis of caffeine in rat striatum (Song and Lunte, 1999). Caffeine is eliminated by active transport across the blood-brain barrier by a saturable carrier-mediated process. Caffeine Ed values were obtained by nnf measurements in animals with steady-state levels of caffeine produced by continuous intravenous infusion. Lower Ed values were obtained when the background caffeine concentration was about 50 mM than for background levels of about 5 mM. These levels were respectively above and below the 15–25 mM levels for saturation of the transport system by the competitive inhibitor, adenine. Consequently, the concentration dependence

of Ed was ascribed to saturation of the efﬂux mechanism. In this same study, no concentration non-linearity was observed in Ed values obtained by retrodialysis of caffeine solutions over the range of 1 mM–1 mM (Song and Lunte, 1999). One relevant difference between the retrodialysis and nnf conditions was that signiﬁcant caffeine levels were present in the plasma for the latter, but not the former, experiments. Another retrodialysis study of blood-brain barrier transport (Sun et al., 2001) also found apparently contradictory observations between transporter substrate and inhibitor pairs. These authors suggested that inhibition of efﬂux transport does not necessarily produce a reduction in Ed. Rather, the effect on Ed may depend on a number of factors, such as the localization and mechanism of the transporter. The conclusions were based on elaboration of the steady-state model of Section II.C.1 illustrating that both agreement and lack of agreement with the predictions of the present simpliﬁed models provide support for favoring quantitative microdialysis approaches and further reﬁnement of the models.

IV.D. Start-up transients As emphasized above, spatial concentration proﬁles are an inherent feature of in vivo microdialysis. The development of the spatial proﬁles over time introduces unavoidable initial transients in the extraction efﬁciency, even for situations in which the distant ECS concentration is a constant. The transients are modiﬁed when the distant ECS concentration varies over time as well. The effect of the two sources of transients will be considered in turn. Both kinds of transients are associated with the addition or removal of analyte from the tissue and, hence, are ‘‘mass transfer’’ transients. The initial variations for constant CN e will be termed ‘‘startup’’ transients, while those for time varying CN e will be discussed as ‘‘pharmacokinetic’’ transients (Bungay et al., 2001). The association between the development of spatial proﬁles and the transient in Ed is illustrated by the simulations in Fig. 8 for a sampling case. Each radial proﬁle displayed is for a speciﬁc time

157

Fig. 8. Start-up transient in the extraction efﬁciency, Ed[t], occurs as the spatial concentration proﬁles develop in the membrane and tissue. This mass transfer transient is illustrated with simulated radial concentration proﬁles for various times after starting to perfuse a probe that is sampling from a tissue with a constant distant analyte concentration, CN e . The analyte is being cleared from the tissue at a rate characterized by the weighted elimination rate constant, K ¼ 1 mL/(h g tissue). Other parameter values for the simulation are given in Table 2. The steady-state concentration proﬁle at t-N is the same as the corresponding proﬁle in Fig. 7. Initially, the ECS concentration is uniform, /CeS ¼ CN e , and the instantaneous initial extraction efﬁciency, Ed[0], is the same as the well-stirred in vitro value, Ews d , as predicted in Eq. (82). Adapted from Bungay et al. (2001, Fig. 2, p. 363).

following a change in the microdialysis protocol. The analyte concentration in ECS is assumed to be uniform at t ¼ 0, so the protocol change could represent the probe implantation or the start of probe perfusion, but the same kind of transient occurs with a change of analyte concentration in the perfusate solution, Cin d . To satisfy the quasisteady-state assumption in Section II.B, the proﬁles in the annulus ﬂuid and membrane at t ¼ 0 are fully developed with respect to Cin d and the uniform Ce. As a result, the proﬁle at t ¼ 0 is the same as that for well-stirred conditions in vitro and the instantaneous initial extraction efﬁciency is Ed[0] ¼ Ews d , as required by Eq. (82). The extraction efﬁciency rapidly decreases as a non-uniform concentration proﬁle develops in the ECS. Proﬁle development occurs through loss of analyte from the tissue by diffusional exchange between the probe and tissue and by the clearance mechanisms. The loss occurs from the ECS and the cellular phase for analytes that distribute into both

Fig. 9. The start-up transient in the extraction efﬁciency is a strong function of the weighted elimination rate constant, K. The magnitudes of the efﬁciencies at steady-state, Ed[t-N], indicated to the right of the plots are the same as the corresponding values in Fig. 7. Adapted from Bungay et al. (2001, Fig. 1, p. 363).

compartments. The compartmental mass balances for the transient model assume rapid equilibration between compartments. The relative proportion of analyte in the two compartments is prescribed by the partition coefﬁcient, Kte. A larger value of Kte implies more analyte to be removed from the tissue and, hence a slower transient. This is the intuitive reason that under transient conditions the relevant diffusion and elimination parameters become the weighted forms, D ¼ Dt/Kte and K ¼ kt/Kte. In the discussions above of the steady-state behavior, the focus is on Dt and kt because Kte derives from unsteady accumulation terms that do not appear in the steady-state mass balance equations. Both D and K inﬂuence the speed of the transients. However, as indicated in Sections IV.B and IV.C, the range of kt values is several orders of magnitude larger than for the Dt values anticipated in microdialysis studies. Also, it is more common to encounter alteration in kt, rather than Dt, values during the course of microdialysis experiments. The inﬂuence of K on transients is illustrated in Fig. 9. In this example, while a 100fold variation in K produces about a threefold variation in the steady-state extraction efﬁciency, the effect on the speed of the transients is much more dramatic. For the value of K ¼ 10 mL/(h g tissue), Ed approaches to within 5% of its steadystate value within 7 min, which is less than a

158

common dialysate sampling interval of 10 min. This behavior may also be representative of sampling in many non-neural tissues in which exchange between the ECS and blood may occur with rate constants of this order. However, in neural tissue with the low permeability of the microvasculature to non-transported hydrophilic solutes, transients may be much slower. For the K value of 0.1 mL/(h g tissue) in Fig. 9, Ed still differs from its steady-state value by more than 10% after 4 h. Although the time course for Ed is not exponential, 1/K appears to provide an indication of the transient time scale. However, the long slow approach to steady-state for a system with low K suggests that it would be difﬁcult experimentally to determine the steady-state Ed value.

IV.E. Pharmacokinetic transients Variations over time in CN e will induce changes in Cout . However, the changes involve adjustment in d the entire spatial proﬁle of Ce concentrations, which introduces a time lag. This implies concomitant changes over time in Ed that are superimposed on the start-up transients discussed in the previous section. If Ed varies over time, then, in general, it cannot be the same as the single value obtained from a probe calibration performed under steady-state conditions. A corollary conclusion is that the steady-state Ed value is the only generally valid product obtainable from most calibration techniques. These conclusions are supported by simulations generated by the transient microdialysis model. The model, in turn, can aid in estimating the magnitude of the differences to be expected between steady-state and transient Ed values. This will be illustrated by simulations for the idealized depictions in Fig. 10 of two common pharmacokinetic experiments in which analyte is administered intravenously. In simulation A in Fig. 10, the analyte is given by a loading dose followed by a continuous infusion. The infusion is assumed to be programmed to achieve a unit step change in the plasma-free concentration. For linear exchange, the step change between two constant levels in the plasma

Fig. 10. Analyte arterial plasma, C pA ½t; and distant extracellular, CN e [t], free concentration–time proﬁles for two idealized pharmacokinetic simulations. (A) i.v. infusion: Intravenous loading dose with programmed infusion producing a step change in C pA at t ¼ 0 and an exponential rise over time in CN e . Symmetric tissue-to-plasma exchange is assumed with the rate constant values, Kinﬂux ¼ Kefﬂux ¼ 1 mL/(h g tissue) so that at steady-state, C1 e ¼ C pA : Other parameter values are given in Table 2. (B) i.v. bolus: Intravenous bolus administration with step increase in C pA followed by monoexponential decay with a rate constant of b ¼ 1 hr1. Except for b, the parameter values are the same as for the iv infusion case. (C) Simulation for the pharmacokinetic transients in extraction efﬁciency for the experiments of (A) and (B). The weighted elimination rate constant represents analyte loss from the ECS to blood as the only clearance process. The iv infusion experiment leads to a steady-state with a constant distant ECS concentration and a constant extraction efﬁciency indicated by the dotted horizontal line. However, since the distant ECS concentration decays to zero in the iv bolus case, the extraction efﬁciency continues to monotonically decrease toward zero over the course of the entire experiment. Adapted from Bungay et al. (2001, Figs. 3 and 4, p. 364).

can be seen to result in an exponential change over time in the distant ECS-free concentration due to microvessel permeation. The plasma-to-ECS exchange is assumed to be symmetric and other

159

elimination processes in the tissue are neglected, so that at steady-state the distant ECS-free concentration is the same as the plasma-free concentration. The constant plasma concentration is speciﬁed by zero values for the plasma decay rate constants in Eq. (16). Simulation B assumes the same unit step change, but then the plasma-free concentration decays monoexponentially with a rate constant of b ¼ 1 hr1. The response in the distant ECS-free concentration exhibits a peak and a delayed decay typical of a linear lumped compartment pharmacokinetic model. The microdialysis extraction efﬁciencies calculated from the intravenous (i.v.) infusion and i.v. bolus simulations are displayed in Fig. 10C. The simulations are for the particular value of K ¼ 1 mL/(h g tissue). In comparing the two simulations, there is an important qualitative difference between the two time courses for Ed. A steadystate is achieved in the iv infusion case with the value of Ed ¼ 0.17 that is the same as the corresponding example in Fig. 9. However, the steadystate in the iv bolus case corresponds to the complete elimination of the analyte. The corresponding response in Ed is a continuous slow decay toward zero. There are a number of inferences that may be drawn from this comparison. The ﬁrst is that different time courses for CN e result in different time courses for Ed. Another is that steady-state probe calibrations are generally inappropriate for pharmacokinetic experiments, since Ed may vary throughout the experiment. A third is that the Ed transient obtained from a retrodialysis calibrator is a start-up transient for the calibrator, which generally does not correspond to the Ed transient for the analyte. The start-up transient for K ¼ 1 mL/(h g tissue) in Fig. 9 is faster than both of the experiment transients in Fig. 10C, which suggests that any kind of pharmacokinetic transient superimposed onto a start-up transient will probably yield a slower overall transient. Of the four calibration techniques described in Section III.A, only dynamic nnf would seem to be generally valid in pharmacokinetic experiments. The ﬁfth approach of slow perfusion is a much simpler means for ECS concentration estimation provided the times characteristic of the transients in CN are sufﬁciently slow compared e with the sampling interval.

IV.F. Area-under-the-curve determinations The preceding section highlights the problems in estimating time varying ECS concentrations when Ed6¼1, because of the difﬁculties in determining the transient Ed values. However, often the time integral of the ECS concentration sufﬁces in pharmacokinetic applications, instead of concentrations at discrete time points. The full areaunder-the-curve (AUC) integral is deﬁned as Z 1 AUC01 C1 (97) e dt. 0

For a pure sampling situation (Cdin ¼ 0), the true concentration in the integrand is related to the out dialysate concentration by CN e [t] ¼ Cd [t]/Ed[t]. Since in general Ed[t]6¼Ed[N] under transient conditions, an apparent ECS concentration calculated from the ratio, Cout d [t]/Ed[N], would differ from the true concentration at most time points. However, for analytes exhibiting concentration linearity, the following equality of the true and apparent concentration–time integrals has been found to hold (Bungay et al., 2001), Z 1 out AUC01 ¼ C d ½t=E d ½t dt 0Z 1 ¼ C out ½t dt =E d ½1, ð98Þ d 0

that is, the AUC integrals agree, even though the instantaneous true and apparent concentrations disagree. This behavior is illustrated in Fig. 11 for a simulation of the iv bolus experiment in Fig. 10. At early times, the apparent concentration provides an overestimate of the true concentration, but then the curves cross and the apparent concentration underestimates the true concentration. According to Eq. (98), the procedure for obtaining an AUC can be simpliﬁed. The second integral can be approximated by pooling a singledialysate sample collected continuously over a sufﬁciently long time interval, 0rtrT, Z 1 Z T out C out dt C out d d dt ¼ T ðC d Þpooled . 0

0

(99)

160

Fig. 11. Concentration–time proﬁles (solid curves) of the analyte in the ECS, CN e , and plasma, Cp, for the iv bolus experiment of Fig. 10B. The dashed curve is the ratio of the instantaneous outﬂow concentration from the probe, Cout d [t], to the extraction efﬁciency that would be obtained by a separate steady-state calibration technique, Ess d . This apparent ECS concentration differs from the true ECS concentration, Cout d /Ed[t], because the instantaneous true extraction efﬁciency, Ed[t], would vary continuously throughout the experiment, as shown in Fig. 10C. However, the complete concentration–time integrals, AUC0N, would be the same for both apparent and true ECS concentrations. Since the rate constants for plasma-to-ECS inﬂux and efﬂux are equal, the AUC0N for the plasma is also the same as that for ECS curves. Adapted from Bungay et al. (2001, Fig. 5, p. 365).

A steady-state calibration value, E d ½1; then sufﬁces for the calculation of the truncated value, AUC0Tapp. Although the apparent and true values for the complete AUC0N are equal, the truncated values differ. For the example in Fig. 11, the integrals for the apparent and true ECS concentrations are 93.5 and 91.0% of AUC0–N when truncated at t ¼ 4 h.

V. Illustrative application to dopamine microdialysis Quantitative in vivo microdialysis provides the capability for more than the determination of analyte concentrations in the tissue ECS. The experimental approaches in the preceding section in themselves yield additional measures of the state of the system under study. Utilization of the predictive expressions in Section II can supply additional insight into underlying physiologic and pharmacokinetic mechanisms. The following

observations in the microdialysis of DA serve to illustrate this added beneﬁt. As for any other analyte, microdialysis theory predicts that the Ed for DA should be sensitive to the rate at which it is removed from the ECS and insensitive to the rate at which it is supplied. The predominant mechanisms for DA supply and removal in the ECS of brain dopaminergic regions are usually neuronal release and uptake, respectively. In accord with the predictions, Smith and Justice (1994) found that Ed was unchanged when either DA synthesis or release was inhibited in NAc of the rat, whereas Ed declined when DA uptake was inhibited. The Ed values were obtained from steady-state nnf measurements. Inhibiting chemical conversion of DA did not signiﬁcantly affect Ed suggesting that the role of this clearance mechanism is secondary to that of uptake. Each of these interventions altered the DA apparent ECS concentration obtained from the intercept at the point-of-nnf: Ceapp decreased with inhibition of DA synthesis or release, while Ceapp increased with inhibition of DA uptake or metabolism. All of the concentration difference plots appeared to be linear. Thus, the observations were consistent with the prediction from the mathematical models that Ed and Ceapp vary independently. These observations suggest an association between Ed and DA uptake. However, the sensitivity of this relationship is limited by the fact that both the tissue and the probe inﬂuence Ed. If the probe contribution is dominant, then alterations in the tissue may be weakly reﬂected in the changes in Ed. As noted in Sections II.C.3 and III.E, the importance of the probe can be assessed by in vitro determination of Ews d . When the in vivo Ed values are close to Ews , the tissue contribution is small and d large changes in the rate of uptake may cause only small changes in Ed. Furthermore, the sensitivity depends on the direction in which the change occurs. Since Ews d imposes an upper bound on Ed, if the two values are close, than an increase in Ed and the corresponding increase in the uptake rate may not be detectable. This was one of the conclusions of Tang et al. (2003) who found no statistically signiﬁcant difference between in vivo and stirred in vitro Ed values for DA by nnf in NAc of freely moving rats. Under these circumstances, decreases

161

in the in vivo Ed may still provide a marker of decreased uptake. However, the likely insensitivity of the relation implies that a substantial decrease in the uptake rate would be required to produce a noticeable drop in Ed. A number of additional supporting studies indeed have demonstrated that the Ed is lowered by interventions that reduce the rate of clearance of DA and other neurotransmitters: 6-hydroxydopamine lesioning resulted in a 50% decrease in the Ed for DA (rat NAc; Parsons et al., 1991b); genetic deletion of the DA transporter resulted in a 46% decrease in the Ed for DA (Jones et al., 1998); inhibition of norepinephrine uptake by desipramine lowered the norepinephrine Ed by about 26% and inhibition of serotonin uptake by paroxetine reduced the serotonin Ed by 25% (rat Nac; Cosford et al., 1996); and inhibition of the enzymatic degradation of acetylcholine by probe perfusion with neostigmine produced a concentration-dependent reduction in the acetylcholine Ed of up to 32% (rat striatum; Vinson and Justice, 1997). Moreover, contrary to the ﬁndings of Tang et al. (2003), the following interventions that promote DA uptake resulted in observable increases in Ed: 1 day of abstinence after 10 days of cocaine administration (rat Nac; Parsons et al., 1991a); repeated, intermittent cocaine administration (rat medial prefrontal cortex; Chefer et al., 2000); acute administration of the D3-preferring agonist (+)PD128907 (mouse Nac; Zapata and Shippenberg, 2002); and pharmacological inactivation or genetic deletion of kappa-opioid receptors (mouse Nac; Chefer et al., 2005, 2006). It is noteworthy that in most of these studies, the increase in DA uptake suggested by the increase in DA Ed was conﬁrmed by other measurements of DA transporter activity like rotating disk electrode voltammetry and 3 H–DA uptake in synaptosomes. Thus, Ed has proven to be a useful indicator of changes associated with altered clearance rates. In contrast, Ed by itself can only serve as a qualitative measure of rate of clearance. It is not possible to discern the degree to which uptake is altered from the change in the magnitude of Ed per se. The sensitivity of the association between Ed and clearance rate, and the ability to detect increases as well as decreases in these quantities, may

depend on the animal species, the brain region, implantation technique, or other factors. The sensitivity is clearly a function of operating conditions, since it can be abolished by lowering the perfusion rate sufﬁciently (Section III.A.5.). Quantitative microdialysis approaches that incorporate the predictive mathematical models provides the means for presenting experimental outcomes in terms of direct measures of the underlying processes, such as the elimination rate constant, kt. Not only is kt direct, it is also a much more sensitive measure of clearance rate than Ed. One of the additional beneﬁts of expressing clearance rates by kt is that the microdialysis values are then amenable to comparison with those obtained by alternative methods, such as fast scan cyclic voltammetry (Bungay et al., 2003). Furthermore, at least for steady-state measurements, kt can be calculated with common software, such as Microsoft Excel. The combined use of the steady-state model of Section II with the nnf method of Section III will be illustrated by excerpting data from the recent study cited above that examined the augmentation of DA uptake by pharmacological inactivation of kappa-opioid receptors (Chefer et al., 2006). In this study, mice received vehicle (sterile saline) or nor-binaltorphimine (nor-BNI), a long-acting kappa-opioid receptor antagonist, by subcutaneous injection 24 h prior to microdialysis experiments. For DA nnf measurements in vivo, probes implanted in NAc were perfused in random order with solutions of 0, 2, 5, 10, and 20 nM DA at a ﬂow rate of Qd ¼ 0.6 mL/min. DA nnf measurements in vitro were obtained for the same probes immersed in a vigorously stirred artiﬁcial cerebrospinal ﬂuid maintained at 371C. Values for the perfusate and steady-state dialysate concentrations were used to prepare the plots in Fig. 12A. Linear regressions yielded slopes of Ed ¼ 0.36 7 0.01 (N ¼ 9) and Ed ¼ 0.42 7 0.005 (N ¼ 10) for the vehicle-treated and nor-BNI-treated animals, respectively, while for the probes in vitro, Ews d ¼ 0.4970.02 (N ¼ 10). All three values were signiﬁcantly different from one another indicating sufﬁcient sensitivity for Ed to reﬂect the enhanced DA uptake rate following nor-BNI treatment. There was no signiﬁcant difference in the apparent ECS concentrations (nnf intercepts): 4.571.0 nM

162

Fig. 12. Steady-state no-net-ﬂux (nnf) microdialysis of dopamine (DA) in the nucleus accumbens (NAc) of mice who received either the kappa-opioid receptor antagonist, norbinaltorphimine (nor-BNI) or sterile saline (vehicle) by subcutaneous injection. (A) The slopes, Ed, of the plots from the norBNI-treated animals (solid line) and vehicle-treated animals (dashed line) are signiﬁcantly different from each other and from the value, Ews d , for the same probes immersed in vigorously stirred medium in vitro at 371C (dot-dashed line). The point-ofout nnf in vivo intercepts, Cnnf ¼ Cin are not signiﬁcantly d ¼ Cd different. Adapted from Figs. 1 and 2 of Chefer et al. (2006). (B) The nor-BNI treatment augmented the rate of DA uptake as reﬂected in the increase in Ed and in the uptake rate constant, kt. Tissue permeabilities to DA, Pto ; were calculated from the Ed and Ews d values using Eq. (100). A ﬁvefold increase in kt values was obtained from this plot of steady-state model Eq. (43) or the approximate empirical equation shown in the ﬁgure.

for the vehicle-treated and 4.870.7 nM for the nor-BNI-treated animals. Values for kt can be obtained by ﬁrst calculating DA tissue permeabilities from the following expression derived by combining Eqs. (45), (47), and (64), 1 ro Pto ro Qd 1 1 ¼ , ln½1 E d Dt Dt So ln½1 E ws d (100)

in which the external surface area was So ¼ 0.75 mm2 for probes of outer radius, ro ¼ 0.12 mm, and membrane length, ‘m ¼ 1 mm. If DA diffuses only through the ECS, Eqs. (13) and (95) lead to an estimate for the tissue diffusion coefﬁcient of Dt ¼ fe De ¼ fe Dd/l2 ¼ 6.7 103 mm2/min using the values of fe ¼ 0.35 (Dykstra et al., 1992) and Dd ¼ 7.6 106 cm2/s and l ¼ 1.54 (Rice, 2000). Then, Eq. (100) yields values for the non-dimensional ratio, ro Pto =Dt ; of 18.8 and 43.1 for the vehicle- and nor-BNI-treated mice, respectively. The corresponding estimates for the dimensionless elimination modulus, Y, are 18.3 and 42.6. These can be obtained by iteration using the exact relation given as the ﬁrst form of Eq. (43). An algebraic alternative is to use Y ro Pto =Dt 0:5: The exact relation agrees closely with this approximate expression for Yb1; the two relations are indistinguishable in Fig. 12B. In this range, the value of the Bessel function ratio in Eq. (43) is close to unity. This corresponds to the situation shown for the neurotransmitter in Fig. 7B for which the diffusion in the tissue is conﬁned to a region adjacent to the probe that is thin compared with the probe radius, ro. For this condition, the inﬂuence of the cylindrical geometry of the probe is unimportant. Finally, the elimination rate constants can be calculated by rearranging the deﬁnition for Y in Eq. (42), kt ¼ Dt (Y/ro)2 ¼ 2.6 and 14.2/s, for the vehicle- and nor-BNI-treated mice, respectively. The magnitudes of these values are uncertain because they depend on the estimate used for Dt, which was not measured during the experiment. However, the ratio of the kt values is much less affected, since the parameters cancel out when Y-N leaving only the measured quantities, Ed and Ews d , 2 1 1 ln½1E ln½1E ws ðkt Þnor-BNI d d vehicle 2 ðkt Þvehicle 1 1 ln½1E d ln½1E ws d nor-BNI ¼ 5:3 for Y 1. ð101Þ Thus, with kt as a direct measure, the results suggest that the uptake rate was increased ﬁvefold by the nor-BNI treatment. By contrast, the

163

corresponding increase in Ed from 0.36 to 0.42 was only 18%. In conclusion, the studies described above show that nnf microdialysis is an effective in vivo technique for monitoring increases as well as decreases in DA uptake in speciﬁc brain regions. The ﬁnding that both post-treatment and control Ed in vivo values were less than the well-stirred in vitro value demonstrates that an assessment of both increase and decrease in DA uptake is feasible with this method. Determination of the rate constant, kt, provides information regarding the magnitude of changes in clearance rates compared with the qualitative indication provided by Ed. This was illustrated for DA uptake as one speciﬁc example of the application of quantitative microdialysis. Additional studies are needed to evaluate the use of kt for more precise quantiﬁcation of changes in the dynamics of neurotransmitters and other analytes that occur in response to particular pharmacological or genetic manipulations.

steady-state is slow. Conversely, for a rapidly cleared neurotransmitter steady-state can be achieved quickly and exchange between the perfusate and tissue may be conﬁned to a thin layer surrounding the probe. This suggests that effects of probe implantation trauma are likely to be more pronounced for such neurotransmitters than more slowly cleared analytes. Calculations for quantitative microdialysis based on steady-state probe calibrations can be performed in spread-sheet software. In general, transient experiments are much more difﬁcult to analyze quantitatively because probe calibrations also vary over time in a manner that depends on the time course of the tissue concentration. Furthermore, most probe calibration techniques are only appropriate for steady-state conditions. However, steady-state calibration appears sufﬁcient for transient experiments in the case of AUC concentration–time integration for linearly behaving analytes.

VI. Summary

Acknowledgments

Microdialysis is more than an empirical tool for delivering bioactive substances and monitoring extracellular concentrations. Quantitative approaches provide the means for evaluating parameters that characterize underlying tissue physiological and biochemical processes, especially those that act to remove analyte from ECS. The parameter values should be amenable to direct comparison with values obtainable from other complementary quantitative in vivo techniques. To this end, a coherent mathematical framework has been presented that is applicable to lipophilic, as well as hydrophilic, analytes. Both the probe and the tissue properties inﬂuence microdialysis performance. Usually, the tissue is more inﬂuential for an analyte whose clearance rate is slow to moderately fast. However, the probe properties can limit performance for some solutes, such as neurotransmitters that are rapidly cleared. Clearance rates have other important effects on performance. Slow clearance generally implies that a relatively large volume of tissue is involved in exchanging analyte with the probe and also that the approach to

This research was supported by the Intramural Research Programs of the National Institutes of Health, National Institute on Drug Abuse.

References Abramowitz, M. and Stegun, I.A. (1964) Handbook of Mathematical Functions. National Bureau of Standards (U.S. Government Printing Ofﬁce), Washington, D.C. Amberg, G. and Lindefors, N. (1989) Intracerebral microdialysis: II. Mathematical studies of diffusion kinetics. J. Pharmacol. Methods,, 22: 157–183. Ao, X., Sellati, T.J. and Stenken, J.A. (2004) Enhanced microdialysis relative recovery of inﬂammatory cytokines using antibody-coated microspheres analyzed by ﬂow cytometry. Anal. Chem., 76: 3777–3784. Beagles, K.E., Morrison, P.F. and Heyes, M.P. (1998) Quinolinic acid in vivo synthesis rates, extracellular concentrations, and intercompartmental distributions in normal and immune-activated brain as determined by multiple-isotope microdialysis. J. Neurochem., 70: 281–291. Benveniste, H., Drejer, J., Schousboe, A. and Diemer, N.H. (1987) Regional cerebral glucose phosphorylation and blood ﬂow after insertion of a microdialysis ﬁber through the dorsal hippocampus in the rat. J. Neurochem., 49: 729–734.

164 Benveniste, H., Hansen, A.J. and Ottosen, N.S. (1989) Determination of brain interstitial concentrations by microdialysis. J. Neurochem., 52: 1741–1750. Benveniste, H. and Hu¨ttemeier, P.C. (1990) Microdialysis – theory and application. Prog. Neurobiol., 35: 195–215. Bird, R.B., Stewart, W.E. and Lightfoot, E.N. (1960) Transport Phenomena. John Wiley & Sons, New York. Bodenlenz, M., Schaupp, L.A., Druml, T., Sommer, R., Wutte, A., Schaller, H.C., Sinner, F., Wach, P. and Pieber, T.R. (2005) Measurement of interstitial insulin in human adipose and muscle tissue under moderate hyperinsulinemia by means of direct interstitial access. Am. J. Physiol. Endocrinol. Metab., 289: E296–E300. Bouw, M.R. and Hammarlund-Udenaes, M. (1998) Methodological aspects of the use of a calibrator in in vivo microdialysis-further development of the retrodialysis method. Pharm. Res., 15: 1673–1679. Brunner, M., Joukhadar, C., Schmid, R., Erovic, B., Eichler, H. and Mu¨ller, M. (2000) Validation of urea as an endogenous reference compound for the in vivo calibration of microdialysis probes. Life Sci., 67: 977–984. Bungay, P.M., Dedrick, R.L., Fox, E. and Balis, F.M. (2001) Probe calibration in transient microdialysis in vivo. Pharm. Res., 18: 361–366. Bungay, P.M. and Gonzales, R.A. (1996) Pressure-enhanced delivery of solutes via microdialysis. Soc. Neurosci. Abstr., 22: 2076. Bungay, P.M., Morrison, P.F. and Dedrick, R.L. (1990) Steady-state theory for quantitative microdialysis of solutes and water in vivo and in vitro. Life Sci., 46: 105–119. Bungay, P.M., Newton-Vinson, P., Isele, W., Garris, P.A. and Justice, J.B., Jr. (2003) Microdialysis of dopamine interpreted with quantitative model incorporating probe implantation trauma. J. Neurochem., 86: 932–946. Cano-Cebria´n, M.J., Zornoza, T., Polache, A. and Granero, L. (2005) Quantitative in vivo microdialysis in pharmacokinetic studies: some reminders. Curr. Drug Metab., 6: 83–90. Chaurasia, C.S. (1999) In vivo microdialysis sampling: theory and applications. Biomed. Chromatogr., 13: 317–332. Chefer, V.I., Czyzyk, T., Bolan, E.A., Moron, J., Pintar, J.E. and Shippenberg, T.S. (2005) Endogenous kappa-opioid receptor systems regulate mesoaccumbal dopamine dynamics and vulnerability to cocaine. J. Neurosci., 25: 5029–5037. Chefer, V.I., Moron, J.A., Hope, B., Rea, W. and Shippenberg, T.S. (2000) Kappa-opioid receptor activation prevents alterations in mesocortical dopamine neurotransmission that occur during abstinence from cocaine. Neuroscience, 101: 619–627. Chefer, V., Zapata, A., Shippenberg, T.S. and Bungay, P.M. (2006) Quantitative no-net-ﬂux microdialysis permits detection of increases and decreases in dopamine uptake in mouse nucleus accumbens. J. Neurosci. Methods, 155: 187–193. Chen, K.C. (2003) Insensitivity of the microdialysis zero-netﬂux method to nonlinear uptake and release processes. Neurosci. Res., 46: 251–256. Chen, K.C. (2005a) Preferentially impaired neurotransmitter release sites not their discreteness compromise the validity

of microdialysis zero-net-ﬂux method. J. Neurochem., 92: 29–45. Chen, K.C. (2005b) Evidence on extracellular dopamine level in rat striatum: implications for the validity of quantitative microdialysis. J. Neurochem., 92: 46–58. Chen, K.C. (2006) Effects of tissue trauma on the characteristics of microdialysis zero-net-ﬂux method sampling neurotransmitters. J. Theor. Biol., 238: 863–881. Chen, K.C., Ho¨istad, M., Kehr, J., Fuxe, K. and Nicholson, C. (2002a) Theory relating in vitro and in vivo microdialysis with one or two probes. J. Neurochem., 81: 108–121. Chen, K.C., Ho¨istad, M., Kehr, J., Fuxe, K. and Nicholson, C. (2002b) Quantitative dual-probe microdialysis: mathematical model and analysis. J. Neurochem., 81: 94–107. Chen, R., Warden, J.T. and Stenken, J.A. (2004) Microdialysis sampling combined with electron spin resonance for superoxide radical detection in microliter samples. Anal. Chem., 76: 4734–4740. Clapp-Lilly, K.L., Roberts, R.C., Duffy, L.K., Irons, K.P., Hu, Y. and Drew, K.L. (1999) An ultrastructural analysis of tissue surrounding a microdialysis probe. J. Neurosci. Methods, 90: 129–142. Cle´ment, R., Malinovsky, J., Dollo, G., Le, C.P., Chevanne, F. and Le, V.R. (1998) In vitro and in vivo microdialysis calibration using retrodialysis for the study of the cerebrospinal distribution of bupivacaine. J. Pharma. Biomed. Anal., 17: 665–670. Clough, G.F. (2005) Microdialysis of large molecules. AAPS J., 7: E686–E692. Collins, M.C. and Ramirez, W.F. (1979) Mass-transport through polymeric membranes. J. Phys. Chem., 83: 2294–2301. Cosford, R.J., Vinson, A.P., Kukoyi, S. and Justice, J.B. Jr. (1996) Quantitative microdialysis of serotonin and norepinephrine: pharmacological inﬂuences on in vivo extraction fraction. J. Neurosci. Methods, 68: 39–47. de Lange, E.C., Bouw, M.R., Mandema, J.W., Danhof, M., de Boer, A.G. and Breimer, D.D. (1995) Application of intracerebral microdialysis to study regional distribution kinetics of drugs in rat brain. Br. J. Pharmacol., 116: 2538–2544. de Lange, E.C.M., de Boer, A.G. and Breimer, D.D. (2000) Methodological issues in microdialysis sampling for pharmacokinetic studies. Adv. Drug Deliv. Rev., 45: 125–148. Dykstra, K.H., Arya, A., Arriola, D.M., Bungay, P.M., Morrison, P.F. and Dedrick, R.L. (1993) Microdialysis study of zidovudine (AZT) transport in rat brain. J. Pharmacol. Exp. Ther., 267: 1227–1236. Dykstra, K.H., Hsiao, J.K., Morrison, P.F., Bungay, P.M., Mefford, I.N., Scully, M.M. and Dedrick, R.L. (1992) Quantitative examination of tissue concentration proﬁles associated with microdialysis. J. Neurochem., 58: 931–940. Ekblom, M., Ga˚rdmark, M. and Hammarlund-Udenaes, M. (1992) Estimation of unbound concentrations of morphine from microdialysate concentrations by use of nonlinear regression analysis in vivo and in vitro during steady state conditions. Life Sci., 51: 449–460.

165 Elmquist, W.F. and Sawchuk, R.J. (1997) Application of microdialysis in pharmacokinetic studies. Pharm. Res., 14: 267–288. Ettinger, S.N., Poellmann, C.C., Wisniewski, N.A., Gaskin, A.A., Shoemaker, J.S., Poulson, J.M., Dewhirst, M.W. and Klitzman, B. (2001) Urea as a recovery marker for quantitative assessment of tumor interstitial solutes with microdialysis. Cancer Res., 61: 7964–7970. Fenstermacher, J.D. and Rapoport, S.I. (1984) Blood-brain barrier. In: Renkin, E.M., Michel C.C. and Geiger, S.R., (Eds.), The Cardiovascular System. Vol. 4, Part 2. American Physiological Society, Bethesda, MD, pp. 969–1000. Gonzales, R.A., McNabb, J., Yim, H.J., Ripley, T. and Bungay, P.M. (1998) Quantitative microdialysis of ethanol in rat striatum. Alcohol. Clin. Exp. Res., 22: 858–867. Gonzales, R.A., Tang, A. and Robinson, D.L. (2002) Quantitative microdialysis for in vivo studies of pharmacodynamics. In: Lui, Y. and Lovinger, D.M. (Eds.), Methods in AlcoholRelated Neuroscience Research. CRC Press, Boca Raton, FL, pp. 287–317. Groothuis, D.R., Ward, S., Schlageter, K.E., Itskovich, A.C., Schwerin, S.C., Allen, C.V., Dills, C. and Levy, R.M. (1998) Changes in blood-brain barrier permeability associated with insertion of brain cannulas and microdialysis probes. Brain Res., 803: 218–230. Hickner, R.C., Ekelund, U., Mellander, S., Ungerstedt, U. and Henriksson, J. (1995) Muscle blood ﬂow in cats: comparison of microdialysis ethanol technique with direct measurement. J. Appl. Physiol., 79: 638–647. Ho¨istad, M., Chen, K.C., Nicholson, C., Fuxe, K. and Kehr, J. (2002) Quantitative dual-probe microdialysis: evaluation of [3H] mannitol diffusion in agar and rat striatum. J. Neurochem., 81: 80–93. Jacobson, I., Sandberg, M. and Hamberger, A. (1985) Mass transfer in brain dialysis devices – a new method for the estimation of extracellular amino acids concentration. J. Neurosci. Methods, 15: 263–268. Jones, S.R., Gainetdinov, R.R., Jaber, M., Giros, B., Wightman, R.M. and Caron, M.G. (1998) Profound neuronal plasticity in response to inactivation of the dopamine transporter. Proc. Natl. Acad. Sci. U.S.A., 95: 4029–4034. Justice, J.B. Jr. (1993) Quantitative microdialysis of neurotransmitters. J. Neurosci. Methods, 48: 263–276. Kehr, J. (1993) A survey on quantitative microdialysis: theoretical models and practical implications. J. Neurosci. Methods, 48: 251–261. Klein, E. and Holland, F.F. (1977) Diffusive and hydraulic permeabilities of commercially available cellulosic hemodialysis ﬁlms and hollow ﬁbers. J. Membr. Sci., 2: 349–364. Kulagina, N.V., Shankar, L. and Michael, A.C. (1999) Monitoring glutamate and ascorbate in the extracellular space of brain tissue with electrochemical microsensors. Anal. Chem., 71: 5093–5100. Kume-Kick, J., Mazel, T., Vorˇ ı´ sˇ ek, I., Hrabeˇtova´, S., Tao, L. and Nicholson, C. (2002) Independence of extracellular tortuosity and volume fraction during osmotic challenge in rat neocortex. J. Physiol. –Lond., 542: 515–527.

Larsson, C.I. (1991) The use of an ‘‘internal standard’’ for control of the recovery in microdialysis. Life Sci., 49: PL73–PL78. Lo¨nnroth, P., Jansson, P.A. and Smith, U. (1987) A microdialysis method allowing characterization of intercellular water space in humans. Am. J. Physiol.-Endocrinol. Metab., 253: E228–E231. Lo¨nnroth, P. and Strindberg, L. (1995) Validation of the ‘internal reference technique’ for calibrating microdialysis catheters in situ. Acta Physiol. Scand., 153: 375–380. Lu, Y., Peters, J.L. and Michael, A.C. (1998) Direct comparison of the response of voltammetry and microdialysis to electrically evoked release of striatal dopamine. J. Neurochem., 70: 584–593. Marchand, S., Chenel, M., Lamarche, I. and Couet, W. (2005) Pharmacokinetic modeling of free amoxicillin concentrations in rat muscle extracellular ﬂuids determined by microdialysis. Antimicrob. Agents Chemother., 49: 3702–3706. Mathy, F.X., Lombry, C., Verbeeck, R.K. and Pre´at, V. (2005) Study of the percutaneous penetration of ﬂurbiprofen by cutaneous and subcutaneous microdialysis after iontophoretic delivery in rat. J. Pharm. Sci., 94: 144–152. Michael, A.C., Borland, L.M., Mitala, J.J., Willoughby, B.M. and Motzko, C.M. (2005) Theory for the impact of basal turnover on dopamine clearance kinetics in the rat striatum after medial forebrain bundle stimulation and pressure ejection. J. Neurochem., 94: 1202–1211. Morrison, P.F., Bungay, P.M., Hsiao, J.K., Ball, B.A., Mefford, I.N. and Dedrick, R.L. (1991a) Quantitative microdialysis: analysis of transients and application to pharmacokinetics in brain. J. Neurochem., 57: 103–119. Morrison, P.F., Bungay, P.M., Hsiao, J.K., Mefford, I.N., Dykstra, K.H. and Dedrick, R.L. (1991b) Quantitative microdialysis. In: Robinson, T.E. and Justice, Jr., J.B. (Eds.) Microdialysis in the Neurosciences, Vol. 7. Elsevier, New York, pp. 47–80. Nicholson, C. (2001) Diffusion and related transport mechanisms in brain tissue. Rep. Prog. Phys., 64: 815–884. Olson, R.J. and Justice, J.B. Jr. (1993) Quantitative microdialysis under transient conditions. Anal. Chem., 65: 1017–1022. Parsons, L.H. and Justice, J.B. Jr. (1992) Extracellular concentration and in vivo recovery of dopamine in the nucleus accumbens using microdialysis. J. Neurochem., 58: 212–218. Parsons, L.H. and Justice, J.B. Jr. (1994) Quantitative approaches to in vivo brain microdialysis. Crit. Rev. Neurobiol., 8: 189–220. Parsons, L.H., Smith, A.D. and Justice, J.B. Jr. (1991a) Basal extracellular dopamine is decreased in the rat nucleus accumbens during abstinence from chronic cocaine. Synapse, 9: 60–65. Parsons, L.H., Smith, A.D. and Justice, J.B. Jr. (1991b) The in vivo microdialysis recovery of dopamine is altered independently of basal level by 6-hydroxydopamine lesions to the nucleus accumbens. J. Neurosci. Methods, 40: 139–147. Patlak, C.S. and Fenstermacher, J.D. (1975) Measurements of dog blood-brain transfer constants by ventriculocisternal perfusion. Am. J. Physiol., 229: 877–884.

166 Peters, J.L. and Michael, A.C. (1998) Modeling voltammetry and microdialysis of striatal extracellular dopamine: the impact of dopamine uptake on extraction and recovery ratios. J. Neurochem., 70: 594–603. Plock, N. and Kloft, C. (2005) Microdialysis – theoretical background and recent implementation in applied life-sciences. Eur. J. Pharma. Sci., 25: 1–24. Rice, M.E. (2000) Distinct regional differences in dopaminemediated volume transmission. In: Agnati, L.F., Fuxe, K., Nicholson, C. and Sykova´, E. (Eds.) Volume Transmission Revisited, Vol. 125. Elsevier, Amsterdam, pp. 277–290. Rosdahl, H., Hamrin, K., Ungerstedt, U. and Henriksson, J. (2000) A microdialysis method for the in situ investigation of the action of large peptide molecules in human skeletal muscle: detection of local metabolic effects of insulin. Int. J. Biol. Macromol., 28: 69–73. Sarre, S., Deleu, D., Van Belle, K., Ebinger, G. and Michotte, Y. (1995) Quantitative microdialysis for studying the in vivo L-DOPA kinetics in blood and skeletal muscle of the dog. Pharm. Res., 12: 746–750. Scheller, D. and Kolb, J. (1991) The internal reference technique in microdialysis: a practical approach to monitoring dialysis efﬁciency and to calculating tissue concentration from dialysate samples. J. Neurosci. Methods, 40: 31–38. Sjo¨strand, M., Holma¨ng, A. and Lo¨nnroth, P. (1999) Measurement of interstitial insulin in human muscle. Am. J. Physiol. Endocrinol. Metab., 276: E151–E154. Smith, A.D. and Justice, J.B. Jr. (1994) The effect of inhibition of synthesis, release, metabolism and uptake on the microdialysis extraction fraction of dopamine. J. Neurosci. Methods, 54: 75–82. Snyder, K.L., Nathan, C.E., Yee, A. and Stenken, J.A. (2001) Diffusion and calibration properties of microdialysis sampling membranes in biological media. Analyst, 126: 1261–1268. Song, Y. and Lunte, C.E. (1999) Comparison of calibration by delivery versus no net ﬂux for quantitative in vivo microdialysis sampling. Anal. Chim. Acta, 379: 251–262. Sta˚hle, L. (1994) Zidovudine and alovudine as cross-wise recovery internal standards in microdialysis experiments? J. Pharmacol. Toxicol. Methods, 31: 167–169. Sta˚hle, L. (2000) On mathematical models of microdialysis: geometry, steady-state models, recovery and probe radius. Adv. Drug Deliv. Rev., 45: 149–167. Stenken, J.A. (1999) Methods and issues in microdialysis calibration. Anal. Chim. Acta, 379: 337–358. Stenken, J.A., Topp, E.M., Southard, M.Z. and Lunte, C.E. (1993) Examination of microdialysis sampling in a well-characterized hydrodynamic system. Anal. Chem., 65: 2324–2328. Strindberg, L. and Lo¨nnroth, P. (2000) Validation of an endogenous reference technique for the calibration of microdialysis catheters. Scand. J. Clin. Lab. Invest., 60: 205–211. Sun, H.Y., Bungay, P.M. and Elmquist, W.F. (2001) Effect of capillary efﬂux transport inhibition on the determination of probe recovery during in vivo microdialysis in the brain. J. Pharmacol. Exp. Ther., 297: 991–1000.

Sun, L. and Stenken, J.A. (2003) Improving microdialysis extraction efﬁciency of lipophilic eicosanoids. J. Pharma. Biomed. Anal., 33: 1059–1071. Tang, A., Bungay, P.M. and Gonzales, R.A. (2003) Characterization of probe and tissue factors that inﬂuence interpretation of quantitative microdialysis experiments for dopamine. J. Neurosci. Methods, 126: 1–11. Terasaki, T., Deguchi, Y., Kasama, Y., Pardridge, W.M. and Tsuji, A. (1992) Determination of in vivo steady-state unbound drug concentration in the brain interstitial ﬂuid by microdialysis. Int. J. Pharma., 81: 143–152. Tong, S. and Yuan, F. (2002) An equivalent length model of microdialysis sampling. J. Pharma. Biomed. Anal., 28: 269–278. Van Belle, K., Dzeka, T., Sarre, S., Ebinger, G. and Michotte, Y. (1993) In vitro and in vivo microdialysis calibration for the measurement of carbamazepine and its metabolites in rat brain tissue using the internal reference technique. J. Neurosci. Methods, 49: 167–173. Van Belle, K., Sarre, S., Ebinger, G. and Michotte, Y. (1995) Brain, liver and blood distribution kinetics of carbamazepine and its metabolic interaction with clomipramine in rats: a quantitative microdialysis study. J. Pharmacol. Exp. Ther., 272: 1217–1222. Vinson, P.N. and Justice, J.B. Jr. (1997) Effect of neostigmine on concentration and extraction fraction of acetylcholine using quantitative microdialysis. J. Neurosci. Methods, 73: 61–67. Wallgren, F., Amberg, G., Hickner, R.C., Ekelund, U., Jorfeldt, L. and Henriksson, J. (1995) A mathematical model for measuring blood ﬂow in skeletal muscle with the microdialysis ethanol technique. J. Appl. Physiol., 79: 648–659. Wang, Y., Wong, S.L. and Sawchuk, R.J. (1993) Microdialysis calibration using retrodialysis and zero-net ﬂux: application to a study of the distribution of zidovudine to rabbit cerebrospinal ﬂuid and thalamus. Pharm. Res., 10: 1411–1419. Wang, Y.-F., Wong, S.L. and Sawchuk, R.J. (1991) Comparison of in vitro and in vivo calibration of microdialysis probes using retrodialysis. Curr. Sep., 10: 88. Westerink, B.H. and De Vries, J.B. (1988) Characterization of in vivo dopamine release as determined by brain microdialysis after acute and subchronic implantations: methodological aspects. J. Neurochem., 51: 683–687. Wong, S.L., Wang, Y. and Sawchuk, R.J. (1992) Analysis of zidovudine distribution to speciﬁc regions in rabbit brain using microdialysis. Pharm. Res., 9: 332–338. Wozniak, K.M., Pert, A., Mele, A. and Linnoila, M. (1991) Focal application of alcohols elevates extracellular dopamine in rat brain: a microdialysis study. Brain Res., 540: 31–40. Xie, R. and Hammarlund-Udenaes, M. (1998) Blood-brain barrier equilibration of codeine in rats studied with microdialysis. Pharm. Res., 15: 570–575. Xie, R., Hammarlund-Udenaes, M., de Boer, A.G. and de Lange, E.C. (1999) The role of P-glycoprotein in blood-brain barrier transport of morphine: transcortical microdialysis

167 studies in mdr1a (–/–) and mdr1a (+/+) mice. Br. J. Pharmacol., 128: 563–568. Yang, H., Peters, J.L. and Michael, A.C. (1998) Coupled effects of mass transfer and uptake kinetics on in vivo microdialysis of dopamine. J. Neurochem., 71: 684–692. Yang, H., Peters, J.L. and Michael, A.C. (1999) Simultaneous microdialysis and voltammetry: evidence for suprananomolar basal extracellular dopamine in rat striatum. In: Proceedings of the 8th International Conference on In vivo Methods, Stony Brook, NY, pp. 81–82. Yang, H., Wang, Q. and Elmquist, W.F. (1996) Fluconazole distribution to the brain: a crossover study in

freely-moving rats using in vivo microdialysis. Pharm. Res., 13: 1570–1575. Yang, H., Wang, Q. and Elmquist, W.F. (1997) The design and validation of a novel intravenous microdialysis probe: application to ﬂuconazole pharmacokinetics in the freely-moving rat model. Pharm. Res., 14: 1455–1460. Zapata, A. and Shippenberg, T.S. (2002) D3 receptor ligands modulate extracellular dopamine clearance in the nucleus accumbens. J. Neurochem., 81: 1035–1042. Zhao, Y., Liang, X. and Lunte, C.E. (1995) Comparison of recovery and delivery in vitro for calibration of microdialysis probes. Anal. Chim. Acta, 316: 403–410.

CHAPTER 2.3

Automation of blood and microdialysis sampling: combinatorial pharmacology Peter T. Kissinger1,2, 2

1 Bioanalytical Systems Inc., West Lafayette, IN, USA Department of Chemistry, Purdue University, West Lafayette, IN, USA

Abstract: Microdialysis has progressed from an inspired academic curiosity in the early 1970s to a widely used biomedical research tool 35 years later. This chapter reviews the evolution of the instrumentation from laboratory built manual apparatus to fully automated systems today that are used in both preclinical and clinical research. A systems view is presented to explain how the various components such as in vivo probes, pumps, fraction collectors, and animal containment systems can be integrated to provide a variety of data in parallel. Microdialysis was the inspiration for modernizing blood sampling and fully automating dosing to provide a pharmacological response in a tightly controlled manner while eliminating inﬂuence of the ﬁght and ﬂight stress response in laboratory animals. The value for biomedical research is to provide a variety of chemical, physiological, and psychological data from the same mammal at the same time. This reduces the number of animals required, improves the quality of the data obtained, and reduces pain and suffering on the part of the animals and stress on the part of the investigator. These new developments are primarily used within the commercial biopharmaceutical research community and thus relatively little has been published in the conventional peer reviewed literature. This chapter is focused strictly on the tools of in vivo sampling. Other chapters in this book are devoted more to the pharmacological (esp. neuroscience) conclusions that can be drawn from using these tools. blood-brain barrier penetration, neurotransmitter (and other biomarkers) response, and accompanying behavioral and physiological data. There would be no meaning to these experiments without bioanalytical chemistry, and to achieve an optimum experiment one must integrate an animal model with dosing and then sampling. This is followed by various sample storage and preparation steps. Finally, there is the choice of an analytical approach, typically a separation step integrated with a physical detector. In 2005, liquid chromatography/mass spectrometry (LC/ MSMS) has become dominant while there remains a place for liquid chromatography/electrochemistry (LC/EC) and liquid chromatography/ﬂuorescence (LC/F) as well as various immunochemical approaches.

I. Introduction My assignment in this chapter is to focus on the laboratory instrumentation that can be implemented for the purpose of sampling laboratory animals in the awake and freely moving state. The goal of laboratory experiments is to get data. The value of data is to make decisions. Thus we will discuss getting better data to make better decisions. The general focus is preclinical in vivo pharmacology and the pharmacokinetic and pharmacodynamic information needed to support mechanistic conclusions about such topics as drug metabolism, drug disposition, Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

169

DOI: 10.1016/S1569-7339(06)16009-9 Copyright 2007 Elsevier B.V. All rights reserved

170

It is helpful here to distinguish between continuous ﬂow methods and methods that are segmented in batches. This is important to how automation is best achieved and coordinated. Continuous ﬂow always causes problems in synchronization of the bioanalytical approach, which typically requires a longer time to achieve than the desired data rate from the biology. This problem is solved by collecting discrete samples and processing them off-line. There are a few cases where continuous monitoring is possible, but these have been limited to major components such as glucose (enzyme electrodes) or oxygen (optical or electrochemical oxygen sensors). This chapter will focus only on discrete samples. Both microdialysis and blood are continuous ﬂow and both are generally studied by taking an aliquot out of circulation. While microdialysis received a ﬂurry of attention in the early 1970s, it is fair to say that it did not achieve much popularity until Prof. Urban Ungerstedt and his pioneering team at the Karolinska Institutet concluded that ‘‘this is hard work; there must be a better way to do this! Collecting samples by hand is not ideal.’’ Table 1 presents a rough historical perspective of how automation evolved over about 30 years. In the very beginning, the animal models were all anesthetized and rigidly held in a sterotaxic frame. At ﬁrst, animals did not survive an acute microdialysis experiment and samples were collected by changing out microvials by the experimenter standing close at hand. Thirty years later most animals are awake, freely moving, and nearly everything is going on without a human present. Dosing is automatically programmed;

blood, microdialysates, and even bile can be serially collected under computer control; and behavior can be monitored along with selected physiological parameters such as blood pressure and body temperature. What can be accomplished with an automatically sampled animal progressed hand-in-hand with improved bioanalytical approaches that reduced the volume of sample required to achieve the needed lower limit of quantitation (LLOQ). In the 1970s, for many drugs this was 5 mL, while today 5 mL is sufﬁcient in many cases with the latest mass spectrometry methods. This is an improvement of one thousand fold in just 30 years. From the perspective of a mouse it means that serial sampling of a single awake mouse is feasible. Previously, the sacriﬁce of several mice at each time point was required, considering the circulating blood volume of an average mouse is about 1.5 mL. Correspondingly, the average rat has a blood volume of about 30 mL. In earlier times, a few samples could induce hypovolemic shock to the animal. Today, multiple samples of 200 mL are quite convenient and sufﬁcient. With respect to brain microdialysis, LC/EC was quite advanced in the late 1970s for the unique oxidizable transmitters (catecholamines and 5-HT). Even then it was possible to manually process 10 mL serial microdialysate samples collected by hand. In those days there were no autosamplers that could deal with such small volumes and thus collecting samples, extracting them and injecting them in a chromatograph were labor-intensive activities. Most neuroscience laboratories were then using chart recorders for chromatographic data, measuring

Table 1. Automation of microdialysis and related data acquisition channels

Consciousness Restraint Ex vivo connection Collection Dosing Blood collection Body temperature Blood pressure EEG Behavior measure

Late 1970s

Late 1980s

Late 1990s

2005

Anesthetized Stereotaxic frame Direct Manual Manual Manual Yes No Yes No

Awake Freely moving Swivel Automatic Manual None No No No No

Awake Freely moving Swivel Automatic Manual None No No No Yes

Awake Freely moving Direct Automatic Automatic Automatic Yes Yes Yes Yes

171

peak heights with rulers and contrasting them with peaks for standards. Today much of this work can be done in an automated fashion, including acquisition of the data and preparation of a report. The following sections of the chapter are organized according to the components of an automated system, beginning ﬁrst with the animal model.

II. Animal containment Sampling a rodent requires implants of sampling devices under anesthesia. These include microdialysis probes, which may be very ﬂexible in peripheral tissue or rigidly placed in the brain using a guide cannula cemented to the skull. Additional catheters may be placed for venous or arterial sampling of blood and perhaps bile as well. In the earlier days when animals did not recover from the anesthesia, it was not necessary to be very concerned with how these sampling devices would survive over time. Most experiments were completed in a few hours. Exteriorization of a central nervous system (CNS) dialysis probe, for example, was quite simple being held on the sterotaxic frame throughout the sampling process. Later, as awake animals began to be studied, attention was paid to how animals recovered from the anesthesia and some effort was required to prevent them getting access to the implant. At the same time, liquid swivels with sufﬁciently low dead volume were introduced to enable connection to the laboratory environment for collection. Balanced beams and round plastic cages without wire were developed to aid animal comfort and avoid physical damage to implants. It quickly became apparent that acquiring data from several channels simultaneously would be desirable. These might include data from more than one dialysis probe or from physiology transducers for temperature, pressure, heart rate, or EEGs. To simultaneously monitor a drug in peripheral circulation with a neurotransmitter response in the brain required either microdialysis in blood vessels or sampling whole blood itself. Such experiments are not possible with liquid swivels and several new approaches were explored in the late 1990s. The most popular of these to date remains the Raturns system by which animal

motion was countered with a responsive cage to not allow twisting of wires or tubes exteriorized from the back of the animal’s neck. This greatly increases the reliability of microdialysis, infusion, biosensor, and electrophysiology experiments by allowing studies of cannulated and wired awake rodents without the limitations of liquid swivels and commutators. II.A. How the Raturn works Swivels and commutators are replaced with an optical switch. As the animal travels in the chamber on an energized turntable the switch signals the cage to incrementally rotate in the opposite direction. Thus the Raturn can interactively respond to the animal’s rotational movements, allowing the animal to move freely while preventing the twisting or breakage of attached lines. Actually the device is based on the orientation of the animal’s head and not its position in the cage. A clear understanding cannot be obtained from words or pictures, although one can try (Bohs et al., 2000). It is necessary to see this apparatus to fully understand the concept. The instrument does not cause the animal to move. With the Raturn, single or multiple contiguous lengths of tubing and/or wires can connect the animal to external syringe pumps, fraction collectors, stimulators, ampliﬁers, etc. II.B. Cages When awake animal microdialysis studies began in the 1980s, the group at the Karolinska Institutet and CMA/microdialysis recognized that conventional rectangular animal cages with wire tops were not optimal. Instead, they conceived of a round bottom cage as shown in Fig. 1 with a balance beam that could accommodate a liquid swivel (or today the sensors for Raturn). Once microdialysis extended from the brain to many peripheral tissues, it became meaningful to combine it with drug metabolism studies where distribution between urine and feces are of interest along with blood sampling and microdialysis. Even more modern is the search for biomarkers of both pharmacology and toxicology and metabolomics of urine

172

Fig. 1. Round bottom, open top cage with balance beam was introduced for in vivo sampling in the 1980s.

using high-resolution NMR and mass spectrometry of urine specimens. Fig. 2 illustrates a metabolism cage which allows for these experiments in combination for an awake freely moving rat. The same experiments can be accomplished with a mouse, but by using a scaled down cage consistent with the much smaller volume of urine that is produced. Mouse and rat metabolism cages are interchangeable on the same apparatus. In fact, the entire apparatus is designed to be modular so that a variety of systems can be built up from compatible parts in a Lego block manner. Fig. 3 illustrates a more complete bench top system and Fig. 4 shows a cart that some refer to as a four bedroom rodent residence. Academic groups sometimes manufacture their own systems as an economy measure. That makes good sense. Many rats have enjoyed life in a plastic ﬂowerpot from a local discount store while they contribute to science via microdialysis. II.C. Behavioral monitoring Turning behavior has been evaluated in rodents for over 30 years. Rearing is also a good measure. The balance beam in the Raturn allows for both. Software we developed in our laboratory tracks both frequency of events and their duration. There are

Fig. 2. Metabolism cages permit the collection of urine and feces while microdialysates and blood samples are automatically collected.

speciﬁc rotational models for CNS pharmacology that predispose an animal to one direction or the other. These models are readily accommodated. Simply tracking general activity over a diurnal cycle and comparing with respect to control animals and dose gives an early indication of CNS activity (blood-brain barrier [BBB] penetration) that might be either pharmacological or toxicological. III. Automated dosing Properly dosing animals is a very important step for pharmacokinetic and pharmacodynamic studies. Manual dosing partly defeats the beneﬁt of automatic body ﬂuid sampling, especially for the early post-dose time points. Different types of automated dosing can be performed, and by eliminating animal-human interaction, the data quality improves. III.A. Intravenous To do this right, we had to develop a new type of infusion pump that allowed for programming

173

Fig. 3. A commercial bench top approach to organize the components for an automated pharmacology system.

concentration of a single drug over time (e.g., apply a ‘‘loading dose’’), or to apply two drugs to explore an in vivo interaction. One of our goals was to be able to initiate a protocol at any time over a rat’s diurnal cycle. Continuous infusion also provides a convenient route to getting clearance data (CL), which results from the infusion rate divided by the steady-state plasma drug concentration (Ro/Css). Fig. 5 shows hardware that includes three stepping motor-driven syringes and reservoirs for physiological saline and up to two dosing solutions, allowing for a dose to be initiated at any time. Bolus doses, continuous infusion or combinations of these are permitted.

III.B. Gastric, intraperitoneal, and duodenal A PO (by mouth) dose requires a gavage. Anyone who has seen a gavage procedure will recognize that this is extremely stressful for the animal. It makes a lot of sense to ask how this stress can distort the data being sought. We have not yet ﬁgured out how to get a rodent to reliably take a tablet on command, and thus PO dosing requires handling the animal. In contrast, gastric dosing via catheter is possible, and we do that as

Fig. 4. A wheeled cart with four animal stations. Many commercial vivaria contain no laboratory benches and study rooms require mobile components. Electrical power and computer network connections are supplied from the ceiling.

an alternative. Intraperitoneal automated dosing is equally straightforward. Dosing directly into the duodenum, by bypassing much of the digestive tract, is very feasible and can provide useful new information when contrasted with both gastric and intravenous (IV) dosing. Detailed protocols are available from the author. IV. Microdialysis pumps Microdialysis is normally always a continuous rather than intermittent process. Thus an unintelligent pump can work ﬁne as part of an automated process as long as it keeps pumping. There are many adequate pumps on the market that perform very

174

Fig. 5. A commercial, programmable infusion pump for continuous microdialysis and drug dosing (bolus and continuous) through various routes (intravenous, gastric, intraperitoneal, and duodenal).

well whether priced at 500 or 5,000 h. In contrast, there is a simple way to make all of these pumps work very poorly by using too large a syringe at too slow a ﬂow rate. Syringe pumps normally operate at a controlled linear velocity established by a motor, a gearbox, and a screwdriver. For many infusion applications, one can ﬁll the pump less frequently by using a larger syringe at a lower ﬂow rate. There will be considerable averaging, or integrating, of a highly variable instantaneous ﬂow rate over time. This occurs due to mechanical imprecision in the gearbox and screwdriver. A 1% volume precision over an hour can be far worse over a few minutes and extremely poor on the time scale of a few seconds. Some syringe pumps will even include a cycle, as the screw turns, that can show periodic ﬂow variations over 10 min. With microdialysis, the concentration recovery of the probe is ﬂow dependent. Thus it can be very important to ensure that what looks like a biological cycle is not the result of inadequate matching of the syringe size, the screw motion, and the volume ﬂow rate. The sample collection volume is also important in this respect. Microdialysates collected every 30 min are less subject to ﬂow-induced recovery variations than are those samples collected every 5 min. The worst case will be for continuous sensors in the exit tubing from the microdialysis probe. If ﬂow sensitive, such sensors can result in noise that is attributable more to the mechanical components in the pump than to either the biology

or the sensor itself. The lesson here is that using a 25 mL syringe to pump at 1 mL/min can be the source of much trouble. At the very least, uniformity of pump ﬂow should be veriﬁed by weight in sealed vials, selecting aliquots consistent with the dialysis sample volume to be collected. If you are told that syringe pumps are inherently pulseless and continuous, do not trust that notion because it can be very wrong. The temptation to use larger syringes is often the case with automation to ensure that the ﬂuid is supplied for 12 or even 24 h. A better solution can be to select a pump with a smaller syringe that automatically reﬁlls from a reservoir. A 1 mL syringe allows for a 1,000 min (16.6 h) experiment at 1 mL/min. Thus it is rarely necessary to use a syringe of larger volume. Automated systems can allow for actuated start/stop of the pump. There are very few situations where this provides much of a practical advantage.

V. Microdialysis probes As mentioned above, in the early days of microdialysis, the animals were anesthetized and thus there was no need to consider probes for long-term use. When awake animals began to be used in the 1980s, dialysis probes were either cemented to the cranium with dental adhesive or placed in guide cannulae that were cemented in place. Automation then required a locking mechanism to prevent the animal from pulling out the probe and also a mechanical strain relief scheme (‘‘tether’’) to protect the tubing from being pulled against both the probe end and the swivel end. As swivels became unnecessary, the tubing could be continuous between a zero dead volume connection at the probe and the fraction collector. There are numerous probes available commercially that vary mechanically and by choice of membrane. These are indicated throughout other chapters in this volume. Fig. 6 illustrates a probe widely used in mouse CNS and Fig. 7 shows a linear probe for peripheral tissue. Other chapters in this volume will undoubtedly discuss the time over which the probe is viable for quality measurements. Generally speaking, viable samples are not collected until

175

Fig. 6. One probe and guide design used in mouse CNS.

after the damage from initial probe placement subsides in chemical terms (several hours). Some describe this as ‘‘resealing the blood-brain barrier’’ and the sweeping away of small molecules that have been mechanically released into the extracellular space by the foreign object.

VI. Sampling physiological ﬂuid VI.A. Blood Over the last several years, we developed an automated blood sampler, a robotic system that collects serial blood samples from awake and freely moving animals (Zhu et al., 2000; He et al., 2001; Long et al., 2001; Tian et al., 2002; Gunaratna et al., 2004; Zhu et al., 2005). This instrument also collects urine and feces, providing additional ADME (absorption, distribution, metabolism, elimination) information from the same subject. The automation adds another parameter by monitoring animal activity during the experiment, providing a look at possible behavioral anomalies associated with the drug much earlier in the screening process. The automated system generates data for the entire PK/PD (pharmacokinetics/pharmacodynamics) experiment using one animal. Fig. 8 outlines

Fig. 7. A linear probe for peripheral soft tissue.

one of the several schemes showing how the system takes each sample and prepares for subsequent samples. Each animal has its own independent sampling protocol, start–stop time, and collection facilities. All functions are controlled by a single notebook computer. Animal damage to catheters is eliminated by afﬁxing the catheter to a tether assembly mounted to a counter-balanced arm that keeps the catheter out of the animals’ reach or view. This is identical in all respects to the apparatus used for microdialysis. The catheter tubing is protected from being twisted by animal movement through use of the Raturn system as described above. Positional inﬂuences on catheter placement may occur during repeated handling of the animal, which may cause a catheter tip to shift within the blood vessel. This is unavoidable during manual blood sampling as a rat is removed from its home cage, placed into a restraining device, and then returned to the home cage. Once an animal is installed, the only time it is touched again is during a drug administration procedure that necessitates handling (e.g., gavage). Dosing methods such as

176

SALINE

RAT

COLLECTOR

1

SALINE

RAT

COLLECTOR

SALINE

RAT

COLLECTOR

SALINE

RAT

COLLECTOR

SALINE

RAT

COLLECTOR

2

SALINE

RAT

COLLECTOR

3

4

SALINE

RAT

COLLECTOR

5

6

SALINE

RAT

COLLECTOR

Whole Blood Saline Blood diluted with saline

7

WASTE

8

Fig. 8. One scheme showing in sequential steps how the automated system collects each blood sample and prepares for subsequent samples.

177

IV infusions or use of gastric or duodenal catheters require no contact with the animal and are highly recommended. VI.B. Bile sampling Rats have many endearing features and one of these is the absence of a gallbladder. The bile duct leading from the liver to the duodenum is thus available for timely kinetic studies of an important mechanism for disposition of xenobiotic metabolites, including prospects for enterohepatic recirculation in which bile components are reabsorbed across the intestinal membrane. In traditional bile sampling, the rat is anesthetized or restrained and bile is collected continuously from a cannulated duct. It is also possible in such circumstances to use an in-line microdialysis probe, like that shown in Fig. 9, which allows for continuous return of bile to the duodenum while small molecules are recovered, in part, through the dialysis membrane for subsequent fraction collection. While this approach is interesting, it is not very practical for the higher molecular weight drugs that are commonly encountered in drug development today. It also is a relatively fragile approach that takes real expertise to maintain patency. Traditional swivels can also be used with the bile generally collected at or below the elevation of the animal, using gravity as an advantage, or at least not a disadvantage. The technology has recently advanced to permit collecting bile automatically with awake, freely moving rats without interruption of the ﬂuid line by a liquid swivel. The Raturn makes this possible with a very simple method involving continuous bile ﬂow (Fig. 10) into an advancing fraction collector. This can also be a simple method to

Fig. 9. An in-line microdialysis probe allowing the sampled physiological ﬂuid (e.g., bile) to ﬂow continuously while dialysates is collected ex vivo.

collect a few fractions manually for the purpose of identifying metabolites in bile. In contrast, a more sophisticated approach allows for periodic sampling with intervening return of bile ﬂow to the animal and supplementation of bile salts, leaving the normal physiology intact (Fig. 11). It thus becomes possible to simultaneously correlate disposition of metabolites in bile with systemic drug in circulation in the same awake and unrestrained animal.

VI.C. Fraction collection The ﬁrst fraction collectors were highly sophisticated biological entities known as graduate students, laboratory technicians, and postdocs. Their reliability was somewhat variable depending on their state of consciousness and need for a cup of coffee. In the mid-1980s, Ungerstedt’s laboratory and CMA/microdialysis became very aware of this problem and undertook the design of several alternative fraction collectors from modest room temperature devices to more sophisticated, refrigerated units capable of handling smaller volume than had ever been achieved in a commercial fraction collector. Even then, the standard for

Fig. 10. Bile ﬂow is traditionally collected into a vial replaced manually or into an advancing fraction collector below the level of the animal. The Raturn makes this process easy for a freely moving animal.

178 Recirculation Phase

V1

V2

V3

Collection Phase

V4

V1

Bile Salt Solution Bile

V2

V3

V4

overflow waste sample collection

Fig. 11. Bile sampling that allows for periodic sampling with return of bile ﬂow and supplementation of bile salts.

collection was the 300 mL glass autosampler vial, and this is still the case today, 20 years later. These vials can be capped in various ways to prevent sample loss by evaporation. Even with manual collection, addition of acid and/or antioxidants was a common procedure to help stabilize fragile compounds from air oxidation. Fraction collectors can enable the collection from several probes simultaneously (usually two). They are programmable from software resident on laptop computers and are quite reliable in a mechanical sense. It took many years of engineering to achieve an optimum result, allowing for conﬁdence in overnight sampling. As systemic blood sampling from indwelling catheters became achievable simultaneously with microdialysis, the same technology was easily adapted to collection of whole blood (serum) and plasma. In the later case, the use of anticoagulants became an option. VI.D. Automating collection from ultrafiltration probes

mouth. BASi ﬁrst commercialized these devices and it was natural to explore means of automation. The original devices used vacutainers as the source of both vacuum and the collection tube. Later, miniature peristaltic pumps were considered as a vacuum source, as were larger vacuum reservoirs connected to a mechanical vacuum pump. In the latter case, a fraction collector can be placed between the vacuum source and the sampling probe to facilitate automated collection of fractions. This is illustrated in Fig. 12. The same refrigerated micro fraction collector used for blood samples and dialysates is useful for ultraﬁltration, lowering the cost. While ultraﬁltration sampling tends to be both slower and physiologically larger than for microdialysis, it does not require any external ﬂuid and is especially interesting for high recovery of small hydrophilic substances like electrolytes and sugars. Fig. 13 illustrates a fraction collector widely used for both microdialysis and blood sampling. VII. Bioanalytical chemistry

In vivo ultraﬁltration probes were pioneered in the 1990s (see chapter 2.6) with an original focus on glucose, but then interesting applications to electrolytes and drugs became clear. We even published on using them in human applications, particularly to saliva with probes used in the

VII.A. Online monitoring Because microdialysis is a continuous ﬂow methodology, there is a temptation to make measurements online, but there are also great limitations in

179

better qualitative and quantitative performance when samples are isolated, this generalization is nearly always true. For one thing, it is easier to optimize the analytical performance when it does not depend on the peculiar limitations of physiological ﬂuid in terms of ionic strength and pH. Development of online sensors is rarely justiﬁed when existing tools are readily available for immediate use. Calibration is also far easier to manage for off-line systems. Another factor here is that in many institutions, laboratory animals and analytical chemistry are rigidly isolated from each other by either tradition or regulatory considerations.

VII.B. Discrete sample processing Fig. 12. For in vivo ultraﬁltration sampling of peripheral tissue (chapter 2.6), a programmed fraction collector can be placed between the vacuum source and the in vivo sampling probe (shown here in a vial).

Fig. 13. One example of a commercial peltier cooled fraction collector adaptable to microdialysis, ultraﬁltration, bile, and blood collection.

doing so. Very few examples have been published and these tend to focus on glucose and lactate enzyme sensors, mass spectrometry (see chapters 3.1, 3.2 and 3.4), or capillary electrophoresis (see chapter 3.4). Our laboratory has participated in such studies more for fun than out of necessity. Doing something ‘‘just to see if we could do it’’ has always been a motivation of mine. While it is a generalization to say that analytical chemistry provides

The discovery phase of pharmaceutical research involves synthesizing libraries and screening them in vitro for effectiveness, metabolism, and potential toxicity. Following those screens, the muchreduced number of compounds needs to be tested in mammalian species. The pharmaceutical industry and the FDA do not yet have full conﬁdence in proceeding to clinical trials based on in vitro data or predictive software alone. Animal studies remain of primary importance to compound evaluations, resulting in large numbers of samples. Our goal was to be able to screen animal samples in a high-throughput format using 96-well plates, similar to investments in sample preparation for LC/MSMS and immunoassays. Fig. 14 illustrates one format for a pharmacokinetics protocol utilizing eight rats with ten samples for each. Robotics for 96-well plates is available in virtually all laboratories pursuing drug development. Such plates can be added to, shaken or centrifuged, and the samples transferred to other plates. Samples on 96-well plates can be processed automatically by liquid–liquid extraction or by solid phase extraction. Excellent 96-well autosamplers for LC/MSMS can then manage multiplewell plates for completely automatic processing overnight. Today, virtually all pharmaceutical research laboratories have implemented such sample preparation automation systems. These are sometimes unaffordable in academic environments, but the cost is rapidly coming down.

180

Fig. 14. Samples for a bioavailability screen using eight animals can be processed in one 96-well plate. Animals can be automatically dosed one day IV and the next day via gastric catheter. Thus a complete oral bioavailability experiment can be performed with each single animal. Dose ranging studies can also be automated to explore if the pharmacokinetics are linear or more complex.

VII.C. Imaging One of the most elegant bioanalytical techniques in recent years is whole body imaging of laboratory animal models, although restrained or anesthetized. This subject is too vast for the current chapter, but it is worth noting that magnetic, radiochemical, and optical imaging schemes can be carried out with simultaneous sampling of physiological ﬂuids.

VIII. Conclusion While we often marvel at the rapid advances in technology, the truth is quite different. Important advances take a lot of time. From Delgado’s pioneering ideas at Yale in the early 1970s to Ungerstedt’s more practical embellishments in the late 1970s, microdialysis continues to evolve. Over some 35 years, opportunities presented themselves from unanticipated sources, including electroanalytical chemistry, immunoassays, chromatography, capillary electrophoresis, and LC/MSMS.

Without these bioanalytical advances, microdialysis would have no practical value. Automation became affordable over this period of time. Handling and preserving very small volumes of ﬂuids required utilization of exquisitely precise capillary tubes, stepping motors, peltier coolers, and digital controllers. As they became available, microdialysis naturally advanced. The rise of automated blood sampling arose directly from microdialysis. Customer complaints about ‘‘membrane recovery’’ and about problems with ‘‘sticky drugs and peptides’’ and the desire for an ‘‘instantaneous sample’’ drove us to add blood to the repertoire. A conservative cry from pharma that ‘‘we have always sampled blood’’ was even more persuasive. Thus microdialysis begat automated blood sampling, which begat automated awake animal PK/PD and a series of associated tools for physiological and behavioral measures. While predicting maturity is not without risk, today microdialysis, ultraﬁltration, and blood sampling can all be highly automated with good reliability. Other chapters in this volume are oriented toward speciﬁc applications, but it is so easy today to type a few key words in an online information resource and recover literature resources in a few seconds. This too was not available when microdialysis began and we had to spend much time in the library. The role of literature reviews has thus also changed and long lists of reference citations are not very compelling.

Acknowledgments It has been my very good fortune to have talented and collegial colleagues and customers. A large team of graduate students, postdocs, engineers, bioanalytical chemists, and laboratory animal technicians contributed to this effort. Major contributions were made by the curious customers who asked if the impossible today can become the reality tomorrow. These selfless visionaries create the challenge that drives our imagination. They have come from big pharma, discovery companies, medical schools, and research institutes. Automated sampling and combinatorial pharmacology arose from the eco system that life science business

181

has become as the old silos crumble. The credit is collectively shared. The author declares a conﬂict of interest. He is the founder of Bioanalytical Systems, Inc. (BASi) and this company designs, manufactures, sells and carries out contract research using the apparatus described in this chapter.

References Bohs, C., Cregor, M., Gunaratna, G. and Kissinger, C. (2000) Culex automated blood sampler part II: managing freelymoving animals and monitoring their activity. Curr. Sep., 18(4): 147–151. Gunaratna, P.C., Kissinger, P.T., Kissinger, C.B. and Gitzen, J.F. (2004) An automated blood sampler for simultaneous sampling of systemic blood and brain microdialysates for drug absorption, distribution, metabolism, and elimination studies. J. Pharmacol. Toxicol. Methods, 49: 57–64. He, H., Kramp, R., Ramos, L. and Bakhtiar, R. (2001) A preliminary study on the feasibility of an automated blood-sampling system in conjunction with liquid chromatography/mass

spectrometry. Rapid Commun. Mass Spectrom., 15(18): 1768–1772. Long, H., Zhu, Y.X., Cregor, M., Tian, F.F., Coury, L., Kissinger, C.B. and Kissinger, P.T. (2001) Liquid chromatography with multi-channel electrochemical detection for the determination of epigallocatechin gallate in rat plasma utilizing an automated blood sampling device. J. Chromatogr. B, 763: 47–51. Tian, F.F., Zhu, Y.X., Long, H., Cregor, M., Xie, F.M., Kissinger, C.B. and Kissinger, P.T. (2002) Liquid chromatography coupled with multi-channel electrochemical detection for the determination of daidzin in rat blood sampled by an automated blood sampling system. J. Chromatogr. B, 772: 173–177. Zhu, Y.X., Chiang, H., Wulster-Radclife, M., Hilt, R., Wong, P., Kissinger, C.B. and Kissinger, P.T. (2005) Liquid chromatography/tandem mass spectrometry for the determination of carbamazepine and its main metabolite in rat plasma utilizing an automated blood sampling system. J. Pharm. Biomed. Anal., 38: 119–125. Zhu, Y.X., Huang, T.H., Cregor, M., Long, H., Kissinger, C.B. and Kissinger, P.T. (2000) Liquid chromatography with multi-channel electrochemical detection for the determination of trans-resveratrol in rat blood utilizing an automated blood sampling device. J. Chromatogr. B, 740: 129–133.

CHAPTER 2.4

Dopamine–acetylcholine interactions in the brain studied by in vivo microdialysis George G. Nomikos Amgen Inc., Cambridge Research Center, Neuroscience, Cambridge, MA, USA

Abstract: Over the past 20 years, in vivo microdialysis methodologies have provided invaluable information on interactions between the dopaminergic and the cholinergic systems in the brain. Such knowledge has helped us to conﬁgure the dynamic interrelationship of these two major neurotransmitter systems, in the whole animal, under physiological and simulated disease states and promote the discovery and development of therapeutic agents. Changes in dopaminergic neurotransmission through dopamine (DA) D1and D2-like receptors affect acetylcholine (ACh) efﬂux, and changes in cholinergic neurotransmission through muscarinic and nicotinic receptors inﬂuence DA efﬂux in the brain. Thus, dopaminergic or cholinergic neurotransmission dysfunctions associated with neuropsychiatric disease states modeled in experimental animals should inﬂuence ACh and DA concentrations, respectively, in the interstitial/extracellular ﬂuid that is sampled by microdialysis. In turn, pharmacological treatments that restore dopaminergic/cholinergic neurotransmission should affect ACh/DA efﬂux in a predictable, traceable manner. Consequently, in vivo microdialysis sampling and related analyses offer a rational, experimental approach in evaluating the pathophysiology and treatment of diseases in neuropsychiatry by providing markers of neurobiological homeostatic mechanisms and relevant drug actions.

actions of medications used for their treatment. The development of various in vivo microdialysis sampling and analytical methods over the last 20 years has advanced our understanding of how dopaminergic and cholinergic neurotransmission elements interface with one another in brain regions of experimental animals under steady-state conditions and upon exposure to behavioral or pharmacological stimuli. In turn, this knowledge has provided a neurobiological frame to view the underlying mechanisms of neuropsychiatric disorders as well as to develop neurochemical markers to gauge pharmacological and, ultimately, therapeutic drug actions. Early in vivo microdialysis studies underlined the significance of certain methodological conditions for optimal detection of physiologically

I. Introduction The classical neurotransmitters dopamine (DA) and acetylcholine (ACh) interact at different functional levels in distinct regions of the brain, notably the basal ganglia, the neocortex, and the hippocampus, to regulate motoric, affective, and cognitive processes. The dynamic interrelationship of these neurotransmitter systems seems to play a major role in the pathophysiology of neuropsychiatric disorders, such as schizophrenia, depression, Parkinson’s disease, and substance dependence, as well as in the neuropharmacological

Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

183

DOI: 10.1016/S1569-7339(06)16010-5 Copyright 2007 Elsevier B.V. All rights reserved

184

relevant concentrations of neurotransmitters in the interstitial ﬂuid and in the microdialysate, as well as in revealing DA/ACh interactions, primarily in the basal ganglia of conscious rats (e.g., Westerink et al., 1988, 1990; Bertorelli and Consolo, 1990; Bertorelli et al., 1990; Damsma et al., 1990a, b; De Boer et al., 1990, 1992; Imperato et al., 1994b). Subsequent studies highlighted the relevance of acetylcholinesterase inhibitors in the perfusion solution to examine the role of the dopaminergic tone and the ﬁne balance between DA-D1 and -D2 receptors in regulating ACh efﬂux in the striatum (Str) (De Boer and Abercrombie, 1996; Acquas and Fibiger, 1998; Acquas and Di Chiara, 1999a, b). Dual microdialysis experiments, combined with local administration of receptor selective compounds, further characterized the neuroanatomical substrates implicated in the studied DA/ACh interactions (e.g., Nisell et al., 1994a, b; Yamamuro et al., 1995; Abercrombie and DeBoer, 1997; Schilstrom et al., 1997, 1998; Enrico et al., 1998; Moore et al., 1999; Arnold et al., 2000, 2001; Moss et al., 2003; Tzavara et al., 2003b; Zackheim and Abercrombie, 2005). In addition, behavioral assessments in parallel to the neurochemical measurements added another dimension in the multifaceted relationship between dopaminergic and cholinergic mechanisms further validating its functional significance in the whole animal (e.g., Hurd et al., 1990; Imperato et al., 1991; Mark et al., 1992, 1995, 1999; Yamamuro et al., 1995; Hildebrand et al., 1998; Schilstrom et al., 1998; Neigh et al., 2004). Recently, in vivo microdialysis experiments in genetically engineered mice with targeted deletions of speciﬁc receptors gave a new insight in the evaluation of the dynamics of DA/ ACh interactions and their relevance to the pathophysiology and treatment of neuropsychiatric disorders (e.g., Picciotto et al., 1998; Gerber et al., 2001; Marubio et al., 2003; Laplante et al., 2004; Tzavara et al., 2004). In essence, the evolutionary history of the in vivo microdialysis sampling technique with its versatility, robustness, and applicability, in conjunction with advances in analytical methods, tracks the scientiﬁc information gained over the years on the functional interplay of these two major neurotransmitter systems in the brain.

II. DA/ACh neurotransmission elements and microdialysis methodological variables A basic coverage of dopaminergic and cholinergic neurotransmission elements follows below; this will also set the background for the actual microdialysis work discussed and depicted throughout this section. Dopaminergic neurons primarily project from the midbrain, that is, the substantia nigra (SN) and the ventral tegmental area (VTA), to the forebrain, mainly the Str, the nucleus accumbens (NAcc), and the frontal cortex (FC). Dopamine exerts its actions through D1-like (D1, D5) and D2-like (D2, D3, D4) receptors, while its actions in the extracellular compartment are terminated by active reuptake through the dopamine transporter (DAT) and enzymatic catabolism through monoaminoxidase (MAO) and catecholO-methyltransferase (COMT) activity. Cholinergic interneurons are found in the basal ganglia, in both the Str and the NAcc; cholinergic neurons also project from the basal forebrain and the medial septum to the neocortex and the hippocampus, respectively; additional cholinergic afferent projections are found in the midbrain (including the VTA and the SN) originating from the brainstem, the pedunculopontine tegmentum (PPT), and the laterodorsal tegmentum (LDT). ACh exerts its actions through the metabotropic muscarinic receptors (M1–M5) or the ionotropic nicotinic receptors (nAChRs) that are either hetero- or homomeric complexes constituted from different a or b subunits, whereas its actions in the interstitial ﬂuid are terminated through the enzymatic activity of acetylcholinesterase (AChE). Initially, in vivo microdialysis studies examined the effects of compounds that affect dopaminergic or cholinergic neurotransmission administered systemically or topically through the probe on extracellular striatal concentrations of ACh or DA, respectively, followed by studies in the other main projection areas, the FC and the hippocampus (see above). As aforementioned, dual in vivo microdialysis experiments offered the advantage of measuring neurotransmitter changes in, for example, the nerve terminal regions in response to compounds administered in the respective somatodendritic regions, or in other distal areas of

185

the brain that transsynaptically affect neurotransmission. In some of these studies, concurrent evaluation of behavioral responses of the conscious, freely-moving animals provided additional information on DA/ACh dynamic changes (see above). Through systematic, methodological work, undertaken in a few laboratories around the globe, it became clear that the optimal conditions to run in vivo microdialysis studies to assessing DA/ACh interactions, and actually for any microdialysis study, should include: (i) physiological concentrations of Ca2+ (1.2–1.3 mM) in the perfusion solution, (ii) adequate post-implantation intervals (20–48 h), and (iii) running experiments in conscious animals to avoid the effects of anesthetics on the measured neurotransmitters (e.g., Westerink et al., 1988, 1990; Damsma et al., 1990a; De Boer et al., 1990, 1992; Damsma and Fibiger, 1991; Zis et al., 1991; Imperato et al., 1994b; Sato et al., 1996). In addition, Imperato et al. (1996) highlighted the importance of genetic background in the neural organization of DA/ACh interactions in limbic regions of the brain of mice. One issue that has received considerable attention revolves around the inclusion of AChE inhibitors in the perfusion solution that are often used to boost basal ACh efﬂux in the extracellular ﬂuid to levels that are readily detectable. Through an intriguing interchange of scientiﬁc data and information obtained by at least three to four research teams, it was unequivocally demonstrated that the concentration of an AChE inhibitor in the perfusion solution could quantitatively and even qualitatively inﬂuence the manner in which dopaminergic agents regulate ACh efﬂux in the brain (De Boer and Abercrombie, 1996; Acquas and Fibiger, 1998; Acquas and Di Chiara, 1999a, b). In fact, and especially after the increased sensitivity of the analytical systems for measuring ACh in the microdialysate, it became apparent that inclusion of an AChE inhibitor in the perfusion solution might not be necessary. It was shown, nonetheless, that physiologically relevant changes in ACh efﬂux were only moderately affected by the level of AChE inhibition, and that for some classes of compounds, for example, the antipsychotic drugs, only the magnitude, but not the direction, of the evoked changes in ACh efﬂux was inﬂuenced by

the presence of an AChE inhibitor in the perfusion solution (e.g., Moor et al., 1998; Ichikawa et al., 2002; Shirazi-Southall et al., 2002; Johnson et al., 2005). Clearly, caution should be exercised in interpreting results obtained with high concentrations of AChE inhibitors, as much as in determining whether pharmacological or physiological challenges have any effect on ACh efﬂux in their absence, given the pronounced enzymatic activity of AChE in eliminating neurotransmitter overﬂow. Changes in dopaminergic neurotransmission via D1- or D2-like receptors affect ACh efﬂux either at the level of the cholinergic nerve terminal regions (e.g., hippocampus, cortex, midbrain, Str) or the somatodendritic regions (e.g., basal forebrain, medial septum, Str) through either a direct action or indirectly through a transsynaptic mechanism that involves either proximal or distal neuronal circuits. In an analogous manner, changes in cholinergic neurotransmission via nAChRs or muscarinic receptors also inﬂuence DA efﬂux in the respective regions of the brain through direct or indirect, polysynaptic mechanisms. The changes in dopaminergic or cholinergic neurotransmission are effectuated either by pharmacological means with compounds that acutely or chronically modify neurotransmission, or by genetic manipulation of a speciﬁc target involved in neurotransmission, typically a receptor, and the consequences on basal and evoked ACh or DA efﬂux are evaluated, respectively. Accordingly, the work presented and discussed throughout this chapter focuses on threads of data on a speciﬁc topic, where in vivo microdialysis studies provided key information; supportive evidence obtained with other methodologies is given in the original contributions.

III. Dopaminergic regulation of ACh efﬂux in the brain III.A. DA-mediated effects on steady-state, basal ACh concentrations Interestingly, the effects of altered dopaminergic neurotransmission on steady-state, basal concentrations of ACh have not systematically been

186

examined with quantitative methods of in vivo microdialysis (e.g., Vinson and Justice, 1997) that allow for a more accurate calculation of the neurotransmitter concentrations in the extracellular compartment than conventional approaches. Using the no-net-ﬂux technique Laplante et al. (2004) found a significant reduction in basal hippocampal ACh levels in mice with genetic deletions of the DA-D5 gene. These results support the notion that DA acting through D1-like receptors exerts a stimulatory action on ACh release in the hippocampus (see below). The results of parallel studies assessing DA/ACh balance in the hippocampus or other parts of the brain of genetically engineered mice with other DA receptor, DAT, or catabolizing enzyme (MAO, COMT) deletions have not as yet been disclosed. Depletion of DA content in the Str brought about with injections of 6-hydroxydopamine (6-OHDA) or a combination of reserpine and a-methyl-para-tyrosine in rats have produced variable effects on extracellular striatal concentrations of ACh, as assessed by conventional microdialysis (e.g., Robertson et al., 1992; De Boer et al., 1993; Russi et al., 1993; Anderson et al., 1994; HerreraMarschitz et al., 1994; Imperato et al., 1994b; Cadoni et al., 1995; Johnson and Bruno, 1995; Ikarashi et al., 1997a; Zackheim and Abercrombie, 2005). It appears that the extent and the neurochemical/neuroanatomical speciﬁcity of the depletions, as well as the in vivo microdialysis variables mentioned above, bear a major impact on the results of these studies. Similarly, the consequences of depleting DA content in the FC or the hippocampus of rats with midbrain 6-OHDA injections on basal ACh efﬂux (e.g., Nilsson et al., 1992; Day et al., 1994) need to be examined further, taking into consideration all the methodological variables outlined above. Repeated administration of compounds that increase or decrease dopaminergic neurotransmission, at least acutely, variably affect basal ACh efﬂux in the brain, depending on the regimen followed, the withdrawal interval (time between the last injection and the actual experiment) and the long-term consequences on DA efﬂux. Thus, chronic treatment with the DA reuptake inhibitor, nomifensine, increased basal ACh efﬂux in the

Str of conscious rats (Hernandez et al., 2006), whereas in animals withdrawn from chronic administration of amphetamine, nicotine, ethanol, or phencyclidine (that stimulate DA efﬂux acutely) or in animals receiving repeated injections of the DAD1 receptor agonist, dihydrexidine, no difference in basal ACh efﬂux in the Str, the NAcc, the FC, or the hippocampus, respectively, compared with control animals was found (Bickerdike and Abercrombie, 1997; Rada et al., 2001, 2004; Wade and Nomikos, 2005). Similarly, no sustained effect on basal ACh efﬂux in the Str or the NAcc of animals receiving repeated administration of neuroleptic drugs that potently block DA-D2 receptors was shown (Imperato et al., 1994a; Osborne et al., 1994), whereas, a persistent decrease in basal ACh efﬂux was detected after repeated administration of diazepam that acutely suppresses DA release (Rada and Hoebel, 2005). In conclusion, the results of studies on the effects of changes in dopaminergic neurotransmission on steady-state, basal ACh efﬂux in the brain point to a stimulatory action of DA through D1-like receptors, although various experimental conditions and neurotransmission-related adaptations may have a significant inﬂuence on the obtained results.

III.B. Effects of changes in dopaminergic neurotransmission through D1/D2 receptors on ACh efflux Recent in vivo microdialysis data that have been obtained under more optimal methodological conditions than in earlier studies (see above) support the notion that DA reciprocally regulates ACh efﬂux in the brain through a DA-D1 receptor-mediated stimulatory action and a DA-D2 receptor elicited inhibitory effect (e.g., Damsma et al., 1990a, b, 1991; Day and Fibiger, 1992, 1993; De Boer et al., 1990, 1992; Russi et al., 1993; Imperato et al., 1995; De Boer and Abercrombie, 1996; Abercrombie and DeBoer, 1997; Ikarashi et al., 1997a, b; Acquas and Fibiger, 1998; Acquas and Di Chiara, 1999a, b). Within the NAcc/Str complex, the facilitatory effect of dopaminergic neurotransmission through DA-D1 receptors on ACh efﬂux seems to involve a composite multisynaptic

187

action with changes in other neurotransmitter systems that are initiated either at an intrastriatal or most likely at an extrastriatal site and impinge upon the cholinergic interneurons (Fig. 1; Bertorelli and Consolo, 1990; Damsma et al., 1990a, b, 1991; De Boer et al., 1990, 1992; Robertson et al., 1992; Zocchi and Pert, 1993; Anderson et al., 1994; Imperato et al., 1994a, b; Steinberg et al., 1995, 1998; Consolo et al., 1996, 1999; De Boer and Abercrombie, 1996; Abercrombie and DeBoer, 1997; Acquas et al., 1997; Ikarashi et al., 1997a, b; Acquas and Fibiger, 1998; Keys and Mark, 1998; Acquas and Di Chiara, 1999a, b; Whitehead et al., 2001; FC

DA-D1R/D2R NMDAR/AMPAR NAcc/Str

GABAAR

VTA/SN

Glu GABA DA

Fig. 1. Schematic drawing of DA/ACh interactions at the level of the basal ganglia (NAcc/Str), as revealed by in vivo microdialysis studies (e.g., Bertorelli and Consolo, 1990; Damsma et al., 1990a, b, 1991; De Boer et al., 1992, 1996; Robertson et al., 1992; Russi et al., 1993; Zocchi and Pert, 1993; Anderson et al., 1994; Imperato et al., 1994b; Mandel et al., 1994; Steinberg et al., 1995, 1998; Consolo et al., 1996; De Boer and Abercrombie, 1996; Acquas et al., 1997; Abercrombie and DeBoer, 1997; Ikarashi et al., 1997a, b; Acquas and Fibiger, 1998; Keys and Mark, 1998; Acquas and Di Chiara, 1999a, b; Zackheim and Abercrombie, 2005). DA released in the NAcc/Str affects ACh efﬂux directly through stimulation of D2 receptors that are localized on cholinergic interneurons and indirectly through stimulation of proximal and distal D1 receptors and transsynaptic neuronal circuits that involve, among others, changes in GABA and Glu release (from neurons originating in the FC and the thalamus), causing a decrease and an increase in ACh efﬂux, respectively. Abbreviations: DA, dopamine; ACh, acetylcholine; GABA, g-aminobutyric acid; Glu, glutamate; NAcc/Str, nucleus accumbens/striatum; VTA/SN, ventral tegmental area/ substantia nigra; FC, frontal cortex; DA-D1R/D2R, dopamine D1/D2 receptors; NMDAR/AMPAR, N-methyl-D-aspartate/ a-amino-3-hydroxy-5-methyl-4-isoxazolepropionate receptors; GABAAR, GABAA receptors.

Zackheim and Abercrombie, 2005). The distal sites that are involved in the tonic DA-D1 regulation of striatal ACh efﬂux include the parafascicular thalamic nucleus, the FC, and the SN (pars reticulata), possibly through an increase in glutamatergic neurotransmission via N-methyl-Daspartate (NMDA)/a-amino-3-hydroxy-5-methyl4-isoxazolepropionate (AMPA) receptors, whereas a local striatal neuronal circuit through an increase in g-aminobutyric acid (GABA)/neurokininA/ substanceP neurotransmission (via GABAA/ tachykinin receptors) also seems to be implicated (see Fig. 1; Damsma et al., 1991; Zocchi and Pert, 1993; Anderson et al., 1994; Consolo et al., 1996; Abercrombie and DeBoer, 1997; Acquas et al., 1997; Steinberg et al., 1995, 1998; Zackheim and Abercrombie, 2005). In addition, DA seems to exert a tonic inhibitory control over spontaneous ACh efﬂux within the NAcc/Str complex that is directly mediated by DA-D2 receptors (Fig. 1; Bertorelli and Consolo, 1990; Damsma et al., 1990a, b, 1991; De Boer et al., 1992; Robertson et al., 1992, 1993; Russi et al., 1993; Imperato et al., 1994a, b; Mandel et al., 1994; De Boer and Abercrombie, 1996; Abercrombie and DeBoer, 1997; Acquas et al., 1997; Ikarashi et al., 1997a, b; Acquas and Fibiger, 1998; Keys and Mark, 1998; Acquas and Di Chiara, 1999a, b). Similarly to DA-D2 receptors, stimulation of DA-D3 receptors with selective ligands appears to mediate an inhibitory action on striatal ACh efﬂux (Sato et al., 1994; Millan et al., 2004). Changes in dopaminergic neurotransmission, through DA-D1 receptors, affect ACh efﬂux within the hippocampus and the FC, similarly to the Str (i.e., increases in dopaminergic tone result in enhanced ACh efﬂux), although the precise site(s) and mechanism(s) of action have not been fully elucidated (Nilsson et al., 1992; Day and Fibiger, 1992, 1993, 1994; Imperato et al., 1993; Acquas et al., 1994; Day et al., 1994; Imperato et al., 1996; Steele et al., 1997; Arnold et al., 2001; Laplante et al., 2004). The role of DA-D2 receptors in regulating ACh efﬂux in the neocortex and the hippocampus appears to be somewhat inconspicuous with a subtle facilitatory action under certain experimental conditions (e.g., Day and Fibiger, 1992, 1994; Imperato et al., 1994a; Umegaki et al., 2001).

188

Interestingly, recent data with a selective DA-D3 receptor antagonist reveal a stimulatory action of this compound on cortical ACh efﬂux, although the exact site and mechanism of action remain unclear (Lacroix et al., 2003). The transsynaptic mechanisms at the level of the extended basal forebrain, where the cholinergic neurons originate, and at the nerve terminal regions of the FC, the hippocampus, and even the NAcc that participate in regulation of DA/ACh interactions in the brain have recently received considerable attention given their importance in cognitive functions and behavioral neuroadaptative processes implicated in pathophysiological states (e.g., Yamamuro et al., 1995; Acquas et al., 1998; Izaki et al., 1998; NailBoucherie et al., 1998; Moore et al., 1999; Arnold et al., 2000, 2001; Laplante et al., 2004; Neigh et al., 2004; Fujishiro et al., 2005; del Arco and Mora 2005; Zmarowski et al., 2005). In this regard, altered cholinergic neurotransmission in the brain elicited by dependence-producing agents that primarily affect other monoaminergic/dopaminergic functions seems to underlie their deleterious effects on mood, cognition, and impulse control (e.g., Hurd et al., 1990; Guix et al., 1992; Imperato et al., 1993; Taguchi et al., 1993, 1998; Zocchi and Pert, 1994; Bickerdike and Abercrombie, 1997; Mark et al., 1999; Nelson et al., 2000; Rada et al., 2001a, b, 2004; Tzavara et al., 2003a; Rada and Hoebel, 2005; Nair and Gudelsky, 2006; Pisanu et al., 2006); these ACh-related dys- or deregulations may provide important clues in identifying the neuronal processes involved in substance dependence and comorbid psychiatric illnesses. Importantly, various psychotropic compounds that are commonly used to treat neuropsychiatric disorders characterized by cognitive, affective, and motoric dysfunctions, appear to affect ACh efﬂux in the brain through neuronal mechanisms that involve dopaminergic neurotransmission elements. For example, atomoxetine, a selective norepinephrine reuptake inhibitor and the ﬁrst non-stimulant medication for attention-deficithyperactiviy-disorder (ADHD), which increases DA efﬂux selectively in the FC, has been shown to increase ACh efﬂux selectively in the FC, as well, through DA-D1 receptor stimulation (Tzavara et al., 2005); this effect is postulated to play a

crucial role in atomoxetine’s procognitive actions in the clinic. Earlier studies on DA/ACh interaction also provided important leads in delineating the mechanism(s) of action of pharmacological and other neurobiological treatments in neuropsychiatry (e.g., Zis et al., 1991, 1992; De Boer et al., 1993; Di Chiara et al., 1993; Imperato et al., 1993, 1994a; Russi et al, 1993; Osborne et al., 1994; Acquas and Fibiger, 1996). In summary, changes in dopaminergic neurotransmission affect ACh efﬂux in the brain in a receptor- and region-selective manner through effects mediated locally or distally via either a direct action or a multisynaptic neuronal circuit. This intricate regulation appears to be of critical significance for our understanding of the pathophysiological mechanisms implicated in neuropsychiatric disease states and the mode of action of related pharmacological treatments.

IV. Cholinergic regulation of DA efﬂux in the brain IV.A. ACh-mediated effects on steady-state, basal DA concentrations Similarly to the results discussed above (see Section III.A), the effects of major pharmacological or genetic manipulations of cholinergic neurotransmission elements on extracellular concentrations of DA in the brain have not systematically been assessed with quantitative microdialysis methods (e.g., Parsons et al., 1991a, b; Crippens et al., 1993; Smith and Weiss, 1999; He and Shippenberg, 2000; Shippenberg et al., 2000). In an early microdialysis study, Herrera-Marschitz et al. (1990) showed a marked drop in extracellular concentrations of cortical and striatal DA after neurochemical lesions of the nucleus basalis, where the majority of cortically projecting cholinergic neurons reside, by standard microdialysis in anesthetized rats. Using a semiquantitative method of assessing extracellular concentrations of DA in the Str of M1 muscarinic receptor deﬁcient mice, a signiﬁcant elevation of striatal DA levels was reported that most likely accounts for the observed spontaneous hyperlocomotion in these animals (Gerber

189 FC M1R M4R ACh

NAcc/Str VTA/SN PPT/LTD ACh

DA

ACh

BF

Fig. 2. Schematic drawing of DA/ACh interactions at the level of the basal ganglia (VTA/SN and NAcc/Str), as revealed by in vivo microdialysis studies (e.g., Gerber et al., 2001; Tzavara et al., 2004). ACh released in the VTA/SN from cholinergic afferents originating in the PPT/LTD tegmental nuclei affect DA neuronal activity and release in the NAcc/Str complex. Cortical (from BF originating cholinergic neurons) and intrastriatal ACh release also affects DA neuronal activity through transsynaptic neuronal circuits that involve changes in glutamatergic (corticostriatal) and peptidergic/GABAergic (striatonigral) neurotransmission in the NAcc/Str and VTA/SN regions, respectively. ACh affects DA function through M1 and M4 muscarinic receptors interspersed with dopaminergic, glutamatergic, and peptidergic/GABAergic neuronal elements in the basal ganglia. Abbreviations: DA, dopamine; ACh, acetylcholine; NAcc/Str, nucleus accumbens/striatum; VTA/SN, ventral tegmental area/substantia nigra; FC, frontal cortex; BF, basal forebrain; PPT/LTD, pedunculopontine/laterodorsal tegmentum; M1R-M1, muscarinic receptors; M4R-M4, muscarinic receptors.

et al., 2001; Fig. 2). Interestingly, a state of overt hyperdopaminergia in the NAcc of mice with invalidation of the M4, but not the M2, muscarinic receptor gene was also indicated following standard microdialysis procedures (Tzavara et al., 2004; Fig. 2). The results of these two studies point to a dynamic regulation of dopaminergic neurotransmission via speciﬁc muscarinic receptors in the brain that may be compromised in pathological conditions characterized by dopaminergic dysfunction. In an analogous manner, it was shown that in mice with genetic deletions of nicotinic a4, but not a6 or b2 subunits, the basal striatal DA levels were twice as high as those observed in wildtype mice (Picciotto et al., 1998; Champtiaux et al., 2003; Marubio et al., 2003); these results indicate that a4-containing nAChRs exert a tonic control on steady-state, striatal DA efﬂux, which is mediated by a heterogeneous population of nAChRs. Evidently, detailed studies assessing the effects of pharmacological or genetic interference with cholinergic neurotransmitter elements on steady-state DA efﬂux in distinct regions of the brain under more optimized methodological conditions are warranted.

Assessments of the effects of chronic administration of compounds affecting cholinergic neurotransmission on basal DA efﬂux in nerve terminal brain regions are essentially restricted to those with nicotine itself. In most of these studies, no difference in the basal efﬂux of DA in the NAcc/ Str or the FC was found in rodents after chronic intermittent or continuous administration of nicotine using standard microdialysis measurements (e.g., Damsma et al., 1989; Janson et al., 1991; Nisell et al., 1996; Marshall et al., 1997; Hildebrand et al., 1998; Reid et al., 1998; Takahashi et al., 1998; Rada et al., 2001; but, see Gaddnas et al., 2002). Interestingly, using the no-net-ﬂux approach, Rahman et al. (2004b) showed a decrease in basal DA efﬂux in the NAcc of rats self-administering nicotine and subjected to spontaneous withdrawal; similar results had previously been obtained with conventional microdialysis methods in animals undergoing precipitated nicotine withdrawal with nicotinic receptor antagonists (see Hildebrand et al., 1998, 1999; Nomikos et al., 1999; Rada et al., 2001). In fact, the study by Hildebrand et al. (1999) further emphasized the importance of nicotinic receptor mechanisms within the VTA, the

190

somatodendritic brain region of the ascending dopaminergic projections to the NAcc and the FC (mesolimbocortical system), for the reinforcing and dependence-producing properties of nicotine (see below; Fig. 3). Also noteworthy is the study by Hatip-Al-Khatib et al. (2005) showing that chronic oral intake of AChE inhibitors increased not only basal ACh efﬂux, but also basal concentrations of DA in the ventral hippocampus that also receives dopaminergic innervation from the VTA, although the contribution of indirect muscarinic or nicotinic receptor stimulation was not determined (see also below). Undoubtedly, further detailed analyses are needed to capture more accurately the neuroadaptive responses of the dopaminergic systems reﬂected

Fig. 3. Schematic drawing of DA/ACh interactions at the level of the midbrain (VTA), as revealed by in vivo microdialysis studies (e.g., Nisell et al., 1994a; Schilstrom et al., 1998a, b, 2000, 2003; Fu et al., 2000a, b; Reid et al., 2000; Wonnacott et al., 2005). DA release in the NAcc is enhanced through stimulation of nAChRs and NMDA receptors localized on dopaminergic cell bodies in the VTA; nAChRs are also localized presynaptically on glutamatergic afferents facilitating Glu release upon stimulation. The nAChRs and NMDA receptors are activated by ACh released from cholinergic neurons originating in the PPT and by Glu released from cortical afferents, respectively (see also Blaha and Winn, 1993; Pidoplichko et al., 2004). Nicotine enhances DA release in the NAcc by stimulating nAChRs receptors in the VTA, an effect that underlies its reinforcing and dependence-producing properties. Abbreviations: DA, dopamine; ACh, acetylcholine; Glu, glutamate; NAcc, nucleus accumbens; VTA, ventral tegmental area; PPT, pedunculopontine tegmentum; FC, frontal cortex; nAChRs, nicotinic acetylcholine receptors; NMDAR, N-methyl-D-aspartate receptors.

at the level of basal DA efﬂux, in response to chronic administration of compounds that affect cholinergic neurotransmission, through muscarinic or nicotinic receptors, under physiologically appropriate conditions.

IV.B. Effects of changes in cholinergic neurotransmission through muscarinic or nicotinic receptors on DA efflux Enhanced cholinergic neurotransmission as a result of systemic or local administration of AChE inhibitors affects DA efﬂux in the brain in a compound-, concentration-, route-, brain region-, and procedure-dependent manner, although overall it appears that an increase in DA efﬂux is found in most cases (e.g., Damsma et al., 1988; Westerink et al., 1990; Baldwin et al., 1991; Blaha and Winn, 1993; Takahashi et al., 1993; Cuadra et al., 1994; Mori et al., 1995; O’Connor et al., 1995; Zhu et al., 1995, 1996; Giacobini et al., 1996; Warpman et al., 1996; Shearman et al., 2006). The stimulatory effects of one of these compounds, tacrine, on striatal DA efﬂux seem to involve activation of both muscarinic and nicotinic receptors (Warpman et al., 1996), although it remains to be determined whether this is due to its direct interaction with cholinergic receptors or indirectly through the elevated ACh efﬂux. Altered cholinergic neurotransmission through muscarinic receptors regulates DA efﬂux in the brain depending on the selectivity of the compound used, the receptor type involved, and the site of DA/ ACh interaction assessed. Thus, local stimulation of presynaptic M1 muscarinic receptors was shown to enhance and activation of M2 muscarinic receptors to inhibit DA efﬂux in the Str of the rat through a local neuronal circuit (de Klippel et al., 1993; Smolders et al., 1997). The precise sequence of events partaking in the tonic/phasic regulation of striatal DA efﬂux through changes in muscarinic receptor-related neurotransmission are not yet clear. Non-selective muscarinic receptor stimulation within the VTA was also shown to result in an enhanced efﬂux of DA not only locally, but also in the NAcc and the FC (Gronier et al., 2000). It is speculated that such an effect is mediated through M1

191

muscarinic receptor activation, as systemic administration of relatively selective M1 receptor agonists was later demonstrated to enhance cortical efﬂux of DA (Perry et al., 2001; Li et al., 2005). In addition, distal muscarinic receptor stimulation at the level of ventral hippocampus was demonstrated to enhance DA efﬂux in the NAcc, an effect probably mediated via M2/M4, but not M1, muscarinic receptor activation (Moss et al., 2003). It needs to be emphasized that the lack of selective compounds for the muscarinic receptors limits our ability to draw a definite conclusion on their speciﬁc involvement in DA efﬂux control. Further evidence on the regulation of DA efﬂux in the brain through altered cholinergic neurotransmission was provided with the use of muscarinic receptor knockout (KO) mice (see also above). These studies revealed that M1 and M4, but not M2, muscarinic receptors homeostatically regulate dopaminergic function in the NAcc/Str, as their genetic invalidation effectuates a state of hyperdopaminergia and enhanced responsiveness to psychostimulants (Gerber et al., 2001; Tzavara et al., 2004; Fig. 2). Taken together, the results of the studies mentioned above indicate that changes in cholinergic neurotransmission through muscarinic receptors modulate DA efﬂux in the brain in a dynamic manner in which the brain region, the speciﬁc muscarinic receptors, the tone of the cholinergic system, and the particular neuronal elements involved represent critical contributing factors. Importantly, muscarinic receptor dysfunction has grave neurobiological consequences for the homeostatic regulation of dopaminergic neurotransmission in the brain that is associated with pathophysiological mechanisms in neuropsychiatry and the therapeutic actions of drugs used in the clinic. Early in vivo microdialysis studies provided unequivocal evidence that stimulation of nAChRs with nicotine itself administered systemically or locally in dopaminergic projection areas increases DA efﬂux and metabolism in diverse regions of the brain (e.g., Damsma et al., 1988, 1989, 1990a, b; Mifsud et al., 1989; Brazell et al., 1991; Benwell and Balfour, 1992; Nakamura et al., 1992; Toth et al., 1992; Benwell et al., 1993). Subsequent studies afﬁrmed that stimulation of nAChRs within the VTA/SN somatodendritic region is, for the most part, responsible for nicotine’s stimulatory actions

on DA efﬂux in the NAcc/Str nerve terminal regions (Yoshida et al., 1993; Nisell et al., 1994a, b; Fu et al., 2000a, b; Wonnacott et al., 2005), effects that are directly relevant to nicotine’s reinforcing and dependence-producing properties. Clearly, though, nicotinic receptors in the dopaminergic nerve terminal regions also participate in the dynamic regulation of DA efﬂux in a physiologically relevant manner (Mifsud et al., 1989; Toth et al., 1992; Marshall et al., 1997; Fu et al., 2000b; Wonnacott et al., 2005). Other nicotinic receptor ligands appear to differentially regulate DA efﬂux in the brain, probably due to their interaction with distinct nAChRs (Mirza et al., 1996; Sacaan et al., 1996; Summers et al., 1996; Sziraki et al., 1998; Lecca et al., 2000; Seppa and Ahtee, 2000; Cohen et al., 2003; Rao et al., 2003; Bednar et al., 2004; Janhunen and Ahtee, 2004); this differentiation may offer the potential of using such ligands in nicotine dependence, as surrogates with low abuse liability. Repeated stimulation of nAChRs with either nicotine or other nicotinic receptor ligands also affected DA efﬂux in the brain in a compound-, region-, regimen-, animal strain-, and context-dependent manner (Damsma et al., 1989; Benwell and Balfour, 1992, 1998; Benwell et al., 1995; Nisell et al., 1996; Reid et al., 1996, 1998; Shoaib and Shippenberg, 1996; Cadoni and Di Chiara, 2000; Iyaniwura et al., 2001; Ferrari et al., 2002; Gaddnas et al., 2002; Rahman et al., 2003, 2004a, b; Bednar et al., 2004; Visanji et al., 2006). Pharmacological studies with selective nAChR ligands brought to light the receptor mechanisms and the local neuronal circuits involved in the stimulatory actions of nicotine on DA efﬂux in the brain (Schilstrom et al., 1998, 2000, 2003; Sziraki et al., 1998; Fu et al., 2000a, b; Reid et al., 2000; Pidoplichko et al., 2004). Thus, homo- (a7) and heteromeric (containing a4 and b2 subunits) nAChRs in the VTA/SN are implicated in the regulation of DA neuronal activity either directly or indirectly through changes in glutamatergic and GABAergic neurotransmission (Fig. 3); these neuronal events underlie the heightened activity of midbrain DA neurons and enhanced DA efﬂux in the forebrain structures that initiate and sustain nicotine intake and dependence. Additional studies with genetically modiﬁed mice attested the

192

speciﬁc role of nAChRs in regulating DA neuronal activity and release in the somatodendritic and nerve terminal dopaminergic regions (Picciotto et al., 1998; Champtiaux et al., 2003; Marubio et al., 2003). Collectively, this work indicated that combinations of a6 or a4 with b2 nAChRs may mediate the endogenous cholinergic modulation of DA release at the projection areas, whereas somatodendritic (non-a6)a4b2 heteromeric nAChRs most likely contribute to nicotine-shaped behavioral responses and dependence; these data also present valuable information about the nAChRmediated tonic/phasic regulation of DA efﬂux in the brain, as the selectivity of nicotinic ligands for various nAChRs that could be used for this purpose is admittedly rather limited. As discussed above, spontaneous or precipitated nicotine withdrawal resulted in marked decreases in DA efﬂux in the NAcc and related dopaminergic areas (Hildebrand et al., 1998; Nomikos et al., 1999; Panagis et al., 2000; Rada et al., 2001; Gaddnas et al., 2002), substantiating further the importance of nicotine’s interaction with dopaminergic neurotransmission elements in the brain for its dependence-producing qualities (Pontieri et al., 1996). In summary, in vivo microdialysis approaches have facilitated our understanding of how changes in cholinergic neurotransmission through nAChRs affect DA functioning in the brain in an instrumental manner that broadens our view of DA/ ACh-related pathophysiology in disease states and therapeutics.

sheds light on the underlying pathophysiological mechanisms; consequently, the gained information supports target validation efforts to rationally design, discover, and develop compounds to address unmet medical needs within the focus indications. For example, Gerber et al. (2001) and Tzavara et al. (2004) showed that the M1 and the M4 muscarinic receptor KO mice were spontaneously hyperactive and hyperdopaminergic and exhibited an enhanced responsiveness to psychostimulants, constituting an a forteriori experimental animal model for such unphysiological neurobiological responses. These results suggest that M1 and M4 muscarinic receptor-related neurotransmission might be dysregulated in pathological conditions, where an imbalanced dopaminergic function is encountered, such as schizophrenia, ADHD, substance dependence, or bipolar disorders, and, accordingly, compounds that restore muscarinic receptor function through M1 or M4 muscarinic receptors might be useful in the treatment of these nosological entities. Interestingly, Perry et al. (2001) and Li et al. (2005) reached similar conclusions using another experimental approach but still based on, among other methodologies, in vivo microdialysis to monitor extracellular concentrations of DA in response to relatively selective M4 and M1 muscarinic receptor ligands, respectively. These seemingly unrelated studies duly exemplify the applicability and usefulness of the in vivo microdialysis methodologies as a heuristic, experimental tool to unravel the neurobiological mechanisms of neuropsychiatric diseases and guide pertinent pharmacological research leading to their treatment.

V. Conclusions Acknowledgments In vivo microdialysis studies have offered compelling neurochemical evidence for the dynamic interrelationship between the dopaminergic and cholinergic systems in the brain. Changes in dopaminergic neurotransmission are associated with speciﬁc modiﬁcations in interstitial concentrations of ACh, and inversely, changes in cholinergic neurotransmission are accompanied by selective alterations in DA efﬂux in distinct regions of the brain. Thus, monitoring DA/ACh interactions with microdialysis in established or putative experimental animal models of neuropsychiatric disorders

I want to thank all my colleagues, who have contributed with their efforts and ideas to the work compiled and presented here. To the dear memory of Geert. References Abercrombie, E.D. and DeBoer, P. (1997) Substantia nigra D1 receptors and stimulation of striatal cholinergic interneurons by dopamine: a proposed circuit mechanism. J. Neurosci., 17: 8498–8505.

193 Acquas, E. and Di Chiara, G. (1999a) Dopamine D1 receptormediated control of striatal acetylcholine release by endogenous dopamine. Eur. J. Pharmacol., 383: 121–127. Acquas, E. and Di Chiara, G. (1999b) Local application of SCH 39166 reversibly and dose-dependently decreases acetylcholine release in the rat striatum. Eur. J. Pharmacol., 383: 275–279. Acquas, E. and Fibiger, H.C. (1996) Chronic lithium attenuates dopamine D1-receptor mediated increases in acetylcholine release in rat frontal cortex. Psychopharmacology (Berl.), 125: 162–167. Acquas, E. and Fibiger, H.C. (1998) Dopaminergic regulation of striatal acetylcholine release: the critical role of acetylcholinesterase inhibition. J. Neurochem., 70: 1088–1093. Acquas, E., Day, J.C. and Fibiger, H.C. (1994) The potent and selective dopamine D1 receptor agonist A-77636 increases cortical and hippocampal acetylcholine release in the rat. Eur. J. Pharmacol., 260: 85–87. Acquas, E., Wilson, C. and Fibiger, H.C. (1997) Nonstriatal dopamine D1 receptors regulate striatal acetylcholine release in vivo. J. Pharmacol. Exp. Ther., 281: 360–368. Anderson, J.J., Kuo, S., Chase, T.N. and Engber, T.M. (1994) Dopamine D1 receptor-stimulated release of acetylcholine in rat striatum is mediated indirectly by activation of striatal neurokinin1 receptors. J. Pharmacol. Exp. Ther., 269: 1144–1151. Arnold, H.M., Fadel, J., Sarter, M. and Bruno, J.P. (2001) Amphetamine-stimulated cortical acetylcholine release: role of the basal forebrain. Brain Res., 894: 74–87. Arnold, H.M., Nelson, C.L., Neigh, G.N., Sarter, M. and Bruno, J.P. (2000) Systemic and intra-accumbens administration of amphetamine differentially affects cortical acetylcholine release. Neuroscience, 96: 675–685. Baldwin, H.A., De Souza, R.J., Sarna, G.S., Murray, T.K., Green, A.R. and Cross, A.J. (1991) Measurements of tacrine and monoamines in brain by in vivo microdialysis argue against release of monoamines by tacrine at therapeutic doses. Br. J. Pharmacol., 103: 1946–1950. Bednar, I., Friberg, L. and Nordberg, A. (2004) Modulation of dopamine release by the nicotinic agonist epibatidine in the frontal cortex and the nucleus accumbens of naive and chronic nicotine treated rats. Neurochem. Int., 45: 1049–1055. Benwell, M.E. and Balfour, D.J. (1992) The effects of acute and repeated nicotine treatment on nucleus accumbens dopamine and locomotor activity. Br. J. Pharmacol., 105: 849–856. Benwell, M.E. and Balfour, D.J. (1998) The inﬂuence of lobeline on nucleus accumbens dopamine and locomotor responses to nicotine in nicotine-pretreated rats. Br. J. Pharmacol., 125: 1115–1119. Benwell, M.E., Balfour, D.J. and Birrell, C.E. (1995) Desensitization of the nicotine-induced mesolimbic dopamine responses during constant infusion with nicotine. Br. J. Pharmacol., 114: 454–460. Benwell, M.E., Balfour, D.J. and Lucchi, H.M. (1993) Inﬂuence of tetrodotoxin and calcium on changes in extracellular dopamine levels evoked by systemic nicotine. Psychopharmacology (Berl.), 112: 467–474.

Bertorelli, R. and Consolo, S. (1990) D1 and D2 dopaminergic regulation of acetylcholine release from striata of freely moving rats. J. Neurochem., 54: 2145–2148. Bertorelli, R., Hallstrom, A., Hurd, Y.L., Karlsson, A., Consolo, S. and Ungerstedt, U. (1990) Anaesthesia effects on in vivo acetylcholine transmission; comparisons of radioenzymatic and HPLC assays. Eur. J. Pharmacol., 175: 79–83. Bickerdike, M.J. and Abercrombie, E.D. (1997) Striatal acetylcholine release correlates with behavioral sensitization in rats withdrawn from chronic amphetamine. J. Pharmacol. Exp. Ther., 282: 818–826. Blaha, C.D. and Winn, P. (1993) Modulation of dopamine efﬂux in the striatum following cholinergic stimulation of the substantia nigra in intact and pedunculopontine tegmental nucleus-lesioned rats. J. Neurosci., 13: 1035–1044. Brazell, M.P., Mitchell, S.N. and Gray, J.A. (1991) Effect of acute administration of nicotine on in vivo release of noradrenaline in the hippocampus of freely moving rats: a dose-response and antagonist study. Neuropharmacology, 30: 823–833. Cadoni, C. and Di Chiara, G. (2000) Differential changes in accumbens shell and core dopamine in behavioral sensitization to nicotine. Eur. J. Pharmacol., 387: R23–R25. Cadoni, C., Pinna, A., Russi, G., Consolo, S. and Di Chiara, G. (1995) Role of vesicular dopamine in the in vivo stimulation of striatal dopamine transmission by amphetamine: evidence from microdialysis and Fos immunohistochemistry. Neuroscience, 65: 1027–1039. Champtiaux, N., Gotti, C., Cordero-Erausquin, M., David, D.J., Przybylski, C., Lena, C., Clementi, F., Moretti, M., Rossi, F.M., Le Novere, N., McIntosh, J.M., Gardier, A.M. and Changeux, J.P. (2003) Subunit composition of functional nicotinic receptors in dopaminergic neurons investigated with knock-out mice. J. Neurosci., 23: 7820–7829. Cohen, C., Bergis, O.E., Galli, F., Lochead, A.W., Jegham, S., Biton, B., Leonardon, J., Avenet, P., Sgard, F., Besnard, F., Graham, D., Coste, A., Oblin, A., Curet, O., Voltz, C., Gardes, A., Caille, D., Perrault, G., George, P., Soubrie, P. and Scatton, B. (2003) SSR591813, a novel selective and partial alpha4beta2 nicotinic receptor agonist with potential as an aid to smoking cessation. J. Pharmacol. Exp. Ther., 306: 407–420. Consolo, S., Baronio, P., Guidi, G. and Di Chiara, G. (1996) Role of the parafascicular thalamic nucleus and N-methylD-aspartate transmission in the D1-dependent control of in vivo acetylcholine release in rat striatum. Neuroscience, 71: 157–165. Consolo, S., Caltavuturo, C., Colli, E., Recchia, M. and Di Chiara, G. (1999) Different sensitivity of in vivo acetylcholine transmission to D1 receptor stimulation in shell and core of nucleus accumbens. Neuroscience, 89: 1209–1217. Crippens, D., Camp, D.M. and Robinson, T.E. (1993) Basal extracellular dopamine in the nucleus accumbens during amphetamine withdrawal: a ‘no net ﬂux’ microdialysis study. Neurosci. Lett., 164: 145–148. Cuadra, G., Summers, K. and Giacobini, E. (1994) Cholinesterase inhibitor effects on neurotransmitters in rat cortex in vivo. J. Pharmacol. Exp. Ther., 270: 277–284.

194 Damsma, G. and Fibiger, H.C. (1991) The effects of anaesthesia and hypothermia on interstitial concentrations of acetylcholine and choline in rat striatum. Life Sci., 48: 2469–2474. Damsma, G., Day, J. and Fibiger, H.C. (1989) Lack of tolerance to nicotine-induced dopamine release in the nucleus accumbens. Eur. J. Pharmacol., 168: 363–368. Damsma, G., de Boer, P., Westerink, B.H. and Fibiger, H.C. (1990a) Dopaminergic regulation of striatal cholinergic interneurons: an in vivo microdialysis study. Naunyn Schmiedebergs Arch. Pharmacol., 342: 523–527. Damsma, G., Robertson, G.S., Tham, C.S. and Fibiger, H.C. (1991) Dopaminergic regulation of striatal acetylcholine release: importance of D1 and N-methyl-D-aspartate receptors. J. Pharmacol. Exp. Ther., 259: 1064–1072. Damsma, G., Tham, C.S., Robertson, G.S. and Fibiger, H.C. (1990b) Dopamine D1 receptor stimulation increases striatal acetylcholine release in the rat. Eur. J. Pharmacol., 186: 335–338. Damsma, G., Westerink, B.H., de Vries, J.B. and Horn, A.S. (1988) The effect of systemically applied cholinergic drugs on the striatal release of dopamine and its metabolites, as determined by automated brain dialysis in conscious rats. Neurosci. Lett., 89: 349–354. Day, J. and Fibiger, H.C. (1992) Dopaminergic regulation of cortical acetylcholine release. Synapse, 12: 281–286. Day, J. and Fibiger, H.C. (1993) Dopaminergic regulation of cortical acetylcholine release: effects of dopamine receptor agonists. Neuroscience, 54: 643–648. Day, J.C. and Fibiger, H.C. (1994) Dopaminergic regulation of septohippocampal cholinergic neurons. J. Neurochem., 63: 2086–2092. Day, J.C., Tham, C.S. and Fibiger, H.C. (1994) Dopamine depletion attenuates amphetamine-induced increases of cortical acetylcholine release. Eur. J. Pharmacol., 263: 285–292. De Boer, P. and Abercrombie, E.D. (1996) Physiological release of striatal acetylcholine in vivo: modulation by D1 and D2 dopamine receptor subtypes. J. Pharmacol. Exp. Ther., 277: 775–783. De Boer, P., Abercrombie, E.D., Heeringa, M. and Westerink, B.H. (1993) Differential effect of systemic administration of bromocriptine and L-dopa on the release of acetylcholine from striatum of intact and 6-OHDA-treated rats. Brain Res., 608: 198–203. De Boer, P., Damsma, G., Fibiger, H.C., Timmerman, W., de Vries, J.B. and Westerink, B.H. (1990) Dopaminergic– cholinergic interactions in the striatum: the critical signiﬁcance of calcium concentrations in brain microdialysis. Naunyn Schmiedebergs Arch. Pharmacol., 342: 528–534. De Boer, P., Damsma, G., Schram, Q., Stoof, J.C., Zaagsma, J. and Westerink, B.H. (1992) The effect of intrastriatal application of directly and indirectly acting dopamine agonists and antagonists on the in vivo release of acetylcholine measured by brain microdialysis. The importance of the post-surgery interval. Naunyn Schmiedebergs Arch. Pharmacol., 345: 144–152.

De Boer, P., Heeringa, M.J. and Abercrombie, E.D. (1996) Spontaneous release of acetylcholine in striatum is preferentially regulated by inhibitory dopamine D2 receptors. Eur. J. Pharmacol., 317: 257–262. de Klippel, N., Sarre, S., Ebinger, G. and Michotte, Y. (1993) Effect of M1- and M2-muscarinic drugs on striatal dopamine release and metabolism: an in vivo microdialysis study comparing normal and 6-hydroxydopamine-lesioned rats. Brain Res., 630: 57–64. del Arco, A. and Mora, F. (2005) Glutamate–dopamine in vivo interaction in the prefrontal cortex modulates the release of dopamine and acetylcholine in the nucleus accumbens of the awake rat. J. Neural Transm., 112: 97–109. Di Chiara, G., Carboni, E., Morelli, M., Cozzolino, A., Tanda, G.L., Pinna, A., Russi, G. and Consolo, S. (1993) Stimulation of dopamine transmission in the dorsal caudate nucleus by pargyline as demonstrated by dopamine and acetylcholine microdialysis and Fos immunohistochemistry. Neuroscience, 55: 451–456. Enrico, P., Bouma, M., de Vries, J.B. and Westerink, B.H. (1998) The role of afferents to the ventral tegmental area in the handling stress-induced increase in the release of dopamine in the medial prefrontal cortex: a dual-probe microdialysis study in the rat brain. Brain Res., 779: 205–213. Ferrari, R., Le Novere, N., Picciotto, M.R., Changeux, J.P. and Zoli, M. (2002) Acute and long-term changes in the mesolimbic dopamine pathway after systemic or local single nicotine injections. Eur. J. Neurosci., 15: 1810–1818. Fu, Y., Matta, S.G., Gao, W., Brower, V.G. and Sharp, B.M. (2000a) Systemic nicotine stimulates dopamine release in nucleus accumbens: re-evaluation of the role of N-methyl-Daspartate receptors in the ventral tegmental area. J. Pharmacol. Exp. Ther., 294: 458–465. Fu, Y., Matta, S.G., Gao, W. and Sharp, B.M. (2000b) Local [alpha]-bungarotoxin-sensitive nicotinic receptors in the nucleus accumbens modulate nicotine-stimulated dopamine secretion in vivo. Neuroscience, 101: 369–375. Fujishiro, H., Umegaki, H., Suzuki, Y., Oohara-Kurotani, S., Yamaguchi, Y. and Iguchi, A. (2005) Dopamine D2 receptor plays a role in memory function: implications of dopamine–acetylcholine interaction in the ventral hippocampus. Psychopharmacology (Berl.), 182: 253–261. Gaddnas, H., Piepponen, T.P. and Ahtee, L. (2002) Mecamylamine decreases accumbal dopamine output in mice treated chronically with nicotine. Neurosci. Lett., 330: 219–222. Gerber, D.J., Sotnikova, T.D., Gainetdinov, R.R., Huang, S.Y., Caron, M.G. and Tonegawa, S. (2001) Hyperactivity, elevated dopaminergic transmission, and response to amphetamine in M1 muscarinic acetylcholine receptor-deﬁcient mice. Proc. Natl. Acad. Sci. U.S.A., 98: 15312–15317. Giacobini, E., Zhu, X.-D., Williams, E. and Sherman, K.A. (1996) The effect of the selective reversible acetylcholinesterase inhibitor E2020 on extracellular acetylcholine and biogenic amine levels in rat cortex. Neuropharmacology, 35: 205–211. Gronier, B., Perry, K.W. and Rasmussen, K. (2000) Activation of the mesocorticolimbic dopaminergic system by stimulation of muscarinic cholinergic receptors in the ventral tegmental area. Psychopharmacology (Berl.), 147: 347–355.

195 Guix, T., Hurd, Y.L. and Ungerstedt, U. (1992) Amphetamine enhances extracellular concentrations of dopamine and acetylcholine in dorsolateral striatum and nucleus accumbens of freely moving rats. Neurosci. Lett., 138: 137–140. Hatip-Al-Khatib, I., Iwasaki, K., Yoshimitsu, Y., Arai, T., Egashira, N., Mishima, K., Ikeda, T. and Fujiwara, M. (2005) Effect of oral administration of zanapezil (TAK-147) for 21 days on acetylcholine and monoamines levels in the ventral hippocampus of freely moving rats. Br. J. Pharmacol., 145: 1035–1044. He, M. and Shippenberg, T.S. (2000) Strain differences in basal and cocaine-evoked dopamine dynamics in mouse striatum. J. Pharmacol. Exp. Ther., 293: 121–127. Hernandez, L.F., Segovia, G. and Mora, F. (2006) Changes in dopamine and acetylcholine in striatum of the awake rat after chronic treatment with a dopamine uptake blocker. Brain Res. Bull., 68: 448–452. Herrera-Marschitz, M., Goiny, M., Utsumi, H., Ferre, S., Hakansson, L., Nordberg, A. and Ungerstedt, U. (1990) Effect of unilateral nucleus basalis lesion on cortical and striatal acetylcholine and dopamine release monitored in vivo with microdialysis. Neurosci. Lett., 110: 172–179. Herrera-Marschitz, M., Luthman, J. and Ferre, S. (1994) Unilateral neonatal intracerebroventricular 6-hydroxydopamine administration in rats: II. Effects on extracellular monoamine, acetylcholine and adenosine levels monitored with in vivo microdialysis. Psychopharmacology (Berl.), 116: 451–456. Hildebrand, B.E., Nomikos, G.G., Hertel, P., Schilstrom, B. and Svensson, T.H. (1998) Reduced dopamine output in the nucleus accumbens but not in the medial prefrontal cortex in rats displaying a mecamylamine-precipitated nicotine withdrawal syndrome. Brain Res., 779: 214–225. Hildebrand, B.E., Panagis, G., Svensson, T.H. and Nomikos, G.G. (1999) Behavioral and biochemical manifestations of mecamylamine-precipitated nicotine withdrawal in the rat: role of nicotinic receptors in the ventral tegmental area. Neuropsychopharmacology, 21: 560–574. Hurd, Y.L., Weiss, F., Koob, G. and Ungerstedt, U. (1990) The inﬂuence of cocaine self-administration on in vivo dopamine and acetylcholine neurotransmission in rat caudate-putamen. Neurosci. Lett., 109: 227–233. Ichikawa, J., Dai, J., O’Laughlin, I.A., Fowler, W.L. and Meltzer, H.Y. (2002) Atypical, but not typical, antipsychotic drugs increase cortical acetylcholine release without an effect in the nucleus accumbens or striatum. Neuropsychopharmacology, 26: 325–339. Ikarashi, Y., Takahashi, A., Ishimaru, H., Arai, T. and Maruyama, Y. (1997a) Regulation of dopamine D1 and D2 receptors on striatal acetylcholine release in rats. Brain Res. Bull., 43: 107–115. Ikarashi, Y., Takahashi, A., Ishimaru, H., Arai, T. and Maruyama, Y. (1997b) Suppression of cholinergic activity via the dopamine D2 receptor in the rat striatum. Neurochem. Int., 30: 191–197. Imperato, A., Obinu, M.C., Carta, G., Mascia, M.S., Casu, M.A., Dazzi, L. and Gessa, G.L. (1994a) Neuroleptics cause stimulation of dopamine D1 receptors and their desensitization after chronic treatment. Eur. J. Pharmacol., 264: 55–60.

Imperato, A., Obinu, M.C., Dazzi, L. and Gessa, G.L. (1994b) Does dopamine exert a tonic inhibitory control on the release of striatal acetylcholine in vivo? Eur. J. Pharmacol., 251: 271–279. Imperato, A., Obinu, M.C. and Gessa, G.L. (1993) Effects of cocaine and amphetamine on acetylcholine release in the hippocampus and caudate nucleus. Eur. J. Pharmacol., 238: 377–381. Imperato, A., Obinu, M.C. and Gessa, G.L. (1995) Is in vivo striatal acetylcholine output under a tonic inhibitory control by dopamine? J. Neural. Transm. Suppl., 45: 91–102. Imperato, A., Obinu, M.C., Mascia, M.S., Casu, M.A., Zocchi, A., Cabib, S. and Puglisi-Allegra, S. (1996) Strain-dependent effects of dopamine agonists on acetylcholine release in the hippocampus: an in vivo study in mice. Neuroscience, 70: 653–660. Imperato, A., Puglisi-Allegra, S., Casolini, P. and Angelucci, L. (1991) Changes in brain dopamine and acetylcholine release during and following stress are independent of the pituitaryadrenocortical axis. Brain Res., 538: 111–117. Iyaniwura, T., Wright, A.E. and Balfour, D. (2001) Evidence that mesoaccumbens dopamine and locomotor responses to nicotine in the rat are inﬂuenced by pretreatment dose and strain. Psychopharmacology (Berl.), 158: 73–79. Izaki, Y., Hori, K. and Nomura, M. (1998) Dopamine and acetylcholine elevation on lever-press acquisition in rat prefrontal cortex. Neurosci. Lett., 258: 33–36. Janhunen, S. and Ahtee, L. (2004) Comparison of the effects of nicotine and epibatidine on the striatal extracellular dopamine. Eur. J. Pharmacol., 494: 167–177. Janson, A.M., Meana, J.J., Goiny, M. and Herrera-Marschitz, M. (1991) Chronic nicotine treatment counteracts the decrease in extracellular neostriatal dopamine induced by a unilateral transection at the mesodiencephalic junction in rats: a microdialysis study. Neurosci. Lett., 134: 88–92. Johnson, B.J. and Bruno, J.P. (1995) Dopaminergic modulation of striatal acetylcholine release in rats depleted of dopamine as neonates. Neuropharmacology, 34: 191–203. Johnson, D.E., Nedza, F.M., Spracklin, D.K., Ward, K.M., Schmidt, A.W., Iredale, P.A., Godek, D.M. and Rollema, H. (2005) The role of muscarinic receptor antagonism in antipsychotic-induced hippocampal acetylcholine release. Eur. J. Pharmacol., 506: 209–219. Keys, A.S. and Mark, G.P. (1998) D1 and D2 dopamine receptor mediation of amphetamine-induced acetylcholine release in nucleus accumbens. Neuroscience, 86: 521–531. Lacroix, L.P., Hows, M.E., Shah, A.J., Hagan, J.J. and Heidbreder, C.A. (2003) Selective antagonism at dopamine D3 receptors enhances monoaminergic and cholinergic neurotransmission in the rat anterior cingulate cortex. Neuropsychopharmacology, 28: 839–849. Laplante, F., Sibley, D.R. and Quirion, R. (2004) Reduction in acetylcholine release in the hippocampus of dopamine D5 receptor-deﬁcient mice. Neuropsychopharmacology, 29: 1620–1627. Lecca, D., Shim, I., Costa, E. and Javaid, I. (2000) Striatal application of nicotine, but not of lobeline, attenuates dopamine release in freely moving rats. Neuropharmacology, 39: 88–98.

196 Li, Z., Huang, M., Ichikawa, J., Dai, J. and Meltzer, H.Y. (2005) N-desmethylclozapine, a major metabolite of clozapine, increases cortical acetylcholine and dopamine release in vivo via stimulation of M1 muscarinic receptors. Neuropsychopharmacology, 30: 1986–1995. Mandel, R.J., Leanza, G., Nilsson, O.G. and Rosengren, E. (1994) Amphetamine induces excess release of striatal acetylcholine in vivo that is independent of nigrostriatal dopamine. Brain Res., 653: 57–65. Mark, G.P., Hajnal, A., Kinney, A.E. and Keys, A.S. (1999) Self-administration of cocaine increases the release of acetylcholine to a greater extent than response-independent cocaine in the nucleus accumbens of rats. Psychopharmacology (Berl.), 143: 47–53. Mark, G.P., Rada, P., Pothos, E. and Hoebel, B.G. (1992) Effects of feeding and drinking on acetylcholine release in the nucleus accumbens, striatum, and hippocampus of freely behaving rats. J. Neurochem., 58: 2269–2274. Mark, G.P., Weinberg, J.B., Rada, P.V. and Hoebel, B.G. (1995) Extracellular acetylcholine is increased in the nucleus accumbens following the presentation of an aversively conditioned taste stimulus. Brain Res., 688: 184–188. Marshall, D.L., Redfern, P.H. and Wonnacott, S. (1997) Presynaptic nicotinic modulation of dopamine release in the three ascending pathways studied by in vivo microdialysis: comparison of naive and chronic nicotine-treated rats. J. Neurochem., 68: 1511–1519. Marubio, L.M., Gardier, A.M., Durier, S., David, D., Klink, R., Arroyo-Jimenez, M.M., McIntosh, J.M., Rossi, F., Champtiaux, N., Zoli, M. and Changeux, J.P. (2003) Effects of nicotine in the dopaminergic system of mice lacking the alpha4 subunit of neuronal nicotinic acetylcholine receptors. Eur. J. Neurosci., 17: 1329–1337. Mifsud, J.C., Hernandez, L. and Hoebel, B.G. (1989) Nicotine infused into the nucleus accumbens increases synaptic dopamine as measured by in vivo microdialysis. Brain Res., 478: 365–367. Millan, M.J., Di Cara, B., Hill, M., Jackson, M., Joyce, J.N., Brotchie, J., McGuire, S., Crossman, A., Smith, L., Jenner, P., Gobert, A., Peglion, J.L. and Brocco, M. (2004) S32504, a novel naphtoxazine agonist at dopamine D3/D2 receptors: II. Actions in rodent, primate, and cellular models of antiparkinsonian activity in comparison to ropinirole. J. Pharmacol. Exp. Ther., 309: 921–935. Mirza, N.R., Pei, Q., Stolerman, I.P. and Zetterstrom, T.S. (1996) The nicotinic receptor agonists (-)-nicotine and isoarecolone differ in their effects on dopamine release in the nucleus accumbens. Eur. J. Pharmacol., 295: 207–210. Moor, E., Schirm, E., Jacso, J. and Westerink, B.H. (1998) Effects of neostigmine and atropine on basal and handlinginduced acetylcholine output from ventral hippocampus. Neuroscience, 82: 819–825. Moore, H., Fadel, J., Sarter, M. and Bruno, J.P. (1999) Role of accumbens and cortical dopamine receptors in the regulation of cortical acetylcholine release. Neuroscience, 88: 811–822. Mori, F., Cuadra, G. and Giacobini, E. (1995) Metrifonate effects on acetylcholine and biogenic amines in rat cortex. Neurochem. Res., 20: 1081–1088.

Moss, S., Sharott, A., Goodhead, L.H. and Mitchell, S.N. (2003) Role of muscarinic receptors in the activation of the ventral subiculum and the consequences for dopamine release in the nucleus accumbens. Eur. J. Pharmacol., 460: 117–125. Nail-Boucherie, K., Dourmap, N., Jaffard, R. and Costentin, J. (1998) The speciﬁc dopamine uptake inhibitor GBR 12783 improves learning of inhibitory avoidance and increases hippocampal acetylcholine release. Cogn. Brain Res., 7: 203–205. Nair, S. and Gudelsky, G. (2006) 3,4-Methylenedioxymethamphetamine enhances the release of acetylcholine in the prefrontal cortex and dorsal hippocampus of the rat. Psychopharmacology (Berl.), 184: 182–189. Nakamura, S., Goshima, Y., Yue, J.L., Miyamae, T. and Misu, Y. (1992) Transmitter-like 3,4-dihydroxyphenylalanine is tonically released by nicotine in striata of conscious rats. Eur. J. Pharmacol., 222: 75–80. Neigh, G.N., Arnold, H.M., Rabenstein, R.L., Sarter, M. and Bruno, J.P. (2004) Neuronal activity in the nucleus accumbens is necessary for performance-related increases in cortical acetylcholine release. Neuroscience, 123: 635–645. Nelson, C.L., Sarter, M. and Bruno, J.P. (2000) Repeated pretreatment with amphetamine sensitizes increases in cortical acetylcholine release. Psychopharmacology (Berl.), 151: 406–415. Nilsson, O.G., Leanza, G. and Bjorklund, A. (1992) Acetylcholine release in the hippocampus: regulation by monoaminergic afferents as assessed by in vivo microdialysis. Brain Res., 584: 132–140. Nisell, M., Nomikos, G.G., Hertel, P., Panagis, G. and Svensson, T.H. (1996) Condition-independent sensitization of locomotor stimulation and mesocortical dopamine release following chronic nicotine treatment in the rat. Synapse, 22: 369–381. Nisell, M., Nomikos, G.G. and Svensson, T.H. (1994a) Systemic nicotine-induced dopamine release in the rat nucleus accumbens is regulated by nicotinic receptors in the ventral tegmental area. Synapse, 16: 36–44. Nisell, M., Nomikos, G.G. and Svensson, T.H. (1994b) Infusion of nicotine in the ventral tegmental area or the nucleus accumbens of the rat differentially affects accumbal dopamine release. Pharmacol. Toxicol., 75: 348–352. Nomikos, G.G., Hildebrand, B.E., Panagis, G. and Svensson, T.H. (1999) Nicotine withdrawal in the rat: role of alpha7 nicotinic receptors in the ventral tegmental area. Neuroreport, 10: 697–702. O’Connor, W.T., Drew, K.L. and Ungerstedt, U. (1995) Differential cholinergic regulation of dopamine release in the dorsal and ventral neostriatum of the rat: an in vivo microdialysis study. J. Neurosci., 15: 8353–8361. Osborne, P.G., O’Connor, W.T., Beck, O. and Ungerstedt, U. (1994) Acute versus chronic haloperidol: relationship between tolerance to catalepsy and striatal and accumbens dopamine, GABA and acetylcholine release. Brain Res., 634: 20–30. Panagis, G., Hildebrand, B.E., Svensson, T.H. and Nomikos, G.G. (2000) Selective c-fos induction and decreased dopamine release in the central nucleus of amygdala in rats displaying a mecamylamine-precipitated nicotine withdrawal syndrome. Synapse, 35: 15–25.

197 Parsons, L.H., Smith, A.D. and Justice, J.B. Jr. (1991a) Basal extracellular dopamine is decreased in the rat nucleus accumbens during abstinence from chronic cocaine. Synapse, 9: 60–65. Parsons, L.H., Smith, A.D. and Justice, J.B. Jr. (1991b) The in vivo microdialysis recovery of dopamine is altered independently of basal level by 6-hydroxydopamine lesions to the nucleus accumbens. J. Neurosci. Methods, 40: 139–147. Perry, K.W., Nisenbaum, L.K., George, C.A., Shannon, H.E., Felder, C.C. and Bymaster, F.P. (2001) The muscarinic agonist xanomeline increases monoamine release and immediate early gene expression in the rat prefrontal cortex. Biol. Psychiatry, 49: 716–725. Picciotto, M.R., Zoli, M., Rimondini, R., Lena, C., Marubio, L.M., Pich, E.M., Fuxe, K. and Changeux, J.P. (1998) Acetylcholine receptors containing the [beta]2 subunit are involved in the reinforcing properties of nicotine. Nature, 391: 173–177. Pidoplichko, V.I., Noguchi, J., Areola, O.O., Liang, Y., Peterson, J., Zhang, T. and Dani, J.A. (2004) Nicotinic cholinergic synaptic mechanisms in the ventral tegmental area contribute to nicotine addiction. Learn. Mem., 11: 60–69. Pisanu, A., Acquas, E., Fenu, S. and Di Chiara, G. (2006) Modulation of [Delta]9-THC-induced increase of cortical and hippocampal acetylcholine release by [mu] opioid and D1 dopamine receptors. Neuropharmacology, 50: 661–670. Pontieri, F.E., Tanda, G., Orzi, F. and Di Chiara, G. (1996) Effects of nicotine on the nucleus accumbens and similarity to those of addictive drugs. Nature, 382: 255–257. Rada, P. and Hoebel, B.G. (2005) Acetylcholine in the accumbens is decreased by diazepam and increased by benzodiazepine withdrawal: a possible mechanism for dependency. Eur. J. Pharmacol., 508: 131–138. Rada, P., Jensen, K. and Hoebel, B. (2001a) Effects of nicotine and mecamylamine-induced withdrawal on extracellular dopamine and acetylcholine in the rat nucleus accumbens. Psychopharmacology (Berl.), 157: 105–110. Rada, P., Johnson, D.F., Lewis, M.J. and Hoebel, B.G. (2004) In alcohol-treated rats, naloxone decreases extracellular dopamine and increases acetylcholine in the nucleus accumbens: evidence of opioid withdrawal. Pharmacol. Biochem. Behav., 79: 599–605. Rahman, S., Zhang, J. and Corrigall, W.A. (2003) Effects of acute and chronic nicotine on somatodendritic dopamine release of the rat ventral tegmental area: in vivo microdialysis study. Neurosci. Lett., 348: 61–64. Rahman, S., Zhang, J. and Corrigall, W.A. (2004a) Local perfusion of nicotine differentially modulates somatodendritic dopamine release in the rat ventral tegmental area after nicotine preexposure. Neurochem. Res., 29: 1687–1693. Rahman, S., Zhang, J., Engleman, E.A. and Corrigall, W.A. (2004b) Neuroadaptive changes in the mesoaccumbens dopamine system after chronic nicotine self-administration: a microdialysis study. Neuroscience, 129: 415–424. Rao, T.S., Reid, R.T., Correa, L.D., Santori, E.M., Gardner, M.F., Sacaan, A.I., Lorrain, D. and Vernier, J.M. (2003) In vivo pharmacological characterization of (+/ )-4-[2-(1methyl-2-pyrrolidinyl)ethyl]thiophenol hydrochloride (SIB1553A), a novel cholinergic ligand: microdialysis studies. Brain Res., 986: 71–81.

Reid, M., Ho, L. and Berger, S. (1996) Effects of environmental conditioning on the development of nicotine sensitization: behavioral and neurochemical analysis. Psychopharmacology (Berl.), 126: 301–310. Reid, M.S., Fox, L., Ho, L.B. and Berger, S.P. (2000) Nicotine stimulation of extracellular glutamate levels in the nucleus accumbens: neuropharmacological characterization. Synapse, 35: 129–136. Reid, M.S., Ho, L.B. and Berger, S.P. (1998) Behavioral and neurochemical components of nicotine sensitization following 15-day pretreatment: studies on contextual conditioning. Behav. Pharmacol., 9: 137–148. Robertson, G.S., Hubert, G.W., Tham, C.S. and Fibiger, H.C. (1992) Lesions of the mesotelencephalic dopamine system enhance the effects of selective dopamine D1 and D2 receptor agonists on striatal acetylcholine release. Eur. J. Pharmacol., 219: 323–325. Robertson, G.S., Tham, C.S., Wilson, C., Jakubovic, A. and Fibiger, H.C. (1993) In vivo comparisons of the effects of quinpirole and the putative presynaptic dopaminergic agonists B-HT 920 and SND 919 on striatal dopamine and acetylcholine release. J. Pharmacol. Exp. Ther., 264: 1344–1351. Russi, G., Girotti, P., Cadoni, C., Di Chiara, G. and Consolo, S. (1993) Neuroleptics increase striatal acetylcholine release by a sequential D-1 and D-2 receptor mechanism. Neuroreport, 4: 1335–1338. Sacaan, A.I., Menzaghi, F., Dunlop, J.L., Correa, L.D., Whelan, K.T. and Lloyd, G.K. (1996) Epibatidine: a nicotinic acetylcholine receptor agonist releases monoaminergic neurotransmitters: in vitro and in vivo evidence in rats. J. Pharmacol. Exp. Ther., 276: 509–515. Sato, K., Ueda, H., Okumura, F. and Misu, Y. (1994) 6OHDA-induced lesion of the nigrostriatal dopaminergic neurons potentiates the inhibitory effect of 7-OHDPAT, a selective D3 agonist, on acetylcholine release during striatal microdialysis in conscious rats. Brain Res., 655: 233–236. Sato, K., Wu, J., Kikuchi, T., Wang, Y., Watanabe, I. and Okumura, F. (1996) Differential effects of ketamine and pentobarbitone on acetylcholine release from the rat hippocampus and striatum. Br. J. Anaesth., 77: 381–384. Schilstrom, B., Fagerquist, M.V., Zhang, X., Hertel, P., Panagis, G., Nomikos, G.G. and Svensson, T.H. (2000) Putative role of presynaptic alpha7* nicotinic receptors in nicotine stimulated increases of extracellular levels of glutamate and aspartate in the ventral tegmental area. Synapse, 38: 375–383. Schilstrom, B., Nomikos, G.G., Nisell, M., Hertel, P. and Svensson, T.H. (1997) N-methyl-D-aspartate receptor antagonism in the ventral tegmental area diminishes the systemic nicotine-induced dopamine release in the nucleus accumbens. Neuroscience, 82: 781–789. Schilstrom, B., Rawal, N., Mameli-Engvall, M., Nomikos, G.G. and Svensson, T.H. (2003) Dual effects of nicotine on dopamine neurons mediated by different nicotinic receptor subtypes. Int. J. Neuropsychopharmacol., 6: 1–11.

198 Schilstrom, B., Svensson, H.M., Svensson, T.H. and Nomikos, G.G. (1998) Nicotine and food induced dopamine release in the nucleus accumbens of the rat: putative role of [alpha]7 nicotinic receptors in the ventral tegmental area. Neuroscience, 85: 1005–1009. Seppa, T. and Ahtee, L. (2000) Comparison of the effects of epibatidine and nicotine on the output of dopamine in the dorsal and ventral striatum of freely-moving rats. Naunyn Schmiedebergs Arch. Pharmacol., 362: 444–447. Shearman, E., Rossi, S., Szasz, B., Juranyi, Z., Fallon, S., Pomara, N., Sershen, H. and Lajtha, A. (2006) Changes in cerebral neurotransmitters and metabolites induced by acute donepezil and memantine administrations: a microdialysis study. Brain Res. Bull., 69: 204–213. Shippenberg, T.S., Hen, R. and He, M. (2000) Region-speciﬁc enhancement of basal extracellular and cocaine-evoked dopamine levels following constitutive deletion of the serotonin1B receptor. J. Neurochem., 75: 258–265. Shirazi-Southall, S., Rodriguez, D.E. and Nomikos, G.G. (2002) Effects of typical and atypical antipsychotics and receptor selective compounds on acetylcholine efﬂux in the hippocampus of the rat. Neuropsychopharmacology, 26: 583–594. Shoaib, M. and Shippenberg, T.S. (1996) Adrenalectomy attenuates nicotine-induced dopamine release and locomotor activity in rats. Psychopharmacology (Berl.), 128: 343–350. Smith, A.D. and Weiss, F. (1999) Ethanol exposure differentially alters central monoamine neurotransmission in alcoholpreferring versus -nonpreferring rats. J. Pharmacol. Exp. Ther., 288: 1223–1228. Smolders, I., Bogaert, L., Ebinger, G. and Michotte, Y. (1997) Muscarinic modulation of striatal dopamine, glutamate, and GABA release, as measured with in vivo microdialysis. J. Neurochem., 68: 1942–1948. Steele, T.D., Hodges, J.R., Levesque, T.R. and Locke, K.W. (1997) D1 agonist dihydrexidine releases acetylcholine and improves cognitive performance in rats. Pharmacol. Biochem. Behav., 58: 477–483. Steinberg, R., Rodier, D., Souiclhac, J., Bougault, I., EmondsAlt, X., Soubrie, P. and Le Fur, G. (1995) Pharmacological characterization of tachykinin receptors controlling acetylcholine release from rat striatum: an in vivo microdialysis study. J. Neurochem., 65: 2543–2548. Steinberg, R., Souilhac, J., Rodier, D., Alonso, R., EmondsAlt, X., Le Fur, G. and Soubrie, P. (1998) Facilitation of striatal acetylcholine release by dopamine D1 receptor stimulation: involvement of enhanced nitric oxide production via neurokinin-2 receptor activation. Neuroscience, 84: 511–518. Summers, K.L., Lippiello, P. and Giacobini, E. (1996) A microdialysis study of the effects of the nicotinic agonist RJR2403 on cortical release of acetylcholine and biogenic amines. Neurochem. Res., 21: 1181–1186. Sziraki, I., Sershen, H., Benuck, M., Hashim, A. and Lajtha, A. (1998) Receptor systems participating in nicotine-speciﬁc effects. Neurochem. Int., 33: 445–457.

Taguchi, K., Atobe, J., Kato, M., Chuma, T., Chikuma, T., Shigenaga, T. and Miyatake, T. (1998) The effect of methamphetamine on the release of acetylcholine in the rat striatum. Eur. J. Pharmacol., 360: 131–137. Taguchi, K., Hagiwara, Y., Suzuki, Y. and Kubo, T. (1993) Effects of morphine on release of acetylcholine in the rat striatum: an in vivo microdialysis study. Naunyn Schmiedebergs Arch. Pharmacol., 347: 9–13. Takahashi, A., Ishimaru, H., Ikarashi, Y. and Maruyama, Y. (1993) Intraventricular injection of neostigmine increases dopaminergic and noradrenergic nerve activities: hyperglycemic effects and neurotransmitters in the hypothalamus. Neurosci. Lett., 156: 54–56. Takahashi, H., Takada, Y., Nagai, N., Urano, T. and Takada, A. (1998) Effects of nicotine and footshock stress on dopamine release in the striatum and nucleus accumbens. Brain Res. Bull., 45: 157–162. Toth, E., Sershen, H., Hashim, A., Vizi, E.S. and Lajtha, A. (1992) Effect of nicotine on extracellular levels of neurotransmitters assessed by microdialysis in various brain regions: role of glutamic acid. Neurochem. Res., 17: 265–271. Tzavara, E.T., Bymaster, F.P., Davis, R.J., Wade, M.R., Perry, K.W., Wess, J., McKinzie, D.L., Felder, C. and Nomikos, G.G. (2004) M4 muscarinic receptors regulate the dynamics of cholinergic and dopaminergic neurotransmission: relevance to the pathophysiology and treatment of related central nervous system pathologies. FASEB J., 18: 1410–1412. Tzavara, E.T., Bymaster, F.P., Overshiner, C.D., Davis, R.J., Perry, K.W., Wolff, M., McKinzie, D.L., Witkin, J.M. and Nomikos, G.G. (2005) Procholinergic and memory enhancing properties of the selective norepinephrine uptake inhibitor atomoxetine. Mol. Psychiatry,, 11: 187–195. Tzavara, E.T., Davis, R.J., Perry, K.W., Li, X., Salhoff, C., Bymaster, F.P., Witkin, J.M. and Nomikos, G.G. (2003a) The CB1 receptor antagonist SR141716A selectively increases monoaminergic neurotransmission in the medial prefrontal cortex: implications for therapeutic actions. Br. J. Pharmacol., 138: 544–553. Tzavara, E.T., Wade, M. and Nomikos, G.G. (2003b) Biphasic effects of cannabinoids on acetylcholine release in the hippocampus: site and mechanism of action. J. Neurosci., 23: 9374–9384. Umegaki, H., Munoz, J., Meyer, R.C., Spangler, E.L., Yoshimura, J., Ikari, H., Iguchi, A. and Ingram, D.K. (2001) Involvement of dopamine D2 receptors in complex maze learning and acetylcholine release in ventral hippocampus of rats. Neuroscience, 103: 27–33. Vinson, P.N. and Justice, J. (1997) Effect of neostigmine on concentration and extraction fraction of acetylcholine using quantitative microdialysis. J. Neurosci. Methods, 73: 61–67. Visanji, N.P., Mitchell, S.N., O’Neill, M.J. and Duty, S. (2006) Chronic pre-treatment with nicotine enhances nicotineevoked striatal dopamine release and [alpha]6 and [beta]3 nicotinic acetylcholine receptor subunit mRNA in the substantia nigra pars compacta of the rat. Neuropharmacology, 50: 36–46.

199 Wade, M. and Nomikos, G. (2005) Tolerance to the procholinergic action of the D1 receptor full agonist dihydrexidine. Psychopharmacology (Berl.), 182: 393–399. Warpman, U., Zhang, X. and Nordberg, A. (1996) Effect of tacrine on in vivo release of dopamine and its metabolites in the striatum of freely moving rats. J. Pharmacol. Exp. Ther., 277: 917–922. Westerink, B.H., de Boer, P. and Damsma, G. (1990) Dopamine–acetylcholine interaction in the striatum studied by microdialysis in the awake rat: some methodological aspects. J. Neurosci. Methods, 34: 117–124. Westerink, B.H., Hofsteede, H.M., Damsma, G. and de Vries, J.B. (1988) The significance of extracellular calcium for the release of dopamine, acetylcholine and amino acids in conscious rats, evaluated by brain microdialysis. Naunyn Schmiedebergs Arch. Pharmacol., 337: 373–378. Whitehead, K.J., Rose, S. and Jenner, P. (2001) Involvement of intrinsic cholinergic and GABAergic innervation in the effect of NMDA on striatal dopamine efﬂux and metabolism as assessed by microdialysis studies in freely moving rats. Eur. J. Neurosci., 14: 851–860. Wonnacott, S., Sidhpura, N. and Balfour, D.J. (2005) Nicotine: from molecular mechanisms to behaviour. Curr. Opin. Pharmacol., 5: 53–59. Yamamuro, Y., Hori, K., Tanaka, J., Iwano, H. and Nomura, M. (1995) Septo-hippocampal cholinergic system under the discrimination learning task in the rat: a microdialysis study with the dual-probe approach. Brain Res., 684: 1–7. Yoshida, M., Yokoo, H., Tanaka, T., Mizoguchi, K., Emoto, H., Ishii, H. and Tanaka, M. (1993) Facilitatory modulation of mesolimbic dopamine neuronal activity by a [mu]-opioid agonist and nicotine as examined with in vivo microdialysis. Brain Res., 624: 277–280.

Zackheim, J. and Abercrombie, E.D. (2005) Thalamic regulation of striatal acetylcholine efﬂux is both direct and indirect and qualitatively altered in the dopamine-depleted striatum. Neuroscience, 131: 423–436. Zhu, X.D., Cuadra, G., Brufani, M., Maggi, T., Pagella, P.G., Williams, E. and Giacobini, E. (1996) Effects of MF-268, a new cholinesterase inhibitor, on acetylcholine and biogenic amines in rat cortex. J. Neurosci. Res., 43: 120–126. Zhu, X.D., Giacobini, E. and Hornsperger, J.M. (1995) Effect of MDL 73,745 on acetylcholine and biogenic amine levels in rat cortex. Eur. J. Pharmacol., 276: 93–99. Zis, A.P., Nomikos, G.G., Brown, E.E., Damsma, G. and Fibiger, H.C. (1992) Neurochemical effects of electrically and chemically induced seizures: an in vivo microdialysis study in the rat hippocampus. Neuropsychopharmacology, 7: 189–195. Zis, A.P., Nomikos, G.G., Damsma, G. and Fibiger, H.C. (1991) In vivo neurochemical effects of electroconvulsive shock studied by microdialysis in the rat striatum. Psychopharmacology (Berl.), 103: 343–350. Zmarowski, A., Sarter, M. and Bruno, J.P. (2005) NMDA and dopamine interactions in the nucleus accumbens modulate cortical acetylcholine release. Eur. J. Neurosci., 22: 1731–1740. Zocchi, A. and Pert, A. (1993) Increases in striatal acetylcholine by SKF-38393 are mediated through D1 dopamine receptors in striatum and not the frontal cortex. Brain Res., 627: 186–192. Zocchi, A. and Pert, A. (1994) Alterations in striatal acetylcholine overﬂow by cocaine, morphine, and MK-801: relationship to locomotor output. Psychopharmacology (Berl.), 115: 297–304.

CHAPTER 2.5

Microdialysis as a platform for multidisciplinary strategies J. Urenjak and T.P. Obrenovitch Pharmacy, School of Life Sciences, University of Bradford, Bradford, UK

Abstract: Microdialysis is a versatile in vivo sampling technique, with the unique capability to provide information on the biochemistry of the cellular microenvironment within a speciﬁc region of the brain or other organs. Microdialysis may also be used to deliver biochemicals, drugs and toxins directly to the site under study. As such, microdialysis by itself allows the combination of biochemistry, pharmacology, and experimental pathology in a single in vivo preparation. This basic multidisciplinary potential can be supplemented by a number of monitoring techniques, whether electrode based or optic ﬁber based, to collect information on other relevant biochemical or physiological variables. Finally, microdialysis can be associated with imaging methods such as magnetic resonance imaging (MRI). In this chapter, we make clear and illustrate with selected applications the unique capability of microdialysis to provide a versatile platform for the assembly of a wide range of multidisciplinary strategies that can be tailored to speciﬁc study objectives. In many cases, the interpretation of the data may be complicated by the fact that the different methods may not involve precisely the same tissue area or compartment. However, this potential difﬁculty can be dealt with, and once resolved it can lead to a better understanding of the biological processes under study. disciplines can be brought in, either applied locally (but independently from microdialysis) or more globally, to achieve a truly multidisciplinary and powerful experimental strategy. The primary purpose of this chapter is to make clear, and illustrate with selected applications, that microdialysis can be used as the core technique for a variety of in vivo multidisciplinary strategies. The monitoring of several variables from the same tissue site favors a more accurate and reliable interpretation of each separate data set, and reduces the number of animals required to test speciﬁc hypotheses. At the end of the chapter, we also discuss some limitations and pitfalls inherent to microdialysis within the context of multidisciplinary strategies. Notes:

Microdialysis is widely recognized as a versatile in vivo sampling technique, with the unique capability to provide information on the biochemistry of the cellular microenvironment (i.e., extracellular ﬂuid composition) within a speciﬁc region of the brain and other organs. Microdialysis may also be\ used to deliver drugs directly to the site under study, which avoids any possible interference of the drugs with other organs, and in brain studies circumvents the potential impermeability of the blood-brain barrier to drug prototypes. As an extension to reverse microdialysis for drug delivery, by changing the composition of the perfusion medium or by adding speciﬁc agents to the medium, it may be possible to induce a relevant pathological condition, or to reproduce some aspect(s) of such a condition. Finally, a number of methods from various

Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

201

Although some of the microdialysis-based multidisciplinary strategies may be applicable DOI: 10.1016/S1569-7339(06)16011-7 Copyright 2007 Elsevier B.V. All rights reserved

202

to other organs, this chapter is essentially focused on in vivo studies of the central nervous system (CNS). Only selected methodological combinations based on microdialysis are considered and discussed herein. Many other combinations are possible, and it is clear that the optimal one for any given purpose will have to be selected carefully on the basis of the study objectives and practical considerations.

I. Pharmacology, biochemistry, and behavioral studies The combination of any two of these disciplines, or of the three together, is discussed only brieﬂy here because other chapters in this handbook are dedicated to these topics (Sections on pharmacology and animal behavior). This combination has been very useful for investigations into the neurochemical control of feeding (Fig. 1), addiction (Fu et al., 2001), stress/anxiety (Swanson et al., 2004), and sexual/maternal behaviors (Da Costa et al., 1996).

I.A. Pharmacology – the drug delivery issue Although emphasis is placed on drug delivery via the perfusion medium in Fig. 1, this is only one way to bring a pharmacological component into the experimental strategy. Reverse microdialysis may be optimal for some studies, for example, when drugs are anticipated to act presynaptically or on the inactivation of released neurotransmitter

PHARMACOLOGY Microdialysis used for the delivery of drug(s) via the perfusion medium.

(i.e., enzymatic degradation and uptake). It also enables the sequential application of different drugs and concentrations to the site under study, and this can be carried out over an extended period, even in freely moving animals. However, reverse microdialysis may be inappropriate for a number of reasons as follows, whether separate or combined: (i) Behavioral effects may only occur when the drug acts on a brain region that is larger than the site that can be ‘inﬂuenced’ by microdialysis (i.e., up to 1 mm away from the probe axis; Benveniste and Huttemeier, 1990), and in some studies drug intake may be part of the behavioral test (e.g., self-administration of addictive drugs). Neurochemical changes are often to be expected in other brain regions than that challenged pharmacologically. In all these cases, systemic or intracerebroventricular drug administration may be considered. (Fu et al., 2001; Price and Lucki, 2001). In some studies, it may be pertinent to use two separate microdialysis probes, one for drug delivery and the other as a sampling device (Alex et al., 2005; Zackheim and Abercrombie, 2005). (ii) Some drugs may not be suitable for administration via the microdialysis probe because of their size relative to the dialysis membrane cutoff (e.g., neuropeptides), their adsorption to the dialysis ﬁber and other components of the microdialysis system (e.g., high-precision syringes and inlet tubing), their poor solubility in aqueous media or simply their cost because microdialysis

BIOCHEMISTRY Microdialysis sampling; Study of changes in the extracellular fluid composition.

BEHAVIORAL TESTS

Illustrative application – Serotonergic (5-HT) control of feeding: Delivery of a subtype selective 5-HT agonist to specific brain regions; monitoring of the extracellular changes in 5-HT and its metabolism in these regions; and measure of changes in food intake (See legend for reference).

Fig. 1. Diagram of one of the most common microdialysis-based multidisciplinary strategy. Reference for illustrative application Hikiji et al., 2004.

203

delivery implies that only a fraction of the drug in the perfusion medium is transferred to the tissue surrounding the dialysis ﬁber (estimated to be o10% of the perfusion medium content and dependent on the drug; Benveniste and Huttemeier, 1990). In any of these cases, and if systemic and intracerebroventricual administration are not suitable, a microinjection cannula positioned in a region adjacent to the microdialysis ﬁber may be used to microinject the drug precisely at the sampling site (Masuo et al., 1993). With both medium-sized molecules (e.g., endothelin-1, MW 2492; Fig. 4) and small molecules (e.g., glutamate receptor agonist AMPA, and MW 186) we have found that similar effects (assessed by local monitoring at the microdialysis site) can be achieved by microdialysis delivery and such a microinjection cannula. (iii) Inherently, reverse microdialysis does not allow a very sudden and brief rise in the effective drug concentration at tissue level. This potential problem, seldom envisaged, may be a pitfall in studies involving agonists designed to act on receptors that desensitize very rapidly (e.g., nicotinic acetylcholine receptors). A suitable pulse application of drug at the microdialysis site may be achievable by using a microcannula connected to a pressure ejection system (Rogers, 1985). With some compounds (e.g., nitric oxide) an alternative may be offered by ‘caged’ molecules, that is, compounds that are transformed to their biologically active state by illumination with appropriate wavelength and intensity (Godwin et al., 1997). However, in studies that combine pharmacology and biochemistry it is likely that microdialysis sampling would then become the limiting factor, despite the much better time resolution that is now achievable with advanced analytical techniques. Voltammetric methods may be more appropriate in these conditions (Sotty et al., 1998; Parikh et al., 2004). (iv) When biochemicals and drugs are perfused through a microdialysis probe, one often

assumes that, providing they permeate the dialysis membrane, they will reach their molecular target within the area ‘inﬂuenced’ by microdialysis. That is likely to occur with drugs designed to act on receptors distributed on the external side of cellular membranes, but not necessarily true when the expected site of drug action is intracellular. In addition, drug access via the extracellular ﬂuid may not be optimal for some components of the brain cytoarchitecture (e.g., brain arterioles endothelium and smooth muscle, as these blood vessels are tightly surrounded by astrocyte endfeet).

II. Pharmacology and/or induction of a pathological condition, biochemistry and electrophysiology Recording some relevant electrophysiological variable(s) is most useful to verify, independently from the biochemical data, that the pharmacological treatment (e.g., concentration of drug in the perfusion medium) or the induced pathophysiological condition is effective and appropriate at the site studied by microdialysis. Conversely, the electrophysiological variable(s) may be the primary one, and the biochemical data used to verify the effectiveness of the pharmacological treatment (Fig. 2). Reverse microdialysis may be used also as a convenient way to apply different drugs, or several concentrations of a single drug, consecutively at the site studied by electrophysiology. For all such studies, optimal validity of the electrophysiological component implies that the electrode system should be located as close as possible to the dialysis ﬁber, which may be achieved as follows: (i) incorporation of the electrode within the microdialysis ﬁber (Obrenovitch et al., 1993a); (ii) a section of the metallic body of the microdialysis probe, just above the dialysis ﬁber, is used as electrode (Pena and Tapia, 1999); and (iii) the electrode system is adjacent to the external surface of the dialysis ﬁber (Bourne and Fosbraey, 2000). Some investigators successfully combined microdialysis and electrophysiological recordings in behaving animals (Ludvig et al., 1994, 2000; Kittner

204

INDUCTION OF CNS PATHOLOGY This may be achieved via the microdialysis probe, or independently. PHARMACOLOGY Microdialysis used for drug delivery

BIOCHEMISTRY Microdialysis sampling

ELECTROPHYSIOLOGY

Illustrative application – Pharmacology of cortical spreading depression (CSD): Elicitation of CSD at the microdialysis site by perfusion of high-potassium medium; delivery of a phosphodiesterase inhibitor to increase the concentration of cGMP; monitoring of the extracellular changes in cGMP to verify the efficacy of the drug treatment; and recording of CSD as a negative shift of the DC-potential, using an electrode incorporated within the microdialysis fibre (See legend for reference). Fig. 2. Diagram illustrating the possibility of using microdialysis to combine four different disciplines, simultaneously, to the same speciﬁc brain region. Reference for illustrative application Wang et al., 2004.

et al., 2004), but this is only possible with some electrophysiological signals and electrode/microdialysis conﬁgurations, because the ﬁne input line connected to the microdialysis probe enhances any electrical noise by acting as an effective high-impedance antenna (Obrenovitch et al., 1993a).

II.A. Electrophysiological variables Electroencephalography (EEG) is the most common electrophysiological variable recorded concurrently with microdialysis, especially for studies into epilepsies (Obrenovitch et al., 1996; Bourne and Fosbraey, 2000; Tian et al., 2005) and cerebral ischemia (Obrenovitch et al., 1993b). This strategy was also used successfully for the study of sleep, but in this case EEG electrodes were generally implanted remotely from the microdialysis probe (Penalva et al., 2003; Thakkar et al., 2003; Hong et al., 2005; Crouzier et al., 2006). Note that in a number of studies of epilepsies (Richards et al., 2000; Slezia et al., 2004) and brain ischemia (Martinez-Tica and Zornow, 2000) the EEG was also recorded remotely from the microdialysis probe, which may be appropriate depending on the study objectives and the pathophysiology investigated. The extracellular direct current (DC) potential is another pertinent electrophysiological variable, because this signal can provide relevant information on depolarization shifts occurring in the cell population surrounding the microdialysis ﬁber,

and negative DC potential shifts are key features of several important neurological conditions (Fig. 3). In cerebral ischemia, a sustained negative shift of the DC potential indicates anoxic depolarization (i.e., the loss of cellular ionic homeostasis subsequent to energy failure), which makes it a very useful variable to assess the severity of ischemia precisely at the microdialysis site (Obrenovitch et al., 1993b; Bruhn et al., 2003). The concurrent monitoring of this signal with microdialysis allowed showing that the sudden ischemic efﬂux of neurotransmitters and other compounds occurs with anoxic depolarization (Obrenovitch and Richards, 1995; Takata et al., 2005). A transient, negative DC shift is the hallmark of spreading depression (SD), that is, the propagating wave of cellular depolarization that is the underlying cause of the aura in classical migraine (Lauritzen, 1994), and an important component of the pathophysiology of focal brain injuries, whether ischemic or traumatic (Strong et al., 2002). In our hands, microdialysis electrodes constitute an unrivalled tool for investigations into the pharmacology and biochemistry of SD (Fig. 2). Monitoring the DC potential may be also pertinent in studies of epilepsies, as paroxysmal depolarization appears on the DC potential as brief negative shifts (Fig. 3). Finally, monitoring the DC potential may be a very useful signal for in vivo pharmacological studies of depolarizing drugs (e.g., agonists of ionotropic glutamate receptors; Obrenovitch et al., 1994). In comparison to EEG, single-unit extracellular recording is a more reﬁned variable, which also

205

( -)

2 mV

5-min 4-AP (3 mM)

( -)

2 mV

15 sec

( -)

2 mV

15-min 4-AP (3 mM) Fig. 3. Changes in the DC potential produced by perfusion of the potassium channel blocker 4-aminopyridine (4-AP, 3 mM) through a microdialysis probe (1 mm ﬁber length) implanted into the cortex of halothane-anesthetized rats. In this ﬁgure, the polarity of the DC potential signal was inverted (–). Relevant variables for the evaluation of the tissue susceptibility to 4-APinduced convulsant activity may be (i) latency for occurrence of the ﬁrst paroxysmal depolarization shift; (ii) number of depolarizations induced by the 4-AP application; (iii) cumulative area of the depolarization(s) (see trace in insert; shaded area) (Urenjak, J. and Obrenovitch, T.P., unpublished data).

provides a better spatial match with microdialysisbased pharmacology and/or biochemistry. We have demonstrated that the DC signal recorded with a microdialysis electrode is primarily generated by cells adjacent to the dialysis membrane, but that a much larger region contributes to the EEG recorded with the same electrode (Obrenovitch et al., 1993a). Similarly, the ﬁring of hippocampal neurons was completely abolished by perfusion of 1% lidocaine through the microdialysis probe (Ludvig et al., 1994), whereas the EEG recorded via a microdialysis electrode was only reduced by lidocaine and other manipulations that are known to suppress neuronal activity (Ludvig et al., 1994; Obrenovitch, T.P. and Urenjak, J., unpublished observations). Studying the ﬁring pattern of neurons in a speciﬁc brain region, together with microdialysis-based pharmacology and/or biochemistry in the same or a different region, is a powerful strategy to determine how different brain regions interfere with each other (Page and Abercrombie, 1999; Lee et al., 2004) and how a speciﬁc neuronal activity inﬂuences a

particular behavior (Lodge and Grace, 2005) or brain function (Brazhnik et al., 2004). Evoked potentials also provide a variable that is more reﬁne than the EEG. They are especially suitable for the investigation of afferent innervations between brain regions. In many studies, the recording electrode was in a region different from the microdialysis implantation site (Zhang de et al, 2005), but evoked potentials can be recorded within the area that can be inﬂuenced by reverse microdialysis (Oldford and Castro-Alamancos, 2003; Crochet et al., 2005) or sampled by microdialysis (Bronzino et al., 1999; Jay et al., 1999). Finally, simultaneous microdialysis and intracellular electrophysiological recording was also performed successfully (West and Grace, 2004). II.B. Induction of a pathological condition, or of a relevant experimental change Microdialysis can be used directly to produce or mimic a pathological condition at the site of study, simply by adding a relevant toxin or agent to the perfusion medium. A variety of neurological disorders have been investigated in this way: (i) Seizure activity, with picrotoxin or 4-aminopyridine as convulsant agents (SierraParedes and Sierra-Marcuno, 1996; Tian et al., 2005) (see also Fig. 3); (ii) Ischemia, with the potent vasoconstrictor endothelin-1 (ET-1) (see illustrative application in Fig. 4); (iii) Parkinson’s disease, with the mitochondrial complex I inhibitor, 1-methyl-4-phenylpyridinium ion (MPP+) (Rollema et al., 1990; Smith and Bennett, 1997; Staal and Sonsalla, 2000; Wu et al., 2000); (iv) Huntington’s disease, with the excitotoxin quinolinic acid or the mitochondrial toxin 3-nitropropionic acid (Reynolds et al., 1999; Blum et al., 2003); (v) Alzheimer’s disease, by reverse microdialysis of b-amyloid protein (Harkany et al., 2000). Reverse microdialysis may also be used to reproduce only one speciﬁc aspect of the neuropathology

206

CNS PATHOLOGY Induced via the microdialysis probe PHARMACOLOGY Microdialysis used for drug delivery

ADDITIONAL VARIABLE(S) measured at the microdialysis site, but via independent method(s)

BIOCHEMISTRY Microdialysis sampling

ELECTROPHYSIOLOGY

Illustrative application – Pathophysiology and pharmacology of cerebral ischemia Focal ischemia was produced by perfusion of the potent vasoconstrictor, endothelin-1 (ET-1) through the microdialysis probe implanted in the cerebral cortex of an anesthetized rat. The data below show: Reduction in local cerebral blood flow (local CBF) monitored at ET-1 application site by laser Doppler flowmetry; changes in electrical activity (EEG) and occurrence of anoxic depolarization (AD) detected with microdialysis electrode; histological damage assessed radially from the microdialysis fiber axis.

% 100

Local CBF

Extent of radial neuronal damage

Fiber track

ET-1 100 µM 0 %

EEG amplitude

100

0

10 mV

DC potential (-) AD

-10

0

10

Histological damage produced by microdialysis perfusion of 100 µM ET-1 for 60 min, assessed by eosinhematoxylin staining. The extent of lesion was 582 ± 33 µm (mean ± SEM, n = 3)

20 min

Fig. 4. Superimposition of an independent method to examine another physiological or biochemical variable, within the region studied with microdialysis. In this illustrative application, the independent method was laser Doppler ﬂowmetry (LDF), used to examine the change in local blood ﬂow produced by perfusion of the potent vasoconstrictor ET-1 through the microdialysis probe (see Fig. 5, diagram B for a diagram of the probe/LDF combination). All the traces presented above were obtained from a single, representative experiment, with a 1-mm ﬁber length microdialysis probe implanted in the cortex of halothane-anesthetized rats. The histology was carried out in separate experiments. Note that biochemical changes (e.g., increased extracellular lactate, neurotransmitter efﬂux) and drugs effects could also be studied in these experiments. This model of ‘micro’ focal ischemia was designed to be used in conjunction with genetic tools that can produce discrete changes in gene expression, such as virus-mediated gene transfer and antisense oligonucleotides (Martin, M. et al., unpublished data).

under study. For example, if we consider brain ischemia, mitochondrial toxins such as malonate (Nixdorf et al., 2001) and rotenone (Santiago et al., 1995) may be used to inhibit energy metabolism

(i.e., chemical ischemia) at the sampling site, whereas the Na+/K+-ATPase blocker ouabain may be used to simulate anoxic depolarization (Fairbrother et al., 1990; Dobolyi et al., 2000).

207

Simple changes of the perfusion medium composition may also be relevant, such as a buffered, acidic perfusion medium to impose extracellular acidosis (Waterfall et al., 1996; Urenjak et al., 1997), or a hypo-osmolar medium to induce cell swelling (Taylor et al., 1995; Scheller, 2000). For some studies, it may be pertinent to use microdialysis for the local delivery of selective neurotoxins, such as the dopaminergic toxin 6-hydroxydopamine (6-OHDA; Lorrain et al., 1998; Ferger et al., 2001) and the serotonergic toxin 3,4-methylenedioxymethamphetamine (MDMA or ecstasy; Nair and Gudelsky, 2006). The latter approach, however, is essentially suitable for investigations into the early effects of the neurotoxin. Another pertinent application of reverse microdialysis, likely to become increasingly popular as our understanding of the genome improves rapidly, is the possibility to apply antisense oligonucleotides (ODNs), selectively to the site under study. Antisense ODNs are short chains of 20 nucleotides that target a speciﬁc mRNA sequence of complementary bases, to ultimately block the synthesis of a speciﬁc protein. This application of microdialysis is well described and discussed by Thakkar et al. (2003). According to these investigators, although only o1% of the antisense diffused through the dialysis membrane, reverse microdialysis of antisense ODNs has several advantages over microinjection techniques, including the ability to deliver very low and constant concentrations of antisense, which is likely to reduce the probability of any neurotoxic damage. All these elements clearly indicate that reverse microdialysis is a versatile approach to reproduce a neurological condition, or a speciﬁc abnormality that is associated with such a condition, in a relevant brain region. However, the validity of the ensuing preparation still very much depends on the hypothesis to be tested, and most of the time it will not reﬂect the complexity and the true neuropathogenesis of the corresponding disease. Beside the possibility offered by microdialysis to produce or mimic a neurological abnormality, microdialysis can obviously be applied to a wide range of models in which a neurological condition is produced independently (Handbook section on

models of CNS pathology). This includes a variety of models of neurological disorders, which are increasingly supplemented by the availability of mice with targeted mutations. Several studies using microdialysis have been carried out in mice with a mutation that is speciﬁc to a human disease, such as the following: mice transfected with a mutant Cu,Zn-superoxide dismutase (SOD1) gene from humans with familial amyotrophic lateral sclerosis (ALS) (Bogdanov et al., 1998; Tovar-Y-Romo and Tapia, 2006); transgenic models of Huntingon’s disease (Petersen et al., 2002; Gianfriddo et al., 2004); transgenic model of early-onset familial Parkinson’s disease (Goldberg et al., 2003); mice overexpressing amyloid-b, as a model of early-onset forms of familial Alzheimer’s disease (Cirrito et al., 2003). Many more studies used knockout mice to examine whether a speciﬁc protein is involved in a particular brain function or CNS pathology (only as examples, see Shimizu-Sasamata et al., 1998; Bortolozzi et al., 2004; Morishima et al., 2005). Although constitutive gene knockout mice are an invaluable research tool, it is important to recognize their potential limitations, which include (Beglopoulos and Shen, 2004): (i) the biological effects of the genetic modiﬁcation may be compensated by adaptive changes as the organism develops; (ii) lack of speciﬁcity to brain areas of interest; and (iii) as the targeted gene(s) are altered throughout the organism and in all cell types, peripheral effects may occur and result in unwanted interferences. Inducible knockouts avoid some of these constraints, but their availability is limited. Accordingly, for microdialysis-based experiments, it may be pertinent to consider using as alternative genetic tools the antisense technology outlined above, or virus-mediated gene transfer, that is, viral vectors that can be used to introduce or upregulate speciﬁc gene(s) in a selected region of the brain of adult mice. As the antisense method, viral gene transfer reduces the risk of compensation for the genetic change, and it is organ speciﬁc. In addition, cell type speciﬁcity is easier to achieve with viral vectors than with antisense. With the viral gene transfer method, the distribution of the genetic change is restricted to a small brain region (Gerdes et al., 2000), but it is large enough for

208

microdialysis applications as the studies of Sanchez-Pernaute et al. (2001) and Hirooka and Sakai (2004) showed.

III. Monitoring of physiological and/or biochemical variables, with methods independent of microdialysis III.A. Measurements focused on the microdialysis site Microdialysis can be supplemented by a wide range of methods that can provide independent information on the site under study. In vivo electrochemistry may be used to examine changes in electroactive compounds, essentially neurotransmitters and their metabolites (i.e., monoamines), ascorbic acid and uric acid. This strategy allowed investigators to examine how neurotransmitter release and uptake may be altered by microdialysis (Borland et al., 2005), and how relevant compounds administered by reverse microdialysis inﬂuenced extracellular monoamine levels monitored by in vivo voltammetry (Moghaddam et al., 1990). Within this context, the advantage of voltammetry over microdialysis is that it can provide a better time and spatial resolution. In the future, one would expect microbiosensors (or enzyme-selective microelectrodes) to extend this strategy to a wider range of endogenous compounds such as glucose, lactate, glutamate, and choline. Within this category, electrochemical nitric oxide (NO) sensors are noteworthy because of the rapid developments in the ﬁeld of NO-cGMP signaling (Cherian et al., 2000; Heinzen and Pollack, 2002, 2004). Finally, local changes in tissue oxygen (PtO2) can be monitored with amperometry (Lowry et al., 1998), and Osborne et al. (2001) used a gold collar electroplated directly onto the metal shaft of the microdialysis probe as working electrode. Ion-selective electrodes (i.e., microelectrodes for the recording of changes in extracellular K+, Ca2+, H+ etc.) have also been combined successfully with microdialysis, for example, in studies of spreading depression (Moghaddam et al., 1990; Herreras and Somjen, 1993) and anoxic depolarization (Perez-Pinzon et al., 1993; Dohmen et al., 2005). They were also used to determine how

far from the dialysis ﬁber the perfusion of Ca2+-free medium impacts on extracellular Ca2+ (Benveniste et al., 1989). Electrochemical techniques and ion-selective electrodes may be used also to monitor exogenous compounds administered as relevant tracers. In the ﬁeld of cerebral ischemia, several studies have combined microdialysis with repeated measures of local cerebral blood ﬂow by the hydrogen clearance technique (i.e., amperometric detection of local brain tissue hydrogen with platinum electrodes) (Lowry et al., 1998; Bhardwaj et al. 2000). Tetramethylammonium (TMA+) is a useful tracer to investigate changes in extracellular space diffusion parameters as its concentration can be monitored with TMA+-selective electrodes (Nicholson, 1993; Nicholson et al., 2000). The current optical ﬁber technology and associated light reﬂectance imaging/monitoring already provide methods that can be combined with microdialysis. So far, the most relevant and readily available is laser Doppler ﬂowmetry (LDF; see Fig. 4), designed to give information on (relative) local changes in cerebral blood ﬂow (Bogaert et al., 2000). The imaging of intrinsic optical signals (i.e., changed in the reﬂectance of the tissue itself) can provide information on the neuronal activity in the vicinity of a microdialysis probe (Poe et al., 1996), but one would expect this approach to be improved by using novel, voltage(i.e., cell membrane potential) sensitive ﬂuorescent tracers (Grinvald and Hildesheim, 2004).

III.B. Microdialysis and imaging of the living brain The imaging methods to which we refer in this section are essentially magnetic resonance imaging (MRI) and positron emission tomography (PET). The obvious advantage of combining microdialysis with one of these brain imaging techniques is that both strategies allow investigators to acquire sequential data over an extended period of time. In some studies, MRI was used to examine changes in relevant variables (e.g., ADC, apparent diffusion coefﬁcient; fMRI, functional MRI through imaging of regional blood ﬂow or deoxyhemoglobin) speciﬁcally in the region sampled

209

with microdialysis (Forman et al., 1998; Harris et al., 2000), or in the brain area where drugs were applied by reverse microdialysis (Benveniste et al., 1992). In other studies, pharmacological MRI (phMRI) was used to collect information while the effectiveness of a systemic pharmacological challenge was tested at brain level with microdialysis (Schwarz et al., 2004). Although the combination of microdialysis with proton nuclear magnetic resonance spectroscopy (1H-MRS) appears less versatile than that with MRI, it is a pertinent strategy for the examination of sequential changes in the level of 1H-MRS-visible metabolites in both total tissue (MRS) and extracellular ﬂuid (microdialysis). This strategy is especially relevant to N-acetylaspartate (NAA) as this compound is the most visible in 1H-MR brain spectra, and NAA may play an important role in the volume regulation of neurons (Alessandri et al., 2000). The same view and comments apply to 31P-MRS and 15N-MRS (Kintner et al., 1999; Kanamori and Ross, 2005). Note that nonmetallic probes need to be used whenever microdialysis is to be combined with magnetic resonance methods in vivo (Mason and Romano, 1995). The combination of microdialysis and PET has been used in a number of studies of ischemic or traumatic brain injury in nonhuman primates (Frykholm et al., 2005), presumably because of the relatively low spatial resolution of PET, and the relevance of these studies to multimodal brain monitoring in neurocritical care units (Vespa et al., 2005). Similarly, as PET may be used to assess the distribution of neurotransmitter receptors and their alteration by neurological disorders in the human brain, the microdialysis/PET combination was also applied to this ﬁeld of study (Zimmer et al., 2002; Tsukada et al., 2004).

IV. Limitations and pitfalls of multidisciplinary strategies centered on microdialysis It is clear that any multidisciplinary strategy including microdialysis will be limited by the potential problems associated with this method, which include (i) the area of study encompasses the ﬁber/ tissue injury interface; (ii) difﬁculties in measuring ‘true’ extracellular concentrations; (iii) marked

concentration gradient of the pharmacological challenge when the drug is delivered by reverse microdialysis; (iv) low time and spatial resolution relative to synaptic events and brain cytoarchitecture, respectively. As all these limitations of microdialysis are dealt with in other chapters, we focus here on a single question that is speciﬁc to multidisciplinary strategies when they are applied together to the same brain region: Can one correlate all the information gathered with the different methods? Or, in other words: Do all the variables considered reflect the status of the same tissue area? This is a critical issue within the context of multidisciplinary strategies, because the primary aim of this approach is actually to monitor signals and gather data from the same tissue area. Unfortunately, as the following examples illustrate, that is seldom the case. We have already mentioned that, with microdialysis electrodes, the region contributing to the EEG is much larger than that contributing to the DC potential (Obrenovitch et al., 1993a). In our hands, the best combination with regard to similarity of the tissue involved in associated methods is drug delivery by reverse microdialysis and DC potential recording, because the DC potential reﬂects primarily the cellular ionic homeostasis of a ring of cells that are adjacent or very close to the ﬁber surface. Although the very selective genesis of the DC potential signal is favorable to the microdialysis electrode/drug delivery combination, this is not the case when microdialysis is used as a sampling technique in a brain region where heterogeneous changes may develop. For example, with a microdialysis electrode (4-mm ﬁber length) implanted into the striatum of rats subjected to middle cerebral artery occlusion (i.e., a model of focal ischemia) the microdialysis glutamate data indicated a severe ischemia with anoxic depolarization in this region – paradoxically, the DC potential showed recurrent, peri-lesion spreading depression indicating that anoxic depolarization did not occur in the ventral striatum (Wahl et al., 1994). Accordingly, probes with a short ﬁber-length (1 mm) should be favored when microdialysis sampling is to be combined with DC potential recording.

210 % 100

A Local cerebral blood flow (lCBF)

AD

ET-1 100 µM 0

0

10

20 min

SD

Microdialysis electrode ET-1 B

C lCBF

Recording electrode pr ob

e

5 min

LD F

D MD

CORTEX

IC

LDF field pSD

Fig. 5. Illustration of the problem that may occur when several methods are applied to the same brain region; that is, data interpretation may be difﬁcult because the different methods do not actually involve identical tissue areas or compartments. In these experiments, 100 mM of the potent vasoconstrictor endothelin-1 (ET-1) was perfused via a microdialysis probe implanted in the cortex of halothane-anesthetized rats to produce a ‘micro’ focal ischemia (see Fig. 4 for complementary data). Diagram B, shows that the microdialysis probe incorporated an electrode for the recording of the EEG and DC potential, and that local cerebral blood ﬂow (lCBF) was monitored at the same site by LDF. Trace A, shows the changes in lCBF as ET-1 was applied to the region under study, with an unexpected, multiphasic change occurring with anoxic depolarization (AD). Peri-lesion spreading depression is known to occur in the vicinity of the ischemic core, and this was conﬁrmed with this model. Trace C, shows the pattern of change in lCBF produced by spreading depression (SD) when it is elicited in normal cortex (transient reduction of lCBF immediately followed by hyperemia; Obrenovitch et al., 2004). On the basis of this information, we propose that the lCBF changes shown in trace A reﬂect the combination of two different components that occur in adjacent areas of the small brain region explored by LDF (see diagram D for the cross section of tissue studied by microdialysis and LDF). The progressive reduction of lCBF reﬂects the vasoconstriction produced by ET-1 in the tissue adjacent to the microdialysis ﬁber (MD), leading to the formation of an ischemic core (IC) around the MD. When anoxic depolarization occurs in this region (IC), a wave of perilesion spreading depression (pSD) is elicited in the area adjacent to IC, and detected by LDF as the biphasic change in lCBF that is circled in trace A.

An interesting illustration of how a difference in the areas contributing to the monitored signals may complicate their interpretation is presented in Fig. 5; that is, the complex pattern of changes in local blood ﬂow monitored by LDF when anoxic depolarization occurred with focal ischemia produced by microdialysis delivery of the vasoconstrictor endothelin-1 (ET-1) (see also Fig. 4). These data clearly show that even with an LDF probe designed to provide information on changes in blood ﬂow in a very small region (0.25 mm optic ﬁber separation) the LDF signal was markedly inﬂuenced by a larger region than the ischemic core at the time of anoxic depolarization. Although we place emphasis on this potential problem, it should not deter investigators to embark in multidisciplinary strategies assembled around microdialysis. Indeed, we have found consistently that such a multidisciplinary approach constitutes a powerful and robust tool, and interpretation difﬁculties can be resolved by running pertinent control experiments. V. Conclusions Microdialysis, by itself, allows investigators to combine neurochemistry, neuropharmacology, and experimental neurology in a single in vivo preparation, by sampling the extracellular ﬂuid of the region under study while exposing it to relevant agents delivered via the probe. This basic multidisciplinary strategy can be supplemented by a number of monitoring techniques, whether electrode based or optic ﬁber based. Finally, microdialysis can be associated with imaging methods such as MRI and PET. As such, microdialysis constitutes a versatile platform for the assembly of a wide range of truly multidisciplinary strategies that can be tailored to speciﬁc study objectives. In many cases, the interpretation of the data may be complicated by the fact that the different methods may not involve precisely the same brain region or tissue compartment, but this potential difﬁculty can be dealt with, and once resolved it often leads to a better understanding of the biological processes under study.

211

References Alessandri, B., al-Samsam, R., Corwin, F., Fatouros, P., Young, H.F. and Bullock, R.M. (2000) Acute and late changes in N-acetyl-aspartate following diffuse axonal injury in rats: an MRI spectroscopy and microdialysis study. Neurol. Res., 22: 705–712. Alex, K.D., Yavanian, G.J., McFarlane, H.G., Pluto, C.P. and Pehek, E.A. (2005) Modulation of dopamine release by striatal 5-HT2C receptors. Synapse, 55: 242–251. Beglopoulos, V. and Shen, J. (2004) Gene-targeting technologies for the study of neurological disorders. Neuromolecular Med., 6: 13–30. Benveniste, H., Hansen, A.J. and Ottosen, N.S. (1989) Determination of brain interstitial concentrations by microdialysis. J. Neurochem., 52: 1741–1750. Benveniste, H., Hedlund, L.W. and Johnson, G.A. (1992) Mechanism of detection of acute cerebral ischemia in rats by diffusion-weighted magnetic resonance microscopy. Stroke, 23: 746–754. Benveniste, H. and Huttemeier, P.C. (1990) Microdialysis: theory and application. Prog. Neurobiol., 35: 195–215. Bhardwaj, A., Northington, F.J., Carhuapoma, J.R., Falck, J.R., Harder, D.R., Traystman, R.J. and Koehler, R.C. (2000) P-450 epoxygenase and NO synthase inhibitors reduce cerebral blood ﬂow response to N-methyl-D-aspartate. Am. J. Physiol. Heart Circ. Physiol., 279: H1616–H1624. Blum, D., Galas, M.C., Pintor, A., Brouillet, E., Ledent, C., Muller, C.E., Bantubungi, K., Galluzzo, M., Gall, D., Cuvelier, L., Rolland, A.S., Popoli, P. and Schiffmann, S.N. (2003) A dual role of adenosine A2A receptors in 3-nitropropionic acid-induced striatal lesions: implications for the neuroprotective potential of A2A antagonists. J. Neurosci., 23: 5361–5369. Bogaert, L., Scheller, D., Moonen, J., Sarre, S., Smolders, I., Ebinger, G. and Michotte, Y. (2000) Neurochemical changes and laser Doppler ﬂowmetry in the endothelin-1 rat model for focal cerebral ischemia. Brain Res., 887: 266–275. Bogdanov, M.B., Ramos, L.E., Xu, Z. and Beal, M.F. (1998) Elevated ‘‘hydroxyl radical’’ generation in vivo in an animal model of amyotrophic lateral sclerosis. J. Neurochem., 71: 1321–1324. Borland, L.M., Shi, G., Yang, H. and Michael, A.C. (2005) Voltammetric study of extracellular dopamine near microdialysis probes acutely implanted in the striatum of the anesthetized rat. J. Neurosci. Methods, 146: 149–158. Bortolozzi, A., Amargos-Bosch, M., Toth, M., Artigas, F. and Adell, A. (2004) In vivo efﬂux of serotonin in the dorsal raphe nucleus of 5-HT1A receptor knockout mice. J. Neurochem., 88: 1373–1379. Bourne, J.A. and Fosbraey, P. (2000) Novel method of monitoring electroencephalography at the site of microdialysis during chemically evoked seizures in a freely moving animal. J. Neurosci. Methods, 99: 85–90. Brazhnik, E., Borgnis, R., Muller, R.U. and Fox, S.E. (2004) The effects on place cells of local scopolamine dialysis are mimicked by a mixture of two speciﬁc muscarinic antagonists. J. Neurosci., 24: 9313–9323.

Bronzino, J.D., Kehoe, P., Hendriks, R., Vita, L., Golas, B., Vivona, C. and Morgane, P.J. (1999) Hippocampal neurochemical and electrophysiological measures from freely moving rats. Exp. Neurol., 155: 150–155. Bruhn, T., Christensen, T. and Diemer, N.H. (2003) Uptake of glutamate is impaired in the cortical penumbra of the rat following middle cerebral artery occlusion: an in vivo microdialysis extraction study. J. Neurosci. Res., 71: 551–558. Cherian, L., Goodman, J.C. and Robertson, C.S. (2000) Brain nitric oxide changes after controlled cortical impact injury in rats. J. Neurophysiol., 83: 2171–2178. Cirrito, J.R., May, P.C., O’Dell, M.A., Taylor, J.W., Parsadanian, M., Cramer, J.W., Audia, J.E., Nissen, J.S., Bales, K.R., Paul, S.M., DeMattos, R.B. and Holtzman, D.M. (2003) In vivo assessment of brain interstitial ﬂuid with microdialysis reveals plaque-associated changes in amyloid-beta metabolism and half-life. J. Neurosci., 23: 8844–8853. Crochet, S., Chauvette, S., Boucetta, S. and Timofeev, I. (2005) Modulation of synaptic transmission in neocortex by network activities. Eur. J. Neurosci., 21: 1030–1044. Crouzier, D., Baubichon, D., Bourbon, F. and Testylier, G. (2006) Acetylcholine release, EEG spectral analysis, sleep staging and body temperature studies: a multiparametric approach on freely moving rats. J. Neurosci. Methods. Da Costa, A.P., Guevara-Guzman, R.G., Ohkura, S., Goode, J.A. and Kendrick, K.M. (1996) The role of oxytocin release in the paraventricular nucleus in the control of maternal behaviour in the sheep. J. Neuroendocrinol., 8: 163–177. Dobolyi, A., Reichart, A., Szikra, T., Nyitrai, G., Kekesi, K.A. and Juhasz, G. (2000) Sustained depolarisation induces changes in the extracellular concentrations of purine and pyrimidine nucleosides in the rat thalamus. Neurochem. Int., 37: 71–79. Dohmen, C., Kumura, E., Rosner, G., Heiss, W.D. and Graf, R. (2005) Extracellular correlates of glutamate toxicity in short-term cerebral ischemia and reperfusion: a direct in vivo comparison between white and gray matter. Brain Res., 1037: 43–51. Fairbrother, I.S., Arbuthnott, G.W., Kelly, J.S. and Butcher, S.P. (1990) In vivo mechanisms underlying dopamine release from rat nigrostriatal terminals: I. Studies using veratrine and ouabain. J. Neurochem., 54: 1834–1843. Ferger, B., Themann, C., Rose, S., Halliwell, B. and Jenner, P. (2001) 6-Hydroxydopamine increases the hydroxylation and nitration of phenylalanine in vivo: implication of peroxynitrite formation. J. Neurochem., 78: 509–514. Forman, S.D., Silva, A.C., Dedousis, N., Barbier, E.L., Fernstrom, J.D. and Koretsky, A.P. (1998) Simultaneous glutamate and perfusion fMRI responses to regional brain stimulation. J. Cereb. Blood Flow Metab., 18: 1064–1070. Frykholm, P., Hillered, L., Langstrom, B., Persson, L., Valtysson, J. and Enblad, P. (2005) Relationship between cerebral blood ﬂow and oxygen metabolism, and extracellular glucose and lactate concentrations during middle cerebral artery occlusion and reperfusion: a microdialysis and positron emission tomography study in nonhuman primates. J. Neurosurg., 102: 1076–1084.

212 Fu, Y., Matta, S.G., Kane, V.B. and Sharp, B.M. (2001) Norepinephrine secretion in the hypothalamic paraventricular nucleus of rats during unlimited access to self-administered nicotine: an in vivo microdialysis study. J. Neurosci., 21: 8979–8989. Gerdes, C.A., Castro, M.G. and Lowenstein, P.R. (2000) Strong promoters are the key to highly efﬁcient, noninﬂammatory and noncytotoxic adenoviral-mediated transgene delivery into the brain in vivo. Mol. Ther., 2: 330–338. Gianfriddo, M., Melani, A., Turchi, D., Giovannini, M.G. and Pedata, F. (2004) Adenosine and glutamate extracellular concentrations and mitogen-activated protein kinases in the striatum of Huntington transgenic mice. Selective antagonism of adenosine A2A receptors reduces transmitter outﬂow. Neurobiol. Dis., 17: 77–88. Godwin, D.W., Che, D., O’Malley, D.M. and Zhou, Q. (1997) Photostimulation with caged neurotransmitters using ﬁber optic light guides. J. Neurosci. Methods, 73: 91–106. Goldberg, M.S., Fleming, S.M., Palacino, J.J., Cepeda, C., Lam, H.A., Bhatnagar, A., Meloni, E.G., Wu, N., Ackerson, L.C., Klapstein, G.J., Gajendiran, M., Roth, B.L., Chesselet, M.F., Maidment, N.T., Levine, M.S. and Shen, J. (2003) Parkin-deﬁcient mice exhibit nigrostriatal deﬁcits but not loss of dopaminergic neurons. J. Biol. Chem., 278: 43628–43635. Grinvald, A. and Hildesheim, R. (2004) VSDI: a new era in functional imaging of cortical dynamics. Nat. Rev. Neurosci., 5: 874–885. Harkany, T., Abraham, I., Timmerman, W., Laskay, G., Toth, B., Sasvari, M., Konya, C., Sebens, J.B., Korf, J., Nyakas, C., Zarandi, M., Soos, K., Penke, B. and Luiten, P.G. (2000) Beta-amyloid neurotoxicity is mediated by a glutamatetriggered excitotoxic cascade in rat nucleus basalis. Eur. J. Neurosci., 12: 2735–2745. Harris, N.G., Zilkha, E., Houseman, J., Symms, M.R., Obrenovitch, T.P. and Williams, S.R. (2000) The relationship between the apparent diffusion coefﬁcient measured by magnetic resonance imaging, anoxic depolarization, and glutamate efﬂux during experimental cerebral ischemia. J. Cereb. Blood Flow Metab., 20: 28–36. Heinzen, E.L. and Pollack, G.M. (2002) Use of an electrochemical nitric oxide sensor to detect neuronal nitric oxide production in conscious, unrestrained rats. J. Pharmacol. Toxicol. Methods, 48: 139–146. Heinzen, E.L. and Pollack, G.M. (2004) Pharmacodynamics of morphine-induced neuronal nitric oxide production and antinociceptive tolerance development. Brain Res., 1023: 175–184. Herreras, O. and Somjen, G.G. (1993) Analysis of potential shifts associated with recurrent spreading depression and prolonged unstable spreading depression induced by microdialysis of elevated K+ in hippocampus of anesthetized rats. Brain Res., 610: 283–294. Hikiji, K., Inoue, K., Iwasaki, S., Ichihara, K. and Kiriike, N. (2004) Local perfusion of mCPP into ventromedial hypothalamic nucleus, but not into lateral hypothalamic area and frontal cortex, inhibits food intake in rats. Psychopharmacology (Berl.), 174: 190–196.

Hirooka, Y. and Sakai, K. (2004) Adenovirus-mediated nitric oxide synthase gene transfer into the nucleus tractus solitarius in conscious rats. Methods Mol. Biol., 279: 187–200. Hong, Z.Y., Huang, Z.L., Qu, W.M., Eguchi, N., Urade, Y. and Hayaishi, O. (2005) An adenosine A receptor agonist induces sleep by increasing GABA release in the tuberomammillary nucleus to inhibit histaminergic systems in rats. J. Neurochem., 92: 1542–1549. Jay, T.M., Zilkha, E. and Obrenovitch, T.P. (1999) Long-term potentiation in the dentate gyrus is not linked to increased extracellular glutamate concentration. J. Neurophysiol., 81: 1741–1748. Kanamori, K. and Ross, B.D. (2005) Suppression of glial glutamine release to the extracellular ﬂuid studied in vivo by NMR and microdialysis in hyperammonemic rat brain. J. Neurochem., 94: 74–85. Kintner, D.B., Anderson, M.E., Sailor, K.A., Dienel, G., Fitzpatrick, J.H. Jr. and Gilboe, D.D. (1999) In vivo microdialysis of 2-deoxyglucose 6-phosphate into brain: a novel method for the measurement of interstitial pH using 31PNMR. J. Neurochem., 72: 405–412. Kittner, H., Krugel, U., Hoffmann, E. and Illes, P. (2004) Modulation of feeding behaviour by blocking purinergic receptors in the rat nucleus accumbens: a combined microdialysis, electroencephalographic and behavioural study. Eur. J. Neurosci., 19: 396–404. Lauritzen, M. (1994) Pathophysiology of the migraine aura. The spreading depression theory. Brain, 117: 199–210. Lee, C.R., Abercrombie, E.D. and Tepper, J.M. (2004) Pallidal control of substantia nigra dopaminergic neuron ﬁring pattern and its relation to extracellular neostriatal dopamine levels. Neuroscience, 129: 481–489. Lodge, D.J. and Grace, A.A. (2005) Acute and chronic corticotropin-releasing factor 1 receptor blockade inhibits cocaine-induced dopamine release: correlation with dopamine neuron activity. J. Pharmacol. Exp. Ther., 314: 201–206. Lorrain, D.S., Matuszewich, L. and Hull, E.M. (1998) 8-OHDPAT inﬂuences extracellular levels of serotonin and dopamine in the medial preoptic area of male rats. Brain Res., 790: 217–223. Lowry, J.P., Demestre, M. and Fillenz, M. (1998) Relation between cerebral blood ﬂow and extracellular glucose in rat striatum during mild hypoxia and hyperoxia. Dev. Neurosci., 20: 52–58. Ludvig, N., Nguyen, M.C., Botero, J.M., Tang, H.M., Scalia, F., Scharf, B.A. and Kral, J.G. (2000) Delivering drugs, via microdialysis, into the environment of extracellularly recorded hippocampal neurons in behaving primates. Brain Res. Brain Res. Protoc., 5: 75–84. Ludvig, N., Potter, P.E. and Fox, S.E. (1994) Simultaneous single-cell recording and microdialysis within the same brain site in freely behaving rats: a novel neurobiological method. J. Neurosci. Methods, 55: 31–40. Martinez-Tica, J.F. and Zornow, M.H. (2000) Effects of adenosine agonists and an antagonist on excitatory transmitter release from the ischemic rabbit hippocampus. Brain Res., 872: 110–115.

213 Mason, P.A. and Romano, W.F. (1995) Recovery characteristics of a rigid, nonmetallic microdialysis probe for use in an electromagnetic ﬁeld. Bioelectromagnetics, 16: 113–118. Masuo, Y., Matsumoto, Y., Tokito, F., Tsuda, M. and Fujino, M. (1993) Effects of vasoactive intestinal polypeptide (VIP) and pituitary adenylate cyclase activating polypeptide (PACAP) on the spontaneous release of acetylcholine from the rat hippocampus by brain microdialysis. Brain Res., 611: 207–215. Moghaddam, B., Gruen, R.J., Roth, R.H., Bunney, B.S. and Adams, R.N. (1990) Effect of L-glutamate on the release of striatal dopamine: in vivo dialysis and electrochemical studies. Brain Res., 518: 55–60. Morishima, Y., Miyakawa, T., Furuyashiki, T., Tanaka, Y., Mizuma, H. and Nakanishi, S. (2005) Enhanced cocaine responsiveness and impaired motor coordination in metabotropic glutamate receptor subtype 2 knockout mice. Proc. Natl. Acad. Sci. U.S.A., 102: 4170–4175. Nair, S.G. and Gudelsky, G.A. (2006) 3,4-Methylenedioxymethamphetamine enhances the release of acetylcholine in the prefrontal cortex and dorsal hippocampus of the rat. Psychopharmacology (Berl.), 184: 182–189. Nicholson, C. (1993) Ion-selective microelectrodes and diffusion measurements as tools to explore the brain cell microenvironment. J. Neurosci. Methods, 48: 199–213. Nicholson, C., Chen, K.C., Hrabetova, S. and Tao, L. (2000) Diffusion of molecules in brain extracellular space: theory and experiment. Prog. Brain Res., 125: 129–154. Nixdorf, W.L., Burrows, K.B., Gudelsky, G.A. and Yamamoto, B.K. (2001) Enhancement of 3,4-methylenedioxymethamphetamine neurotoxicity by the energy inhibitor malonate. J. Neurochem., 77: 647–654. Obrenovitch, T.P. and Richards, D.A. (1995) Extracellular neurotransmitter changes in cerebral ischaemia. Cerebrovasc. Brain Metab. Rev., 7: 1–54. Obrenovitch, T.P., Richards, D.A., Sarna, G.S. and Symon, L. (1993a) Combined intracerebral microdialysis and electrophysiological recording: methodology and applications. J. Neurosci. Methods, 47: 139–145. Obrenovitch, T.P., Urenjak, J., Richards, D.A., Ueda, Y., Curzon, G. and Symon, L. (1993b) Extracellular neuroactive amino acids in the rat striatum during ischaemia: comparison between penumbral conditions and ischaemia with sustained anoxic depolarisation. J. Neurochem., 61: 178–186. Obrenovitch, T.P., Urenjak, J. and Zilkha, E. (1994) Intracerebral microdialysis combined with recording of extracellular ﬁeld potential: a novel method for investigation of depolarizing drugs in vivo. Br. J. Pharmacol., 113: 1295–1302. Obrenovitch, T.P., Urenjak, J. and Zilkha, E. (1996) Evidence disputing the link between seizure activity and high extracellular glutamate. J. Neurochem., 66: 2446–2454. Obrenovitch, T.P., Wang, M., Urenjak, J., Butler, M.J. and Dreier, J.P. (2004) The role(s) or nitric oxide during cortical spreading depression. In: Krieglstein, J. and Klumpp, S. (Eds.), Pharmacology of Cerebral Ischemia 2004. Medpharm, Stuttgart, pp. 157–166.

Oldford, E. and Castro-Alamancos, M.A. (2003) Input-speciﬁc effects of acetylcholine on sensory and intracortical evoked responses in the ‘‘barrel cortex’’ in vivo. Neuroscience, 117: 769–778. Osborne, P.G., Li, X., Li, Y. and Han, H. (2001) Oxygen-sensing microdialysis probe for in vivo use. J. Neurosci. Res., 63: 224–232. Page, M.E. and Abercrombie, E.D. (1999) Discrete local application of corticotropin-releasing factor increases locus coeruleus discharge and extracellular norepinephrine in rat hippocampus. Synapse, 33: 304–313. Parikh, V., Pomerleau, F., Huettl, P., Gerhardt, G.A., Sarter, M. and Bruno, J.P. (2004) Rapid assessment of in vivo cholinergic transmission by amperometric detection of changes in extracellular choline levels. Eur. J. Neurosci., 20: 1545–1554. Pena, F. and Tapia, R. (1999) Relationships among seizures, extracellular amino acid changes, and neurodegeneration induced by 4-aminopyridine in rat hippocampus: a microdialysis and electroencephalographic study. J. Neurochem., 72: 2006–2014. Penalva, R.G., Lancel, M., Flachskamm, C., Reul, J.M., Holsboer, F. and Linthorst, A.C. (2003) Effect of sleep and sleep deprivation on serotonergic neurotransmission in the hippocampus: a combined in vivo microdialysis/EEG study in rats. Eur. J. Neurosci., 17: 1896–1906. Perez-Pinzon, M.A., Nilsson, G.E. and Lutz, P.L. (1993) Relationship between ion gradients and neurotransmitter release in the newborn rat striatum during anoxia. Brain Res., 602: 228–233. Petersen, A., Puschban, Z., Lotharius, J., NicNiocaill, B., Wiekop, P., O’Connor, W.T. and Brundin, P. (2002) Evidence for dysfunction of the nigrostriatal pathway in the R6/ 1 line of transgenic Huntington’s disease mice. Neurobiol. Dis., 11: 134–146. Poe, G.R., Nitz, D.A., Rector, D.M., Kristensen, M.P. and Harper, R.M. (1996) Concurrent reﬂectance imaging and microdialysis in the freely behaving cat. J. Neurosci. Methods, 65: 143–149. Price, M.L. and Lucki, I. (2001) Regulation of serotonin release in the lateral septum and striatum by corticotropin-releasing factor. J. Neurosci., 21: 2833–2841. Reynolds, N.C. Jr., Lin, W., Cameron, C.M. and Roerig, D.L. (1999) Extracellular perfusion of rat brain nuclei using microdialysis: a method for studying differential neurotransmitter release in response to neurotoxins. Brain Res. Brain Res. Protoc., 4: 124–131. Richards, D.A., Morrone, L.A. and Bowery, N.G. (2000) Hippocampal extracellular amino acids and EEG spectral analysis in a genetic rat model of absence epilepsy. Neuropharmacology, 39: 2433–2441. Rogers, R.C. (1985) An inexpensive picoliter-volume pressure ejection system. Brain Res. Bull., 15: 669–671. Rollema, H., Johnson, E.A., Booth, R.G., Caldera, P., Lampen, P., Youngster, S.K., Trevor, A.J., Naiman, N. and Castagnoli, N. Jr. (1990) In vivo intracerebral microdialysis studies in rats of MPP+ analogues and related charged species. J. Med. Chem., 33: 2221–2230.

214 Sanchez-Pernaute, R., Harvey-White, J., Cunningham, J. and Bankiewicz, K.S. (2001) Functional effect of adeno-associated virus mediated gene transfer of aromatic L-amino acid decarboxylase into the striatum of 6-OHDA-lesioned rats. Mol. Ther., 4: 324–330. Santiago, M., Granero, L., Machado, A. and Cano, J. (1995) Complex I inhibitor effect on the nigral and striatal release of dopamine in the presence and absence of nomifensine. Eur. J. Pharmacol., 280: 251–256. Scheller, D.K. (2000) Extracellular taurine as a parameter to monitor cerebral insults ‘on-line’: time courses and mechanisms as studied in vivo. Adv. Exp. Med. Biol., 483: 265–272. Schwarz, A.J., Zocchi, A., Reese, T., Gozzi, A., Garzotti, M., Varnier, G., Curcuruto, O., Sartori, I., Girlanda, E., Biscaro, B., Crestan, V., Bertani, S., Heidbreder, C. and Bifone, A. (2004) Concurrent pharmacological MRI and in situ microdialysis of cocaine reveal a complex relationship between the central hemodynamic response and local dopamine concentration. Neuroimage, 23: 296–304. Shimizu-Sasamata, M., Bosque-Hamilton, P., Huang, P.L., Moskowitz, M.A. and Lo, E.H. (1998) Attenuated neurotransmitter release and spreading depression-like depolarizations after focal ischemia in mutant mice with disrupted type I nitric oxide synthase gene. J. Neurosci., 18: 9564–9571. Sierra-Paredes, G. and Sierra-Marcuno, G. (1996) Microperfusion of picrotoxin in the hippocampus of chronic freely moving rats through microdialysis probes: a new method of induce partial and secondary generalized seizures. J. Neurosci. Methods, 67: 113–120. Slezia, A., Kekesi, A.K., Szikra, T., Papp, A.M., Nagy, K., Szente, M., Magloczky, Z., Freund, T.F. and Juhasz, G. (2004) Uridine release during aminopyridine-induced epilepsy. Neurobiol. Dis., 16: 490–499. Smith, T.S. and Bennett, J.P. Jr. (1997) Mitochondrial toxins in models of neurodegenerative diseases. I: in vivo brain hydroxyl radical production during systemic MPTP treatment or following microdialysis infusion of methylpyridinium or azide ions. Brain Res., 765: 183–188. Sotty, F., Souliere, F., Brun, P., Chouvet, G., Steinberg, R., Soubrie, P., Renaud, B. and Suaud-Chagny, M.F. (1998) Differential effects of neurotensin on dopamine release in the caudal and rostral nucleus accumbens: a combined in vivo electrochemical and electrophysiological study. Neuroscience, 85: 1173–1182. Staal, R.G. and Sonsalla, P.K. (2000) Inhibition of brain vesicular monoamine transporter (VMAT2) enhances 1-methyl-4-phenylpyridinium neurotoxicity in vivo in rat striata. J. Pharmacol. Exp. Ther., 293: 336–342. Strong, A.J., Fabricius, M., Boutelle, M.G., Hibbins, S.J., Hopwood, S.E., Jones, R., Parkin, M.C. and Lauritzen, M. (2002) Spreading and synchronous depressions of cortical activity in acutely injured human brain. Stroke, 33: 2738–2743. Swanson, C.J., Perry, K.W. and Schoepp, D.D. (2004) The mGlu2/3 receptor agonist, LY354740, blocks immobilization-induced increases in noradrenaline and dopamine release in the rat medial prefrontal cortex. J. Neurochem., 88: 194–202.

Takata, K., Takeda, Y., Sato, T., Nakatsuka, H., Yokoyama, M. and Morita, K. (2005) Effects of hypothermia for a short period on histologic outcome and extracellular glutamate concentration during and after cardiac arrest in rats. Crit. Care Med., 33: 1340–1345. Taylor, D.L., Davies, S.E., Obrenovitch, T.P., Doheny, M.H., Patsalos, P.N., Clark, J.B. and Symon, L. (1995) Investigation into the role of N-acetylaspartate in cerebral osmoregulation. J. Neurochem., 65: 275–281. Thakkar, M.M., Winston, S. and McCarley, R.W. (2003) A1 receptor and adenosinergic homeostatic regulation of sleep-wakefulness: effects of antisense to the A1 receptor in the cholinergic basal forebrain. J. Neurosci., 23: 4278–4287. Tian, G.F., Azmi, H., Takano, T., Xu, Q., Peng, W., Lin, J., Oberheim, N., Lou, N., Wang, X., Zielke, H.R., Kang, J. and Nedergaard, M. (2005) An astrocytic basis of epilepsy. Nat. Med., 11: 973–981. Tovar-Y-Romo, L.B. and Tapia, R. (2006) Cerebral neurons of transgenic ALS mice are vulnerable to glutamate release stimulation but not to increased extracellular glutamate due to transport blockade. Exp. Neurol. Tsukada, H., Nishiyama, S., Fukumoto, D., Ohba, H., Sato, K. and Kakiuchi, T. (2004) Effects of acute acetylcholinesterase inhibition on the cerebral cholinergic neuronal system and cognitive function: functional imaging of the conscious monkey brain using animal PET in combination with microdialysis. Synapse, 52: 1–10. Urenjak, J., Zilkha, E., Gotoh, M. and Obrenovitch, T.P. (1997) Effect of acidotic challenges on local depolarizations evoked by N-methyl-D-aspartate in the rat striatum. Life Sci., 61: 523–535. Vespa, P., Bergsneider, M., Hattori, N., Wu, H.M., Huang, S.C., Martin, N.A., Glenn, T.C., McArthur, D.L. and Hovda, D.A. (2005) Metabolic crisis without brain ischemia is common after traumatic brain injury: a combined microdialysis and positron emission tomography study. J. Cereb. Blood Flow Metab., 25: 763–774. Wahl, F., Obrenovitch, T.P., Hardy, A.M., Plotkine, M., Boulu, R. and Symon, L. (1994) Extracellular glutamate during focal cerebral ischaemia in rats: time course and calcium dependency. J. Neurochem., 63: 1003–1011. Wang, M., Urenjak, J., Fedele, E. and Obrenovitch, T.P. (2004) Effects of phosphodiesterase inhibition on cortical spreading depression and associated changes in extracellular cyclic GMP. Biochem. Pharmacol., 67: 1619–1627. Waterfall, A.H., Singh, G., Fry, J.R. and Marsden, C.A. (1996) Acute acidosis elevates malonaldehyde in rat brain in vivo. Brain Res., 712: 102–106. West, A.R. and Grace, A.A. (2004) The nitric oxide-guanylyl cyclase signaling pathway modulates membrane activity States and electrophysiological properties of striatal medium spiny neurons recorded in vivo. J. Neurosci., 24: 1924–1935. Wu, W.R., Zhu, Z.T. and Zhu, X.Z. (2000) Differential effects of L-deprenyl on MPP+- and MPTP-induced dopamine overﬂow in microdialysates of striatum and nucleus accumbens. Life Sci., 67: 241–250.

215 Zackheim, J. and Abercrombie, E.D. (2005) Thalamic regulation of striatal acetylcholine efﬂux is both direct and indirect and qualitatively altered in the dopamine-depleted striatum. Neuroscience, 131: 423–436. Zhang, deX., Williamson, J.M., Wu, H.Q., Schwarcz, R. and Bertram, E.H. (2005) In situ-produced 7-chlorokynurenate has different effects on evoked responses in rats with limbic

epilepsy in comparison to naive controls. Epilepsia, 46: 1708–1715. Zimmer, L., Mauger, G., Le Bars, D., Bonmarchand, G., Luxen, A. and Pujol, J.F. (2002) Effect of endogenous serotonin on the binding of the 5-HT1A PET ligand 18F-MPPF in the rat hippocampus: kinetic beta measurements combined with microdialysis. J. Neurochem., 80: 278–286.

CHAPTER 2.6

Ultraslow microﬁltration and microdialysis for in vivo sampling: principle, techniques, and applications Kirsten D. Huinink and Jakob Korf Department of Psychiatry, University Medical Centre Groningen, University of Groningen, Groningen, The Netherlands

Abstract: Microﬁltration (MF), often referred to as ultraﬁltration (UF), is a sampling method that is an alternative to microdialysis (MD). The method is underexposed and deserves more attention. In this review, we summarize the applications of MF. First, we explain the principles of MF and outline the similarities and differences between MD and MF. Next we discuss the in vivo biocompatibility of the hollow ﬁbers. Then we summarize what has been done in the ﬁeld of MF for the last 5 years. In addition, we will discuss some applications of ultraslow microdialysis (usMD). The analytical (detection) devices will be highlighted shortly as this topic deserves a report on its own. was recognized for the ﬁrst time. It was suggested by Janle-Swain et al. (1987) that MF could be used in hospitalized patients in whom biochemical parameters have to be monitored frequently, without the necessity of repeated blood sampling. In vivo MF has been performed in human subjects (Ash et al., 1993; Tiessen et al., 1999; Cheng et al., 2000) and in animals including dogs (Janle-Swain et al., 1987), cat (Janle et al., 1992a), rats (Linhares and Kissinger, 1992, 1993a, b; Moscone et al., 1996; Schneiderheinze and Hogan, 1996; Kaptein et al., 1997, 1998; Janle and Kissinger, 1998; LeegsmaVogt et al., 2001, 2003b, 2004b; Kissinger et al., 2003), mice (Janle et al., 1992b; Janle and Kissinger, 1993), chicken (Savenije et al., 2003), sheep (Imsilp et al., 2000; Janle and Sojka, 2000; Sojka et al., 2000; Janle et al., 2001), pigs (Tiessen et al., 2001), and horses (Spehar et al., 1998). Most in vivo experiments with MF are performed subcutaneously, but intravenous application of MF (Kaptein et al., 1997; Leegsma-Vogt et al., 2001, 2003b, 2004b; Tiessen et al., 2001; Savenije et al., 2003) of saliva (Schramm and Smith, 1991;

I. Introduction and scope The ﬁrst reports on microdialysis (MD) appeared in the late seventies and early eighties of the last century, more than 5 years before methods and applications of microﬁltration (MF) were reported. We preferred to speak about MF instead of ultraﬁltration (UF) as this conforms more to MD terminology. Pioneering studies on MD were described by Delgado et al. (1972), Brodin et al. (1983), Ungerstedt (1986), and their coworkers. MD is now a well-established technique to monitor metabolites, energy substrates, neurotransmitters, and hormones of experimental animals from a wide variety of sites, with an early emphasis on cerebral applications. The clinical potential of MD was recognized from the early onset and, in particular, its application to monitor glucose in diabetes has been pursued ever since. In 1987, the potential of in vivo UF or MF as a clinical device Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

217

DOI: 10.1016/S1569-7339(06)16012-9 Copyright 2007 Elsevier B.V. All rights reserved

218

Table 1. An overview of microﬁltration sample analysis Analyte

Hollow Fiber Membrane Type

Polymer

Loop (L) or Concentric (con) in cm Glucose

Theophylline Acetaminophen Glucose Glucose Theophylline Acetaminophen Sodium Potassium Calcium Phosphates Glucose Glucose Cefazolin Glucose Insulin Myoglobulin Albumin Proteins Sodium

3L

3 3 3 3 3 1 3 3 3

L L L L L L L L L

20

7 12 16 – 2 2 12 16 16

Diameter (mm)

o.d.

–

i.d.

200

MWCO (kDa)

Site of withdrawal

Animal

Analysis method

Sampling period (Hours)

Ref.

Subcutaneous

Dog

Glucose oxidase strips

7–63 d

(Janle-Swain et al., 1987)

Flow

(kDa)

(ml/min)

30

1–2 (ﬁrst week)

Polyacrylonitrile

310

220

30

2–5

Saliva Subcutaneous

Human Rat

Accucheck II reﬂactance meter Spectrophotometric RP-HPLC-UV detector 4 RP-HPLC-electrochemical 8–12

– – – Polyacrylonitrile

– – – 310

– – – 220

– – – 30

– 0.13–0.16 0.02–0.03 2–3

Subcutaneous Subcutaneous

Cat Mouse

– LC-IFA-enzyme reactor

Subcutaneous

Rat

–

310

220

30

1–3

Subcutaneous

Rat

RP-HPLC-UV detector 4.5 RP-HPLC-electrochemical 2 Ion selective electrode –

(Janle et al., 1992) (Linhares et al., 1993b) (Linhares et al., 1993a)

Spectrophotmetric 3L 3L

– – Polyacrylonitrile – Polyacrylontrile

– –

– –

30 –

0.7–1 –

Subcutaneous Subcutaneous

Human Mouse

Glucose-oxidase –

3L con

– – 12 12 1–1.5

– 340

– 240

– 40

– –

Subcutaneous In vitro

Rat na

LC-FIA-enzyme reactor Capillary electrophoresis

con con 3L

0.5–1 1.5 12

Polysulfone Polyacrylontrile –

1100 340 –

– 240 –

100 40 –

– 5

Subcutaneous In vitro

Rat na

SDS-Page 5 Cardi Compact Ion Meter 1 (Horiba, Japan)

Potassium Chloride Glucose Lactate Acebutolol Cephalothin chloroamphenicol Salicylic acid Acetaminophen Ionized Glucose

30 d 21–48 d

(Linhares et al., 1992)

3L

2

Polyacrylonitrile

–

–

–

–

In vitro

na

HPLC-UV detector

30 d 15, 30, and 50 d 5.5 24 o1

(Ash et al., 1993) (Janle et al., 1993) (Janle et al., 1995) (Schneiderheinze et al., 1996)

(Janle et al., 1996)

Spectrophotmetric Sigma Kit 735-10 BAS LC method glucose oxidase –

(Lam et al., 1996)

24

(Moscone et al., 1996)

HPLC-electrochemical con

4

Acrylonitrilsodium methallyl sulfonate

290

240

–

0.2

Subcutaneous

Rat

Electrochemical

Glucose

con

4

Glucose

con

4

3L

12

Lactate Sodium Potassium Chloride Lactate Glucose

Acrylonitrilsodium methallyl sulfonate Acrylonitrilsodium methallyl sulfonate –

290

240

–

0.1

Intraveneous

Rat

290

240

–

0.1–0.3

Subcutaneous

Rat

320

220

–

5

Subcutaneous

Rat

Ion selective electrode Spectrophotemetric Sigma Kit 735–10 BAS LC method glucose oxidase Ion selective electrode

Sodium Potassium Calcium Glucose

–

–

–

–

–

–

1

Subcutaneous

Horse

con

2.5

Acrylonitrilsodium methallyl sulfonate

340

240

50

0.05

Intramuscalar Subcutaneous

Human

Lactate Glucose

con

2

340

240

–

0.2

In vitro

–

Lactate

con

4

290

240

–

0.1–0.35

In vitro

na

Calcium

3L

12

Acrylonitrilsodium methallyl sulfonate Acrylonitrilsodium methallyl sulfonate –

–

–

40

–

Subcutaneous

Sheep

Sheep na

Magnesium Phosphorus Inﬂammation study Glucose

3L con

2 2

Polyacrylonitrile –

– –

– –

30 –

na 0.1

Intramuscalar Bone Intramuscular In vitro

Calcium

3L

12

–

–

–

40

–

Subcutaneous

Sheep

Pig

FIA-enzyme reactor (electrochemical) FIA-enzyme reactor (electrochemical)

5

(Kaptein et al., 1997)

3

(Kaptein et al., 1998)

25 d 15 d – 16 d –

(Janle et al., 1998)

–

(Spehar et al., 1998)

Spectrophotometric FIA-enzyme reactor (electrochemical)

8 2.5

(Tiessen et al., 1999)

FIA-enzyme reactor (electrochemical) FIA-enzyme reactor (electrochemical) Spectrophotometric/ ionselective electrode Spectrophotometric

–

– FIA-enzyme reactor (electrochemical) Spectrophotometric/ ionselective electrode

5.5

(Rhemrev-Boom et al., 1999) (Cheng et al., 2000)

6.5

(Janle et al., 2000)

na – –

(Imsilp et al., 2000) (Rhemrev-Boom et al., 2001) (Janle et al., 2001)

27 (total)

(Tiessen et al., 2001)

Magnesium Phosphorus Glucose

con

–

Acrylonitrilsodium methallyl sulfonate

340

240

20

–

Intramuscular Bone Intravenous

Lactate Glucose

con

4

Acrylonitrilsodium methallyl sulfonate

290

240

–

0.05

Intravenous

Chicken

FIA-enzyme reactor (electrochemical)

8

(Savenije et al., 2003)

Lactate Carbamazepine Cimetidine na Glucose

3L con con con

2 7 4 –

Polyacrylonitrile – – Acrylonitrilsodium methallyl sulfonate

– 300 – 290

– 150 – 240

– 11.5 5 20

– 5 – 0.1

Subcutaneous Subcutaneous In vitro Skin ﬂap Intravascular

Rat na Rat Rat

– Chemiluminescense na FIA-enzyme reactor (electrochemical)

48 – 8 3–4

(Kissinger et al., 2003) (Wang et al., 2003) (Odland et al., 2003) (Leegsma-Vogt et al., 2001)

8 5

(Leegsma-Vogt et al., 2003b) (Leegsma-Vogt et al., 2004b) (Odland et al., 2004) (Huinink et al., 2005)

Spectrophotometric FIA-enzyme reactor (electrochemical)

Lactate

na Poly-l-lysine Poly-l-tryptophan

con con

5 4

– – Polyethylene (ethylenevinyl 430 alcohol coating) Polysulfone 450

– 330

– 0.3 (1)

320

0.2 (1)

– 0.5, 3.0

Skin ﬂap In vitro

Rat na

na Spectrophotometric

–

219

220

Linhares and Kissinger, 1992; Schramm et al., 1993; Kissinger et al., 2003), bone (Janle et al., 2001), and muscle (Spehar et al., 1998) has also been described. On MD, more than 10,000 experimental studies and numerous reviews have been published since 1982. In contrast, since 1987 only about 40 scientiﬁc papers on MF have been published (Table 1), together with two review articles (Garrison et al., 2002; Leegsma-Vogt et al., 2003a). So far in most studies on MF and MD, perfusion rates of over 1 mL/min, have been used for in vivo sampling. MF estimation of the real recovery of the analyte from tissue is not a real issue because essentially undiluted body ﬂuid is collected. In contrast, from MD performed at relatively high perfusion rates complicated procedures are required to calculate (or estimate) the real concentrations of the analyte in the tissue compartment. One option to overcome this issue is to apply such low rates of perfusion that the concentrations of the analyte in the dialysate and in the tissue compartment become the same. The main focus of the present review is on articles on ultraslow MF (usMF) and ultraslow MD (usMD) that were published between 2000 and 2005.

II. Microﬁltration sampling techniques Two different types of MF probes have been developed of which one is commercially available (Bioanalytical Systems, Lafayette, IN, USA). Basically an MF device consists of a hollow ﬁber membrane that is connected via tubing to a syringe pump. The ﬂow rate/sampling rate is induced by applying a negative pressure over the hollow ﬁber membrane; this can be conveniently accomplished by a syringe pump or monovette/vacutainer. The maximum size of particles that are transported over the hollow ﬁber membrane is limited not only by its molecular weight cut-off (MWCO) value, but also by the conﬁguration and the charge of the analyte. MF samples are undiluted, small (microliter or nanoliter size), and sterile. The commercially available MF probe (Bioanalytical systems) consists of one or more hollow ﬁber loops with various hollow ﬁber lengths. The

smallest loop has a length of 2 cm and the biggest is 12 cm.The sampling rate is relatively high, in the range of 0.5–2 mL/h/cm hollow ﬁber membrane. By using a relatively high negative pressure, ﬂuid and dissolved metabolites are directly pulled from the blood capillaries to the ultraﬁltrate membrane, and thus the ultraﬁltrate closely resembles the plasma water composition of blood (Janle-Swain et al., 1987). We refer to this type of MF as fast MF, for example, the UF-3-12 (Bioanalytical Systems) possesses a total length of 36 cm of hollow ﬁber, so at a high negative pressure the typical ﬁltration rate is between 0.3 and 1.2 mL/min. Disadvantages of fast MF are the relatively large size of the hollow ﬁber that has to be implanted through a 14-gauge cannula with (local) anesthesia, the large hollow ﬁber can cause tissue damage, and the MF sampling construction may not be suitable for intravenous MF sampling, possibly except in the larger blood vessels. Normalrate MF allows the collection of discrete samples of microliter size in small commercial tubes that can conveniently be processed manually for subsequent analysis. A far smaller MF probe was introduced by Moscone et al. (1996), which allows to sample continuously for at least 30 h at ﬂow rates of 200 nL/ min and less. Such samples are too small to handle manually. This MF probe is combined with a storage device, so time proﬁles of analytes in, for example, the subcutaneous or blood compartment can be collected and stored before further analysis, without manual intervention. This integrated device is called the ultraﬁltration collection device (UCD); see Fig. 1. This device consists of a hollow ﬁber probe, a long coil for sampling storage of fused silica or coated glass; as with other materials, evaporation of the collected ﬂuid disturbs ﬂow patterns and a ﬂow creator by means of negative pressure (monovette). Because of the small size of the probe, the probe can be introduced subcutaneously via adapted needles with minimal tissue damage and without anesthesia. The amount of ﬂuid withdrawn from the tissue compartment is usually small relative to the size of that compartment, so the chemical balances of the extracellular ﬂuid (ECF) are not or minimally disturbed and the body ﬂuid is easily replenished. Because

Flow direction

221

2.9 cm

Restriction Fig. 1. Schematic design and photograph of ultraﬁltration collection device (UCD).

MF concentrations are (nearly) identical to ECF concentrations, the measured levels of the analyte reﬂect rather accurately the metabolic changes in the intercellular spaces (Moscone et al., 1996). It appeared, however, that the volume of the ECF may differ between animal species, condition of the organism, and investigated organ. For instance, Kissinger et al. (2003) mention that an anesthesized rat produces almost no ultraﬁltrate, whereas a rat that has free access to water and food and is moving around produces plenty of ultraﬁltrate. The MF sampling technique is comparable to MD, as it too separates chemicals by moving them across a semipermeable membrane, but there are differences. The most important differences are that in MD, the separation is exclusively due to a concentration gradient and diffusion of the analytes, whereas in MF, analytes are transported over the membrane together with the body ﬂuids in which they are dissolved. Owing to the diffusion gradient, in MD technique one has to calculate the recovery of the analyte, which is time-consuming and complicated as equilibrium is almost never reached. A possible solution to this problem is lowering the perfusion rate to reach diffusion equilibrium, so that calibration becomes redundant. This can be accomplished by applying usMD, a technique developed by our laboratory and ﬁrstly described by Kaptein et al. (1998). However, in contrast to MA and usMD, MF and usMF cannot be performed in the brain. Kaptein

et al. (1998) concluded that ﬂow rates of 50–100 nL/min MF (the usual range of usMF) are not possible in the striatum of the conscious rat. With a ﬂow length of only 0.5 mm, the ﬂow rate decreased during measurement and often stopped completely. Apparently, the ﬂuid production and/or supply in the rat brain is too low. They also showed that in contrast to the normal rate MF, sampling with usMF up to 300 nL/min was well possible in the subcutaneous interstitial ﬂuid in the back of the rats. In healthy volunteers, subcutaneous usMF sampling was also possible, for example, for monitoring glucose on-line (Tiessen et al., 1999). From a physical point of view, the ﬂow rate should be as low as possible in usMD. Whether usMD or MF is the best option depends on the in vivo application. Both techniques can use disposable material, and relatively clean samples can be created. When the amount of extracellular tissue ﬂuid is limited, usMD is preferable. A drawback of usMD technique is the necessity of an additional ﬂuid, making it more difﬁcult to keep the sampling system sterile in contrast to MF. In contrast, local application of drug near the probe is easier with MD. The MD and MF sampling technique is also applied as biosensor interface between the body and the sensor. Leegsma-Vogt et al. (2004a) reviewed glucose- and lactate-biosensors and explained how the two biosensors are working and reviewed clinical and experimental results. More differences and/or common characteristics between MF and MD were described by LeegsmaVogt et al. (2003a) and Garrison et al. (2002).

II.A. Pumps for usMF and usMD For the collection of nanoliter samples by usMF and usMD, ﬂow rates as low as 10 nL/min are preferred. In particular, when storage in capillary coils of the UCD is desired, the pumps should generate a pulse-less ﬂow that is constant over several hours. Such low ﬂow rates cannot easily be achieved with mechanical pumps. One option is the use of pumps based on the use of underpressure, combined with a ﬂuid resistance. Such stable ﬂow rates can be realized by using Poiseuille’s law

222

combined with a capillary restriction placed in the UCD. Poiseuille’s law is: v¼

DP p r4 8lZ

Where v ¼ ﬂow (m3/s), l ¼ length (m), r ¼ radius (m) of the restriction tube, DP ¼ pressure (Pa) is the difference in pressure between the tubing ends, and Z ¼ viscosity (Pa/s) of the sampled ﬂuid. Different ﬂow rates can be realized by changing either the applied vacuum or the restriction length or diameter. We tested the usefulness of various underpressure devices, in particular cheap disposable injection syringes. Monovettes were tested (1.2, 5.5, and 9.0 mL) with different underpressure values and different restriction speciﬁcations. Flow rates could be maintained for several hours or even days; values ranging from 10 to 500 nL/min can thus be obtained. Advantages of such pumping devices are that they are very cheap, energy friendly, and light weight, so they can be used in freely moving animals and humans. For instance, we have applied them in freely moving chickens (Savenije et al., 2003), where the devices were installed under the wing for subcutaneous and intravenous sampling. Recently alternative pulse-less pumps have been described. In particular, pumps based on electroosmotic ﬂuid movements are a promising alternative for ﬂuid collection. The principle is that electrical ﬁelds in thin laminar channels induce the movement of electrolytes and concomitant water. Electro-osmotic pumping is based on the drag force exerted on an electrolyte by double-layer charge moving in an electric ﬁeld. Electro-osmotic ﬂow can be switched on/off easily by switching of the voltages that establish the electric ﬁeld. Similarly, switching of ﬂows from and to different liquid lines is possible, without (mechanical) valves, allowing complex sample and reagent manipulation. Like the underpressure systems developed by us, electro-osmotic pumps can be constructed with small (nanoliter) dead volumes. Various designs of electro-osmotic pumps have been described and most of the earlier problems that were associated with these devices have been solved. For example, gas bubbles generated at the

electrodes that provide the electric ﬁeld are prevented from entering and thereby obstructing the main ﬂow channel by decoupling the electrode compartments from the ﬂow channel via porous frits or (nano-) channels (Lazar et al., 2000; Morf et al., 2001), or by preventing electrolysis completely using asymmetric electrode patterns (Mpholo et al., 2003) or an asymmetric AC-voltage (Selvaganapathy et al., 2002). A drawback of the conventional electro-osmotic pumps is that high voltages are required to achieve reasonable ﬂows. Particularly, for the application of systems to be carried on or close to the human body, the accessible voltages should not exceed ca. 50 V, and preferably stay below 10 V. Low-voltage electroosmotic pumps have been described (Takamura et al., 2001), and a theoretical analysis of the experimental results obtained by the latter groups has been provided (Brask et al., 2003).

III. Membrane biofouling and tissue changes The hollow ﬁber membranes, the sampling part of MF and MD, placed in vivo, can elicit a material–tissue/blood interaction. Several articles appeared regarding this topic in the context of MF, MD sampling, and/or biosensors (Clark et al., 2000; Imsilp et al., 2000; Wisniewski et al., 2000, 2001; Wisniewski and Reichert, 2000), and some articles described blood–hollow ﬁber interaction in hemodialysis studies; hollow ﬁbers are then being tested extracorporeal (von Baeyer et al., 1988; Sefton et al., 2000; Ishihara et al., 2002). In the ﬁrst article of Janle on MF (Janle-Swain et al., 1987), the subcutaneously implanted hollow ﬁber in rats was evaluated for tissue reaction. However, the material of the hollow ﬁber was not mentioned. Minimal tissue reaction was noticed and only a small sheath of ﬁbrous connective tissue was formed. According to Burhop et al. (1993) and Hakim and Lowrie (1982), polyacrylonitrile implanted as dialyzer caused little complement activation and was classiﬁed as low-complement activity material. However, the period of sampling in both studies lasted only 3–4 h. MF (UF-3-2 and UF-1-2, BAS) implanted subcutaneously in rats by Janle and Kissinger (1993) showed no

223

macroscopically apparent inﬂammatory response 6 weeks after implantation. Also, no local edema was seen within 2 days after intramuscular implantation of RUF-3-12 in horses (Spehar et al., 1998). The histopathological effect of MF probe UF-3-2 made from polyacrylonitrile was tested in sheep in combination without and with intramuscular injections of procaine and benzathine salts of penicillin G (Imsilp et al., 2000). This group found that implantation of the hollow ﬁber alone caused greater inﬂammatory response than the injections of procaine or procaine plus benzathine penicillin G at remote intramuscular sites. Implantation of the MF probe thus caused inﬂammation and degeneration of the surrounding muscle, microscopically and macroscopically. The intensity of the inﬂammation differs depending on the time elapsed after implantation of the hollow ﬁber probe (8–11 days). The longer the implantation, the greater was the magnitude of the inﬂammation. MF probes made from polyacrylonitrile may not be suitable for intramuscular UF sampling in sheep (Imsilp et al., 2000). Clark et al. (2000) investigated a range of commercial hollow ﬁbers for subcutaneously implanted applications. The tested hollow ﬁbers were made of mixed esters of cellulose (pore size ¼ 0.2 mm; Id ¼ 350 mm; Od ¼ 500 mm), polysulfone (pore size ¼ 0.2 mm; Id ¼ 1,000 mm; Od ¼ 1,800 mm, pore size ¼ 0.65 mm; Id ¼ 750 mm; Od ¼ 1,200 mm, MWCO ¼ 10 kDa; Id ¼ 500 mm; Od ¼ 820 mm), regenerated cellulose (MWCO ¼ 13 kDa; Id ¼ 200 mm; Od ¼ 216 mm), and cellulose diacetate (MWCO ¼ 68 kDa; Id ¼ 195 mm; Od ¼ 255 mm). The hollow ﬁbers were implanted subcutaneously in rats, and after 3, 6, and 12 weeks, the implants with surrounded tissue were harvested. The samples were evaluated histologically for thickness of the ﬁbrotic capsule and for membrane integrity. As early as 3 weeks, the hollow ﬁber membranes made of polysulfone with a pore size of 0.2 mm and cellulose diacetate were extensively degradated and considered as unsuitable. The mixed ester cellulose and regenerated retained their integrity even after 12 weeks. All membranes exhibited typical foreign body response with ﬁbrotic capsule formation. The regenerated cellulose (MWCO ¼ 13 kDa) and polysulfone membranes (MWCO ¼ 10 kDa) had the thinnest ﬁbrotic

capsule formation. Not only is the implanted material important for material–tissue interaction, also the MWCO value and the size of the implanted hollow ﬁber may have additional effect. The advantage of sampling subcutaneously is that it avoids the complement cascade. In MD, substances can be added to the perfusate to reduce or lower the complement cascade. In MF, this cannot be done. Sefton discussed the topic: What really is blood compatibility? (Sefton et al., 2000). In hemodialysis studies, most of the time continuous administration of an anticoagulant is necessary. Heparin is applied most often, forming a complex with ATIII and thrombin, the enzyme that converts ﬁbrinogen into ﬁbrin. von Baeyer et al. (1988) tested the following membrane types extracorporeal: cuprophan, cellulose diacetate, symmetric and asymmetric polyacrylonitrile, polyamide, polysulfone, and polymethylmethacrylate. The cuprophan, cellulose diacetate, and symmetric polyacrylonitrile showed continuous ﬁbrin precipitation including white blood cells. Polysulfone, asymmetric polyacrilonitrile, and polyamide remained clear except of a few white blood cells and platelets. The hemodialysis devices were ﬁlled with polymethylmethacrylate, and the surface was covered with a mural white thrombus. It was proposed that the more rigid surface of symmetric membranes hinders sterically the action of heparin near the membrane wall. Micro-elasticity of asymmetric membranes avoids such inhibition (von Baeyer et al., 1988). Our group applied heparin to enable intravenous blood sampling with MF in rats (Kaptein et al., 1997) and pigs (Sojka et al., 2000). Heparin was not necessary for the intravenous sampling of chickens (Savenije et al., 2003).

IV. Biomedical and clinical application of MF and usMD Blood and urine are the most sampled body ﬂuids to investigate the concentration of the analyte of interest, for example, hormones, metabolites, and ions. Blood samples are relatively easily achieved, and clinicians increasingly relying on fast biochemical analysis urge on continuous monitoring.

224

However, continuous blood sampling poses infectious hazards and risks of internal bleedings because the patient is often heparinized. It also remains the question if the blood values represent tissue values after all concentrations may differ because of different distribution, protein binding, or metabolism. MF is an ideal sampling technique to investigate analyte values in the different tissues and it can be performed continuously. Most of the studies describing MF sample low-molecular-weight (LMW) analytes. For example, glucose and lactate are the most sampled analytes, but minerals like magnesium, calcium, phosphorus, and chloride are sampled as well. To sample small analytes, hollow ﬁber membranes with a MWCOo50 kDa are used; in this manner large particles like proteins and cellular elements are excluded. Macro-analytes such as proteins and biopharmaceutical agents, whose pharmakinetics and metabolism are of interest both systemically and in local tissue, are also interesting to sample. One might consider cytokines and peptide hormones as high-molecular-weight (MW) analytes on the order of 20–90 kDa. In this section the LMW analytes sampled with MF are discussed with two articles, at the end, discussing HWM analyte sampling.

IV.A. MF IV.A.1. Minerals Janle and Sojka (2000) studied calcium and magnesium mineral distribution in bone, muscle, and subcutaneous tissue in sheep by infusing calcium gluconate over a period of 1 h. Peak concentration occurred for all studied minerals after 90 min in bone, muscle, and subcutaneous tissue, but the magnitude of the peak varied between the tissues. The mineral concentrations in muscle and subcutaneous tissue were similar, but the concentration in bone was lower. This group also studied mineral metabolism in sheep, also in bone, muscle, and subcutaneous tissue but now without infusing calcium gluconate (Janle et al., 2001). The calcium concentration was higher in plasma than in bone, muscle, and subcutaneous tissue. The concentration of magnesium was the opposite; this value was lower in plasma.

This study shows differences between plasma and interstitial bone mineral concentrations for calcium, magnesium, and phosphorus. Variation in concentration from one tissue to another also differed for the various bone minerals. For example, magnesium bone interstitial value was greater than subcutaneous interstitial magnesium and calcium bone interstitial value was lower compared with other tissues. Some of these differences are likely to be due to calcium protein binding. IV.A.2. Drugs Carbamazepine is extensively used for the management of epilepsy and psychiatric diseases. It is almost completely metabolized in the body, so only small traces are excreted unchanged in the urine (Camara et al., 2005). Kissinger et al. (2003) sampled this drug and its epoxide metabolite in subcutaneous tissue in rats. They infused carbamezepine intraveneously and withdrew ultraﬁltrate and blood concomitantly. It appeared that the mean concentration in the ultraﬁltrates was less than half of the plasma levels. IV.A.3. Glucose and lactate Lactate is not only an important metabolite in sport medicine, but also an analyte of interest in the intensive care medicine, for example, lactate is released from the relation to the extent of ischemic heart (Jackson et al., 1978; Rosano et al., 1996). High blood lactate is associated with poor prognosis in septic patients (de Boer et al., 1994). Cheng et al. (2000) sampled lactate with continuous MF and stored the sample simultaneously in a collection coil made from polyetheretherketone (PEEK) (length ¼ 6 m; Id ¼ 125 mm). The sample was collected with a monovette syringe disposable pump in combination with a capillary restriction for 6 h. The coil was analyzed afterwards with a lactate sensor. Such setup is ideal as it does not require the direct connection of the sampling system to analytical detection system. It could thus be applied to monitor athletes actively performing sports. Glucose is an important metabolite in, for example, diabetes and stroke management. Tiessen et al. (1999) explored the on-line MF technique

225

subcutaneously in human volunteers, using the same analysis setup as used by Kaptein et al. (1997), but now glucose and lactate were determined simultaneously. It was conﬁrmed that subcutaneous glucose concentration is not clearly exclusively linked to blood levels. Continuous but intravenous MF of lactate and glucose was extended in pigs for several hours, during which ischemia of the heart was induced for various periods of time. A signiﬁcant increase in the myocardial lactate efﬂux was detected within minutes after the cardial arterial ﬂow was restored (Tiessen et al., 2001). Balloon inﬂation experiment to induce ischemic brain damage was also performed in the rat (Leegsma-Vogt et al., 2001), also monitoring arterial and venous glucose and lactate simultaneously, so two MF probes were inserted in a single animal allowing the measurement of arteriovenous differences related to brain energy metabolism. Glucose and lactate concentrations were measured in the jugular vein and the aorta. These analytes were measured under control conditions and the experimental condition, during and after inﬂation of an embolectomy balloon for 2 min.Under control conditions a small net cerebral lactate efﬂux and glucose uptake were observed, whereas during brain injury the release of lactate and glucose uptake were reduced and there was an increased lactate inﬂux at high arterial lactate levels. Savenije sampled ultraslow (50 nL/min) glucose and lactate ultraﬁltrate continuously in vivo in broiler chickens in subcutaneous tissue and intravascular; simultaneously the ultraﬁltrate was stored in a collection coil made from fused silica tubing for 6 h. The intravenous glucose and lactate concentration were 8.170.4 and 2.470.3 mM and subcutaneous 6.770.2 and 2.670.2 mM, respectively, before feeding. Glucose levels rose signiﬁcantly after food was provided 11.470.5 intravenously and 10.670.7 subcutaneously. In the plasma samples the concentration of glucose and lactate did not differ signiﬁcantly from the ultraﬁltrate values (Savenije et al., 2003).

IV.A.4. Therapeutical ultrafiltration Therapeutic UF was described by Odland. This group used MF as a method to remove excess

interstitial tissue ﬂuid, thereby reducing edema. The hypothesis was that reduction of tissue edema improved tissue viability and according to their report, it did (Odland et al., 2004). IV.A.5. High-molecular-weight analytes Microﬁltration allows the sampling of high-MW analytes (Schneiderheinze and Hogan, 1996; Huinink et al., 2005). The latter authors investigated MF sampling in vitro and in vivo subcutaneously in rats and they showed that proteins with an MW up to 68 kDa could be recovered for >90% in vitro. In vivo, in the extracellular space, proteins of 9.8, 13.0, and 26.8 kDa were measured. In an in vitro study (Huinink et al., 2005), the UCD was tested for adsorption to the hollow ﬁber membrane and for diffusion of stored proteins in the collection coil. Two model proteins were tested: poly-L-lysine, a hydrophilic protein, and poly-Ltryptophan, a hydrophobic protein, with an MW of 227–354 and 1–5 kDa, respectively. The hollow ﬁber made from polyethylene coated with ethylenevinylalcohol (Plasmaﬂo OP-05W(L), Asahi Medical Co., Japan) gave near 100% recoveries for both proteins. Filling the collection coil with various poly-L-lysine concentrations also gave good recoveries and insigniﬁcant diffusion even after storage for 6 days at 37 1C. Huinink sampled with a rather fast ﬂow rate of 500 nL/min. This approach is convenient for blood sampling and reconstructing time proﬁles. IV.B. usMD Kaptein et al. were the ﬁrst to describe usMD (1998); this group compared usMD with the MF sampling technique. Two sampling rates of 100 and 300 nL/min were subcutaneously investigated in anesthesized rats. The levels of glucose measured in the efﬂuent of usMD obtained from the rat subcutis were exactly the same as that of the ultraﬁltrate, so the recovery was 100% for glucose and lactate sampled with usMD method. usMD was used as a sample in healthy humans at ﬂow rates of 300 and 100 nL/min (Rhemrev-Boom et al., 2002; Tiessen et al., 2002). Tiessen measured glucose levels during an oral glucose tolerance test (100 g

226

glucose) simultaneously in blood, adipose tissue, and loose connective tissue of the abdominal subcutis. Fasting glucose levels were 2.5270.77 mM in adipose tissue and 4.6770.17 mM in blood. This discrepancy increased further to 6.4071.57 and 11.5971.52 at maximum glucose concentration. Unlike in connective tissue insigniﬁcant different glucose values were found compared with blood levels (Tiessen et al., 2002). Ekberg et al. (2005) also measured glucose in humans ingesting glucose (75 g) using four ﬂow rates: 0.3 ( ¼ usMD), 1, 2, and 5 mL/min in interstitial ﬂuid of subcutaneous adipose tissue. Lactate, pyruvate, and glycerol were also measured. This group mentions that in healthy volunteers the concentrations of substances in interstitial ﬂuid is not known but they assume that concentrations in interstitial ﬂuid are equal to concentration in blood (Wisniewski et al., 2002); however, this may be questioned considering other results (Tiessen et al., 2002). The mean maximum blood glucose level was 7.870.4 mM, and the mean maximum glucose level in interstitial adipose tissue using a ﬂow rate of 0.3 mL/min was 8.970.5 mM. In both studies, the mean maximum concentrations in the adipose tissue were observed 80 min after glucose consumption, but there is a contradiction between the levels observed in blood and adipose tissue. This could be due to different kinetics and/or different ﬂow rates and/or the measurements in the connective tissue. The common conclusion was that the usMD ﬂow rate was preferred in studies of local tissue metabolism measuring true interstitial glucose concentrations. Another study (Lourido et al., 2002) examined the correlation obtained from usMD (ﬂow rate ¼ 0.3 mL/min) of subcutaneous adipose tissue in patients with severe brain injury. The correlation between glucose in blood and subcutaneous adipose tissue was not as good during intensive care as in normal healthy humans. An important factor could be differences in local blood ﬂow. Rosdahl and coworkers identiﬁed a perfusion ﬂow rate at which the interstitial ﬂuid completely equilibrates with the MD perfusion ﬂuid in human skeletal muscle and adipose tissue (Rosdahl et al., 1998). Complete equilibrium was reached in both tissues at a ﬂow rate of 0.16 mL/min, and the measured concentrations of glucose, glycerol, and

urea were in good agreement with the expected tissue-speciﬁc levels. The glucose concentration in adipose tissue (4.9870.14 mM) was equal to plasma (5.0770.07 mM), whereas the concentration in muscle (4.4170.11 mM) was lower than in plasma and adipose tissue. The concentration of lactate was higher in muscle (2.3970.22 mM) than in adipose tissue (1.3070.12 mM). The concentration of glycerol was higher in adipose tissue (233719.7 mM) than in muscle (40.873.0 mM) and in plasma (38.773.97 mM). The urea concentration was equal in both tissues. usMD is also used to monitor cerebral metabolism in patients. A methodological study was performed (Hutchinson et al., 2000) in patients with head injury. A number of issues were studied, like the effect of hollow ﬁber length and the relative recovery using different ﬂow rates (0.3 and 1.0 mL/min). The relative recovery for glucose, lactate, pyruvate, and glutamate were 65, 67, 72, and 69%, respectively, using a ﬂow rate of 0.3 and for 1.0 mL/min these values were 21, 27, 34, and 30%, respectively. Stahl et al. (2001) monitored cerebral biochemical alterations that precede and accompany increases in intracranial pressure. Approaching a complete cessation of cerebral blood ﬂow showed that glucose and pyruvate values decreased, and that lactate, glycerol, glutamate, and lactate/pyruvate ratio (characterized cerebral ischemia) were increased. Values of the healthy brain were obtained with MD (Reinstrup et al., 2000). Trafﬁcking of substrates between cells is considered to meet energy demand of the brain. Accordingly, it has been hypothesized that neuronal energy consumption depends, at least in part, on the release of glucose and lactate of astroglia cells and on subsequent diffusion to neurons via the brain ECS. The importance of an estimation of the concentration of glucose and lactate in the ECS has been recognized by several authors (LundAndersen, 1979; Fellows et al., 1992; Fellows and Boutelle, 1993; Demestre et al., 1997; Fray et al., 1997; Lowry et al., 1998; Shram et al., 1998; McNay and Gold, 1999). Concentrations ranging from 0.35 to 3.3 mM glucose (McNay and Gold, 1999) and 0.35 to 1.1 mM lactate (Demestre et al., 1997; Shram et al., 1998) have been reported.

227

There are, however, no reports on either estimates of the turnover rate of glucose and lactate of the ECS, or of the proportion of energy substrates that is transferred via this compartment. Our group proposed using usMD as a simple approach to estimate these turnover rates based on the loss of, for example, glucose or lactate added to the perfusate at steady-state levels. In practice, the decrease of glucose or lactate in the inﬂux and efﬂux were measured at a constant ﬂow rate of 100 nL/min. This approach is justiﬁed because it has been shown that acute hyperglycemia does not affect glucose cerebral conscious in rats (Duckrow and Bryan, 1987; Orzi et al., 1988). During low rate perfusion, rapidly new steady-state levels are achieved and the difference between the amounts infused and collected per time unit reﬂects the sum of diffusion and consumption of the substrates in the ECS surrounding the MD-probe. Diffusion appeared to be a minor component. The normal ECS compartment of the brain is 18% (LundAndersen, 1979; Lazar et al., 2000). Steady-state levels of glucose and lactate 2–4 days after probe implantation were 0.2370.12 and 0.6670.36 mM, respectively. The turnover rate of lactate was 0.1370.03 mmol/g/min and that of glucose 0.037 0.01 mmol/g/min. These ﬁgures show that less than 10% of the rat striatal energy substrates were transferred via the extracellular space and mostly attributed to lactate trafﬁcking. The values were close to those of the glucose–lactate shuttle estimated from a total energy balance (Attwell and Laughlin, 2001).

V. Analytical detection Samples collected with MD and MF can be collected with conventional manual or automated collection devices. The analysis of nanoliter samples, as obtained with usMF and usMD, deserves particular attention. Only few attempts have been described to combine these micro-perfusion techniques with analytical devices. Promising applications can be expected from a wide variety of analytical techniques such as HPLC (Bengtsson et al., 2005; Tsai, 2005), Microbore HPLC (Leggas et al., 2004), MS (Andren et al., 2002; Bengtsson

et al., 2005; Jakubowski Jennifer et al., 2005; Wilson et al., 2005), capillary electrophoresis (Mader et al., 1998; Kennedy et al., 2002; O’Brien et al., 2004; Parrot et al., 2004), biosensors (Cheregi et al., 1996; Rhemrev-Boom et al., 2001, 2002; Ehwald, 2004; Korf et al., 2003; LeegsmaVogt et al., 2004a; Yao, 2004), microﬂuidics (Cellar et al., 2005; Hsieh and Zahn, 2005; Zahn et al., 2005), and/or the coupling of several systems to each other. For an overview of MF sample analysis see Table 1. VI. Conclusion Microﬁltration is a good alternative sampling technique compared with MD. We emphasized the versatility of usMD and MF. In Table 1, most of the overviewed reports are about sampling glucose and/or lactate with MF. However, other analytes should be considered as well, for example, proteins (small and large), nucleic acids, hormones (bound versus unbound), or cytokines. Apart from the brain different tissues can be sampled. MF can simultaneously sample various sites including the intravenous route. Future research should optimize the technique with the ﬁnal goal of using usMF and usMD in a hospital setting. With the current trend to increase detector sensitivity and to decrease sample size, application of low-rate MD and MF for biomedical research is now becoming a realistic option. The current overview may serve as a reference for low perfusion technologies. References Andren, P., Farmer, T. and Klintenberg, R. (2002) Endogenous release and metabolism of neuropeptides utilizing in vivo microdialysis microelectrospray mass spectrometry. Mass Spectrom. Hyphenated Tech. Neuropept. Res., 193–213. Ash, S.R., Rainier, J.B., Zopp, W.E., Truitt, R.B., Janle, E.M., Kissinger, P.T. and Poulos, J.T. (1993) A subcutaneous capillary ﬁltrate collector for measurement of blood chemistries. ASAIO J., 39: M699–M705. Attwell, D. and Laughlin, S.B. (2001) An energy budget for signaling in the grey matter of the brain. J. Cereb. Blood Flow and Metab.: ofﬁcial journal of the International Society of Cerebral Blood Flow and Metabolism, 21: 1133–1145. Bengtsson, J., Jansson, B. and Hammarlund-Udenaes, M. (2005) On-line desalting and determination of morphine,

228 morphine-3-glucuronide and morphine-6-glucuronide in microdialysis and plasma samples using column switching and liquid chromatography/tandem mass spectrometry. Rapid Commun. Mass Spectrom., 19: 2116–2122. Brask, A., Goranovic, G. and Bruus, H. (2003) Theoretical analysis of the low-voltage cascade electro-osmotic pump. Sens. Actuators B Chem., B92: 127–132. Brodin, E., Lindefors, N. and Ungerstedt, U. (1983) Potassium evoked in vivo release of substance P in rat caudate nucleus measured using a new technique of brain dialysis and an improved substance P-radioimmunoassay. Acta Physiol. Scand. Suppl., 515: 17–20. Burhop, K.E., Johnson, R.J., Simpson, J., Chenoweth, D.E. and Borgia, J. (1993) Biocompatibility of hemodialysis membranes: evaluation in an ovine model. J. Lab. Clin. Med., 121: 276–293. Camara, M.S., Mastandrea, C. and Golier, J.A. (2005) Chemometrics-assisted simple UV-spectroscopic determination of carbamazepine in human serum and comparison with reference methods. J. Biochem. Biophys. Methods, 64: 153–166. Cellar, N., Burns, S., Meiners, J., Chen, H. and Kennedy, R. (2005) Microﬂuidic chip for low-ﬂow push-pull perfusion sampling in vivo with on-line analysis of amino acids. Anal. Chem., 77: 7067–7073. Cheng, C., Kaptein, W.A., Gruendig, B., Yuen, Y., Korf, J. and Renneberg, R. (2000) Continuous lactate measurement that combines a portable ultraﬁltration storage device with an enzyme sensor. Anal. Lett., 33: 2153–2168. Cheregi, M., Matachescu, C., Moscone, D. and Ciucu, A. (1996) Glucose and lactate biosensors coupled with microdialysis probe for continuous monitoring. South. Braz. J. Chem., 4: 9–17. Clark, H., Barbari, T.A., Stump, K. and Rao, G. (2000) Histologic evaluation of the inﬂammatory response around implanted hollow ﬁber membranes. J. Biomed. Mater. Res., 52: 183–192. de Boer, J., Plijter-Groendijk, H., Visser, K.R., Mook, G.A. and Korf, J. (1994) Continuous monitoring of lactate during exercise in humans using subcutaneous and transcutaneous microdialysis. Eur. J. Appl. Physiol. Occup. Physiol., 69: 281–286. Delgado, J.M., DeFeudis, F.V., Roth, R.H., Ryugo, D.K. and Mitruka, B.M. (1972) Dialytrode for long term intracerebral perfusion in awake monkeys. Arch. Int. Pharmacodyn. Ther., 198: 9–21. Demestre, M., Boutelle, M. and Fillenz, M. (1997) Stimulated release of lactate in freely moving rats is dependent on the uptake of glutamate. J. Physiol., 499(Pt 3): 825–832. Duckrow, R.B. and Bryan, R.M. Jr. (1987) Regional cerebral glucose utilization during hyperglycemia. J. Neurochem., 48: 989–993. Ehwald, R. (2004) Afﬁnity assay in the microdialysis cell – a general principle for on-line measuring biosensors. Technisches Messen., 71: 24–28. Ekberg, N.R., Wisniewski, N., Brismar, K. and Ungerstedt, U. (2005) Measurement of glucose and metabolites in subcutaneous adipose tissue during hyperglycemia with microdialysis at various perfusion ﬂow rates. Clin. Chim. Acta, 359: 53–64.

Fellows, L.K. and Boutelle, M.G. (1993) Rapid changes in extracellular glucose levels and blood ﬂow in the striatum of the freely moving rat. Brain Res., 604: 225–231. Fellows, L.K., Boutelle, M.G. and Fillenz, M. (1992) Extracellular brain glucose levels reﬂect local neuronal activity: a microdialysis study in awake, freely moving rats. J. Neurochem., 59: 2141–2147. Fray, A.E., Boutelle, M. and Fillenz, M. (1997) Extracellular glucose turnover in the striatum of unanaesthetized rats measured by quantitative microdialysis. J. Physiol., 504(Pt 3): 721–726. Garrison, K.E., Pasas, S.A., Cooper, J.D. and Davies, M.I. (2002) A review of membrane sampling from biological tissues with applications in pharmacokinetics, metabolism and pharmacodynamics. Eur. J. Pharm. Sci., 17: 1–12. Hakim, R.M. and Lowrie, E.G. (1982) Hemodialysis-associated neutropenia and hypoxemia: the effect of dialyzer membrane materials. Nephron, 32: 32–39. Hsieh, Y. and Zahn, J. (2005) Glucose recovery in a microﬂuidic microdialysis biochip. Sens. Actuators B Chem., B107: 649–656. Huinink, K.D., Venema, K., Roelofsen, H. and Korf, J. (2005) In vitro sampling and storage of proteins with an ultraﬁltration collection device (UCD) and analysis with absorbance spectrometry and SELDI-TOF-MS. Analyst, 130: 1168–1174. Hutchinson, P.J., O’Connell, M.T., Al Rawi, P.G., Maskell, L.B., Kett-White, R., Gupta, A.K., Richards, H.K., Hutchinson, D.B., Kirkpatrick, P.J. and Pickard, J.D. (2000) Clinical cerebral microdialysis: a methodological study. J. Neurosurg., 93: 37–43. Imsilp, K., Whittem, T., Koritz, G.D., Zachary, J.F. and Schaeffer, D.J. (2000) Inﬂammatory response to intramuscular implantation of polyacrylonitrile ultraﬁltration probes in sheep. Vet. Res., 31: 623–634. Ishihara, K., Hasegawa, T., Watanabe, J. and Iwasaki, Y. (2002) Protein adsorption-resistant hollow ﬁbers for blood puriﬁcation. Artif. Organs., 26: 1014–1019. Jackson, G., Atkinson, L., Clark, M., Crook, B., Armstrong, P. and Oram, S. (1978) Diagnosis of coronary artery disease by estimation of coronary sinus lactate. Br. Heart J., 40: 979–983. Jakubowski Jennifer, A., Hatcher Nathan, G. and Sweedler Jonathan, V. (2005) Online microdialysis-dynamic nanoelectrospray ionization-mass spectrometry for monitoring neuropeptide secretion. J. Mass Spectrom., 40: 924–931. Janle, E., Cregor, M. and Sojka, J.E. (2001) Interstitial ﬂuid calcium, magnesium and phosphorus concentrations in bone, muscle and subcutaneous tissue sampled with ultraﬁltration probes. Curr. Sep., 19: 81–85. Janle, E.M., Clark, T. and Ash, S.R. (1992a) Use of an ultraﬁltrate sampling probe to control glucose levels in a diabetic cat: case study. Curr. Sep., 11: 3–6. Janle, E.M. and Kissinger, P.T. (1993) Microdialysis and ultraﬁltration sampling of small molecules and ions from in vivo dialysis ﬁbers. Am. Assoc. Clin. Chem., 14: 159–165. Janle, E.M. and Kissinger, P.T. (1998) Monitoring physiological variables with membrane probes. Acta Astronaut., 43: 87–99.

229 Janle, E.M., Ostroy, S. and Kissinger, P.T. (1992b) Monitoring the progress of streptozotocin diabetes in the mouse with the ultraﬁltrate probe. Curr. Sep., 11: 17–19. Janle, E.M. and Sojka, J.E. (2000) Use of ultraﬁltration probes in sheep to collect interstitial ﬂuid for measurement of calcium and magnesium. Contemp. Top. Lab. Anim. Sci., 39: 47–50. Janle-Swain, E., Van Vleet, J.F. and Ash, S.R. (1987) Use of a capillary ﬁltrate collector for monitoring glucose in diabetics. ASAIO Trans., 33: 336–340. Kaptein, W.A., Kemper, R.H., Ruiters, M.H., Venema, K., Tiessen, R.G. and Korf, J. (1997) Methodological aspects of glucose monitoring with a slow continuous subcutaneous and intravenous ultraﬁltration system in rats. Biosens. Bioelectron., 12: 967–976. Kaptein, W.A., Zwaagstra, J.J., Venema, K. and Korf, J. (1998) Continuous ultraslow microdialysis and ultraﬁltration for subcutaneous sampling as demonstrated by glucose and lactate measurements in rats. Anal. Chem., 70: 4696–4700. Kennedy, R.T., Watson, C.J., Haskins, W.E., Powell, D.H. and Strecker, R.E. (2002) In vivo neurochemical monitoring by microdialysis and capillary separations. Curr. Opin. Chem. Biol., 6: 659–665. Kissinger, C., Peters, S. and Zhu, Y. (2003) New method for automating sample collection from in vivo ultraﬁltration probes. Curr. Sep., 20: 97–102. Korf, J., Tiessen, R.G., Venema, K. and Rhemrev, M.M. (2003) Biosensors for continuous glucose and lactate monitoring. Ned. Tijdschr. Geneeskd., 147: 1204–1208. Lazar, I.M., Ramsey, R.S., Jacobson, S.C., Foote, R.S. and Ramsey, J.M. (2000) Novel microfabricated device for electrokinetically induced pressure ﬂow and electrospray ionization mass spectrometry. J. Chromatogr. A, 892: 195–201. Leegsma-Vogt, G., Janle, E., Ash, S.R., Venema, K. and Korf, J. (2003a) Utilization of in vivo ultraﬁltration in biomedical research and clinical applications. Life Sci., 73: 2005–2018. Leegsma-Vogt, G., Rhemrev-Boom, M.M., Tiessen, R.G., Venema, K. and Korf, J. (2004a) The potential of biosensor technology in clinical monitoring and experimental research. Biomed. Mater. Eng., 14: 455–464. Leegsma-Vogt, G., van der Werf, S., Venema, K. and Korf, J. (2004b) Modeling cerebral arteriovenous lactate kinetics after intravenous lactate infusion in the rat. J. Cereb. Blood Flow Metab., 24: 1071–1080. Leegsma-Vogt, G., Venema, K. and Korf, J. (2003b) Evidence for a lactate pool in the rat brain that is not used as an energy supply under normoglycemic conditions. J. Cereb. Blood Flow Metab., 23: 933–941. Leegsma-Vogt, G., Venema, K., Postema, F. and Korf, J. (2001) Monitoring arterio-venous differences of glucose and lactate in the anesthetized rat with or without brain damage with ultraﬁltration and biosensor technology. J. Neurosci. Res., 66: 795–802. Leggas, M., Zhuang, Y., Welden, J., Self, Z., Waters, C. and Stewart, C. (2004) Microbore HPLC method with online microdialysis for measurement of topotecan lactone and carboxylate in murine CSF. J. Pharm. Sci., 93: 2284–2295.

Linhares, M.C. and Kissinger, P.T. (1992) Capillary ultraﬁltration: in vivo sampling probes for small molecules. Anal. Chem., 64: 2831–2835. Linhares, M.C. and Kissinger, P.T. (1993a) Determination of endogenous ions in intercellular ﬂuid using capillary ultraﬁltration and microdialysis probes. J. Pharm. Biomed. Anal., 11: 1121–1127. Linhares, M.C. and Kissinger, P.T. (1993b) Pharmacokinetic monitoring in subcutaneous tissue using in vivo capillary ultraﬁltration probes. Pharm. Res., 10: 598–602. Lourido, J., Ederoth, P., Sundvall, N., Ungerstedt, U. and Nordstrom, C.H. (2002) Correlation between blood glucose concentration and glucose concentration in subcutaneous adipose tissue evaluated with microdialysis during intensive care. Scand. J. Clin. Lab. Invest, 62: 285–292. Lowry, J.P., Demestre, M. and Fillenz, M. (1998) Relation between cerebral blood ﬂow and extracellular glucose in rat striatum during mild hypoxia and hyperoxia. Dev. Neurosci., 20: 52–58. Lund-Andersen, H. (1979) Transport of glucose from blood to brain. Physiol. Rev., 59: 305–352. Mader, R., Brunner, M., Rizovski, B., Mensik, C., Steger, G., Eichler, H. and Mueller, M. (1998) Analysis of microdialyzates from cancer patients by capillary electrophoresis. Electrophoresis, 19: 2981–2985. McNay, E.C. and Gold, P.E. (1999) Extracellular glucose concentrations in the rat hippocampus measured by zeronet-ﬂux: effects of microdialysis ﬂow rate, strain, and age. J. Neurochem., 72: 785–790. Morf, W.E., Guenat, O.T. and de Rooij, N.F. (2001) Partial electroosmotic pumping in complex capillary systems. Part 1: principles and general theoretical approach. Sens. Actuators B Chem., B72: 266–272. Moscone, D., Venema, K. and Korf, J. (1996) Ultraﬁltrate sampling device for continuous monitoring. Med. Biol. Eng. Comput., 34: 290–294. Mpholo, M., Smith, C.G. and Brown, A.B.D. (2003) Low voltage plug ﬂow pumping using anisotropic electrode arrays. Sens. Actuators B Chem., B92: 262–268. O’Brien, K., Esguerra, M., Miller, R. and Bowser, M. (2004) Monitoring neurotransmitter release from isolated retinas using online microdialysis-capillary electrophoresis. Anal. Chem., 76: 5069–5074. Odland, R.M., Rheuark, D., Ispirescu, S. and Kizziar, R. (2004) Effect of tissue ultraﬁltration on skin ﬂap survival. Otolaryngol. Head Neck Surg., 131: 296–299. Orzi, F., Lucignani, G., Dow-Edwards, D., Namba, H., Nehlig, A., Patlak, C.S., Pettigrew, K., Schuier, F. and Sokoloff, L. (1988) Local cerebral glucose utilization in controlled graded levels of hyperglycemia in the conscious rat. J. Cereb. Blood Flow Metab., 8: 346–356. Parrot, S., Sauvinet, V., Xavier, J., Chavagnac, D., Mouly-Badina, L., Garcia-Larrea, L., Mertens, P. and Renaud, B. (2004) Capillary electrophoresis combined with microdialysis in the human spinal cord: a new tool for monitoring rapid peroperative changes in amino acid neurotransmitters within the dorsal horn. Electrophoresis, 25: 1511–1517.

230 Reinstrup, P., Stahl, N., Mellergard, P., Uski, T., Ungerstedt, U. and Nordstrom, C.H. (2000) Intracerebral microdialysis in clinical practice: baseline values for chemical markers during wakefulness, anesthesia, and neurosurgery. Neurosurgery, 47: 701–709. Rhemrev-Boom, M.M., Korf, J., Venema, K., Urban, G. and Vadgama, P. (2001) A versatile biosensor device for continuous biomedical monitoring. Biosens. Bioelectron., 16: 839–847. Rhemrev-Boom, M.M., Tiessen, R.G., Jonker, A.A., Venema, K., Vadgama, P. and Korf, J. (2002) A lightweight measuring device for the continuous in vivo monitoring of glucose by means of ultraslow microdialysis in combination with a miniaturised ﬂow-through biosensor. Clin. Chim. Acta, 316: 1–10. Rosano, G.M., Kaski, J.C., Arie, S., Pereira, W.I., Horta, P., Collins, P., Pileggi, F. and Poole-Wilson, P.A. (1996) Failure to demonstrate myocardial ischaemia in patients with angina and normal coronary arteries. Evaluation by continuous coronary sinus pH monitoring and lactate metabolism. Eur. Heart J., 17: 1175–1180. Rosdahl, H., Hamrin, K., Ungerstedt, U. and Henriksson, J. (1998) Metabolite levels in human skeletal muscle and adipose tissue studied with microdialysis at low perfusion ﬂow. Am. J. Physiol., 274: E936–E945. Savenije, B., Venema, K., Gerritzen, M.A., Lambooij, E. and Korf, J. (2003) Minimally invasive technique based on ultraslow ultraﬁltration to collect and store time proﬁles of analytes. Anal. Chem., 75: 4397–4401. Schneiderheinze, J.M. and Hogan, B.L. (1996) Selective in vivo and in vitro sampling of proteins using miniature ultraﬁltration sampling probes. Anal. Chem., 68: 3758–3762. Schramm, W. and Smith, R.H. (1991) An ultraﬁltrate of saliva collected in situ as a biological sample for diagnostic evaluation. Clin. Chem., 37: 114–115. Schramm, W., Smith, R.H. and Craig, P.A. (1993) Methods of simpliﬁed saliva collection for the measurement of drugs of abuse, therapeutic drugs, and other molecules. Ann. N.Y. Acad. Sci., 694: 311–313. Sefton, M.V., Gemmell, C.H. and Gorbet, M.B. (2000) What really is blood compatibility? J. Biomater. Sci. Polym. Ed., 11: 1165–1182. Selvaganapathy, P., Ki, Y., Renaud, P. and Mastrangelo, C. (2002) Bubble-free electrokinetic pumping. J. Microelectromech. Syst., 11: 448–453. Shram, N.F., Netchiporouk, L.I., Martelet, C., Jaffrezic-Renault, N., Bonnet, C. and Cespuglio, R. (1998) In vivo voltammetric detection of rat brain lactate with carbon ﬁber microelectrodes coated with lactate oxidase. Anal. Chem., 70: 2618–2622. Sojka, J.E., Adams, S.B., Rohde, C. and Janie, E.M. (2000) Surgical implantation of ultraﬁltration probes in ovine bone and muscle. J. Invest. Surg., 13: 289–294. Spehar, A., Tiedje, L., Sojka, J.E., Janle, E.M. and Kissinger, P.T. (1998) Recovery of endogenous ions from subcutaneous and intramuscular spaces in horses using ultraﬁltrate probes. Curr. Sep., 17: 47–51.

Stahl, N., Mellergard, P., Hallstrom, A., Ungerstedt, U. and Nordstrom, C.H. (2001) Intracerebral microdialysis and bedside biochemical analysis in patients with fatal traumatic brain lesions. Acta Anaesthesiol. Scand., 45: 977–985. Takamura, Y., Onoda, H., Inokuchi, H., Adachi, S., Oki, A. and Horiike, Y. (2001) In: Ramsey, J.M. and van den Berg, A. (Eds.), Proceedings of mTAS. Kluwer Academic Publishers, Dordrecht. Tiessen, R.G., Kaptein, W.A., Venema, K. and Korf, J. (1999) Slow ultraﬁltration for continuous in vivo sampling: application for glucose and lactate in man. Anal. Chim. Acta, 379: 327–335. Tiessen, R.G., Rhemrev-Boom, M.M. and Korf, J. (2002) Glucose gradient differences in subcutaneous tissue of healthy volunteers assessed with ultraslow microdialysis and a nanolitre glucose sensor. Life Sci., 70: 2457–2466. Tiessen, R.G., Tio, R.A., Hoekstra, A., Venema, K. and Korf, J. (2001) An ultraﬁltration catheter for monitoring of venous lactate and glucose around myocardial ischemia. Biosens. Bioelectron., 16: 159–167. Tsai, T. (2005) Concurrent measurement of unbound genistein in the blood, brain and bile of anesthetized rats using microdialysis and its pharmacokinetic application. J. Chromatogr. A, 1073: 317–322. Ungerstedt, U. (1986) Microdialysis, a new bioanalytical sampling technique. Curr. Sep., 7: 43–46. von Baeyer, H., Lajous-Petter, A., Debrandt, W., Hampl, H., Kochinke, F. and Herbst, R. (1988) Surface reactions on blood contact during haemodialysis and haemoﬁltration with various membrane types. J. Neurosci. Methods, 36: 215–229. Wilson, S., Boix, F., Holm, A., Molander, P., Lundanes, E. and Greibrokk, T. (2005) Determination of bradykinin and argbradykinin in rat muscle tissue by microdialysis and capillary column-switching liquid chromatography with mass spectrometric detection. J. Sep. Sci., 28: 1751–1758. Wisniewski, N. and Reichert, M. (2000) Methods for reducing biosensor membrane biofouling. Colloids Surf. B Biointerfaces, 18: 197–219. Wisniewski, N., Klitzman, B., Miller, B. and Reichert, W.M. (2001) Decreased analyte transport through implanted membranes: differentiation of biofouling from tissue effects. J. Biomed. Mater. Res., 57: 513–521. Wisniewski, N., Moussy, F. and Reichert, W.M. (2000) Characterization of implantable biosensor membrane biofouling. Fresenius. J. Anal. Chem., 366: 611–621. Wisniewski, N., Rajamand, N., Adamsson, U., Lins, P.E., Reichert, W.M., Klitzman, B. and Ungerstedt, U. (2002) Analyte ﬂux through chronically implanted subcutaneous polyamide membranes differs in humans and rats. Am. J. Physiol. Endocrinol. Metab., 282: E1316–E1323. Yao, T. (2004) Flow-injection biosensor system for in vivo analysis of biomolecules in the extracellular space of brain. J. Flow Injection Anal., 21: 59–62. Zahn, J., Hsieh, Y. and Yang, M. (2005) Components of an integrated microﬂuidic fevice for continuous glucose monitoring with responsive insulin delivery. Diabetes Technol. Ther., 7: 536–545.

CHAPTER 3.1

Liquid chromatographic methods used for microdialysis: an overview Sophie Sarre and Yvette Michotte Department of Pharmaceutical Chemistry, Vrije Universiteit Brussel, Brussels, Belgium

Abstract: Liquid chromatography (LC) is usually the method of choice for the analysis of microdialysis samples. In this chapter, we describe the characteristics of the dialysate matrix and review the different parameters of a microdialysis experiment that determine the choice of the LC mode. Advantages and disadvantages of miniaturisation are discussed. Although feasible, miniaturised systems remain limited to scientists with ample experience with LC in general. MS/MS detection is gaining interest in the ﬁeld of microdialysis because of its superior sensitivity and selectivity. It is expected that this mode of detection will grow exponentially in the coming years. However, there are still many issues related to this detection mode that require investigation (matrix effects, use of internal standards, linearityy). A critical overview is given of the different existing methods for biogenic amines, amino acids, acetylcholine, peptides and drug compounds and the most interesting developments are cited. Finally, the importance of thorough validation of LC assays for dialysate analysis is discussed.

(Csoregi et al., 1995) or glutamate (Glu) (Zilkha et al., 1995), on-line immunoassays for peptides (Maidment et al., 1991) and drugs or direct mass spectrometric (MS) detection (Deterding et al., 1992). These approaches will not be addressed further here. Interested readers are referred to the review by Davies et al. (2000). Also, besides LC, capillary electrophoresis (CE) is a separation technique that has especially been exploited for the analysis of amino acids. Because of the very small volume of sample required for CE, high-temporal resolution has been achieved after derivatisation of the amino acids and detection with laser-induced ﬂuorescence (LIF) (Paez and Hernandez, 2001; Sauvinet et al., 2003). This approach is discussed further in this chapter. Microdialysis produces relatively clean samples. They are protein-free, making direct injection into the LC-system possible. The membrane excludes enzymes that may be responsible for degradation of certain compounds such as peptides. Also, since

I. Introduction Microdialysis can be used in conjunction with a wide variety of analytical techniques, but liquid chromatography (LC) has usually been the method of choice for dialysate analysis. Indeed, microdialysis provides a highly ﬁltered (protein-free), low volume, aqueous solution of polar analytes. This deﬁnes a perfect sample for LC. Analytes of interest include neurotransmitters, amino acids, neuromodulators such as certain peptides, drugs and their metabolites. Although most applications include a separation method to isolate the analyte(s) of interest from other endogenous compounds, direct analysis by on-line sensors has also been used. Examples include the use of on-line biosensors for the determination of lactate (Volpe et al., 1995), glucose Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

233

DOI: 10.1016/S1569-7339(06)16013-0 Copyright 2007 Elsevier B.V. All rights reserved

234

proteins do not pass the membrane, only the pharmacologically active free fraction of a drug is determined. So, the dialysis process in itself provides a certain sample clean-up, but the matrix still contains a large amount of small molecules and salts, so the overall selectivity of the assay has to be achieved by the LC column and the LC detector. Besides selectivity, sensitivity of the applied analytical method is probably the most important issue for a successful microdialysis experiment. Indeed, the concentration of many neurotransmitters in the extracellular space is quite low (nM–pM range). The fact that what is collected is only a reﬂection (a few percent) of the true extracellular concentration, due to the recovery of the probe, means that routine detection of sub-picogram amounts of neurotransmitters is required in small volumes of dialysate. The determination of metabolites of transmitters is less stringent since they are usually present in 100–1,000 higher concentrations than those of the parent transmitter. Of course, the innervation of the brain area being monitored will also determine the concentration of the neurotransmitter in the dialysate. If a drug is highly protein-bound, the free fraction of a drug becomes even lower in the dialysate so that ultrasensitive systems are required for accurate determination. Pharmacokinetic studies can be limited by this problem, especially for potent drugs. The probe type or characteristics will also determine the sensitivity of the LC method required. The membrane length that can be used, the ﬂow rate, the type of membrane and the diffusion of the compound(s) of interest across the dialysis membrane all determine probe recovery and thus also the concentration that can be measured in the dialysate. Finally, but not less important, the temporal resolution required will inﬂuence the choice of analytical technique. Indeed, the sample volume obtained with microdialysis is time and ﬂowdependent. So the temporal resolution of the sampling is inversely related to volume of the dialysate. This means that if one needs 30 min to obtain sufﬁcient sample for the LC assay, events in 10 min cannot be monitored. Enhanced temporal resolution is usually an important factor in behavioural studies or in pharmacokinetic studies of drugs with a short half-life.

Table 1. Factors determining the required sensitivity of the LC system Concentration of the analyte in the extracellular space

Tissue (brain area, muscle, fat, y) Drug potency Degree of protein binding

Probe recovery

Type of probe Membrane length Flow rate Diffusion characteristics compound of interest

Temporal resolution

Pharmacological study Behavioural study Pharmacokinetic study (t1/2)

The factors that determine the sensitivity of the LC method are summarised in Table 1.

II. LC considerations The physicochemical properties of the analyte(s) determine the choice of the LC mode. Most applications use a reversed-phase or ion-exchange LC system since these are most compatible with the aqueous character of the dialysates. More recently, with the boost in MS applications, hydrophilic interaction chromatography (HILIC) as introduced by Alpert (1990) has grown in success. This mode is especially interesting for polar compounds such as acetylcholine (Ach) and peptides. The stationary phase is more polar (diol, silica or amine) and the mobile phase is aqueous but contains more than 50% organic solvent. The retention of the polar compounds increases with percentage organic solvent in the mobile phase. The absence of buffers in the mobile phase suits MS detection perfectly. There is an array of detection possibilities for the LC analysis of microdialysates. UV detection is not widely used because of its lack of sensitivity. Sometimes this is overcome by using U- or Zshaped cells in miniaturised systems (Van Belle et al., 1995). In most cases, ﬂuorescence, electrochemical (ECD) or MS detection have been used to improve both the sensitivity and selectivity of the methods. Fluorescence detection takes advantage of any native ﬂuorescent properties a compound

235

may have. However, in many cases a derivatisation protocol is necessary to achieve ﬂuorescence. Electrochemical detection can be used in the amperometric or coulometric mode. Amperometric detection has the advantage that the detector cell can easily be adapted to miniaturised systems. This is not the case for coulometric systems. However, the latter have the advantage over amperometric cells that they can pre-oxidise or reduce the sample prior to passing the analytical cell, enhancing the selectivity of the method. This approach has especially been beneﬁcial for the determination of noradrenaline (NAD) in microdialysates (Gobert et al., 1998; Pudovkina et al., 2001). Because of its selectivity and sensitivity, MS detection for quantitative determination is rapidly gaining interest in the microdialysis ﬁeld since it may render certain compounds, such as peptides readily detectable, which was not the case until now. However, the presence of salts and endogenous compounds in the dialysates can hamper sensitivity of these detection systems by ion-suppression or ion-enhancement during the electrospray ionisation (ESI) process (Matuszewski et al., 2003) so that sample preparation including desalting steps need to be introduced. Recently, the currently existing LC–MS methods for the determination of neurotransmitters in microdialysates as well as the future possibilities of these systems have been reviewed and discussed (Zhang and Beyer, 2006). Microdialysis yields many samples from one animal so that the use of automatic injection for the analysis of the microdialysates should always be considered. While automating a system, the option exists to carry out an on-line or off-line analysis. Advantages of on-line analysis are the possibility to improve temporal resolution and the fact that small samples do not have to be handled. Also, the data produced in near real time provide direct feedback with the experiment. However, the temporal resolution depends on the length of the chromatographic run. Quite often, sample is lost between injections as the dialysate is shunted to waste while the separation is being performed. Also, sample splitting is not possible. Many of these problems, however, can be overcome if multi-port valves are used (Davies et al., 2000).

Troubleshooting becomes more cumbersome with on-line systems. Microdialysis is simply not possible if the LC system is not running, whereas with off-line systems samples can be frozen before analysis. Carry-over problems can also occur with on-line systems, especially in miniaturised ones. Finally, especially in the industrial setting, highthroughput analysis is a major issue. Indeed, the small volume of sample obtained from microdialysis does not allow for multiple injections on different LC systems targeted at certain compounds. To this end, attempts are made to separate and determine as many compounds of interest from one dialysate as possible. We published an LCECD method for the simultaneous determination of L-Dopa, dopamine (DA) and their metabolites dihydroxyphenylacetic acid (DOPAC), homovanillic acid (HVA) and 3-O-methyldopa in microdialysates (Sarre et al., 1992a). This method was used to determine the distribution of L-Dopa in the brain of intact and hemi-parkinson rats as well as its neurochemical effects on the dopaminergic system. More recently, McKenzie et al. (2002) described a capillary LC system with ECD for the simultaneous determination of NAD, DA and serotonin (5HT) together with a number of neuroactive amino acids in dialysates from rat striatum. For the simultaneous measurement of a number of amino acids as well as the citrulline/arginine ratio in stroke, we also developed and validated a method for the determination of these compounds in dialysates from striatum and the cortex after endothelin-1 administration (Van Hemelrijck et al., 2005). Although these applications can be of interest, chromatographic runs can become quite long. Minimal separation followed by MS detection with multiple reaction monitoring (MRM) mode could possibly reduce analysis time, because high selectivity is achieved. However, whether full separation of the compounds from the matrix is essential for good selectivity and sensitivity remains to be determined. Another interesting development the last few years has been the use of monolithic columns. These columns have a higher porosity than particulate columns. Therefore, they can be used at higher ﬂow rates at back pressures that are significantly lower than those observed with particulate columns, without significant

236

reductions in effective plate height (Ikegami and Tanaka, 2004). These columns offer new possibilities in the development of rapid and sensitive analysis of microdialysis samples.

III. Miniaturisation: how far can we go? Typically, LC assays for the determination of neurotransmitters or drugs in microdialysates use conventional size columns [internal diameter (i.d.) 4.6 mm] with 5 or 3 mm packing material. For many applications, these assays are sensitive enough, but when lower analyte concentrations are to be determined and/or a higher sampling frequency is required, miniaturised systems are recommended (Wages et al., 1986). In some cases, conventional columns are still used, but adjustments are made to the microdialysis parameters in order to achieve sufﬁcient sensitivity. These changes include the use of a longer dialysis membrane, longer sampling periods (reduced temporal resolution), decreasing the ﬂow rate to enhance recovery, and changing the composition of the perfusion ﬂuid (higher Ca2+ concentrations or addition of reuptake- or enzyme-inhibitors for neurotransmitters). Obviously, the addition of drugs to the perfusion ﬂuid complicates the interpretation of the data obtained. Using microbore LC columns, it is possible to avoid these adjustments. Microbore systems use columns with i.d.’s of 1 mm (or less). The small samples obtained with microdialysis are diluted less during the chromatography step on a microbore system (typically by a factor 20) and thus become easier to detect. Scott (1984), who together with Kucera (1984), introduced the use of small-bore columns in 1976, reviewed the advantages and disadvantages of these columns. Owing to the lower ﬂow rates applied through these columns, there is a large reduction in mobile phase consumption. Microbore columns have a higher mass sensitivity than their larger diameter counterparts (mass sensitivity is inversely proportional to the square of the column radius) and further require less sample volume to achieve the same concentration sensitivity. Other advantages for microdialysis users are the fact that smaller volumes of sample are

injected, so that from one dialysate, usually two different analyses can be run. Alternatively, one could choose to enhance the temporal resolution by more frequent sampling. Finally, microbore systems are more compatible with MS because of the reduced ﬂow rates. The disadvantages of microbore LC are associated with the need for modiﬁcations to the conventional LC equipment (injector, column and detector connections) in order to reduce system dead volume as much as possible since this can lead to severe band broadening. Further miniaturisation to capillary LC (column i.d. 180–320 mm) or nano LC (column i.d. o 100 mm) emphasises these effects even more, making their routine use even more difﬁcult than for microbore systems. Theoretically, concentration sensitivity is lost in capillary and nano LC systems because of the limitation in volume that can be injected on these columns (Kucera, 1984; Chervet et al., 1992). However, nowadays this problem is solved by preconcentrating the analytes (several microlitres) on a trapping column followed by switching to the analytical column (Baseski et al., 2005; Wilson et al., 2005). Miniaturisation of LC systems has proved to drastically improve the overall sensitivity of LC systems, rendering the determination of many compounds in dialysates more reliable. The use of capillary LC and nano LC may be necessary to be able to quantify peptides or other endogenous substances present in picomolar concentrations in dialysates. However, the technical expertise required to carry out such analyses will remain limited to those with many years of experience with such systems. Routine analysis of dialysates using these systems is not yet on the horizon. In the meantime, microbore LC in some laboratories is being automated with more ease (Smolders et al., 1995; Sarre et al., 1997; Gobert et al., 2003). Despite this, many laboratories are hesitant to introduce these systems as they may be more vulnerable and less ‘student proof’ (Westerink, 2000). Nevertheless, conventional LC offers sufﬁcient sensitivity for many analyses. Narrowbore LC, which uses 2 mm i.d. columns, is a nice go-between, since it can already enhance the sensitivity of a conventional system without serious adaptations required to the instrumentation (Feenstra and Botterblom, 1996).

237

IV. Overview of LC methods for analysis of microdialysates It is by no means possible to mention all applications using LC for the analysis of microdialysis samples. We aim here to give the reader an overview of the existing methods and also to cite some interesting developments.

IV.A. Biogenic amines The introduction of the ECD by Kissinger et al. (1973) (for review see Kissinger, 1989) together with ion-pair reversed-phase LC (Johansson et al., 1978) stimulated the development of sensitive LC assays for biogenic amines. Catecholamines and indoleamines were probably therefore the ﬁrst to be determined in brain dialysates. Three decades later this approach remains the mode of choice for the determination of NAD, DA, 5HT and their main metabolites DOPAC, HVA and 5-HIAA. Conventional LC-ECD has limits of detection (LOD’s) of around 30 fmol (1 nM) (Cheng and Kuo, 1995) and is typically used for the determination of DA in striatal and 5HT in hippocampal dialysates together with their metabolites (Westerink et al., 1987; Sharp et al., 1987; Sarre et al., 1992a). By reducing the column diameter to 2 mm, Feenstra and Botterblom (1996) were able to enhance the detection limit sufﬁciently to allow detection of DA in prefrontal cortex. However, to achieve sufﬁcient detection at 5 min sampling intervals, nomifensin, a DA reuptake inhibitor was added to the perfusion ﬂuid and the membrane length of the probe was increased to 4 mm. Similarly, extracellular levels of DA in the much smaller substantia nigra were measured by adding nomifensin to the perfusion ﬂuid, while using conventional columns (Santiago and Westerink, 1991). It was only by using microbore LC-ECD that DA levels could be determined in the substantia nigra without the use of a reuptake inhibitor, even in 6-hydroxydopamine-lesioned rats (Sarre et al., 1998, 2004). This approach is of course physiologically/pharmacologically more relevant. Baseline dialysate concentrations were about 100 pM (LOD: 40 pM). Using the same approach,

the hippocampal DA and 5HT was also readily detectable (Fig. 1A). This way, an important role of monoamines in the control of experimentally induced limbic seizures (Clinckers et al., 2004) was established. Many neuropharmacologists are interested in the concomitant monitoring of the transmitter and its metabolites, since changes in extracellular levels of the metabolites (e.g., DOPAC or 5-HIAA) can be related to changes in neuronal turnover (Zetterstro¨m et al., 1988). As previously stated, the extracellular concentrations of metabolites are usually 100–1,000 higher than those of the parent transmitter. This may hamper the precision of detection of the neurotransmitter, when separation is critical. This can be overcome by either working with dual amperometric cells (Sarre et al., 1992b, 1997), or when using single cell detectors programming changes in the sensitivity of the detector during the chromatographic run. Alternatively, separate LC systems can be set-up for neurotransmitters and their metabolites. This latter approach probably yields the best sensitivity and thus accuracy/precision of results for the different transmitters. This is achieved by increasing the pH of the mobile phase up to 5 and using more polar stationary phases (C8 instead of C18). In these conditions, metabolites elute in the void, allowing more reliable detection of the amines (Fig. 1A). Maybe the most difﬁcult amine to determine is NAD, especially when determined simultaneously with DA and 5HT using ion-pair reversed-phase LC. The ECD is extremely sensitive to changes in composition of mobile phase, limiting its use to isocratic elution. In this mode, the retention order is always NAD, DA and ﬁnally 5HT. Optimised determination of NAD results in long retention times for 5HT, diminishing its sensitive detection. Conversely, optimising the determination of 5HT, usually results in poor separation of the NAD peak from the void and other interfering peaks. In practice, usually two separate LC-ECD systems are applied for determination of NAD and 5HT (Vahabzadeh and Fillenz, 1994; Ciu et al., 1999). Alternatively, the use of a two-electrode system operating in coulometric mode (Gobert et al., 1998) can be used in which a pre-reduction of the eluent is carried out at 90 mV before oxidation

238 A Ringer’s solution

Standard 250pM

Hippocampal dialysate

Time (min)

8

6

ref. fluorescence (mV)

Dialysate prefrontal cortex

2

3

Standard 450pM

ref. fluorescence (mV)

B

2

1

4

2 0

20

40

60

min

1 0

20

40 min

Standard 150pM

Standard 30pM

60

0

20

40

60

min

Dialysate prefrontal cortex

Fig. 1. (A) Chromatograms of a blank (Ringer’s solution), a standard containing 250 pM NAD, DA and 5HT and a dialysate from rat hippocampus obtained using reversed-phase ion-pair microbore LC with amperometric detection according to Sarre et al. (1997). Retention times are about 2, 4 and 8.5 min for NAD, DA and 5HT respectively. Note that using this approach, separation of NAD from the void is insufﬁcient for its reliable measurement. Estimated concentrations of DA and 5HT in the dialysate are 200 and 140 pM respectively. Microdialysis conditions are as described by Clinckers et al. (2004). (B) Chromatograms of a standard containing 450 pM NAD, DA and 5HT and a dialysate from prefrontal cortex obtained using conventional reversed-phase ion-pair LC with coulometric detection according to Gobert et al. (1998). The three amines are separated within 20 min. Note that using this approach, separation of NAD from the void is better than using the amperometric approach as shown in A. Estimated concentrations in the frontal cortex dialysate are 200 pM for NAD and DA and 100 pM for 5HT. Microdialysis conditions are as described by Gobert et al. (2004). [Chromatograms reproduced with permission]. (C) Chromatograms of a standard containing 150 pM NAD, DA and 5HT, a standard containing approximately 30 pM NAD, DA and 5HT and a dialysate from prefrontal cortex obtained using reversed-phase ion-pair microbore LC with ﬂuorimetric detection after derivatisation with benzylamine and 1,2-diphenylethylenediamine according to Yoshitake et al. (2004). The three amines are separated within 60 min. Note the different elution order compared with A and B, as well as the superior sensitivity. Basal levels of 5HT, NAD and DA in the dialysates were 1.5 1.1 and 1.6 nM respectively. Microdialysis conditions are as described by Yoshitake et al. (2004). [Chromatograms reproduced with permission].

at+280 mV (Fig. 1B). Others have also used a similar approach (Pudovkina et al., 2001). This reduces the detection of interfering peaks from the void making the simultaneous determination of the three amines more accurate. Gobert et al. (2004) also reported an amperometric assay for the

determination of these three amines. However, chromatographic separation took a lot longer (34 min instead of 20 min) for NAD to be well separated from the void response. Similar sensitivity compared with the coulometric approach was obtained by introducing a narrow bore

239

column (see Fig. 1 in ‘‘The Role of microdialysis in drug discovery: focus on antipsychotic agents’’ further in this book). Recently, Yoshitake et al. (2004, 2006) have developed a two-step derivatisation with benzylamine and 1,2-diphenylethylenediamine for ultrasensitive ﬂuorescence detection of NAD, DA and 5HT in microdialysates by microbore LC. Using this elegant approach, a better retention is obtained for NAD and the elution order is NAD, 5HT and DA. Depending on the chosen chromatographic conditions, metabolites can also be determined. The simultaneous determination of NAD, DA and 5HT in microdialysates yielded LOD’s that were sub-femtomole and thus similar or even better than those obtained with microbore LC-ECD (Fig. 1C). Single derivatisation of NAD and 5HT with benzylamine gave even better sensitivity with LOD’s around 50 amol (5 pM). These methods are interesting since they allow to detect biogenic amines in knock-out-mice or in brain areas of rats that can only hold 1 mm probes and even allow to study reductions in extracellular levels of these amines after certain pharmacological manipulations. Finally, Hows et al. (2004) described a narrowbore LC/tandem MS assay for the simultaneous determination of DA, NAD, 5HT and cocaine in brain dialysates. All components were separated in 4 min using a mobile phase without an ion-pairing reagent but with addition of formic acid which could give slight retention of the amines on the C18 column. The LOD’s were 200 pM for DA, 1,000 pM for NAD and 900 pM for 5HT. This approach was not sensitive enough to determine NAD and 5HT in dialysates from the nucleus accumbens in baseline conditions, but provided fast analysis and high selectivity for DA and cocaine.

IV.B. Amino acid neurotransmitters Because of their implication in neurodegenerative and neuropsychiatric disorders, amino acid neurotransmitters have always been of interest to the in vivo neuroscientist/neuropharmacologist since their measurement may provide more mechanistic

insight into the disorders or may possibly provide a diagnostic tool. Aspartate (Asp) and Glu are the major excitatory neurotransmitters, whereas GABA and glycine are important inhibitory transmitters. Other non-transmitter amino acids such as alanine (Ala), glutamine (Gln), histidine (His), lysine (Lys), serine (Ser), taurine (Tau) etcy., have been shown to be altered in certain disorders (Dawson et al., 2004; Van Hemelrijck et al., 2005). The concentrations of amino acids in microdialysates are in the nanomolar–micromolar. Since they are not natively ﬂuorescent or electrochemically active, often a pre-column derivatisation is carried out with either ortho-phtalaldehyde (OPA) and a thiol (Lindroth and Mopper, 1979) or 2,3-naphtalene-dicarboxaldehyde (NDA) prior to analysis (McKenzie et al., 2002). When an array of amino acids is determined from one dialysate, derivatisation with OPA-2-mercapto-ethanol followed by gradient elution and ﬂuorescence detection is often used. Several methods are described in the literature (Jarrett et al., 1986; Rowley et al., 1995; Smolders et al., 1995) using this approach. Recently, we developed a similar narrowbore LC method targeted at amino acids associated with cerebral ischaemia (including citrulline and arginine) (Van Hemelrijck et al., 2005) (Fig. 2C). All these references describe the difﬁculty to separate all amino acids from each other and the necessity for adequate validation in terms of selectivity and accuracy. Recently, Dawson et al. (2004) described the separation of 17 amino acids in microdialysates using a monolithic column. The NDA-derivatised amino acids were separated within 10 min and were detected using ﬂuorescence detection, yielding respectable LOD’s (Glu: 90 nM; GABA: 3 nM; Gly: 300 nM). This high-throughput assay allowed neurochemical evaluation of Lpa1 knockout mice. Simple adaptation of the gradient easily transforms an amino acid assay to one targeted at the determination of Glu and Asp. The later eluting amino acids are washed away also allowing highthroughput monitoring of Glu and Asp (Smolders et al., 1995; Van Hemelrijck et al., 2005) (Fig. 2B). For some amino acids such as GABA, the ﬂuorescent derivative with OPA–2-mercaptoethanol is not stable (Kehr and Ungerstedt, 1988; Smith

240

A Ringer’s solution

Standard 50nM

Dialysate striatum

B

Standard 0.3µM Asp - 1µM Glu

Dialysate striatum

GABA

Time (min)

Time (min) Ringer’s solution

Standard

Dialysate striatum

mV

C

Time (min)

Fig. 2. (A) Chromatograms of a blank (Ringer’s solution), a standard containing 50 nM GABA and a dialysate from the rat striatum obtained using reversed-phase microbore LC with amperometric detection after post-column derivatisation with OPA-tert-butylthiol according to Smolders et al. (1995). Retention time is 7 min.Mean dialysate concentrations of GABA in the striatum determined with this assay were 5373 nM (mean7SEM, n ¼ 6). LC and microdialysis conditions are as described by Smolders et al. (1995). (B) Chromatogram of a standard mixture of 0.3 mM aspartate (Asp) and 1 mM glutamate (Glu) and of a dialysate from the rat striatum using a fast analysis with narrowbore LC with ﬂuorescence detection after post-column derivatisation with OPA–2-mercaptoethanol as described by Van Hemelrijck et al. (2005). The dashed line shows the development of mobile phase B (in percentage) during the chromatographic run. LC and microdialysis conditions as described by Van Hemelrijck et al. (2005). (C) Chromatogram of a blank (Ringer’s solution), a standard mixture of eight amino acids (related to cerebral ischaemia) with the following concentrations: 2 mM for Asp, Glu, citrulline (Cit) and arginine (Arg); 16 mM for serine (Ser) and glutamine (Gln); 8 mM for taurine (Tau) and alanine (Ala) and a dialysate from the rat striatum using narrowbore LC with ﬂuorescence detection after post-column derivatisation with OPA–2mercaptoethanol as described by Van Hemelrijck et al. (2005). The dashed line shows the development of mobile phase B (in percentage) during the chromatographic run. Sensitivity and gain are indicated below the chromatogram. LC and microdialysis conditions as described by Van Hemelrijck et al. (2005).

and Sharp, 1994). This can result in co-eluting degradation products rendering the selectivity of the GABA measurement a lot more critical (Rea et al., 2005). Alternatively, OPA-tert-butylthiol gives good stability derivatives and good electrochemical response (but less ﬂuorescent properties). Owing to the relatively long retention time (usually 40 min) of GABA in the chromatogram when

other amino acids are co-determined, ECD of GABA is preferred to achieve sufﬁcient sensitivity. However, gradient elution in this mode results in baseline drift and long equilibration periods. Therefore, isocratic elution of GABA with ECD after OPA-tert-butylthiol derivatisation is an excellent choice when a microdialysis experiment is targeted towards GABA monitoring. Smolders et al.

241

(1995) developed and validated a microbore LC method using this approach, yielding an LOD of 1 nM (Fig. 2A). This allows accurate determination of extracellular GABA levels, even in small brain structures, since the retention time of GABA is only 8 min and all other amino acids are present in the void of the chromatogram. An interesting approach to simultaneously determine Glu and GABA with high-frequency sampling (5 min) was recently described by Zhang et al. (2005). The LC system consisted of two pumps, an electrochemical detector, a six-way switching valve to control the eluate containing GABA or glutamate to be led to the microbore column and a guard column as a primary column to preseparate the dialysate and to protect the analytical column from contamination. LOD’s were 6 nM for GABA and 160 nM for Glu.

IV.C. Acetylcholine The concentrations of Ach and its metabolite choline in the extracellular ﬂuid range from 0.1 to 6 nM and from 0.6 to 5 mM respectively (Uutela et al., 2005). Indeed, the fast enzymatic hydrolysis of Ach in the synaptic cleft results in very low microdialysate concentrations so that highly sensitive and selective analytical methods are required. LC with ECD or ESI tandem mass spectrometry (ESI-MS/MS) has been most commonly used for the analysis of Ach in microdialysates (for review see Tsai, 2000). LC with ECD for Ach is based on the separation of Ach and choline by reversed-phase ion-pair or cation-exchange chromatography at pH 7–8.5 followed by conversion of the eluents to hydrogen peroxide (H2O2) for electrochemical detection. This conversion is carried out with acetylcholine esterase and choline oxidase (Fig. 3A) as introduced by Potter et al. (1983) and further reﬁned in a post-column immobilised enzyme reactor (IMER) by Damsma et al. (1987). When the column deteriorates (cation-exchange or reversedphase), the choline peak can interfere with the Ach peak. Pre-column IMERs containing choline oxidase and catalase (removes the formed H2O2), have successfully been used to eliminate the

response of choline and thus reducing the void response, rendering the detection of Ach more favourable (Tsai et al., 1996). Initial methods used a Pt-electrode at +500 mV vs. Ag/AgCl for the detection of the H2O2. However, this electrode suffers from a rapid drop in initial sensitivity after coating and oxidation of the electrode surface is often observed after multiple injections, so that intensive maintenance of the electrode is required. This problem can partially be overcome by using an internal standard such as butyrylcholine (Kehr et al., 1998). Using the Pt-electrode, LOD’s of 2 nM at best (Xu et al., 1991: Greaney et al., 1993; Carter and Kehr, 1997) have been achieved. In most cases, this sensitivity was not sufﬁcient to be able to detect baseline release of Ach from microdialysates, so that an acetylcholine esterase inhibitor such as neostigmine or physostigmine needed to be added to the perfusion ﬂuid for adequate quantiﬁcation. Great improvement in electrode stability and sensitivity was achieved with the introduction of the horse-radish peroxidase-osmium polymermodiﬁed glassy carbon electrode (Huang et al., 1995) at +100 mV vs. Ag/AgCl. In combination with a microbore column (now all commercially available in an Ach/choline kit from BAS; USA), LOD’s reached 1–0.3 nM, allowing physiological monitoring of basal Ach release in most cases (Gobert et al., 2003; Fig. 3B). Variants to this approach have been published more recently (Dong et al., 2003; Yamamoto et al., 2004) but without significant improvements in LOD’s. To date, the best sensitivity for Ach assays has been achieved with LC–MS/MS. Both reversed-phase ion-pair (heptaﬂuorobutyric acid as volatile ion-pairing agent) (Zhu et al., 2000) and cation-exchange narrowbore LC (Hows et al., 2002) have yielded extreme sensitivity. LOD’s are 0.1–0.03 nM, which is an order of magnitude more sensitive than what was previously achieved with LC-ECD. High-throughput analysis can be achieved, since Ach is determined only after minimal separation from matrix components (analysis time is less than 5 min). An advantage of the use of MS over other detection methods is the identiﬁcation by both retention time and molecular weight. Also, the triple quadrupole mass spectrometer

242 ‘WIRED ENZYME ELECTRODE’ A

HRP/Os/PVP ChOx AchE Oxidase Ach Substrate Ch

Chromatographic Separation

Glassy Carbon Electrode +100 mV

Post Column IMER

Catalase ChOx ACh/H2O ACh/Ch ACh/H2O2

H2O

HH2O 2O 22

Sample Ach/Ch

Pre Column IMER

3000

3000 Blank (Ringer’s) 0.5 nM Ach 2.5 nM Ach 5.0 nM Ach

mV

B

Dialysate Hippocampus Probe 4 mm MAB Flow Rate 1.5 µL/min Injection volume 20 µL

1500

1500

0

0 0

4

8

12

18

0

4

8

12

18

Time (min) Fig. 3. (A) General scheme describing the enzymatic reactions within pre- and post-IMERs used for the detection of Ach with the horse-radish peroxidase-osmium polymer-modiﬁed glassy carbon electrode (Huang et al., 1995) at+100 mV vs. Ag/AgCl. Ach ¼ acetylcholine; Ch ¼ choline; AchE ¼ acetylcholine esterase; ChOx ¼ choline oxidase. (B) Chromatograms of a blank (Ringer’s solution), standards (0.5–5 nM Ach) and a dialysate from hippocampus of the rat in baseline conditions using above-described approach. Separation is carried out on a Shiseido Capcell PAK AQ C18 (150 1 mm column). The mobile phase consists of 75 mM phosphate buffer (pH 7.5), 0.5 mM sodium octanesulphonate, 0.5 mM tetramethylammoniumbromide, 0.5 mM EDTA and 0.0005% Kathon. The ﬂow rate is 75 mL/min and the injection volume is 20 mL. The retention time for Ach is 13 min.The LOD is 0.25 nM. Dialysates are sampled from the hippocampus of the rat using a 4 mm probe (MAB, Sweden), perfused with Ringer’s solution (2.3 mM Ca2+) at a ﬂow rate of 1.5 mL/min. Basal extracellular Ach levels are detected in the absence of any Ach-esterase inhibitors in the perfusion ﬂuid and are 0.3570.04 nM (mean7SEM. for n ¼ 12).

allows speciﬁc product ions to be monitored. This reﬁnes the detection. Furthermore, no baseline drift is observed and the system equilibrates very quickly (Hows et al., 2002). Finally, no conversion of Ach and choline is required prior to detection. Further improvement in sensitivity was recently achieved by Uutela et al. (2005). Using a diol column in HILIC mode, Ach, choline and the

internal standard acetyl-b-methylcholine were separated within 5 min, but were also well separated from other inorganic salts and endogenous components in the dialysate. The LOD was 0.02 nM for Ach. Considering that these authors used a standard bore column, it must be possible to even improve the sensitivity of this assay. So, the future of Ach measurements lies in the use of MS/MS

243

detection. Only, the cost of these detectors is still relatively high hampering the standard use of these systems up till now.

IV.D. Peptides One of the most challenging problems in neurochemical monitoring in vivo is quantifying known and identifying new neuropeptides. Concentrations of peptides in microdialysates are estimated to be 1–100 pM, so attomol levels in microlitre samples need to be detected/quantiﬁed (Maidment et al., 1989). Identiﬁcation of novel peptides has recently reached a new dimension, by coupling in vivo microdialysis to high sensitivity nano LC (25 mm i.d. column) with ESI-quadrupole ion trap MS/MS (Haskins et al., 2001, 2004; Kennedy et al., 2002). This approach has the advantage that only extracellular releasable peptides are measured. The reader is referred to the chapter further in this book for an elaborate description of these assays. A combination of microdialysis and ESI timeof-ﬂight (TOF) MS has recently been proved a powerful tool to study the metabolism and kinetic processes of neuropeptides in vivo. This approach has the advantage that proteolytic cleavage by enzymes does not occur within the dialysate (ex vivo), allowing identiﬁcation of metabolites. Indeed, Nydahl et al. (2003) perfused LVV-hemorphin-7, a decapeptide with opioid activity but also related to the central renin angiotensin system, into the striatum or the blood and discovered differential metabolism at these two sites. Similarly, Klintenberg and Andren (2005) showed altered biotransformation of dynorphin A (1-17) in striatum of intact and 6-hydroxydopamine-lesioned rats. More experience has been achieved with target analysis of selected peptides. For target analysis of neuropeptides in microdialysates, radioimmunoassay (RIA) is often used. Although RIA provides high sensitivity, it has limited speciﬁcity due to cross reactivity. Emmett et al. (1995) were the ﬁrst to detect extracellular concentrations of neurotensin and met-enkephalin in the striatum of rats using nano LC-microESI-MS/MS with picomolar

sensitivity. A similar approach for monitoring substance P in brain dialysates (Andren and Caprioli, 1995) was used and their method for neurotensin was improved later (Andren and Caprioli, 1999). The speciﬁcity of the method for the different neuropeptides is achieved with tandem mass spectrometry (MS/MS). After ionisation, the ﬁrst mass analyser of the MS selects the precursor ion, which is then fragmented in a collision chamber to form fragments (product ions). The second analyser selectively measures the intensity of one or more of the product ions which only originate from the precursor peptide. This enhances the speciﬁcity of the assay but also diminishes background signals. More recently, Baseski et al. (2005) used nano LCESI quadrupole ion trap MS in MS3 mode to determine enkephalins in microdialysates. The use of an extra fragmentation step is not a standard mode for determination of target species, since with each stage of MS both chemical noise and signal decrease. These authors however, showed that LOD’s were similar in MS2 and MS3 mode (3 pM), but that in MS3 mode precision of injection of standards of 60 pM was significantly better than that observed in MS2 mode (o5% vs. 20%). In all the above described methods, the LC column is mainly used for concentration of the samples and desalting. Little attention is paid to separation, because most of the selectivity is achieved by the MS detection. However, these valuable studies present no extensive validation data in terms of accuracy, linearity and reproducibility. The use of internal standards or isotopically labelled internal standards has not been considered either. Furthermore, although well described in plasma, matrix effects (ion suppression and ion enhancement) due to microdialysate components have not been investigated, but can have an effect on overall sensitivity of these methods (Matuszewski et al., 2003). Finally, sticking of the peptides to microdialysis and/or LC components which can result in carry-over problems is merely discussed. Future research should address these issues, so that assays are obtained that can be used in routine to achieve reliable physiological/pharmacological conclusions. Two recent studies describing assays for the determination of non-central peptides in dialysates

244

are worth mentioning, since they show the feasibility of the use of LC–MS for quantitative determination of target peptides. Wilson et al. (2005) described the use of microbore LC-ESI-TOF MS for the determination of bradykinin and arg-bradykinin in dialysates from rat muscle. Here (Tyr8)bradykinin was added to the dialysate prior to injection. Owing to the large injection volumes (450 mL) detection of 100 pM peptide were easily achieved. Also, possible matrix effects that may inﬂuence ionisation are discussed. Another wellperformed study is the use of narrowbore LC–MS/ MS for the determination of prostaglandin E2 and D2 in spinal microdialysates (Schmidt et al., 2005). These peptides were extracted from the dialysates by liquid–liquid extraction with deuterated prostaglandin D2 as an internal standard. This assay is well validated. The limit of quantiﬁcation was found to be about 70 pM for prostaglandin E2.

IV.E. Drugs and other compounds Microdialysis has been demonstrated to be applicable in many different organs for studying the tissue speciﬁc pharmacokinetics of a large variety of drugs. This has been reviewed recently by Plock and Kloft (2005), so the reader is referred to this review for a more complete overview of the most recent applications. As already mentioned in the introduction, the use of microdialysis in pharmacokinetic studies is limited by the sensitivity of the analytical technique. Indeed, only the free fraction of the drug is measured, so potent drugs that are highly protein-bound become difﬁcult to detect. The half-life of the drug will determine the temporal resolution required to obtain sufﬁcient kinetic data. Thus, higher ﬂow rates decrease the overall recovery of the probe, again putting more strain on the analytics. Usually, reversed-phase LC combined with UV, ﬂuorescence, ECD or MS are used. Using a conventional LC column, unbound ranitidine (Huang et al., 2005) in blood and bile was monitored in dialysates with UV detection at 315 nm.For telitromycin in muscle and adipose tissue dialysates, ﬂuorescence detection yielded

LOD’s of 15 ng/mL (Traunmuller et al., 2005). Some authors use narrowbore LC to enhance sensitivity. This way, the brain distribution of stavudine was evaluated after intranasal administration using on-line LC with UV detection (Yang et al., 2005) and morphine in brain dialysates was determined with ECD with a LOD of 0.5 ng/mL (Groenendaal et al., 2005). Often, further miniaturisation of the LC system and coupling of the microdialysis system on-line are required to reach sufﬁcient sensitivity and to manage the smaller samples with more ease. Microbore LC-UV has been shown to enhance sensitivity for oxcarbazepine (OXC) in blood and hippocampal dialysates (Van Belle et al., 1995) (Fig. 4A), ceftriaxone in dialysates from CSF (Owens et al., 1999), omeprazole in blood, brain and bile dialysates (Cheng et al., 2002), ﬂuconazole in dermal microdialysates (Mathy et al., 2003), theofylline in rat blood and brain dialysates (Tsai and Liu, 2004) and salvianolic acid B, the herbal ingredient from Salvia miltiorrhiza, in blood and bile dialysates (Chen et al., 2005). Depending on the component, LOD’s were obtained between 5 and 100 ng/mL. Limits lower than 1 ng/mL were obtained when combining microbore LC with electrochemical (McLaughlin et al., 2000) or ﬂuorescence detection (Leggas et al., 2004) for the anti-tumour agent tirapazamine and topotecan lacton respectively. Capillary LC (i.d. o 200 mm) has not been described combined with standard detectors. Schmidt and Martinsson (1998) measured a cholinesterase inhibitor NXX-066 with capillary LC-LIF resulting in a limit of quantiﬁcation (LOQ) of 100 pM corresponding to 2 fmol injected and Bergstrom and Markides (2002) coupled a 200 mm column to MS for the determination of ropivacaine and its metabolite in blood dialysates (LOQ: 100 pM). A deuterated internal standard was used. Recently, there has been a boost in new methods for microdialysates with the availability of triple quad MS. Both the selectivity and sensitivity of MS/MS allows the detection of drugs in dialysates even with conventional or narrowbore columns. Examples include the determination of amphetamine in rat brain dialysates (Fuh et al., 2004) with a LOD of 1 ng/mL and the enantiomeric separation of cetirizine with a lower LOQ of 0.25 ng/mL

245

Fig. 4. (A) Chromatogram of a standard of 25 ng/mL oxcarbazepine (OXC) and its metabolite 10,11-dihydro-10-hydroxycarbamazepine (MHD) obtained with reversed-phase microbore LC with UV detection using a U-shaped optical cell (Van Belle et al., 1995). For automation of this assay, an external standard (ES) was used (carbamazepine-10,11-epoxide). (B) MRM chromatogram (B1) and SRM chromatogram (B2) of a standard of 1 ng/mL OXC and its metabolite MHD obtained with reversed-phase microbore LC with MS/MS detection. For this application, methylcarbamazepine (m-CBZ) was used as internal standard for probe calibration. The following mass transitions were followed: m/z 255-194 (MHD), m/z 253-180 (OXC) and m/z 251.1-193 (m-CBZ). (C) MRM chromatogram (C1) and SRM chromatogram (C2) of a dialysate obtained from the hippocampus of the rat, after administration of 10 mg/kg OXC i.p. For microdialysis and LC conditions see Lanckmans et al. (2006). [MRM ¼ multiple reaction monitoring; SRM ¼ single reaction monitoring; TIC ¼ total ion count].

(Gupta et al., 2005). An interesting application is the determination of morphine and its metabolites with column-switching followed by HILIC in conventional mode and tandem MS (Bengtsson et al., 2005). This well-validated study used a deuterated internal standard. The LOQ was 0.5 ng/ mL for morphine, which is similar to that obtained with narrowbore LC and ECD (Groenendaal et al., 2005). Finally, a microbore LC–MS/MS assay was recently published from our lab for the determination of OXC and its metabolite

10,11-dihydro-10-hydroxycarbamazepine (MHD) in rat brain dialysates (Lanckmans et al., 2005) (Fig. 4B and C). Methylcarbamazepine (m-CBZ) was used as internal standard for probe calibration. The LOD achieved was 0.04 ng/mL and the validated LOQ was 1 ng/mL resulting in an increase in sensitivity of nearly an order of magnitude compared with our previously published method for OXC using microbore LC and UV detection (Van Belle et al., 1995) (Fig. 4A). This method was used to study the distribution of OXC

246

in the hippocampus of the rat. Levels of OXC and its metabolite were detectable until 4 h after OXC administration. MS/MS detection is on the up rise, but many aspects need further investigation in order to achieve more insights into its possibilities and limitations for target analysis of drugs and other compounds in microdialysates.

V. Conclusions In general, the analysis of dialysates often turns out to be the bottleneck in a microdialysis experiment. However, over the last three decades, major advances have been made in analytical chemistry in order to render analysis of microdialysis samples available for all (neuro)-scientists. First time users of microdialysis and LC for the analysis of microdialysates are advised to start with conventional or narrowbore LC applications as these systems are more robust. Miniaturised systems are feasible, but are still limited to scientists with ample experience with LC in general. MS/MS detection is gaining in interest in the ﬁeld of microdialysis because of its superior sensitivity and selectivity. It is expected that this mode of detection will grow exponentially in the coming years. However, there are still many issues related to this detection mode that require investigation (matrix effects, use of internal standards, linearityy). Many LC methods have been published for the determination of endogenous and exogenous compounds in microdialysates. However, there is a clear lack of well-validated assays. Indeed, a method that is not proved selective, precise and accurate can result in incorrect determination of concentrations of substances in the dialysates. This can affect the pharmacological response and can lead to wrong pharmacological conclusions. It may also explain differences in pharmacological data between laboratories. Validation should also include data on method robustness and data on stability of analytes or derivatisation products, possible sticking of analytes, etc. In general, more care should be taken when developing or setting up LC assays before their general use for analysis of microdialysis samples.

References Alpert, A.J. (1990) Hydrophilic-interaction chromatography for the separation of peptides, nucleic acids and other polar compounds. J. Chromatogr., 499: 177–196. Andren, P.E. and Caprioli, R.M. (1995) In vivo metabolism of substance P in rat striatum utilizing microdialysis/liquid chromatography/micro-electrospray mass spectrometry. J. Mass Spectrom., 30: 817–824. Andren, P.E. and Caprioli, R.M. (1999) Determination of extracellular release of neurotensin in discrete rat brain regions utilizing in vivo microdialysis/electrospray mass spectrometry. Brain Res., 845: 123–129. Baseski, H.M., Watson, C.J., Cellar, N.A., Shackman, J.G. and Kennedy, R.T. (2005) Capillary liquid chromatography with MS3 for the determination of enkephalins in microdialysis samples from the striatum of anesthetized and freely-moving rats. J. Mass Spectrom., 40(2): 146–153. Bengtsson, J., Jansson, B. and Hammarlund-Udenaes, M. (2005) On-line desalting and determination of morphine, morphine-3-glucuronide and morphine-6-glucuronide in microdialysis and plasma samples using column switching and liquid chromatography/tandem mass spectrometry. Rapid Commun. Mass Spectrom., 19: 2116–2122. Bergstrom, S.K. and Markides, K.E. (2002) On-line coupling of microdialysis to packed capillary column liquid chromatography-tandem mass spectrometry demonstrated by measurement of free concentrations of ropivacaine and metabolite from spiked plasma samples. J. Chromatogr. B., 775(1): 79–87. Carter, A.J. and Kehr, J. (1997) Microbore high-performance liquid chromatographic method for measuring acetylcholine in microdialysis samples: optimizing performance of platinum electrodes. J. Chromatogr. B., 692: 207–212. Chen, Y.F., Jaw, I., Shiao, M.S. and Tsai, T.H. (2005) Determination and pharmacokinetic analysis of salvianolic acid B in rat blood and bile by microdialysis and liquid chromatography. J. Chromatogr. A., 1088(1–2): 140–145. Cheng, F.C., Ho, Y.F., Hung, L.C., Chen, C.F. and Tsai, T.H. (2002) Determination and pharmacokinetic proﬁle of omeprazole in rat blood, brain and bile by microdialysis and high-performance liquid chromatography. J. Chromatogr. A., 949(1–2): 35–42. Cheng, F. and Kuo, J. (1995) High-performance liquid chromatographic analysis with electrochemical detection of biogenic amines using microbore columns. J. Chromatogr. B., 665: 1–13. Chervet, J.P., van Soest, R.E.J. and Salzmann, J.P., 1992. Recent advances in capillary liquid chromatography. Enhance detectability in bioanalysis. LCGC publication 28. Ciu, M., Feng, Y., McAdoo, D.J. and Willis, W.D. (1999) Periaqueductal gray stimulation-induced inhibition of nociceptive dorsal horn neurons in rats is associated with the release of norepinephrine, serotonin, and amino acids. J. Pharmacol. Exp. Ther., 289: 868–876. Clinckers, R., Smolders, I., Meurs, A., Ebinger, G. and Michotte, Y. (2004) Anticonvulsant action of endogenous

247 hippocampal dopamine and serotonin is independently mediated by D2 and 5HT1A receptors. J. Neurochem., 89(4): 834–843. Csoregi, E., Laurell, T., Katakis, I., Heller, A. and Gorton, P. (1995) On-line glucose monitoring by using microdialysis sampling and amperometric detection based on ‘wired’ glucose oxidase in carbon paste. Mikrochim. Acta., 121: 31–40. Damsma, G., Westerink, B.H.C., Imperato, A., Rollema, H., de Vries, J.B. and Horn, A.S. (1987) Automated brain dialysis of acetylcholine in freely moving rats: detection of basal acetylcholine. Life Sci., 41(7): 873–876. Davies, M.I., Cooper, J.D., Desmond, S.S., Lunte, C.E. and Lunte, S.M. (2000) Analytical considerations for microdialysis sampling. Adv. Drug. Deliv. Rev., 45: 169–188. Dawson, L.A., Organ, A.J., Winter, P., Lacroix, L.P., Shilliam, C.S., Heidbreder, C. and Shah, A.J. (2004) Rapid highthroughput assay for the measurement of amino acids from microdialysates and brain tissue using monolithic C-18bonded reversed-phase columns. J. Chromatgr. B., 807: 235–241. Deterding, L.J., Dix, K., Burka, L.T. and Tomer, K.B. (1992) On-line coupling of in vivo microdialysis with tandem mass spectrometry. Anal. Chem., 64: 2636–2641. Dong, Y., Wang, L., Shangguan, D., Zhao, R. and Liu, G. (2003) Improved method for the routine determination of acetylcholine and choline in brain microdialysate using horseradish peroxidase column as the immobilized enzyme reactor. J. Chromatogr. B., 788: 193–198. Emmett, M.R., Andren, P.E. and Caprioli, R.M. (1995) Speciﬁc molecular mass detection of endogenously released neuropeptides using in vivo microdialysis/mass spectrometry. J. Neurosci. Methods, 62: 141–147. Feenstra, M.G. and Botterblom, M.H. (1996) Rapid sampling of extracellular dopamine in the rat prefrontal cortex during food consumption, handling and exposure to novelty. Brain Res., 742(1–2): 17–24. Fuh, M.R., Haung, C.H., Wu, T.Y., Lin, S.L. and Pan, W.H. (2004) Determination of amphetamine in rat brain by in vivo microdialysis and ion-pairing liquid chromatography with electrospray tandem mass spectrometry. Rapid Commun. Mass Spectrom., 18(15): 1711–1714. Gobert, A., Billiras, R., Cistarelli, L. and Millan, M.J. (2004) Quantiﬁcation and pharmacological characterization of dialysate levels of noradrenaline in the striatum of freely-moving rats: release from adrenergic terminals and modulation by a2-autoreceptors. J. Neurosci. Methods, 140: 141–152. Gobert, A., Di Cara, B., Cistarelli, L. and Millan, M.J. (2003) Piribedil enhances frontocortical and hippocampal release of acetylcholine in freely moving rats by blockade of alpha 2A-adrenoceptors: a dialysis comparison to talipexole and quinelorane in the absence of acetylcholinesterase inhibitors. J. Pharmacol. Exp. Ther., 305(1): 338–346. Gobert, A., Rivet, J.M., Audinot, V., Newman-Tancredi, A., Cistarelli, L. and Millan, M.J. (1998) Simultaneous quantiﬁcation of serotonin, dopamine and noradrenaline levels in single frontal cortex dialysates of freely moving rats reveals a

complex pattern of reciprocal auto- and heteroreceptor-mediated control of release. Neuroscience, 84(2): 413–429. Greaney, M.D., Marshall, D.L., Bailey, B.A. and Acworth, I.N. (1993) Improved method for the routine analysis of acetylcholine release in vivo: quantitation in the presence and absence of esterase inhibitor. J. Chromatogr., 622(2): 125–135. Groenendaal, D., Blom-Roosmalen, M.C., Danhof, M. and Lange, E.C. (2005) High-performance liquid chromatography of nalbuphine, butorphanol and morphine in blood and brain microdialysate samples: application to pharmacokinetic/pharmacodynamic studies in rats. J. Chromatogr. B., 822(1–2): 230–237. Gupta, A., Jansson, B., Chatelain, P., Massingham, R. and Hammarlund-Udenaes, M. (2005) Quantitative determination of cetirizine enantiomers in guinea pig plasma, brain tissue and microdialysis samples using liquid chromatography/tandem mass spectrometry. Rapid Commun. Mass Spectrom., 19(12): 1749–1757. Haskins, W.E., Wang, Z., Watson, C.J., Rostand, R.R., Witowski, S.R., Powell, D.H. and Kennedy, R.T. (2001) Capillary LC-MS2 at attomole level for monitoring and discovering endogenous peptides in microdialysis samples collected in vivo. Anal. Chem., 73: 5005–5014. Haskins, W.E., Watson, C.J., Cellar, N.A., Powell, D.H. and Kennedy, R.T. (2004) Discovery and neurochemical screening of peptides in brain extracellular ﬂuid by chemical analysis of in vivo microdialysis samples. Anal. Chem., 76: 5523–5533. Hows, M.E.P., Lacroix, L., Heidbreder, C., Organ, A.J. and Shah, A.J. (2004) High-performance liquid chromatography/ tandem mass spectrometric assay for the simultaneous measurement of dopamine, norepinephrine, 5-hydroxytryptamine and cocaine in biological samples. J. Neurosci. Methods, 138: 123–132. Hows, M.E., Organ, A.J., Murray, S., Dawson, L.A., Foxton, R., Heidbreder, C., Hughes, Z.A., Lacroix, L. and Shah, A.J. (2002) High-performance liquid chromatography/tandem mass spectrometry assay for the rapid high sensitivity measurement of basal acetylcholine from microdialysates. J. Neurosci. Methods, 121(1): 33–39. Huang, S.M., Tsai, T.R., Yeh, P.H. and Tsai, T.H. (2005) Measurement of unbound ranitidine in blood and bile of anesthetized rats using microdialysis coupled to liquid chromatography and its pharmacokinetic application. J. Chromatogr. A., 1073(1–2): 297–302. Huang, T., Yang, L., Gitzen, J., Kissinger, P.T., Vreeke, M. and Heller, A. (1995) Detection of basal acetylcholine in rat brain microdialysate. J. Chromatogr. B., 670: 323–327. Ikegami, T. and Tanaka, N. (2004) Monolithic columns for high-efﬁciency HPLC separations. Curr. Opin. Chem. Biol., 8: 527–533. Jarrett, H.W., Cooksy, K.D., Ellis, B. and Anderson, J.M. (1986) The separation of o-phtalaldehyde derivatives of amino acids by reversed-phase chromatography on octylsilica columns. Anal. Biochem., 153: 189–198. Johansson, I.N., Wahkund, K.G. and Schill, G. (1978) Reversed phase ion-pair chromatography of drugs and related organic compounds. J. Chromatogr., 149: 281–296.

248 Kehr, J. and Ungerstedt, U. (1988) Fast HPLC estimation of g-aminobutyric acid in microdialysis perfusates: effect of nipecotic and 3-mercaptopropionic acids. J. Neurochem., 51: 1308–1310. Kehr, J., Dechent, P., Kato, T. and Ogren, S.O. (1998) Simultaneous determination of acetylcholine, choline and physostigmine in microdialysis samples from rat hippocampus by microbore liquid chromatography/electrochemistry on peroxidase redox polymer coated electrodes. J. Neurosci. Methods, 83(2): 143–150. Kennedy, R.T., Watson, C.J., Haskins, W.E., Powell, D.H. and Strecker, R.E. (2002) In vivo neurochemical monitoring by microdialysis and capillary separations. Curr. Opin. Chem. Biol., 6: 659–665. Kissinger, P.T. (1989) Biomedical applications of liquid chromatography–electrochemistry. Review. J. Chromatogr., 448(1): 31–52. Kissinger, P.T., Refshauge, C., Dreiling, R. and Adams, R.N. (1973) An electrochemical detector for liquid chromatography with picogram sensitivity. Anal. Lett., 6: 465–477. Klintenberg, R. and Andren, P.E. (2005) Altered extracellular striatal in vivo biotransformation of the opioid neuropeptide dynorphin A(1-17) in the unilateral 6-OHDA rat model of Parkinson’s disease. J. Mass Spectrom., 40: 261–270. Kucera, P. (Ed.). (1984) Microcolumn High Performance Liquid Chromatography. Elsevier, Amsterdam. Lanckmans, K., Clinckers, R., Van Eeckhaut, A., Sarre, S., Smolders, I. and Michotte, Y. (2005) Use of microbore LCMS/MS for the quantiﬁcation of oxcarbazepine and its active metabolite in rat brain microdialysis samples. J. Chromatogr. B., 831(1–2): 205–212. Leggas, M., Zhuang, Y., Welden, J., Self, Z., Waters, C.M. and Stewart, C.F. (2004) Microbore HPLC method with on-line microdialysis for measurement of topotecan lactone and carboxylate in murine CSF. J. Pharm. Sci., 93(9): 2294–2295. Lindroth, P. and Mopper, K. (1979) High performance liquid chromatographic determination of subpicomole amounts of amino acids by precolumn ﬂuorescence derivatisation with o-phtaldialdehyde. Anal. Chem., 51: 1667–1674. Maidment, N.T., Brumbaugh, D.R., Rudolph, V.D., Erdelyi, E. and Evans, C.J. (1989) Microdialysis of extracellular endogenous opioid peptides from rat brain in vivo. Neuroscience, 33: 549–557. Maidment, N.T., Siddall, B.J., Rudolph, V.R., Erdelyi, E. and Evans, C.J. (1991) Dual determination of extracellular cholecystokinin and neurotensin fragments in rat forebrain: microdialysis combined with a sequential multiple antigen radio-immunoassay. Neuroscience, 45(1): 81–93. Mathy, F.X., Vroman, B., Ntivunwa, D., De Winne, A.J., Verbeeck, R.K. and Preat, V. (2003) On-line determination of ﬂuconazole in blood and dermal rat microdialysates by microbore high-performance liquid chromatography. J. Chromatogr. B., 787(3): 323–331. Matuszewski, B.K., Constanzer, M.L. and Chavez-Eng, C.M. (2003) Strategies for the assessment of matrix effect in quantitative bioanalytical methods based on HPLC-MS/MS. Anal. Chem., 75: 3019–3030.

McKenzie, J.A.M., Watson, C.J., Rostand, R.D., German, I., Witowski, S.R. and Kennedy, R.T. (2002) Automated capillary liquid chromatography for simultaneous determination of neuroactive amines and amino acids. J. Chromatogr. A., 962: 105–115. McLaughlin, K.J., Faibuschevich, A.A. and Lunte, C.E. (2000) Microdialysis sampling with on-line microbore HPLC for the determination of tirapazamine and its reduced metabolites in rats. Analyst, 125(1): 105–110. Nydahl, K.S., Pierson, J., Nyberg, F., Caprioli, R.M. and Andren, P.E. (2003) In vivo processing of LVV-hemorphin-7 in rat brain and blood utilizing microdialysis combined with electrospray mass spectrometry. Rapid Commun. Mass Spectrom., 17(8): 838–844. Owens, H.M., Destache, C.J. and Dash, A.K. (1999) Simple liquid chromatographic method for the analysis of the blood brain barrier permeability characteristics of ceftriaxone in an experimental rabbit meningitis model. J. Chromatogr. B., 728(1): 97–105. Paez, X. and Hernandez, L. (2001) Biomedical applications of capillary electrophoresis with laser-induced ﬂuorescence detection. Biopharm. Drug Dispos., 22(7–8): 273–289. Plock, N. and Kloft, C. (2005) Microdialysis – theoretical background and recent implementation in applied lifesciences. Eur. J. Pharm. Sci., 25: 1–24. Potter, P.E., Meek, J.L. and Neff, N.H. (1983) Acetylcholine and choline in neuronal tissue measured by HPLC with electrochemical detection. J. Neurochem., 41(1): 188–194. Pudovkina, O.L., Kawahara, Y., de Vries, J. and Westerink, B.H.C. (2001) The release of noradrenaline in the locus coeruleus and prefrontal cortex studied with dual-probe microdialysis. Brain Res., 906(1–2): 38–45. Rea, K., Cremers, T.I.F.H and Westerink, B.H.C. (2005) HPLC conditions are critical for detection of GABA by microdialysis. J. Neurochem., 94: 672–679. Rowley, H.L., Martin, K.F. and Marsden, C.A. (1995) Determination of in vivo amino acid neurotransmitters by highperformance liquid chromatography with o-phtalaldehydesulphite derivatisation. J. Neurosci. Methods, 57: 93–99. Santiago, M. and Westerink, B.H.C. (1991) Characterization and pharmacological responsiveness of dopamine release recorded by microdialysis in the substantia nigra of conscious rats. J. Neurochem., 57: 738–747. Sarre, S., Michotte, Y., Herregodts, P., Deleu, D., De Klippel, N. and Ebinger, G. (1992a) HPLC with electrochemical detection for analysis of L-DOPA, catecholamines and their metabolites in rat brain dialysates. J. Chromatogr., 575: 207–212. Sarre, S., Marvin, C.A., Ebinger, G. and Michotte, Y. (1992b) Microbore liquid chromatography with dual electrochemical detection for the determination of serotonin and 5-hydroxyindoleacetic acid in rat brain dialysates. J. Chromatogr., 582: 29–34. Sarre, S., Thorre´, K., Smolders, I. and Michotte, Y., 1997. Microbore liquid chromatography analysis of monoamine transmitters. In: Rayne, R.C., (Ed.), Methods in Molecular

249 Biology, Vol. 72. Neurotransmitter Methods, Humana Press, Totowa, pp. 185–197. Sarre, S., Vandeneede, D., Ebinger, G. and Michotte, Y. (1998) Biotransformation of L-DOPA to dopamine in the substantia nigra of freely moving rats: effect of dopamine receptor agonists and antagonists. J. Neurochem., 70: 1730–1739. Sarre, S., Yuan, H., Jonkers, N., Van Hemelrijck, A., Ebinger, G. and Michotte, Y. (2004) In vivo characterization of somatodendritic dopamine release in the substantia nigra of 6-hydroxydopamine lesioned rats. J. Neurochem., 90: 29–39. Sauvinet, V., Parrot, S., Benturquia, N., Bravo-Moraton, E., Renaud, B. and Denoroy, L. (2003) In vivo simultaneous monitoring of gamma-aminobutyric acid, glutamate, and L-aspartate using brain microdialysis and capillary electrophoresis with laser-induced ﬂuorescence detection: analytical developments and in vitro/in vivo validations. Electrophoresis, 24(18): 3187–3196. Schmidt, R., Coste, O. and Geisslinger, G. (2005) LC-MS/MSanalysis of prostaglandin E2 and D2 in microdialysis samples of rats. J. Chromatogr. B., 826: 188–197. Scott, R.P.W. (1984) Small Bore Liquid Chromatography Columns: Their Properties and Uses. Wiley, New York, Toronto, Singapore. Schmidt, S. and Martinsson, S. (1998) In vivo/in vitro microdialysis coupled to capillary LC-LIF. Biomed. Chromatogr., 12: 109–110. Sharp, T., Zetterstro¨m, T., Series, H.G., Carlsson, A., Grahame-Smith, D.G. and Ungerstedt, U. (1987) HPLC-EC analysis of catechols and indoles in rat brain dialysates. Life Sci., 41: 869–872. Smith, S.E. and Sharp, T. (1994) Measurement of GABA in rat brain microdialysates using o-phtaldialdehyde-sulphite derivatization and high-performance liquid chromatography with electrochemical detection. J. Chromatogr., 652: 228–233. Smolders, I., Sarre, S., Ebinger, G. and Michotte, Y. (1995) The analysis of excitatory, inhibitory and other amino acids in rat brain dialysates using microbore liquid chromatography. J. Neurosci. Methods, 57: 47–53. Traunmuller, F., Gattringer, R., Zeitlinger, M.A., Graninger, W., Muller, M. and Joukhadar, C. (2005) Determination of telithromycin in human plasma and microdialysates by highperformance liquid chromatography. J. Chromatogr. B., 822(1–2): 133–136. Tsai, T. (2000) Separation methods used in the determination of choline and acetylcholine. J. Chromatogr. B., 747: 111–122. Tsai, T., Cham, T., Chen, K., Chen, C. and Tsai, T. (1996) Determination of acetylcholine by on-line microdialysis coupled with pre- and post-column enzyme reactors with electrochemical detection. J. Chromatogr. B., 678: 151–155. Tsai, T.H. and Liu, M.C. (2004) Determination of unbound theophylline in rat blood and brain by microdialysis and liquid chromatography. J. Chromatogr. A., 1032(1–2): 97–101. Uutela, P., Reinila¨, R., Piepponen, P., Ketola, R.A. and Kostiainen, R. (2005) Analysis of acetylcholine and choline in microdialysis samples by liquid chromatography/tandem

mass spectrometry. Rapid Commun. Mass Spectrom., 19: 2950–2956. Vahabzadeh, A. and Fillenz, M. (1994) Comparison of stressinduced changes in noradrenergic and serotonergic neurons in rat hippocampus using microdialysis. Eur. J. Neurosci., 6: 1205–1212. Van Belle, K., Verfaillie, I., Ebinger, G. and Michotte, Y. (1995) Liquid chromatographic assay using a microcolumn coupled to a U-shaped optical cell for high-sensitivity ultraviolet absorbance detection of oxcarbazepine and its major metabolite in microdialysates. J. Chromatogr. B., 672(1): 97–102. Van Hemelrijck, A., Sarre, S., Smolders, I. and Michotte, Y. (2005) Determination of amino acids associated with cerebral ischaemia in rat brain microdialysates using narrowbore liquid chromatography and ﬂuorescence detection. J. Neurosci. Methods, 144: 63–71. Volpe, G., Moscone, D., Compagnone, D. and Palleschi, G. (1995) In vivo continuous monitoring of L-lactate coupling subcutaneous microdialysis and an electrochemical biocell. Sensors Actuators B., 24–25: 138–141. Wages, S.A., Church, W.H. and Justice, J.B. Jr. (1986) Sampling considerations of on-line microbore liquid chromatography of brain dialysate. Anal. Chem., 58: 1649–1656. Westerink, B.H.C. (2000) Analysis of biogenic amines in microdialysates of the brain. Review. J. Chromatogr. B., 747: 21–32. Westerink, B.H.C., Damsma, G., Rollema, H., de Vries, J.B. and Horn, A.S. (1987) Scope and limitations of in vivo brain dialysis: a comparison of its application to several neurotransmitter systems. Life Sci., 41: 1763–1776. Wilson, S.R., Boix, F., Holm, A., Molander, P., Lundanes, E. and Greibrokk, T. (2005) Determination of bradykinin and arg-bradykinin in rat muscle tissue by microdialysis and capillary column-switching liquid chromatography with mass spectrometric detection. J. Sep. Sci., 28(14): 1751–1758. Xu, M., Nakamura, Y., Yamamoto, T., Natori, K., Irie, T., Utsumi, H. and Kato, T. (1991) Determination of basal acetylcholine release in vivo by rat brain dialysis with a Ushaped cannula: effect of SM-10888, a putative therapeutic drug for Alzheimer’s disease. Neurosci. Lett., 123(2): 179–182. Yamamoto, K., Sato, K., Chikuma, T. and Kato, T. (2004) A highly sensitive and stable detection of acetylcholine by HPLC-osmium-horseradish peroxidase redox polymer electrode coated on a gold radial ﬂow ring disk. Anal. Chim. Acta., 521: 209–213. Yang, Z., Huang, Y., Gan, G. and Sawchuk, R.J. (2005) Microdialysis evaluation of the brain distribution of stavudine following intranasal and intravenous administration to rats. J. Pharm. Sci., 94(7): 1577–1588. Yoshitake, T., Kehr, J., Todoroki, K., Nohta, H. and Yamaguchi, M. (2006) Derivatization chemistries for determination of serotonin, norepinephrine and dopamine in brain microdialysis samples by liquid chromatography with ﬂuorescence detection, Biomed. Chromatogr., 20(3): 267–281.

250 Yoshitake, T., Yoshitake, S., Fujino, K., Nohta, H., Yamaguchi, M. and Kehr, J. (2004) High sensitive liquid chromatographic method for the determination of neuronal release of serotonin, noradrenaline and dopamine monitored by microdialysis in rat prefrontal cortex. J. Neurosci. Methods, 140: 163–168. Zetterstro¨m, T., Sharp, T., Collin, A.K. and Ungerstedt, U. (1988) In vivo measurement of extracellular dopamine and DOPAC in rat striatum after various dopamine-releasing drugs; implications for the origin of extracellular DOPAC. Eur. J. Pharmacol., 148: 327–334. Zhang, M.-Y. and Beyer, C.E. (2006) Measurement of neurotransmitters from extracellular ﬂuid in brain by in vivo microdialysis and chromatography-mass spectrometry. J. Pharm. Biomed. Anal., 40: 492–499.

Zhang, S., Takeda, Y., Hagioka, S., Takata, K., Aoe, H., Nakatsuka, H., Yokoyama, M. and Morita, K. (2005) Measurement of GABA and glutamate in vivo levels with high sensitivity and frequency. Brain Res. Prot., 14: 61–66. Zhu, Y., Wong, P.S.H., Cregor, M., Gitzen, J., Coury, L.A. and Kissinger, P.T. (2000) In vivo microdialysis and reverse phase ion pair liquid chromatography/tandem mass spectrometry for the determination and identiﬁcation of acetylcholine and related compounds in rat brain. Rapid. Commun. Mass Spectrom., 14: 1695–1700. Zilkha, E., Obrenovitch, T.P., Koshy, A., Kusakabe, H. and Bennetto, H.P. (1995) Extracellular glutamate: on-line monitoring using microdialysis coupled to enzyme-amperometric analysis. J. Neurosci. Methods, 60: 1–9.

CHAPTER 3.2

Microdialysis coupled with liquid chromatography/mass spectrometry Malin Andersson and Richard M. Caprioli Department of Biochemistry, Mass Spectrometry Research Center, Vanderbilt University School of Medicine, Nashville, TN, USA

Abstract: This chapter describes basic principles and protocols in the ﬁeld of microdialysis combined with liquid chromatography/mass spectrometry (LC/MS). We describe currently used separation methods coupled with MS and its application to in vivo microdialysis in the central nervous system. Newly developed strategies and analytical methods are discussed for the study of drug pharmacokinetics, neuropeptide metabolism, and treatment-induced or disease-speciﬁc effects on the ‘‘neuropeptidome’’. used for the determination of sequences with modest demands on sample quantity and purity. Sequences obtained by MS can then be used to search protein databases for identiﬁcation with high conﬁdence. A schematic overview of a microdialysis experiment followed by LC/MS and MS/MS of a selected peak is illustrated in Fig. 1. The molecular speciﬁcity and ease of coupling MS with highefﬁciency separation techniques has allowed it to become a powerful tool for the rapid and sensitive detection of compounds in complex mixtures. The microdialysate typically contains low-molecular weight species such as neuropeptides that are pre-concentrated and desalted on a reversed-phase LC column and subsequently eluted into the mass spectrometer by an increasing concentration of organic modiﬁer. Commonly, electrospray ionization (ESI) is used to form protonated molecular species that can be measured by the MS analyzer at a given mass-to-charge ratio or ratios (m/z) in cases of multiply charged ions. A single peak may be selected for MS/MS analysis by which it is fragmented within the instrument and the resulting fragments used to verify or establish its sequence. This article reviews some of the LC/MS technology available for the analysis of neurochemicals

I. Introduction Analysis of microdialysate samples with liquid chromatography/mass spectrometry (LC/MS) offers many advantages over traditional analytical techniques such as radioimmunoassays (RIA), enzyme-linked immunosorbent assay (ELISA), and high-performance liquid chromatography (HPLC). MS has the capability of detecting virtually any compound able to cross the microdialysis membrane, whereas RIA, ELISA, and HPLC rely on targeted analyses and speciﬁc chemistries for identiﬁcations. In addition, whereas proteolytic cleavage and post-translational modiﬁcations (PTMs) of neuropeptides are common mechanisms for the ﬁne-tuning regulation of biological activity, these events may be undetectable using immunological methods that target speciﬁc antigenic epitopes. In the past, neuropeptide and protein discovery was based on known biological activities and chemical properties, and novel identiﬁcation was obtained by Edman degradation sequencing following extensive sample collection and puriﬁcation. Today, tandem-MS (MS/MS) is routinely Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

251

DOI: 10.1016/S1569-7339(06)16014-2 Copyright 2007 Elsevier B.V. All rights reserved

252

Fig. 1. Schematic illustration of a typical microdialysis experiment coupled to liquid chromatography/mass spectrometry (LC/MS). The microdialysate typically contains low-molecular weight species such as neuropeptides that is pre-concentrated and desalted on a reversed-phase column and eluted into the mass spectrometer by an increasing concentration of organic modiﬁer. The mass spectrometer can analyze the molecular weight species or can establish and verify structure in the MS/MS mode.

and their changing concentrations during physiological and pathophysiological events. The versatility of MS-based approaches makes it a suitable analytical tool for many microdialysis applications, such as pharmacokinetic studies of centrally acting drugs, neuropeptide metabolism, or discoverybased neuropeptidomics. In addition, the recent

development of micro- and nanoscale LC/MS has enabled signiﬁcantly higher temporal resolution using lower dialysis volumes. Newly developed strategies and analytical methods are also brieﬂy described, since they are expected to aid future studies of treatment-induced or disease-state effects on the neuropeptidome.

253

II. Neuropeptides and mass spectrometry Brain microdialysates typically contain relatively small molecules, such as neuropeptides and neurotransmitters (amino acids and monoamines; Davies et al., 2000). One of the most challenging aspects of neurochemistry is the detection of endogenous neuropeptides due to their low in vivo concentrations ranging from pico- to sub-femtomolar levels that are nearly six orders of magnitude lower than that of neurotransmitters (Kendrick, 1990; Strand, 2003). The temporal resolution is also limited by the sensitivity of the analytical technique used, that is the time needed to collect enough analyte to be detected. However, long collection periods are generally undesirable because neurochemical levels typically change rapidly (milliseconds to minutes). The development of an analytical technology and strategy combining nano-LC with micro-electrospray ionization (micro-ESI) MS has enabled the monitoring and quantiﬁcation of endogenous neuropeptides at attomol levels in microdialysate volumes as small as a few microliters (Andren et al., 1994; Emmett and Caprioli, 1994; Andren and Caprioli, 1995; Emmett et al., 1995). A more detailed description of the micro-ESI/LC/MS technology is provided later in this article. Neuropeptides are often activated from preprohormones through tissue and cell-speciﬁc peptidases and subjected to posttranslational modiﬁcations (PTMs) prior to release (Strand, 2003). In general, over 200 PTMs have been reported to date and known PTMs of neuropeptides include glycosylation, C-terminal amidation, N-terminal acetylation, phosphorylation, and sulfation (Baggerman et al., 2004). MS-based approaches using LC/MS and MS/MS are well-suited for the detection and identiﬁcation of previously unknown PTMs, using for example pattern recognition algorithms such as the scoring algorithm for mass spectral analysis (SALSA; Hansen et al., 2001). One example of a PTM detected by LC–MS/MS under physiological conditions is the recently discovered phosphorylation of serine-48 of the cocaine and amphetamine-regulated transcript (CART) peptide in rat brain tissue (Svensson et al., 2003). The CART peptides are neuropeptides involved in the brain’s reward/reinforcement

circuitry and are conceived to be putative targets for pharmacological drug-development for drug addiction, anxiety, and obesity (Hokfelt et al., 2003). Microdialysis allows for selective metabolic ‘‘snap-shots’’ of neuropeptides and other small molecules over time because the microdialysis membrane excludes larger proteins such as proteases and peptidases. The diffusion of large molecules through the dialysis membrane becomes progressively impaired as molecular weight increases, and the effective molecular cut-off of a 30-kDa microdialysis membrane is typically below 10 kDa in vivo (Stenken, 1999). Microdialysis combined with LC/MS was ﬁrst used for the metabolic mapping of region-speciﬁc degradation pathways in vivo by infusing substance P (SP) peptides into the extracellular space of the striatum while simultaneously collecting metabolites (Andren and Caprioli, 1995). More importantly, the relative abundance of SP metabolic fragments could be quantiﬁed and the pattern of fragmentation found indicated multiple degradation enzymes. This study highlights the importance of an unambiguous analysis of metabolic pathways, especially as some of the fragments found (e.g., SP1–7) are known to share some biological effects of the intact SP but oppose others (Kreeger and Larson, 1993; Zhou et al., 2000). The metabolic conversion of dynorphins has also been extensively investigated using microdialysis and MS-based approaches. The dynorphins are expressed in many brain regions and are involved in various neurological processes such as nociception and drug addiction. Dynorphins have in recent years rendered much attention as putative targets for the treatment of Parkinson’s disease and are also suspected to cause common adverse effects of L-DOPA pharmacotherapy. A disease-state speciﬁc alteration in the metabolic processing of exogenous dynorphin A has been shown in the striatum of a rat model of Parkinson’s disease using microdialysis combined with LC/MS (Klintenberg and Andren, 2005). The different metabolic fragments found in the parkinsonian but not in the intact striatum deserve much attention since they are well in line with ﬁndings of disturbed opioid transmission in both animal models and patients with

254

Parkinson’s disease (e.g., Nisbet et al., 1995; Johansson et al., 2001). MS has also been used for detection of the conversion of synthetic dynorphins having afﬁnity to the kappa-opioid receptor into enkephalins with afﬁnity to the deltaopioid receptor, thus targeting different opioid signaling systems (e.g., Sandin et al., 1997; Prokai et al., 1998; Reed et al., 2003). Furthermore, it has been shown by MS analysis that the conversion of dynorphin B is region-speciﬁc and the local enzymatic milieu results in different enkephalinergic metabolites in the striatum, substantia nigra, and hippocampus (Sandin et al., 1997). III. Liquid chromatography The coupling of LC with MS has centered on chromatographic modes using volatile buffers with reversed-phased liquid chromatography (RP-LC) because of compatibility with soft ionization methods for MS analysis. In addition, the perfusion medium for microdialysis is typically a buffer, such as Ringer’s solution or artiﬁcial cerebrospinal ﬂuid (aCSF), containing millimolar levels of inorganic salts, and so the use of RP-LC is often preferred as the procedure includes removal of salts that interferes with MS. Examples of the common chromatographic modes are reversed-phase (separation of analytes according to their hydrophobicity), cationic-exchange (separation mainly by charge), size-exclusion (separation by molecular weight), capillary electrophoresis (CE, separation by size and charge), and afﬁnity (separation through interactions). Here, we restrict ourselves to the description of a few chromatographic modes that are important for the MS analysis of brain microdialysates. A more detailed discussion of the characteristics and mechanisms of HPLC and CE are discussed elsewhere in this book and in several reviews (Tomer et al., 1994; Kennedy et al., 2002; Wang et al., 2005). III.A. Reversed-phase liquid chromatography Some aspects of RP-LC pertain speciﬁcally for the analysis of microdialysates by MS. In RP-LC, hydrocarbons of different lengths (typically C4, C8,

or C18) are chemically bonded to the silica matrix and provide the hydrophobic interaction sites for separation. C18 columns are usually preferred for the separation of peptides smaller than 5 kDa and C8 columns for peptides and small proteins below 10 kDa, whereas C4 columns are generally recommended for larger proteins. Testing different columns is the only practical way to determinate the optimum separation for the analyte in question, although for samples obtained by microdialysis C8 and C18 resins are generally recommended. The most commonly used mobile phase in RPLC is the organic solvent acetonitrile (ACN), a solvent that is highly volatile and easily removed from the eluate. A gradient of increasing percentage organic solvent elutes the peptides from the column, resulting in higher resolution separation relative to isocratic elutions. An ion-pairing reagent may be added to the mobile phase to maximize separation capacity by increasing the hydrophobicity of molecules. In addition, the ion-pairing agents can also aid in protein unfolding and denaturation that can result in improved resolution. Triﬂuoroacetic acid (TFA) is commonly used as an ion-pairing reagent in reversed phase chromatography, often at concentrations from 0.1% for small peptides up to 0.5% for solubilizing larger proteins. However, TFA may form strong complexes with many proteins and peptides and thus cause a loss of detection sensitivity of the mass spectrometer. A number of alternative ionpairing agents have been used, ranging from bases such as morpholine, to salts like perchlorate and ammonium sulfate, to acids such as perﬂuorinated carboxylic acids (e.g., HFBA), formic acid and acetic acid. In general, formic acid and acetic acid yield a higher MS response compared with TFA, however, the gain in MS sensitivity is accompanied with lower separation performance (Garcia, 2005).

III.B. Ion-exchange, size-exclusion, and affinity chromatography Ion-exchange chromatography, especially strong cationic-exchange resins that separates molecules

255

according to their net positive charge, has been widely used for separations of peptides and proteins in various biological ﬂuids. However, the high salt concentrations commonly used in the mobile phase to elute the analytes often causes ionization suppression along with other undesirable effects. Eluents containing organic salts such as ammonium formate or formic acid in ACN have been employed to avoid these problems (Dear et al., 2000; Loughlin et al., 2001). This approach has been successfully used for the detection of basal levels of acetylcholine in microdialysates from rat prefrontal cortex (Hows et al., 2002). Size-exclusion chromatography separates analytes according to their size because the stationary phase contains porous materials with size-deﬁned cavities that only allows molecules smaller than the pore size to pass through the column. This method has limited resolving power, but has been applied to the analysis of neuropeptides in brain tissue (Nylander et al., 1995). Afﬁnity chromatography is based on the immobilization of a receptor, antibody, or ligand, through which a continuous infusion of the sample through the column allows the elution time to be correlated with the binding afﬁnity of analyte to ligand. Afﬁnity chromatography can be used in combination with MS for screening and discovery of novel interactions, for example, the screening of oligosaccharides against an antibody, peptides against lectin, and an enzyme inhibitor library against protease ligands (Schriemer and Hindsgaul, 1998; Zhang et al., 2001a, b; Lee and Lee, 2004). The addition of a buffer (e.g., 2 mM ammonium acetate and 10% v/v elution buffer in ACN) to the eluent is required for rendering a solution compatible with ionization of analyte by MS. Afﬁnity chromatography has so far not been used in brain microdialysis studies; however, it is conceivable that immobilized receptors may be used for future screening of neuroactive molecules in pathophysiological studies.

III.C. Multidimensional separations Multidimensional LC separations are increasingly favored and have been shown to increase the

separation resolution of complex samples compared with single-dimension chromatography (Giddings, 1984). Multidimensional approaches also allow for the use of salt-containing mobile phases in the ﬁrst dimension, taking advantage of the ability of a reversed-phase LC to desalt the sample in the second dimension. The combined separation mechanisms should be orthogonal to each other in that the separations are based on different mechanisms. This demands that the sampling frequency of the second dimension needs to be greater than the peaks’ widths (in time) emerging from the ﬁrst-dimensional separation. In an ideal orthogonal separation, the resulting peak capacity is the product of both individual separation dimensions. To explore peptides in complex mixtures, a number of approaches combining two or more dimensions of separations have been employed both off-line and on-line. Coupled on-line, the separated peaks have to be continuously fed into the second dimension. For example, a common combination includes a size-exclusion column and a reversed-phase column, however the ﬂow rates needed are different such that separation in the ﬁrst dimension is relatively slow whereas separation with RP-LC can be much higher. This can be overcome in several ways, for example by adding a preconcentration step between the two systems (Clarke et al., 2001) or the use of two different RP-LC columns operating in alternating fashion where one column is loaded with the efﬂuent of the ﬁrst dimension while the second column is running the separation of a previously loaded sample (Opiteck et al., 1997, 1998). Capillary electrophoresis is well-suited for online coupling with LC and is often used in the second dimension of separation because of the fast separation process, ranging from seconds to minutes (Bushey and Jorgenson, 1990; Tong et al., 1999; Kennedy et al., 2002). However, maintaining high sensitivity without peak broadening caused by the introduction of a dead volume in the interface can be troublesome. Interface designs for coupling ﬁrst and second dimensions predominantly include transverse ﬂow gating interfaces to better control sample injections (Lemmo and Jorgenson, 1993).

256

Strong cationic exchange chromatography has also been combined with RP-LC, taking advantage of the desalting washes in the second dimension. This particular separation followed by MS and MS/MS analysis is referred to as the multidimensional protein identiﬁcation technique (MudPIT; Washburn et al., 2001) and has so far been mostly used in ‘‘shot-gun’’ proteomics for proteolytically cleaved protein mixtures.

IV. Mass spectrometry For an analyte to be analyzed in a mass spectrometer it has to be ionized and vaporized into a gas phase. Two ionization methods often used for biological molecules are ESI (Fenn et al., 1989) and matrix-assisted laser desorption/ionization (MALDI; Karas and Hillenkamp, 1988; Hillenkamp et al., 1991). These two ionization technologies have revolutionized the ﬁeld of peptide and protein analysis. Electrospray ionization is based on the production of a ﬁne spray of a volatile solvent that contains the analyte of interest. The spray is generated by electrically charging the liquid to a high voltage producing charged droplets (Fig. 2A) with this charge or charges ultimately ending up on the analyte. Two principle ionization mechanisms have been proposed; the charge residue model (CRM) whereby solvent evaporates and leaves behind a net positively charged molecule in gas phase, and the ion evaporation model (IEM) whereby an ion is formed in the droplet and evaporates from the droplet into gas phase. In fact, both processes may coexist, with that of the CRM model dominating for ions of greater mass, and that of the IEM dominating in the case of small surface-active molecules (Wilm and Mann, 1996; De la Mora, 2000; Kebarle, 2000). Typically, electrospray ionization results in multiply charged ions from proteins and peptides, as they contain many proton-accepting sites. The several charged forms of a molecule will be detected by the mass spectrometer as an envelope of peaks at individual m/z values of the different charge states. A charge-deconvoluting algorithm is then used to derive the molecular weight of the

peptide or protein. Although the process is efﬁcient, the presence of multiply charged ions can have the effect of increasing the complexity of mass spectra, making interpretation more difﬁcult for complex mixtures. To interface capillary LC with low-ﬂow rate ESI-MS, there are several factors to consider, such as solvent properties, pressure, and ﬂow rate. Typical low-ﬂow rate micro-ESI sources are operated between 20 and 500 nL/s and this reduced ﬂow rate results in increased efﬁciency of the transfer of peptide ions from solution to mass analyzer without dilution of the analyte (Emmett and Caprioli, 1994). In addition, the low-ﬂow rate has been shown to increase the stability of spray formation and the applied voltage to the emitter is sufﬁcient to generate a spray, thus reducing the need of a sheath gas to stabilize the spray and addition of heat for evaporation of solvent. The need of handling small sample volumes for on-line coupling of LC with ESI-MS at low-ﬂow rates has spurred the development of capillary and nanoscale LC systems using microbore, nanobore, cap-LC, and nano-LC technology (Rapp and Tallarek, 2003). For microdialysis experiments where the sample volumes are generally kept as low as possible, one of the most successful advances were obtained by packing a microspray emitter with a C18 resin, thus combining micro-ESI with high-efﬁciency capillary LC within a single device (Fig. 2B) (Andren and Caprioli, 1994; Emmett and Caprioli, 1994). Brieﬂy, the construction of a microcapillary ESI tip tapered to about 5 mm to hold the C18-resin avoids the use of frits and minimizes dead volumes that otherwise would cause peak broadening. This is now commonly referred to as micro-ESI/LC, and has been shown to increase the limits of neuropeptide detection in microdialysis samples to atto- and zeptomoles in ml volumes. This system has further evolved with the introduction of solvent gradients to allow for peptide separation (Emmett et al., 1998; Haskins et al., 2001, 2004; Baseski et al., 2005). In addition, an increase in analytical robustness has been demonstrated by integrating a microcapillary preolumn with the micro-ESI/LC device (Yi et al., 2003). A variant of the micro-ESI/LC containing two different stationary phases was developed for

257

Fig. 2. Schematic illustrations of the principles of soft ionization and common mass analyzers. (A) The electrospray ionization process starts with the generation of a ﬁne spray containing the peptides and proteins, and as the droplets are transported toward the mass analyzer the analyte molecules are ionized. (B) The micro-electrospray ionization (micro-ESI)/LC device is a pulled microcapillary ESI emitter that is packed with a reversed-phase C18 resin. On increasing concentrations of organic solvent, the analyte is displaced from the column, dissolved, and ionized for detection in the mass analyzer. (C) A quadrupole mass analyzer transmits ions with a speciﬁc mass-to-charge (m/z) ratio along the axis of the four rods and a full MS scan is obtained by sweeping a mass range over time. (D) A MALDI-TOF mass spectrometer, where a sample is mixed with matrix and deposited on a target plate placed in the ion source. A laser pulse vaporizes the analyte together with matrix and the excited matrix ionizes the analyte in the plume. An accelerating potential is applied, the ions travel in a ﬁeld-free ﬂight tube and are separated according to m/z.

multidimensional chromatography of small volumes of trypsinized protein samples (Gatlin et al., 1998; Link et al., 1999). In this system, the pulled microcapillary is packed with a reversed-phase C18 resin and a second section containing strong cationic exchange particles. The peptide mix is loaded onto the strong cationic exchange column and an iterative process of increasing organic salt concentrations is used to displace the peptides. Each step-wise increase in salt concentration cause discrete peptide fractions to be eluted onto the RP column, the contaminating salts and buffers are washed away, and the peptides eluted into the mass spectrometer using a gradient of increasing organic solvent. Following re-equilibration of the RP column, another peptide fraction is displaced from the strong cationic exchange column by an increase in salt concentration and the process is repeated until total analysis of the original sample is completed.

Liquid chromatography has traditionally been coupled with three types of mass spectrometers; triple quadrupoles, ion traps, and dual-quadrupole-TOF (Qq-TOF) since both separation and mass analyzers are amenable to ESI. Mass analyzers all directly or indirectly measure gas phase ions according to their m/z. For example, the quadrupole mass analyzer consists of four rods or poles in parallel that allows for the selection of an ion of a particular m/z (Fig. 2C). Mass selection is dependent on the stability of the trajectories of ions transiting an oscillating electric ﬁelds applied to the rods. Only ions with a stable trajectory (i.e., the selected m/z) will pass through the quadrupole to the detector, whereas ions of lesser or greater mass will fail to pass through to hit the detector. By sweeping the electrical ﬁeld on the rods over a period of time (usually 500 Da/s), ions of different m/z are sequentially selected and thus a full-scan mass spectrum obtained.

258

Matrix-assisted laser desorption/ionization (Karas and Hillenkamp, 1988; Hillenkamp et al., 1991) has also been used for off-line LC/MS analysis of microdialysates. The analyte is imbedded in a solid matrix that absorbs energy from ultraviolet laser irradiation (Fig. 2D). The matrix along with the analyte is vaporized and in the process, the analyte is protonated. In general, for proteins and peptides, single-positively charged ions are formed and detected at their m/z. Matrices commonly used are 2,5-dihydroxybenzoic acid (DHB), 3,5dimethoxy-4-hydroxycinnamic acid (sinapinic acid), and alpha-cyano-4-hydroxycinnamic acid (a-CHCA). The different matrices have distinct characteristics, for example, sinapinic acid is suitable for the analysis of peptides and proteins over 2 kDa whereas a-CHCA and DHB are more efﬁcient for lower molecular weight compounds and peptides. MALDI is somewhat tolerant to salt contaminations relative to ESI, although traces of ionic detergents or involatile liquid additives (e.g., glycerol or DMSO) can be detrimental to the analysis. The time-of-ﬂight (TOF) mass analyzer requires ions to be produced in bundles and is well-suited for pulsed ion sources utilizing lasers. TOF ion analyzers measure the time it takes for an ion to travel a speciﬁc distance in a ﬁeld-free ﬂight tube (Fig. 2D) and are able to detect ions over 100 kDa. The sensitivity of MALDI-TOF instruments, using delayed extraction and reﬂector technology, is in the attomole to low-femtomole range, with a mass accuracy of 10–50 ppm (Vestal et al., 1995; Russell and Edmondson, 1997). The off-line nature of combining nano-LC with MALDI MS/ MS has recently been shown to have advantages, for example, since the sample consumption is minimal, the sample may be re-analyzed multiple times if needed (Mirgorodskaya et al., 2005). Several studies have successfully combined microdialysis with LC–MALDI MS. For example, microdialysis was used both to infuse the preproenkephalin product peptide E and to collect its metabolites from the brain extracellular space (Zhang et al., 1999). A solid-phase pre-concentration capillary electrophoresis (spPC-CE) separation step was then applied and the eluate containing the MALDI matrix was traced across a target plate and analyzed. Utilizing this

approach, an impressive number of over 75 metabolic fragments of peptide E were observed.

V. Tandem-mass spectrometry for the monitoring of drugs and neuropeptides In general MS/MS refers to methods involving at least two stages of MS in conjunction with either a dissociation process, for example collision-induced dissociation (CID) in a neutral gas such as argon or helium, or a chemical reaction that produces a change in the mass of the precursor or parent ion. The masses of the fragments are subsequently recorded and speciﬁc molecular identity of the precursor can be determined. For example, a peptide sequence can be deduced or PTMs can be studied. One type of mass spectrometer that can perform MS/MS experiments is the triple quadrupole analyzer consisting of three quadrupoles in sequence (Q1, q2, and Q3), where Q1 and Q3 are used as mass analyzers and q2 serves as a collision cell. It is possible to increase the number of fragmentation events by either coupling several mass spectrometers sequentially or by using an ion storage device such as an ion trap and perform sequential MS over time. This allows for multiple dissociation events to be performed in an MSn experiment, where n is the number of generations of ions analyzed. In this process, the number of transmitted ions decreases with each subsequent dissociation event, however the chemical noise may decrease even more, so that the signal-to-noise ratio (SNR) of the molecule of interest increases and thus the level of detection or sensitivity can be signiﬁcantly enhanced (Noble, 1995). Several different scan modes are commonly used in MS/MS experiments. The product scan (sometimes also referred to as a daughter scan) consists of selecting a precursor ion and detecting all the product ions emerging from a fragmentation event. It is also possible to select a speciﬁc product ion and determine all possible precursor ions in a precursor ion scan. This type of scan can only be performed with mass spectrometers such as the triple quadrupole, because the second mass analyzer (Q3) has to be focused on the selected daughter ion while scanning all possible parent ions using the

259

ﬁrst mass analyzer. Similarly, neutral loss scans can be performed where a constant off-set between the ﬁrst and the second mass analyzer enables the detection of all the fragments produced by the loss of that mass. This can be particularly useful in detecting labile PTMs such as phosphorylation of threonine and serine, where the ionization process may cause an elimination of the phosphoric acid and thus a distinct loss of 98 Da compared with the intact mass (Giorgianni et al., 2004). Selected reaction monitoring (SRM) consists of monitoring a speciﬁc collision-induced fragmentation reaction. For this type of scan, both mass analyzers are focused on selected masses. The absence of scanning allows the measurement of the precursor and product ions over longer times and increases the sensitivity for the selected ions. For example, SRM of cocaine will monitor the transition of the parent ion (m/z 304) to the fragment ion with an observed m/z of 182 (Fig. 3A and B). The observed product ion spectra can be plotted as a chromatogram over time, allowing for sensitive and accurate quantiﬁcation (Fig. 3C). The use of SRM greatly reduces the levels of chemical noise in the MS/MS spectrum, thus enhancing the limits of detection. In addition, multiple reactions monitoring (MRM) can be performed sequentially, typically allowing for up to four different transitions to be detected in the same sample. Different measurements can be used to make a relative

quantitation of the ion of interest, for example peak height, area under the peak, or signalto-noise levels of the peak of interest. In a study of basal and potassium-stimulated release of endogenous neurotensin, an external standard curve was generated and displayed linearity over a range from 40 amol to 1.23 fmol levels of neurotensin, with a linear coefﬁcient of >0.99 and a mean variation of less than 17% (Andren and Caprioli, 1999). In this study, 20 scans were averaged over the apex of the fragment ion peak and the SNR was used to estimate the amount of neurotensin. The SRM technique has gained popularity in recent years and has been extensively used for the analysis of brain microdialysates to monitor the pharmacokinetics of drugs, for example the GABA-B antagonist APBP (Andren et al., 1998), cocaine (Chen et al., 2005), ecstasy (Jones et al, 2005), amphetamine (Fuh et al., 2004), the antiepileptic drug oxcarbazepine (Lanckmans et al., 2006), and the mu-opioid antagonist remifentanil (Crespo et al., 2005). SRM has also been used to monitor and quantify the levels of basal and stimulus-induced release of endogenous neuropeptides, for example substance P, neurotensin, and enkephalin (Andren and Caprioli, 1995, 1999; Emmett et al., 1995; Haskins et al., 2001, 2004; Baseski et al., 2005) and neurotransmitters such as acetylcholine (Hows et al., 2002, 2004; Uutela et al., 2005).

Fig. 3. (A) The speciﬁcity of selected reaction monitoring (SRM) is illustrated for their MS/MS analysis of cocaine. (B) The m/z of the intact molecule at m/z 304.36 is selected for fragmentation and a major fragment is observed with m/z of 182.12 as shown in the product ion mass spectrum. (C) A SRM chromatogram can be plotted since the compound is entering the mass spectrometer by a chromatographic process, and the peak produced represents the relative intensity of the fragment ion over time.

260

Fig. 4. (A) Fragmentation of a theoretical peptide and the corresponding nomenclature of the fragment ions as proposed by Roepstorff (1984). (B) A typical peptide MS/MS spectrum.

VI. Neuropeptide identiﬁcation Tandem-mass spectrometry can be used for identiﬁcation of peptides and proteins by the detection of fragments produced by cleavage of a bond in the peptide chain (Fig. 4A and B). The cleavages mainly occur between three types of bonds; Ca-C, C-N, and N-Ca, yielding ion fragments depending on where the positive charge is retained. If the N-terminal fragment is positively charged the fragments are termed an, bn, and cn, respectively. When the charge resides on the C-terminal fragment, the fragments are termed xn, yn, and zn, the subscript ‘‘n’’ representing the number of amino acids contained in the fragment (Fig. 4A; Roepstorff, 1984). The resulting MS/MS spectra is the summation of these processes and contains sufﬁcient information from which the sequence can be deduced (Fig. 4B). In a recent study, 1,500 endogenous peptides were detected in 1 mg brain tissue using a nanoliter-ﬂow LC coupled with ESI Qq-TOF MS (Skold et al., 2002), of which about 10% of the peaks were selected for MS/MS analysis and were subsequently identiﬁed as endogenous peptides, for example stathmins, cyclophilin, prosomatostatin, b-actin, a-synuclein, NADH dehydrogenases, and cytochrome C oxidase VIa-L. Tandem-mass spectrometry analysis on intact high-molecular weight species (e.g., >4 kDa) in general does not produce sufﬁcient fragmentation for sequence analysis and so enzymatic

digestion of proteins using proteases such as trypsin or chymotrypsin is necessary. For example, trypsin speciﬁcally cleaves proteins on the Cterminal side of the amino acids lysine or arginine. The peptide mixture from protease digestion can be subsequently analyzed by LC/MS to record the masses of the tryptic fragments in a type of peptide tryptic map that enables comparison with a database of in silica trypsin-digested proteins. This is called peptide mass ﬁngerprinting and usually allows for a positive identiﬁcation of a protein using 4–6 tryptic fragments with individual masses ranging from 700 to 3,000 Da, determined with an accuracy of 10 ppm (Mann et al., 1993; Pappin et al., 1993; Yates et al., 1993). LC/MS/MS can also be employed for analysis of protease digests in a data-dependent scan mode whereby the instrument is set to ﬁrst collect full-scan mass spectra and second MS/MS spectra by selecting ions of major intensity for subsequent fragmentation and analysis (Mayya et al., 2005). This experiment can be automated to a high degree, including the database searches for both MS/MS data and MS data of tryptic fragments. Discovery and identiﬁcation of neuropeptides in brain dialysates poses several challenges, as many neuropeptides are too low in molecular weight for a tryptic digestion to yield informative fragments and therefore direct sequencing using MS/MS has to be employed. To reduce the number of false positive identiﬁcations of peptides, a number of

261

different software packages may be used in parallel, such as Sequest, Mascot, and Luteﬁsk. Haskins et al. (2004) applied this particular combination of software to analyze potassium-induced release of neuropeptides in striatal microdialysis samples. Data-dependent scanning was used to acquire MS/ MS spectra from samples collected during basal and K+-stimulated conditions. A subsequent subtractive analysis revealed more than 800 MS/MS spectra that were only observed during depolarization. These MS/MS spectra were correlated to sequences using Sequest and Mascot in parallel to search protein databases, followed by de novo sequencing conﬁrmation using the Luteﬁsk software. Both unknown and previously known peptides were identiﬁed, such as proenkephalin A, neurogranin, ﬁbrinogen alpha chain precursor, excitatory amino acid transporter A, and brain acidic membrane protein.

VII. In vivo microdialysis coupled on-line with LC–MS/MS It may be desirable to couple the microdialysis experiment directly on-line with the LC/MS analysis, as freeze thawing of microdialysis samples is known to cause degradation and further loss of peptides and proteins may occur by absorption to the surfaces of storage containers (Paulson et al., 2005). On-line and real-time monitoring using directly coupled microdialysis with MS/MS was ﬁrst demonstrated in 1990, in a study of penicillin pharmacokinetics after an intramuscular injection to a rat (Caprioli and Lin, 1990). The microdialysis probe was implanted into the jugular vein and blood was sampled for 10 min every 20 min over a total period of 4 h. This study also highlights the fact that LC-separation can be omitted at times if experimental conditions allow, as the microdialysate was directly ionized by fast atom bombardment (FAB) MS. Penicillin pharmacokinetics was accurately measured using this approach, as compared with known data obtained by others. One of the difﬁculties of coupling microdialysis with LC/MS on-line is the time required for separation of sample on a column versus the continuous ﬂow of the dialysate. However, the resolving

power of a mass spectrometer reduces the need for chromatographic resolution as shown by the SRM/MRM studies utilizing the micro-ESI/LC system (Andren and Caprioli, 1994, 1995, 1999; Emmett and Caprioli, 1994; Emmett et al., 1995, 1998; Andren et al., 1998). For on-line coupling of in vivo microdialysis to LC/MS, Haskins et al. (2001) used a micro-ESI/LC device and optimized the separation gradient to operate at low-ﬂow rates of 20 nL/min with a chromatographic peak width of 0.05 min. Using this approach, microdialysis samples (2 mL, 0.6 mL/min) were collected every 30 min for 3.5 h from the globus pallidus of anesthetized rats. Basal and K+-stimulated levels of Leu-enkephalin, Met-enkephalin, and an unknown endogenous peptide were monitored in MRM mode, and furthermore, the unknown peptide was identiﬁed as a novel enkephalinergic metabolite from both MS/MS and MS3 sequencing. Baseski et al. (2005) further developed this method and reduced the separation time to 4 min using a steeper elution gradient. This increased the sampling rate by about 30% compared with earlier studies. In addition, a direct comparison was made of SRM in MS/MS and MS3 mode for quantiﬁcation of Leu-enkephalin and Met-enkephalin. The variability of measurement (relative standard deviation) for a 60 pM enkephalin standard in MS3 mode was shown to be less than 5% compared with 20% in MS/MS mode, and subsequently MS3 analysis was used for measuring basal and potassium-stimulated release of enkephalins in rat brain microdialysates.

VIII. Future perspectives Several improvements in the ﬁelds of both chromatography and MS will aid future microdialysis applications. The need for higher temporal resolution to elucidate fast signaling events in the brain will take advantage of new columns designed to increase separation efﬁciency and reduce analysis times. One avenue that is currently under investigation includes reversed-phase silica monolith columns that are compatible with low-ﬂow rate electrospray ion sources and show great promise for high-throughput analysis using high pressure

262

and steep elution gradients (Rieux et al., 2006). Applying micro- or nanoscale separations and instrumentation can minimize inefﬁciencies in the analysis; such as the use of microchips that combines sample clean-up, separation, and ESI emitter into one single device (m-TAS; Manz et al., 1990). Another fast separation process that has gained recent interest is ion mobility spectrometry (IMS) and ﬁeld asymmetric waveform IMS (FAWIMS). The separation of ions depends on the mobility of an ion in an inert gas buffer and an applied electric ﬁeld, and although the peak-capacity is equal to that of standard LC systems, the separation process using IMS combined with FAWIMS is completed in a time span of milliseconds to seconds (Tang et al., 2005). The Fourier transform-ion cyclotron resonance (FT-ICR, FTMS) mass spectrometers have superior mass accuracy (low ppm) and high resolution (105–106), which may accelerate the analysis of exceedingly complex mixtures (Comisaro and Marshall, 1974; Amster, 1996; Marshall, 2000). In addition, the recent development of electron capture dissociation (ECD) to fragment precursor ions, shows great potential in the study of PTMs (Zubarev et al., 2000; Mann and Jensen, 2003). ECD is a mild fragmentation method that allows relatively labile modiﬁcations, such as sulfations and phosphorylation of serines, to be retained on the fragments and subsequently unambiguously assigned in the mass spectrum. Combined with the high-mass accuracy obtained by FTMS, it is thus possible to completely map the precise location of most if not all PTMs, as well as obtaining the sequence of small intact proteins. The coupling of the micro-ESI/LC system to an FT-ICR mass spectrometer has been successfully used for detection and identiﬁcation of neuropeptides in aCSF at a concentration of attomole/microliter levels (Emmett et al., 1998), thus demonstrating that the sensitivity of this approach could be applicable to brain microdialysate samples. Accurate determination of the intact molecular weight of a protein or a peptide is essential for its identiﬁcation; however, due to unknown mRNA editing and PTMs, the correct intact mass often is not sufﬁcient for positive identiﬁcation using current database information alone. A major

(R.M. Caprioli). disadvantage of most proteomic studies is that in each case the peptide or protein needs to be ‘‘re-sequenced’’ to determine its identity, thereby slowing analysis throughput considerably. Therefore, several strategies have been put forward to facilitate the rapid identiﬁcation of peaks in mass spectra. The accurate mass tags (AMT) approach is based the extremely high-mass accuracy that can be obtained using FTMS, thus increasing the probability for correct amino acid assignment. Using an in silica prediction of unique tryptic peptide masses for the entire proteome of a speciﬁc species, followed by a validation of a realworld tryptic digest with both FTMS and LC–MS/ MS makes it possible to create a library of validated AMTs. An unknown sample can then be interrogated against the library of AMT using the combination of both high-precision molecular weight measurement and the LC-retention time (Lipton et al., 2002; Smith et al., 2002). It is feasible that a similar strategy can be applied to the neuropeptidome. One of the more commonly used technologies in proteomics for relative quantiﬁcation is twodimensional difference gel electrophoresis (2-D DIGE), which involves the labeling of samples with CyDyes ﬂuorophores. Two samples are labeled with different dyes (e.g., Cy3 and Cy5, respectively) and a standard control mix of all samples is labeled with a third dye (e.g., Cy2). The labeled samples are combined and run on a single 2-D gel, allowing direct quantitative comparisons. However, this technique is not widely used for microdialysis samples for several reasons; ﬁrst, dialysates contain low concentrations of analytes; second, the mass range of analytes is usually below that of the effective lower molecular weight limits of gel-based approaches (5–10 kDa); and third, the low-throughput of this technique is not wellsuited to the many samples typically collected in microdialysis experiments. Quantitative proteomics using MS takes advantage of chemical derivatization of peptides and proteins in complex mixtures. Isotope-coded afﬁnity tag (ICAT) chemistry is an approach that utilizes a set of heavy and light isotope afﬁnity tags targeting cysteine residues to label two samples, respectively (Gygi et al., 1999). After tryptic

263

digestion, the samples are mixed and LC/MS is used for simultaneous detection of peptide pairs with the expected mass difference arising from the heavy and light isotopes present. Although the ICAT methodology contributes signiﬁcantly to quantitative proteomic analysis, it has limited application to neuropeptidomics because of the lack of cysteine residues in many neuropeptides. A variant of ICAT, ‘‘isobaric tags for relative and absolute quantitation’’ (iTRAQ; Ross et al., 2004), has the advantage that it targets all amines, i.e., the N-terminus and lysine side chain amine. The tags act as reporter groups and are detected only on fragmentation in MS/MS mode, and the design of four different isobaric tags allows for the relative quantiﬁcation of up to four samples to be interrogated simultaneously. Beyond these, a wide variety of chemical derivatization strategies has been developed for quantitative measurements of peptides and this will further complement MS-based analyses of the neuropeptidome (Ong and Mann, 2005). Metabolic labeling by incorporation of speciﬁc isotopes such as 15N provides a more universal labeling of most proteins and peptides and has been mostly employed in cell culture experiments (Washburn et al., 2002). Metabolic labeling of whole animals would be ideal for proteomic studies, and indeed, one rat has been fed 15N-enriched algae over an extended period of time to create a whole-body library of 15N-labeled proteins and peptides (Wu et al., 2004). A variant of this method that is especially well-suited for brain microdialysis experiments has been used to study glutamate metabolism and uptake into glia (Kanamori et al., 2003). In this study, 13C-enriched glucose was infused through reverse microdialysis into the brain, and subsequently a time-course analysis of glutamate synthesis and its metabolism was monitored in vivo. Brain microdialysis and LC/MS strategies will certainly continue to develop in the future, and we can look forward to more detailed time-course studies of drug pharmacokinetics and neuropeptide metabolism. With increases in sensitivity and improvements in temporal resolution comes the possibility of studying neuropeptide transmission associated with behavior in real-time. In addition,

the multianalyte capability of MS detection will provide insights to brain malfunction in psychiatric disorders or neurological diseases, and ultimately lead to the discovery of novel targets for pharmacological interventions.

References Amster, I.J. (1996) Fourier transform mass spectrometry. J. Mass Spectrom., 31: 1325–1337. Andren, P.E. and Caprioli, R.M. (1995) In-vivo metabolism of substance-P in rat striatum utilizing microdialysis liquidchromatography micro-electrospray mass-spectrometry. J. Mass Spectrom., 30: 817–824. Andren, P.E. and Caprioli, R.M. (1999) Determination of extracellular release of neurotensin in discrete rat brain regions utilizing in vivo microdialysis/electrospray mass spectrometry. Brain Res., 845: 123–129. Andren, P.E., Emmett, M.R. and Caprioli, R.M. (1994) Micro-electrospray – zeptomole-attomole per microliter sensitivity for peptides. J. Am. Soc. Mass Spectrom., 5: 867–869. Andren, P.E., Emmett, M.R., DaGue, B.B., Steulet, A.F., Waldmeier, P. and Caprioli, R.M. (1998) Blood-brain barrier penetration of 3-aminopropyl-n-butylphosphinic acid (CGP 36742) in rat brain by microdialysis/mass spectrometry. J. Mass Spectrom., 33: 281–287. Baggerman, G., Verleyen, P., Clynen, E., Huybrechts, J., De Loof, A. and Schoofs, L. (2004) Peptidomics. J. Chromatogr. B. Analyt. Technol. Biomed. Life Sci., 803: 3–16. Baseski, H.M., Watson, C.J., Cellar, N.A., Shackman, J.G. and Kennedy, R.T. (2005) Capillary liquid chromatography with MS3 for the determination of enkephalins in microdialysis samples from the striatum of anesthetized and freely-moving rats. J. Mass Spectrom., 40: 146–153. Bushey, M.M. and Jorgenson, J.W. (1990) Automated instrumentation for comprehensive 2-dimensional high-performance liquid-chromatography capillary zone electrophoresis. Anal. Chem., 62: 978–984. Caprioli, R.M. and Lin, S.N. (1990) On-line analysis of penicillin blood levels in the live rat by combined microdialysis/ fast-atom bombardment mass spectrometry. Proc. Natl. Acad. Sci. U.S.A., 87: 240–243. Chen, Y.F., Chang, C.H., Wang, S.C. and Tsai, T.H. (2005) Measurement of unbound cocaine in blood, brain and bile of anesthetized rats using microdialysis coupled with liquid chromatography and veriﬁed by tandem mass spectrometry. Biomed. Chromatogr., 19: 402–408. Clarke, N.J., Crow, F.W., Younkin, S. and Naylor, S. (2001) Analysis of in vivo-derived amyloid-beta polypeptides by online two-dimensional chromatography-mass spectrometry. Anal. Biochem., 298: 32–39. Comisaro, M. and Marshall, A.G. (1974) Fourier-transform ion-cyclotron resonance spectroscopy. Chem. Phys. Lett., 25: 282–283.

264 Crespo, J.A., Sturm, K., Saria, A. and Zernig, G. (2005) Simultaneous intra-accumbens remifentanil and dopamine kinetics suggest that neither determines within-session operant responding. Psychopharmacology (Berl.), 183: 201–209. Davies, M.I., Cooper, J.D., Desmond, S.S., Lunte, C.E. and Lunte, S.M. (2000) Analytical considerations for microdialysis sampling. Adv. Drug Deliv. Rev., 45: 169–188. Dear, G.J., Plumb, R.S., Sweatman, B.C., Parry, P.S., Roberts, A.D., Lindon, J.C., Nicholson, J.K. and Ismail, I.M. (2000) Use of directly coupled ion-exchange liquid chromatogra phy-mass spectrometry and liquid chromatography-nuclear magnetic resonance spectroscopy as a strategy for polar metabolite identiﬁcation. J. Chromatogr. B Biomed. Sci. Appl., 748: 295–309. Emmett, M.R. and Caprioli, R.M. (1994) Micro-electrospray mass-spectrometry – ultra-high-sensitivity analysis of peptides and proteins. J. Am. Soc. Mass Spectrom., 5: 605–613. Emmett, M.R., Andren, P.E. and Caprioli, R.M. (1995) Speciﬁc molecular mass detection of endogenously released neuropeptides using in vivo microdialysis/mass spectrometry. J. Neurosci. Methods, 62: 141–147. Emmett, M.R., White, F.M., Hendrickson, C.L., Shi, S.D. and Marshall, A.G. (1998) Application of micro-electrospray liquid chromatography techniques to FT-ICR MS to enable high-sensitivity biological analysis. J. Am. Soc. Mass Spectrom., 9: 333–340. Fenn, J.B., Mann, M., Meng, C.K., Wong, S.F. and Whitehouse, C.M. (1989) Electrospray ionization for massspectrometry of large biomolecules. Science, 246: 64–71. Fuh, M.R., Haung, C.H., Lin, S.L. and Pan, W.H. (2004) Determination of free-form amphetamine in rat brain by ionpair liquid chromatography-electrospray mass spectrometry with in vivo microdialysis. J. Chromatogr. A., 1031: 197–201. Garcia, M.C. (2005) The effect of the mobile phase additives on sensitivity in the analysis of peptides and proteins by highperformance liquid chromatography-electrospray mass spectrometry. J. Chromatogr. B. Analyt. Technol. Biomed. Life Sci., 825: 111–123. Gatlin, C.L., Kleemann, G.R., Hays, L.G., Link, A.J. and Yates, J.R. (1998) Protein identiﬁcation at the low femtomole level from silver-stained gels using a new fritless electrospray interface for liquid chromatography microspray and nanospray mass spectrometry. Anal. Biochem., 263: 93–101. Giddings, J.C. (1984) Two-dimensional separations: concept and promise. Anal. Chem., 56: 1258A–1260A 1262A. Giorgianni, F., Beranova-Giorgianni, S. and Desiderio, D.M. (2004) Identiﬁcation and characterization of phosphorylated proteins in the human pituitary. Proteomics, 4: 587–598. Gygi, S.P., Rist, B., Gerber, S.A., Turecek, F., Gelb, M.H. and Aebersold, R. (1999) Quantitative analysis of complex protein mixtures using isotope-coded afﬁnity tags. Nat. Biotechnol., 17: 994–999. Hansen, B.T., Jones, J.A., Mason, D.E. and Liebler, D.C. (2001) SALSA: a pattern recognition algorithm to detect electrophile-adducted peptides by automated evaluation of CID spectra in LC-MS-MS analyses. Anal. Chem., 73: 1676–1683.

Haskins, W.E., Wang, Z., Watson, C.J., Rostand, R.R., Witowski, S.R., Powell, D.H. and Kennedy, R.T. (2001) Capillary LC-MS2 at the attomole level for monitoring and discovering endogenous peptides in microdialysis samples collected in vivo. Anal. Chem., 73: 5005–5014. Haskins, W.E., Watson, C.J., Cellar, N.A., Powell, D.H. and Kennedy, R.T. (2004) Discovery and neurochemical screening of peptides in brain extracellular ﬂuid by chemical analysis of in vivo microdialysis samples. Anal. Chem., 76: 5523–5533. Hillenkamp, F., Karas, M., Beavis, R.C. and Chait, B.T. (1991) Matrix-assisted laser desorption ionization mass-spectrometry of biopolymers. Anal. Chem., 63: A1193–A1202. Hokfelt, T., Bartfai, T. and Bloom, F. (2003) Neuropeptides: opportunities for drug discovery. Lancet Neurol., 2: 463–472. Hows, M.E., Lacroix, L., Heidbreder, C., Organ, A.J. and Shah, A.J. (2004) High-performance liquid chromatography/ tandem mass spectrometric assay for the simultaneous measurement of dopamine, norepinephrine, 5-hydroxytryptamine and cocaine in biological samples. J. Neurosci. Methods, 138: 123–132. Hows, M.E., Organ, A.J., Murray, S., Dawson, L.A., Foxton, R., Heidbreder, C., Hughes, Z.A., Lacroix, L. and Shah, A.J. (2002) High-performance liquid chromatography/tandem mass spectrometry assay for the rapid high sensitivity measurement of basal acetylcholine from microdialysates. J. Neurosci. Methods, 121: 33–39. Johansson, P.A., Andersson, M., Andersson, K.E. and Cenci, M.A. (2001) Alterations in cortical and basal ganglia levels of opioid receptor binding in a rat model of L-DOPA-induced dyskinesia. Neurobiol. Dis., 8: 220–239. Jones, D.C., Duvauchelle, C., Ikegami, A., Olsen, C.M., Lau, S.S., de la Torre, R. and Monks, T.J. (2005) Serotonergic neurotoxic metabolites of ecstasy identiﬁed in rat brain. J. Pharmacol. Exp. Ther., 313: 422–431. Kanamori, K., Kondrat, R.W. and Ross, B.D. (2003) C-13 enrichment of extracellular neurotransmitter glutamate in rat brain – combined mass spectrometry and NMR studies of neurotransmitter turnover and uptake into glia in vivo. Cell. Mol. Biol., 49: 819–836. Karas, M. and Hillenkamp, F. (1988) Laser desorption ionization of proteins with molecular masses exceeding 10000 Daltons. Anal. Chem., 60: 2299–2301. Kebarle, P. (2000) A brief overview of the present status of the mechanisms involved in electrospray mass spectrometry. J. Mass Spectrom., 35: 804–817. Kendrick, K.M. (1990) Microdialysis measurement of in vivo neuropeptide release. J. Neurosci. Methods, 34: 35–46. Kennedy, R.T., Watson, C.J., Haskins, W.E., Powell, D.H. and Strecker, R.E. (2002) In vivo neurochemical monitoring by microdialysis and capillary separations. Curr. Opin. Chem. Biol., 6: 659–665. Klintenberg, R. and Andren, P.E. (2005) Altered extracellular striatal in vivo biotransformation of the opioid neuropeptide dynorphin A(1–17) in the unilateral 6-OHDA rat model of Parkinson’s disease. J. Mass Spectrom., 40: 261–270.

265 Kreeger, J.S. and Larson, A.A. (1993) Substance P-(1–7), a substance P metabolite, inhibits withdrawal jumping in morphine-dependent mice. Eur. J. Pharmacol., 238: 111–115. Lanckmans, K., Clinckers, R., Van Eeckhaut, A., Sarre, S., Smolders, I. and Michotte, Y. (2006) Use of microbore LCMS/MS for the quantiﬁcation of oxcarbazepine and its active metabolite in rat brain microdialysis samples. J. Chromatogr. B. Analyt. Technol. Biomed. Life Sci., 831: 205–212. Lee, W.C. and Lee, K.H. (2004) Applications of afﬁnity chromatography in proteomics. Anal. Biochem., 324: 1–10. Lemmo, A.V. and Jorgenson, J.W. (1993) Transverse ﬂow gating interface for the coupling of microcolumn-LC with Cze in a comprehensive 2-dimensional system. Anal. Chem., 65: 1576–1581. Link, A.J., Eng, J., Schieltz, D.M., Carmack, E., Mize, G.J., Morris, D.R., Garvik, B.M. and Yates, J.R. (1999) Direct analysis of protein complexes using mass spectrometry. Nat. Biotechnol., 17: 676–682. Lipton, M.S., Pasa-Tolic, L., Anderson, G.A., Anderson, D.J., Auberry, D.L., Battista, J.R., Daly, M.J., Fredrickson, J., Hixson, K.K., Kostandarithes, H., Masselon, C., Markillie, L.M., Moore, R.J., Romine, M.F., Shen, Y., Stritmatter, E., Tolic, N., Udseth, H.R., Venkateswaran, A., Wong, K.-K., Zhao, R. and Smith, R.D. (2002) From the cover: global analysis of the Deinococcus radiodurans proteome by using accurate mass tags. Proc. Natl. Acad. Sci. U.S.A., 99: 11049–11054. Loughlin, A.F., Skiles, G.L., Alberts, D.W. and Schaefer, W.H. (2001) An ion exchange liquid chromatography/mass spectrometry method for the determination of reduced and oxidized glutathione and glutathione conjugates in hepatocytes. J. Pharm. Biomed. Anal., 26: 131–142. Mann, M. and Jensen, O.N. (2003) Proteomic analysis of posttranslational modiﬁcations. Nat. Biotechnol., 21: 255–261. Mann, M., Hojrup, P. and Roepstorff, P. (1993) Use of massspectrometric molecular-weight information to identify proteins in sequence databases. Biol. Mass Spectrom., 22: 338–345. Manz, A., Graber, N. and Widmer, H.M. (1990) Miniaturized total chemical-analysis system–a novel concept for chemical sensing. Sensors and Actuators B-Chemical, 1: 244–248. Marshall, A.G. (2000) Milestones in Fourier transform ion cyclotron resonance mass spectrometry technique development. Int. J. Mass Spectrom., 200: 331–356. Mayya, V., Rezaul, K., Cong, Y.S. and Han, D. (2005) Systematic comparison of a two-dimensional ion trap and a three-dimensional ion trap mass spectrometer in proteomics. Mol. Cell. Proteomics, 4: 214–223. Mirgorodskaya, E., Braeuer, C., Fucini, P., Lehrach, H. and Gobom, J. (2005) Nanoﬂow liquid chromatography coupled to matrix-assisted laser desorption/ionization mass spectrometry: sample preparation, data analysis, and application to the analysis of complex peptide mixtures. Proteomics, 5: 399–408. Nisbet, A.P., Foster, O.J.F., Kingsbury, A., Eve, D.J., Daniel, S.E., Marsden, C.D. and Lees, A.J. (1995) Preproenkephalin and preprotachykinin messenger-RNA expression in normal human basal ganglia and in Parkinsons-disease. Neuroscience, 66: 361–376.

Noble, D. (1995) MS/MS ﬂexes its muscles. Anal. Chem., 67: A265–A269. Nylander, I., Tan-No, K., Winter, A. and Silberring, J. (1995) Processing of prodynorphin-derived peptides in striatal extracts. Identiﬁcation by electrospray ionization mass spectrometry linked to size-exclusion chromatography. Life Sci., 57: 123–129. Ong, S.E. and Mann, M. (2005) Mass spectrometry-based proteomics turns quantitative. Nat. Chem. Biol., 1: 252–262. Opiteck, G.J., Lewis, K.C., Jorgenson, J.W. and Anderegg, R.J. (1997) Comprehensive on-line LC/LC/MS of proteins. Anal. Chem., 69: 1518–1524. Opiteck, G.J., Ramirez, S.M., Jorgenson, J.W. and Moseley, M.A. III. (1998) Comprehensive two-dimensional high-performance liquid chromatography for the isolation of overexpressed proteins and proteome mapping. Anal. Biochem., 258: 349–361. Pappin, D.J.C., Hojrup, P. and Bleasby, A.J. (1993) Rapid identiﬁcation of proteins by peptide-mass ﬁngerprinting. Curr. Biol., 3: 327–332. Paulson, L., Persson, R., Karlsson, G., Silberring, J., Bierczynska-Krzysik, A., Ekman, R. and Westman-Brinkmalm, A. (2005) Proteomics and peptidomics in neuroscience. Experience of capabilities and limitations in a neurochemical laboratory. J. Mass Spectrom., 40: 202–213. Prokai, L., Kim, H.S., Zharikova, A., Roboz, J., Ma, L., Deng, L. and Simonsick, W.J. Jr. (1998) Electrospray ionization mass spectrometric and liquid chromatographic-mass spectrometric studies on the metabolism of synthetic dynorphin A peptides in brain tissue in vitro and in vivo. J. Chromatogr. A., 800: 59–68. Rapp, E. and Tallarek, U. (2003) Liquid ﬂow in capillary (electro)chromatography: generation and control of microand nanoliter volumes. J. Sep. Sci., 26: 453–470. Reed, B., Zhang, Y., Chait, B.T. and Kreek, M.J. (2003) Dynorphin A(1–17) biotransformation in striatum of freely moving rats using microdialysis and matrix-assisted laser desorption/ ionization mass spectrometry. J. Neurochem., 86: 815–823. Rieux, L., Lubda, D., Niederlander, H.A., Verpoorte, E. and Bischoff, R. (2006) Fast, high-efﬁciency peptide separations on a 50-mum reversed-phase silica monolith in a nanoLC-MS set-up. J. Chromatogr. A., 1120: 165–172. Roepstorff, P. and Fohlman, J. (1984) Proposal for a common nomenclature for sequence ions in mass-spectra of peptides. Biomed. Mass Spectrom., 11: 601. Ross, P.L., Huang, Y.N., Marchese, J.N., Williamson, B., Parker, K., Hattan, S., Khainovski, N., Pillai, S., Dey, S., Daniels, S., Purkayastha, S., Juhasz, P., Martin, S., BartletJones, M., He, F., Jacobson, A. and Pappin, D.J. (2004) Multiplexed protein quantitation in Saccharomyces cerevisiae using amine-reactive isobaric tagging reagents. Mol. Cell. Proteomics, 3: 1154–1169. Russell, D.H. and Edmondson, R.D. (1997) High-resolution mass spectrometry and accurate mass measurements with emphasis on the characterization of peptides and proteins by matrix-assisted laser desorption/ionization time-of-ﬂight mass spectrometry. J. Mass Spectrom., 32: 263–276.

266 Sandin, J., Tan-No, K., Kasakov, L., Nylander, I., Winter, A., Silberring, J. and Terenius, L. (1997) Differential metabolism of dynorphins in substantia nigra, striatum, and hippocampus. Peptides, 18: 949–956. Schriemer, D.C. and Hindsgaul, O. (1998) Deconvolution approaches in screening compound mixtures. Comb. Chem. High Throughput Screen., 1: 155–170. Skold, K., Svensson, M., Kaplan, A., Bjorkesten, L., Astrom, J. and Andren, P.E. (2002) A neuroproteomic approach to targeting neuropeptides in the brain. Proteomics, 2: 447–454. Smith, R.D., Anderson, G.A., Lipton, M.S., Pasa-Tolic, L., Shen, Y.F., Conrads, T.P., Veenstra, T.D. and Udseth, H.R. (2002) An accurate mass tag strategy for quantitative and high-throughput proteome measurements. Proteomics, 2: 513–523. Stenken, J.A. (1999) Methods and issues in microdialysis calibration. Anal. Chim. Acta., 379: 337–358. Strand, F.L. (2003) Neuropeptides: general characteristics and neuropharmaceutical potential in treating CNS disorders. Prog. Drug Res., 61: 1–37. Svensson, M., Skold, K., Svenningsson, P. and Andren, P.E. (2003) Peptidomics-based discovery of novel neuropeptides. J. Proteome Res., 2: 213–219. Tang, K.Q., Li, F.M., Shvartsburg, A.A., Strittmatter, E.F. and Smith, R.D. (2005) Two-dimensional gas-phase separations coupled to mass spectrometry for analysis of complex mixtures. Anal. Chem., 77: 6381–6388. Tomer, K.B., Moseley, M.A., Deterding, L.J. and Parker, C.E. (1994) Capillary liquid-chromatography mass-spectrometry. Mass Spectrom. Rev., 13: 431–457. Tong, W., Link, A., Eng, J.K. and Yates, J.R. (1999) Identiﬁcation of proteins in complexes by solid phase microextraction multistep elution capillary electrophoresis tandem mass spectrometry. Anal. Chem., 71: 2270–2278. Uutela, P., Reinila, R., Piepponen, P., Ketola, R.A. and Kostiainen, R. (2005) Analysis of acetylcholine and choline in microdialysis samples by liquid chromatography/tandem mass spectrometry. Rapid Commun. Mass Spectrom., 19: 2950–2956. Vestal, M.L., Juhasz, P. and Martin, S.A. (1995) Delayed extraction matrix-assisted laser-desorption time-of-ﬂight massspectrometry. Rapid Commun. Mass Spectrom., 9: 1044–1050. Wang, Y., Balgley, B.M., Rudnick, P.A. and Lee, C.S. (2005) Effects of chromatography conditions on intact protein separations for top-down proteomics. J. Chromatogr. A, 1073: 35–41. Washburn, M.P., Ulaszek, R., Deciu, C., Schieltz, D.M. and Yates, J.R. (2002) Analysis of quantitative proteomic data

generated via multidimensional protein identiﬁcation technology. Anal. Chem., 74: 1650–1657. Washburn, M.P., Wolters, D. and Yates, J.R. III. (2001) Largescale analysis of the yeast proteome by multidimensional protein identiﬁcation technology. Nat. Biotechnol., 19: 242–247. Wilm, M. and Mann, M. (1996) Analytical properties of the nanoelectrospray ion source. Anal. Chem., 68: 1–8. Wu, C.C., MacCoss, M.J., Howell, K.E., Matthews, D.E. and Yates, J.R. (2004) Metabolic labeling of mammalian organisms with stable isotopes for quantitative proteomic analysis. Anal. Chem., 76: 4951–4959. Yates, J.R., Speicher, S., Grifﬁn, P.R. and Hunkapiller, T. (1993) Peptide mass maps – a highly informative approach to protein identiﬁcation. Anal. Biochem., 214: 397–408. Yi, E.C., Lee, H., Aebersold, R. and Goodlett, D.R. (2003) A microcapillary trap cartridge-microcapillary high-performance liquid chromatography electrospray ionization emitter device capable of peptide tandem mass spectrometry at the attomole level on an ion trap mass spectrometer with automated routine operation. Rapid Commun. Mass Spectrom., 17: 2093–2098. Zhang, B., Palcic, M.M., Mo, H., Goldstein, I.J. and Hindsgaul, O. (2001a) Rapid determination of the binding afﬁnity and speciﬁcity of the mushroom polyporus squamosus lectin using frontal afﬁnity chromatography coupled to electrospray mass spectrometry. Glycobiology, 11: 141–147. Zhang, B., Palcic, M.M., Schriemer, D.C., Alvarez-Manilla, G., Pierce, M. and Hindsgaul, O. (2001b) Frontal afﬁnity chromatography coupled to mass spectrometry for screening mixtures of enzyme inhibitors. Anal. Biochem., 299: 173–182. Zhang, H., Stoeckli, M., Andren, P.E. and Caprioli, R.M. (1999) Combining solid-phase preconcentration, capillary electrophoresis and off-line matrix-assisted laser desorption/ ionization mass spectrometry: intracerebral metabolic processing of peptide E in vivo. J. Mass Spectrom., 34: 377–383. Zhou, Q., Le Greves, P., Ragnar, F. and Nyberg, F. (2000) Intracerebroventricular injection of the N-terminal substance P fragment SP(1–7) regulates the expression of the N-methyl-D-aspartate receptor NR1, NR2A and NR2B subunit mRNAs in the rat brain. Neurosci. Lett., 291: 109–112. Zubarev, R.A., Horn, D.M., Fridriksson, E.K., Kelleher, N.L., Kruger, N.A., Lewis, M.A., Carpenter, B.K. and McLafferty, F.W. (2000) Electron capture dissociation for structural characterization of multiply charged protein cations. Anal. Chem., 72: 563–573.

CHAPTER 3.3

Improvement of the temporal resolution of brain microdialysis: sampling in seconds Luis Herna´ndez, Sergio Rossell, Sonia Tucci, Daniel Paredes and Pedro Rada Laboratorio de Fisiologı´a de la Conducta, Escuela de Medicina, Universidad de los Andes, Me´rida, Venezuela

Abstract: Brain microdialysis could provide better insight into brain functions on the condition that its temporal resolution could be improved. Such endeavor requires the creation of analytical techniques that are able to analyze samples whose volumes are in the order of a few nanoliters. In addition, it requires the creation of laboratory methods to handle, in a reproducible manner, nanoliter volumes. In the present chapter, the progress made in both areas during the last 10 years is reviewed. The methods invented so far include capillary-zone electrophoresis with laser-induced ﬂuorescence detection, nanoliter volume reactors, assay tubes that allow carrying out chemical reactions in such volumes, and automatic injection systems to load picoliter volumes. Evidences that it is currently possible to lower the time resolution of brain microdialysis to a few seconds are presented. This chapter also discusses the way in which the improvement of the temporal resolution is inﬂuencing the interpretation of some of the neurochemical phenomena so far studied with brain microdialysis. However, even though brain dialysates contain (in the form of molecules at a very low concentration) the vital information needed to understand the chemical basis of brain functions, in most cases the available analytical techniques cannot detect these molecules. Hence, the use of brain microdialysis as a technique to further understand brain neurotransmission is subjected to the development of analytical techniques that are more sensitive than the currently available ones. Perhaps the most adverse consequence of the poor sensitivity of the analytical techniques has been a severe limitation of the temporal analysis of the behavioral and mental phenomena that have been tackled with brain microdialysis. So, phenomena such as feeding, fear, mating, cognitive activities, pain, and many others have a temporal course that requires collecting microdialysis samples every few seconds (sampling time). Unfortunately, these phenomena have had to be addressed sampling for 10 or more minutes. This is mainly due to the fact that briefer collection times do not

I. Introduction The mechanisms underlying phenomena such as memory, fear, behavior, hate, love, rage, and other expressions of the mind rely on chemical communication between the cells that constitute the brain, that is, neurons and glial cells. Since Otto Loewi (1957) and Henry Dale (1963) discovered chemical transmission, there has been a remarkable progress in the development of techniques for studying neurotransmitters in the brain. One of these techniques, brain microdialysis, was developed in the seventies and the eighties of the last century (Delgado et al., 1972; Hernandez et al., 1983; Johnson and Justice, 1983; Zetterstrom et al., 1983; Imperato and Di Chiara, 1984; Damsma et al., 1988), and over the last 20 years this technique has shed light on various aspects of chemical transmission in the brain of freely moving animals. Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

267

DOI: 10.1016/S1569-7339(06)16015-4 Copyright 2007 Elsevier B.V. All rights reserved

268

provide enough mass of the analyte to make it measurable with the available analytical techniques. In the present chapter, we review the attempts to reduce the sampling time and evaluate how successful they have been.

I.A. Why high temporal resolution in brain microdialysis? The main objective of brain microdialysis is to measure the chemical changes that occur in the brain to correlate them with behavior, drug effects, and other phenomena of interest. Very often these changes are caused by synapses, which release their chemical messengers during brief time intervals. For example, the axon terminals of glutamatergic, gabaergic, and cholinergic neurons (Higgs and Lukasiewicz, 1999; Colgin et al., 2003; Moore et al., 2003) can release glutamate, g-aminobutyric acid, or acetylcholine for a few milliseconds and yet induce, biologically speaking, significant changes in the central excitatory state of other neurons. Such changes may be diluted when the collection times last longer than the time of release. Indeed, it has been shown by Bert et al. (2002) that glutamate changes occurring in 1 min can be completely dampened when samples are pooled (Fig. 1). It is also well known that several neurotransmitters coexist in a given synaptic button and that they could be released at different times (Cesselin and Hamon, 1984; Hokfelt et al., 1992). Moreover, axon terminals of amino acidergic, cholinergic, monoaminergic, and peptidergic neurons converge in the same brain region. These terminals can become active at different times and participate in the same function. Bert et al. (2002) and Tucci et al. (2000) have observed glutamate extracellular concentration increase lasting 1.5 min and dopamine increase lasting 1 h in the same region (see Fig. 2). If the sampling time is too long, there is a risk of erroneously concluding that neurotransmitters simultaneously change their extracellular level. Mark et al. (1992), collecting dialysates each 30 min, observed that acetylcholine and dopamine increased in the nucleus accumbens in fasted and fed rats suggesting that both neurotransmitters

Fig. 1. Microdialysis samples were collected every minute from the striatum and analyzed by CZE–LIFD. NMDA was perfused by reverse microdialysis (black horizontal bar) and dopamine, glutamate, and aspartate increased 1,000, 150, and 250%, respectively (solid circle curves). All three neurotransmitters returned to basal levels during the NMDA perfusion. After the perfusion, glutamate increased again to 150% (bimodal increase) but dopamine remained at basal levels. When the samples were pooled to simulate 10 min sample collection time (gray bars), the increase were attenuated to 400% for DA, 115% for glutamate, and 160% for aspartate and the bimodal increase in glutamate and aspartate disappeared (Bert et al., 2002; reproduced with permission).

269

Fig. 3. Acetylcholine increase in the nucleus accumbens in fastfed rats. Ten minute temporal resolution microdialysis reveals that the feeding peak precedes acetylcholine increase. This might suggest a role for dopamine in food reward and a role for acetylcholine in satiety (Mark et al., 1992; reproduced with permission).

Fig. 2. Microdialysis samples from the nucleus accumbens were collected every 30 s to assess glutamate by CZE–LIFD and every 20 min to evaluate dopamine by HPLC-EC. Electrical stimulation of the medial prefrontal cortex increased glutamate for 30 s and dopamine for more than 2 h. Better temporal resolution for dopamine is needed to determine whether or not dopamine increase was caused by glutamate increase (Tucci et al., 2000; reproduced with permission).

might be involved in reward. Nonetheless, when they collected dialysates each 10 min they found that the peak of food intake preceded the increase of acetylcholine, suggesting that dopamine release was related to food reward and acetylcholine release to the aversive components of satiety. Situations like this one demand improvement of the temporal resolution for brain microdialysis (see Fig. 3). Such improvement can be accomplished by decreasing the sampling time. However, since the ﬂow rates in most of microdialysis experiments are low, sample collection in less than 1 min yields sample volumes between 30 nL and 1 mL. These small volumes originate too hard to solve problems. First, they decrease the mass of biologically interesting chemicals contained in the dialysate

making them difﬁcult to detect. This problem has to be addressed by the creation of new analytical techniques capable of separating and detecting the chemical components of the dialysates in nanoliter samples. Second, evaporation demands the development of new methods to collect nanoliter samples in a reproducible manner. Thus, in the next section we will examine some of the efforts to solve these problems. II. Analytical techniques to improve the temporal resolution of brain microdialysis To improve the temporal resolution of brain microdialysis, the technique used to analyze the dialysates has to fulﬁll two requirements: it must have high mass sensitivity and it should work with nanoliter sample volumes. So far, two techniques have shown capacity to process nanoliter sample volumes: high pressure liquid chromatography (HPLC) with small internal diameter columns and capillary-zone electrophoresis (CZE). HPLC microcolumns are called microbore columns because their internal diameter varies between 500 mm and 1 mm (Wages et al., 1986; Cheng and Kuo, 1995). These columns separate chemical substances in dialysates of 1 mL or less.

270

The regular HPLC pulse pumps that are used with large internal diameter columns tend to deliver a pulsed ﬂow rather than a constant ﬂow. If the detection technique is electrochemical, it is well known that such technique is highly sensitive to ﬂow-rate variations. Therefore, to suppress ﬂow noise, a very constant ﬂow rate is required. When HPLC is used with large internal diameter columns, the column itself acts as an excellent pulse dampener. However, in the case of microbore columns this dampening capacity is lost. This problem has been solved by improving the efﬁciency of pulse dampeners. If the detection technique is ﬂuorescence, ﬂow rate variations do not cause too much noise and the sensitivity improves. Kehr has used o-phthalaldehyde derivatization with ﬂuorescence detection combined with microbore columns to measure glutamate and aspartate in brain dialysates every minute (Kehr, 1998). In any event, microbore columns have been used to measure drugs such as ethanol (Nurmi et al., 1994) and acetaminophen (Chen and Lunte, 1995) in plasma dialysates every minute. Newton and Justice (1994) measured extracellular levels of dopamine in brain dialysates every minute, using microbore columns. When they infused cocaine by reverse microdialysis in anesthetized rats, dopamine increased in the ﬁrst minute and reached a peak at the third minute. Richter et al. (1999) measured glutamate, GABA, adenosine, and serotonin in ventral medulla dialysates, and they found that these neurotransmitters increase in the ﬁrst minute of hypoxia in anesthetized cats. From these and other studies it can be concluded that although microbore columns have improved the time resolution of brain microdialysis to a minute, they do not seem to be able to increase this resolution to a subminute point. Finally, the small sample volume requires to decrease the volume of the detection chamber. Originally, the volumes of the detection chambers varied between 5 and 10 mL. Currently there are commercially available microchambers with volumes as small as 5 nL (Antec, Leyden, the Netherlands). Capillary-zone electrophoresis was ﬁrst attempted by Hjerten (1967) revealed in a rudimentary way by Mikkers et al. (1979), and fully developed by Jorgenson and Lukacs (1981). It has

been widely used for the analysis of microdialysates. Currently, two modalities of detection have been applied: UV–visible and laser-induced ﬂuorescence detection (LIFD). UV–visible detection has limited applications in high-resolution brain microdialysis. This is mainly due to the poor UV absorption of many biologic compounds, particularly in the range of high UV (270–370 nm), which hinders sensitivity of CZE–UV detection. Nonetheless, several exogenous compounds with high UV absorption are easy to detect yielding high temporal resolution to blood microdialysis pharmacokinetic studies. This is the case of trimebutine maleate, a powerful antispasmodic drug that has been measured in blood microdialysates in rabbits by means of CZE–UV absorption with a 200 nM sensitivity (Wang et al., 2005). UV indirect is another detection method in which the running buffer for CZE strongly absorbs UV. The presence of the analyte is detected by a decrease of UV absorption. This method has been successfully used to measure the antibiotic phosphomycin in human plasma microdialysates with a 2 mM sensitivity (Petsch et al., 2004). Another method to improve the sensitivity of UV detection consists in tagging a molecule with low UV absorption with another possessing large UV-absorption capacity. By means of this technique alphatriﬂuordimethylornithine, a chemotherapeutic agent, has been measured in human plasma microdialysates at concentrations as low as 5 mM (Hu et al., 1995). In spite of its low sensitivity, CZE–UV detection is a promising technique because remarkable progress made in sample preconcentration or stacking has increased its sensitivity by a factor of 5,000 (Quirino and Terabe, 1998; Zhao et al., 1998). Laser-induced ﬂuorescence detection improved the sensitivity of CZE making it possible to measure the trace amounts of neurotransmitters commonly found in brain microdialysates. LIFD for CZE was introduced in the eighties by Richard Zare and his collaborators who showed that it was feasible measuring nanomolar concentrations of dansylated amino acids (Gassmann et al., 1985). Three years later, in the laboratory of Norman Dovichi it was shown that high picomolar concentrations of amino acids could be measured with

271

CZE–LIFD (Cheng and Dovichi, 1988; Wu and Dovichi, 1989). In the early nineties the introduction of the collinear geometry took the limits of concentration detection to the low picomolar level (Hernandez et al., 1991). This permitted the technique to be applied to the analysis of brain microdialysates to measure excitatory and inhibitory amino acids in nanoliter sample volumes (Hernandez et al., 1993; Rada et al., 1999). In CZE–LIFD the target molecule is tagged with a ﬂuorescent molecule or in some cases it can be detected by native ﬂuorescence. So far, several dyes have been used including ﬂuorescein isothiocyanate isomer III (Cheng and Dovichi, 1988; Hernandez et al., 1991), naphthalenedicarboxyaldehyde (Oates and Jorgenson, 1989; Bert et al., 1996a, b), and o-phthalaldehyde and ﬂuorescamine (Lada and Kennedy, 1996; Skelley and Mathies, 2003). These dyes bind to the primary amine groups of organic compounds and form ﬂuorescent derivatives. The compounds are then separated by CZE and detected by LIFD.

III. Collection and treatment of small volume samples Evaporation causes large variations in the nanovolume samples collected in very short periods of time. In such volumes the shape of the surface of the sample inﬂuences evaporation decisively. In our laboratory, we did several tests to determine which was the most appropriate geometry to avert evaporation in a container collecting a few nanoliters. We found that conic glass containers made out of hematocrit tubing prevent evaporation better than any other geometry or material (Rada et al., 1998, see Fig. 4). The reason is that in the conic glass container capillary forces create a concave surface that decreases the tension of saturated vapor eliminating saturation or rather attracting vapor to the liquid phase. By contrast, inside round bottom plastic tubes the liquid forms a convex surface, which raises the tension of saturated vapor and promotes evaporation (Landau et al., 1973). Conic bottom glass containers are made by holding a hematocrit tube in a vertical position on a ﬂame. The force of gravity stretches

Fig. 4. If the surface of a 1,000 nL drop is convex, 80% of it will be evaporated at room temperature in 50 min (open circle curve). By contrast, when the surface is concave (solid circle curve) evaporation is suppressed and even there is a trend to condensation due to the capillary condensation phenomenon (Unpublished data).

the melted glass, sealing the tube and creating the conic geometry. If, in addition to the conic shape, the tubes are capped and stored in a humid atmosphere at 4–81C, evaporation is totally eliminated. The samples are collected in the proper tubes and a derivatization mixture is added. Then they are injected into the CZE–LIFD instrument. With this technique we were able to measure glutamate, aspartate, and GABA by brain microdialysis with a temporal resolution from 1 min down to 30 s in freely moving rats. We have found reciprocal changes of glutamate and GABA in the lateral hypothalamus in rats during a meal (Rada et al., 2003). Collecting microdialysis samples every 30 s, a fast glutamate increase was observed in the ﬁrst third of a meal followed by a sharp decrease in the remainder two-thirds. At the same time GABA increases steadily and peaks in the last two-thirds of the meal (Fig. 5). These experiments suggest that feeding neurons of the lateral hypothalamus

272

Fig. 5. Glutamate and GABA were monitored every 30 s in the lateral hypothalamus of fast-fed rats for one single meal. As soon as the meal starts glutamate increases (open circle curve, bottom graph, individual rat) in the ﬁrst 3 min of the meal. After that time, glutamate steadily decreases to levels lower than the basal. In the last minutes GABA increases and peaks when the meal ends (solid circle curve, bottom graph, individual rat). The upper, bar graph shows the neurochemical changes in the whole population of rats; the solid bars correspond to GABA and the open ones to glutamate (Rada et al., 2003; reproduced with permission).

are excited by glutamate during a meal and inhibited by GABA at the end of a meal. We also showed that at the time these neurochemical changes occur in the lateral hypothalamus there is a decrease of glutamate in the nucleus accumbens (Rada et al., 1997). The time course of these neurochemical events is at most 3 min, proving that fast neurochemical changes occur during feeding and probably are more relevant to behavior than long-term neurochemical changes. In

another series of experiments we found that fast glutamate changes occur during conditioned taste aversion (Tucci et al., 1998). In this paradigm, a novel ﬂavor is associated to malaise induced by an intraperitoneal lithium chloride injection. Days later, the animal rejects any food with the novel ﬂavor. By measuring glutamate in microdialysates collected every 30 s, we found that the intraoral injection of a solution with the novel ﬂavor causes a sharp glutamate increase lasting barely 1 min in the amygdala in freely moving conditioned rats. We also found fast glutamate increase in the visual cortex in freely moving rats. When this animal received photic stimulation, glutamate increased within the ﬁrst minute (Reyes et al., 2002). CZE–LIFD technique was further improved by the addition of on-line reactors. Bert et al. (2002) built an on-line reactor to derivatize microdialysates as they were collected every minute. The ﬂuorochrome was naphthalenedicarboxyaldehyde, and by means of the 440 nm line of a helium–cadmium laser they detected catecholamines, glutamate, and aspartate in the dorsal striatum of anesthetized rats (Bert et al., 2002). When they administered NMDA by reverse microdialysis to the dorsal striatum, they observed an increase of glutamate and dopamine, but after the perfusion dopamine decreased and glutamate increased again. When the samples were evaluated every 10 min, dopamine increase was dampened and glutamate did not exhibit the bimodal increase revealed by 1-min resolution microdialysis. These experiments show that an inadequate temporal resolution masks the actual temporal course of synaptic activity. In another modality of on-line reactors, the mixture of dialysate and derivatizing agents was collected in fused silica capillaries. By properly choosing the length and the internal diameter of the collecting tubes it was possible to obtain samples every 6 s. The content of these tubes was transferred by centrifugation to the conic tubes mentioned above. After 16 h the samples were diluted and injected into the CZE–LIFD instrument. With this method, we measured glutamate release in the parietal cortex during occlusion of the Sylvian artery in anesthetized rats (Tucci et al., 1997) and observed that glutamate decreased shortly

273

after clamping the artery in ketamine-anesthetized rats but not in thiopental ones (see Fig. 6). Bert et al. also measured amino acids and dopamine every 10 s in anesthetized rats. When they administered L-trans-pirrolidin 2,4-dicarboxylic acid (PDC, a glutamate transporter blocker) by means of an injection micropipette attached to the microdialysis probe, they observed that glutamate and aspartate increased in the ﬁrst 10 s. Dopamine also increased but after 20 s, suggesting that glutamate causes dopamine release (Bert et al., 2002; Parrot et al., 2003) (Fig. 7). For 1-s temporal resolution (after derivatization into the on-line reactor), the samples were collected into a 16 cm long, 99 mm inside diameter fused-silica tube. Then the tube was cut into 4 mm long pieces containing 32 nL of the mixture and the content of each piece was transferred by centrifugation into conical tubes. After 16 h the mixture was diluted and injected into a CZE–LIFD instrument. With this technique, we were able to stimulate the vibrissae and measure a glutamate increase in the

Fig. 6. Transient occlusion of the Sylvian artery (black, horizontal bar) enhances glutamate extracellular levels in sodium thiopental anesthetized rats (solid circle curve). By contrast, it lowers extracellular levels of glutamate in ketamine anesthetized rats (solid black square curve) (Tucci et al., 1997; reproduced with permission).

Fig. 7. The administration of PDC (a glutamate reuptake blocker) by reverse microdialysis increases extracellular levels of glutamate and aspartate in anesthetized rats (B and C). Shortly after glutamate and aspartate increase, dopamine increases too (A). Calculation of 50% of maximal increase shows that glutamate and aspartate increase precedes dopamine increase by 26 s. This result suggests that glutamate and aspartate stimulate dopamine release (Bert et al., 2002; reproduced with permission).

274

Fig. 8. Moving the vibrissae increases extracellular glutamate for 1 s in the somatosensory cortex in a freely moving rat (Rossell et al., 2003; reproduced with permission).

Fig. 9. Systemic administration of ethyl alcohol increases extracellular level of taurine in the striatum. The taurine increase was larger in the ventral striatum (VS, part A) than in the dorsal striatum (DS, part B) (Smith et al., 2003; reproduced with permission).

275

parietal cortex (primary sensory cortex) in freely moving rats (Rossell et al., 2003) (Fig. 8). Finally, the samples can be collected, derivatized on-line and intermittently injected, by means of a ﬂow-gated interphase, into the CZE–LIFD instrument (Lada and Kennedy, 1995, 1996). The dialysate and the derivatization mixture are mixed into a T tube. The mixture ﬂows into a gate invented by Lemmo and Jorgenson (1993). This gate washes out the mixture by a continuous ﬂow of running buffer. When this ﬂow is turned off, the mixture accumulates in front of the injection end of the capillary and the sample is injected. With this device, Smith et al. (2003) at Robert Kennedy’s laboratory measured taurine in dialysates from the striatum every 11 s in anesthetized rats (Fig. 9). With the same device, Haskins et al. (2004) administered brain-derived non-opioid peptides by reverse microdialysis and measured glutamate, aspartate, taurine, and GABA in anesthetized rats. They found that some of these peptides induce the release of these neurotransmitters in an impulseindependent manner. Lada et al. (1998) electrically stimulated the prefrontal cortex and measured glutamate increase in the striatum every 5 s in anesthetized rats. Recently, Kennedy’s group has adapted its online method to freely moving animals. It achieved glutamate and GABA measurements every 13 s in microdialysates from the striatum in freely moving rats (Presti et al., 2004). In a tour the force experiment they correlated neurochemistry with rearing behavior in mouse mutants that exhibit spontaneous rearing behavior. When the rearing episodes occurred, both glutamate and GABA increased. The present review describes a dramatic improvement of temporal resolution of brain microdialysis over the last 10 years. From a 10 min temporal resolution 10 years ago, nowadays it is feasible to have 1-s temporal resolution in freely moving animals. The technology to handle 30 nL volumes has been developed and it is highly reproducible. In parallel, analytical techniques to process nanoliter samples have been developed too. In the near future, it will be possible to measure picoliter samples and take the temporal resolution of brain microdialysis to the subsecond level.

References Bert, L., Parrot, S., Robert, F., Desvignes, C., Denoroy, L., Suaud-Chagny, M. and Renaud, B. (2002) In vivo temporal sequence of rat striatal glutamate, aspartate and dopamine efﬂux during apomorphine, nomifensine, NMDA and PDC in situ administration. Neuropharmacology, 43: 825–835. Bert, L., Robert, F., Denoroy, L. and Renaud, B. (1996a) Highspeed separation of subnanomolar concentrations of noradrenaline and dopamine using capillary zone electrophoresis with laser-induced ﬂuorescence detection. Electrophoresis, 17: 523–525. Bert, L., Robert, F., Denoroy, L., Stoppini, L. and Renaud, B. (1996b) Enhanced temporal resolution for the microdialysis monitoring of catecholamines and excitatory amino acids using capillary electrophoresis with laser-induced ﬂuorescence detection. Analytical developments and in vitro validations. J. Chromatogr. A., 755: 99–111. Cesselin, F. and Hamon, M. (1984) Possible functional significance of the simultaneous release of several putative neurotransmitters by the same neuron. Ann. Endocrinol. (Paris), 45: 207–213. Chen, A. and Lunte, C. (1995) Microdialysis sampling coupled on-line to fast microbore liquid chromatography. J. Chromatogr. A., 691: 29–35. Cheng, F. and Kuo, J. (1995) High-performance liquid chromatographic analysis with electrochemical detection of biogenic amines using microbore columns. J. Chromatogr. B. Biomed. Appl., 665: 1–13. Cheng, Y. and Dovichi, N. (1988) Subattomole amino acid analysis by capillary zone electrophoresis and laser-induced ﬂuorescence. Science, 242: 562–564. Colgin, L., Kramar, E., Gall, C. and Lynch, G. (2003) Septal modulation of excitatory transmission in hippocampus. J. Neurophysiol., 90: 2358–2366. Dale, H. (1963) Pharmacology during the past sixty years. Annu. Rev. Pharmacol., 3: 1–9. Damsma, G., Biessels, P., Westerink, B., De Vries, J. and Horn, A. (1988) Differential effects of 4-aminopyridine and 2,4-diaminopyridine on the in vivo release of acetylcholine and dopamine in freely moving rats measured by intrastriatal dialysis. Eur. J. Pharmacol., 145: 15–20. Delgado, J., DeFeudis, F., Roth, R., Ryugo, D. and Mitruka, B. (1972) Dialytrode for long term intracerebral perfusion in awake monkeys. Arch. Int. Pharmacodyn. Ther., 198: 9–21. Gassmann, E., Kuo, J. and Zare, R. (1985) Electrokinetic separation of chiral compounds. Science, 230: 813–814. Haskins, W., Watson, C., Cellar, N., Powell, D. and Kennedy, R. (2004) Discovery and neurochemical screening of peptides in brain extracellular ﬂuid by chemical analysis of in vivo microdialysis samples. Anal. Chem., 76: 5523–5533. Hernandez, L., Escalona, J., Joshi, N. and Guzman, N. (1991) Laser-induced ﬂuorescence and ﬂuorescence microscopy for capillary electrophoresis zone detection J. Chromatogr. A., 559: 183–196.

276 Hernandez, L., Joshi, N., Murzi, E., Verdeguer, P., Mifsud, J. and Guzman, N. (1993) Colinear laser-induced ﬂuorescence detector for capillary electrophoresis. Analysis of glutamic acid in brain dialysates. J. Chromatogr. A., 652: 399–405. Hernandez, L., Paez, X. and Hamlin, C. (1983) Neurotransmitters extraction by local intracerebral dialysis in anesthetized rats. Pharmacol. Biochem. Behav., 18: 159–162. Higgs, M. and Lukasiewicz, P. (1999) Glutamate uptake limits synaptic excitation of retinal ganglion cells. J. Neurosci., 19: 3691–3700. Hjerten, S. (1967) Free zone electrophoresis. Chromatogr. Rev., 9: 122–219. Hokfelt, T., Bean, A., Ceccatelli, S., Dagerlind, A., Elde, R., Goldstein, M., Meister, B., Melander, T., Nicholas, A. and Pelto-Huikko, M. et al. (1992) Neuropeptides and classical transmitters. Localization and interaction. Arzneimittelforschung, 42: 196–201. Hu, T., Zuo, H., Riley, C., Stobaugh, J. and Lunte, S. (1995) Determination of alpha-diﬂuoromethylornithine in blood by microdialysis sampling and capillary electrophoresis with UV detection. J. Chromatogr. A., 716: 381–388. Imperato, A. and Di Chiara, G. (1984) Trans-striatal dialysis coupled to reverse phase high performance liquid chromatography with electrochemical detection: a new method for the study of the in vivo release of endogenous dopamine and metabolites. J. Neurosci., 4: 966–977. Johnson, R. and Justice, J. (1983) Model studies for brain dialysis. Brain Res. Bull., 10: 567–571. Jorgenson, J. and Lukacs, K. (1981) Free-zone electrophoresis in glass capillaries. Clin. Chem., 27: 1551–1553. Kehr, J. (1998) Determination of glutamate and aspartate in microdialysis samples by reversed-phase column liquid chromatography with ﬂuorescence and electrochemical detection. J. Chromatogr. B. Biomed. Sci. Appl., 708: 27–38. Lada, M. and Kennedy, R. (1995) Quantitative in vivo measurements using microdialysis on-line with capillary zone electrophoresis. J. Neurosci. Methods, 63: 147–152. Lada, M. and Kennedy, R. (1996) Quantitative in vivo monitoring of primary amines in rat caudate nucleus using microdialysis coupled by a ﬂow-gated interface to capillary electrophoresis with laser-induced ﬂuorescence detection. Anal. Chem., 68: 2790–2797. Lada, M.W., Vickroy, T.W. and Kennedy, R.T. (1998) Evidence for neuronal origin and metabotropic-receptor mediated regulation of extracellular glutamate and aspartate in rat striatum in vivo following electrical stimulation of the prefrontal cortex. J. Neurochem., 70: 617–625. Landau, L., Ajiezer, A. and Lifshitz, E. (1973) Curso General de Fı´ sica: Meca´nica y Fı´ sica Molecular. MIR, Moscu´, pp. 308–309. Lemmo, A. and Jorgenson, J. (1993) Transverse ﬂow gating interface for the coupling of microcolumn LC with CZE in a comprehensive two-dimensional system. Anal. Chem., 65: 1576–1581. Loewi, O. (1957) On the background of the discovery of neurochemical transmission. J. Mt. Sinai. Hosp. N.Y., 24: 1014–1016.

Mark, G., Rada, P., Pothos, E. and Hoebel, B. (1992) Effects of feeding and drinking on acetylcholine release in the nucleus accumbens, striatum, and hippocampus of freely behaving rats. J. Neurochem., 58: 2269–2274. Mikkers, F., Everaerts, F. and Verheggen, P. (1979) High-performance zone electrophoresis. J. Chromatogr. A., 169: 11–20. Moore, K., Nicoll, R. and Schmitz, D. (2003) Adenosine gates synaptic plasticity at hippocampal mossy ﬁber synapses. Proc. Natl. Acad. Sci. U.S.A., 100: 14397–14402. Newton, A. and Justice, J.J. (1994) Temporal response of microdialysis probes to local perfusion of dopamine and cocaine followed with one-minute sampling. Anal. Chem., 66: 1468–1472. Nurmi, M., Kiianmaa, K. and Sinclair, J. (1994) Brain ethanol in AA, ANA, and Wistar rats monitored with one-minute microdialysis. Alcohol, 11: 315–321. Oates, M. and Jorgenson, J. (1989) Determination of naphthalene-2,3-dicarboxaldehyde-labeled amino acids by open tubular liquid chromatography with electrochemical detection. Anal. Chem., 61: 432–435. Parrot, S., Bert, L., Mouly-Badina, L., Sauvinet, V., ColussiMas, J., Lambas-Senas, L., Robert, F., Bouilloux, J., SuaudChagny, M., Denoroy, L. and Renaud, B. (2003) Microdialysis monitoring of catecholamines and excitatory amino acids in the rat and mouse brain: recent developments based on capillary electrophoresis with laser-induced ﬂuorescence detection – a mini-review. Cell. Mol. Neurobiol., 23: 793–804. Petsch, M., Mayer-Helm, B., Sauermann, R., Joukhadar, C. and Kenndler, E. (2004) Capillary electrophoresis analysis of fosfomycin in biological ﬂuids for clinical pharmacokinetic studies. Electrophoresis, 25: 2292–2298. Presti, M., Watson, C., Kennedy, R., Yang, M. and Lewis, M. (2004) Behavior-related alterations of striatal neurochemistry in a mouse model of stereotyped movement disorder. Pharmacol. Biochem. Behav., 77: 501–507. Quirino, J. and Terabe, S. (1998) Exceeding 5000-fold concentration of dilute analytes in micellar electrokinetic chromatography. Science, 282: 465–468. Rada, P., Mendialdua, A., Hernandez, L. and Hoebel, B.G. (2003) Extracellular glutamate increases in the lateral hypothalamus during meal initiation, and GABA peaks during satiation: microdialysis measurements every 30 s. Behav. Neurosci., 117: 222–227. Rada, P., Tucci, S., Murzi, E. and Hernandez, L. (1997) Extracellular glutamate increases in the lateral hypothalamus and decreases in the nucleus accumbens during feeding. Brain Res. Brain Res. Rev., 768: 338–340. Rada, P., Tucci, S., Perez, J., Teneud, L., Chuecos, S. and Hernandez, L. (1998) In vivo monitoring of gabapentin in rats: a microdialysis study coupled to capillary electrophoresis and laser-induced ﬂuorescence detection. Electrophoresis, 19: 2976–2980. Rada, P., Tucci, S., Teneud, L., Paez, X., Perez, J., Alba, G., Garcia, Y., Sacchettoni, S., del Corral, J. and Hernandez, L. (1999) Monitoring gamma-aminobutyric acid in human brain

277 and plasma microdialysates using micellar electrokinetic chromatography and laser-induced ﬂuorescence detection. J. Chromatogr. B. Biomed. Sci. Appl., 735: 1–10. Reyes, E., Rossell, S., Paredes, D., Rada, P., Tucci, S., Gonzalez, L. and Hernandez, L. (2002) Haloperidol abolished glutamate release evoked by photic stimulation of the visual cortex in rats. Neurosci. Lett., 327: 149–152. Richter, D., Schmidt-Garcon, P., Pierreﬁche, O., Bischoff, A. and Lalley, P. (1999) Neurotransmitters and neuromodulators controlling the hypoxic respiratory response in anaesthetized cats. J. Physiol., 514: 567–578. Rossell, S., Gonzalez, L. and Hernandez, L. (2003) One-second time resolution brain microdialysis in fully awake rats. Protocol for the collection, separation and sorting of nanoliter dialysate volumes. J. Chromatogr. B. Analyt. Technol. Biomed. Life Sci., 784: 385–393. Skelley, A.M. and Mathies, R.A. (2003) Chiral separation of ﬂuorescamine labeled amino acids using microfabricated devices for extraterrestrial exploration. J. Chromatogr. A., 1021: 191–199. Smith, A., Watson, C., Kennedy, R. and Peris, J. (2003) Ethanol-induced taurine efﬂux: low dose effects and high temporal resolution. Adv. Exp. Med. Biol., 526: 485–492. Tucci, S., Contreras, Q., Paez, X., Gonzalez, L., Rada, P. and Hernandez, L. (2000) Medial prefrontal transection enhances social interaction. II: neurochemical studies. Brain Res., 887: 259–265.

Tucci, S., Rada, P. and Hernandez, L. (1998) Role of glutamate in the amygdala and lateral hypothalamus in conditioned taste aversion. Brain Res. Brain Res. Rev., 813: 44–49. Tucci, S., Rada, P., Sepulveda, M. and Hernandez, L. (1997) Glutamate measured by 6-s resolution brain microdialysis: capillary electrophoretic and laser-induced ﬂuorescence detection application. J. Chromatogr. B. Biomed. Sci. Appl., 694: 343–349. Wages, S., Church, W. and Justice, J.J. (1986) Sampling considerations for on-line microbore liquid chromatography of brain dialysate. Anal. Chem., 58: 1649–1656. Wang, L., Zhang, Z. and Yang, W. (2005) Pharmacokinetic study of trimebutine maleate in rabbit blood using in vivo microdialysis coupled to capillary electrophoresis. J. Pharm. Biomed. Anal., 39: 399–403. Wu, S. and Dovichi, N. (1989) High-sensitivity ﬂuorescence detector ﬂuorescein isothiocyanate derivatives of amino acids separated by capillary zone electrophoresis J. Chromatogr. A., 480: 141–155. Zetterstrom, T., Sharp, T., Marsden, C. and Ungerstedt, U. (1983) In vivo measurement of dopamine and its metabolites by intracerebral dialysis: changes after d-amphetamine. J. Neurochem., 41: 1769–1773. Zhao, Y., McLaughlin, K. and Lunte, C. (1998) On-column sample preconcentration using sample matrix switching and ﬁeld ampliﬁcation for increased sensitivity of capillary electrophoretic analysis of physiological samples. Anal. Chem., 70: 4578–4585.

CHAPTER 3.4

In vivo peptidomics: discovery and monitoring of neuropeptides using microdialysis and liquid chromatography with mass spectrometry Robert T. Kennedy Department of Chemistry, University of Michigan, Ann Arbor, MI, USA

Abstract: Neuropeptides are challenging molecules to monitor in the extracellular space of the brain because of low concentrations and the presence of many similar sequences due to enzymatic degradation of released peptides.Recent advances in sampling and analysis have improved the ability to detect, monitor, and discover novel neuropeptides. In sampling, use of higher molecular weight cut-off dialysis membranes, and solid supports offer improved recovery. For analysis, the advent of capillary liquid chromatography with tandem mass spectrometry enables sensitive detection (attomole detection limits) of peptides with sequence speciﬁcity. Application of proteomic methods allows novel neuropeptides to be identiﬁed. Combined with peptidomic analysis of tissues, a host of new peptides have been identiﬁed in the brain.

150–175 G protein coupled receptors that do not have known ligands (Katugampola and Davenport, 2003; Wise et al., 2004). Given that peptides are the largest family of signaling molecules, and that peptides are typically found to be the endogenous ligand for orphan receptors, it is reasonable to expect that the ligands for these receptors will usually be peptides (Saito and Civelli, 2005). Furthermore, peptides may act as co-factors or modulators of other receptors providing more opportunities for signaling. Advances in peptide identiﬁcation technology (see Sections II.C. and II.D.) have greatly increased the number of peptides and potential protein precursors that are known to be expressed in the CNS. The discovery of these new peptides provides evidence of active peptides and neurotransmitters that are yet to be identiﬁed. In this chapter we review emerging methods for the detection and identiﬁcation of neuropeptides in vivo with an emphasis on peptide discovery.

I. Introduction I.A. Neuropeptides in the central nervous system Neuropeptides are a group of intercellular signaling molecules present in the central nervous system (CNS) that function as neurotransmitters, neuromodulators, and neurohormones. Over 100 peptide neurotransmitters have been identiﬁed in mammalian systems and they have been implicated in a wide variety of physiological processes including memory and learning, pain transmission, and appetite control. Defects in the regulation of peptides are associated with several diseases including Alzheimer’s and Parkinson’s. In addition to the large suite of known neuropeptides, it is reasonable to expect that many unknown or unrecognized peptides act as CNS signaling molecules. Genomic methods have identiﬁed Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

279

DOI: 10.1016/S1569-7339(06)16016-6 Copyright 2007 Elsevier B.V. All rights reserved

280

I.B. Detection of neuropeptides in dialysate collected in vivo An important approach to studying the regulation and role of peptides involves their measurement in fractions collected by in vivo sampling methods such as microdialysis. Microdialysis permits samples to be taken from living animals with temporal resolution allowing the correlation between extracellular peptide level and behavior, stimuli, or pharmacological manipulations. Monitoring peptides in vivo by this method is difﬁcult because they tend to be present at trace concentrations in microdialysates. Peptide concentrations in brain extracellular ﬂuid are usually less than 1 nM and since the microdialysis probe only recovers a fraction of the extracellular concentration, peptides must be detected at even lower concentrations. As a result of these low levels, most peptide measurements in microdialysates have been made by radioimmunoassay (RIA) or enzyme-linked immunosorbent assay (ELISA) (Maidment et al., 1989, 1991; Consolo et al., 1994; Strecker et al., 2001). These methods have picomolar or better concentration detection limits. Although immunoassay allows detection at the recovered concentration, immunoassay mass detection limits usually prevent good temporal resolution for monitoring peptide concentration changes. The low temporal resolution arises since absolute recovery, deﬁned as mass removed per unit time by the sampling probe, is typically 15–50 amol/min for peptides whereas the mass detection limit for RIA’s are 100–500 amol (Maidment and Evans, 1991). Therefore the temporal proﬁle can only be monitored at 10–15 min, although 30 min intervals are more typical. Measurement of peptides in microdialysis sample is also complicated by the biochemistry of their production and degradation. Neuropeptides are processed from protein precursors (prohormones) within secretory vesicles by enzymes such as prohormone convertases that cleave the protein into active components. Further processing that may occur includes post-translation modiﬁcation such as acylation of the N-terminus and amidation of the C-terminus. Many prohormones contain multiple copies of a given active peptide sequence resulting in the possibility of several peptides with

sequence homology being produced. Furthermore, prohormones frequently contain the sequence of other active neuropeptides. A classic example is proenkephalin which contains the sequence for leu-enkephalin once, for met-enkephalin six times, in addition to the sequences for BAM, Peptide E, and Peptide I. Once released by the neuron, the peptides are degraded by a variety of exo- and endopeptidases. As a result, exocytosis of peptidergic secretory vesicles from a neuron may result in a multitude of peptides in the brain extracellular space with similar sequences and varying degrees of activities. The similarity of sequences creates signiﬁcant difﬁculties for immunological methods of analysis such as RIA and ELISA. Cross-reactivity among the similar peptides is common and typical results obtained by immunoassay must be reported as concentration of ‘‘immunoreactive peptide’’ rather than as the concentration of a speciﬁc peptide. Coupling immunoassay to HPLC alleviates this problem, but the resulting method is laborious and rarely practiced. Despite these limitations, RIA and ELISA are the methods of choice for detecting neuropeptides in dialysate and remarkable successes have been achieved. Nonetheless, the problems listed above severely limit the use of immunoassay. Furthermore, immunoassay is clearly not useful in identifying novel neuropeptides. Advances in capillary liquid chromatography (LC) with mass spectrometry (MS) detection have suggested a new route to both monitoring known peptides and discovering novel peptides in the CNS.

II. Neuropeptide detection and identiﬁcation in vivo using microdialysis combined with liquid chromatography with mass spectrometry (LC–MS) II.A. Detection of known endogenous peptides by LC– MS in dialysates II.A.1. Early work in endogenous neuropeptide detection Neuropeptide detection in the brain was revolutionized by Caprioli who ﬁrst demonstrated the possibility of detecting low-level peptides in microdialysates by tandem mass spectrometry (MS2)

281

(Emmett et al., 1995; Andren et al., 1996; Andren and Caprioli, 1999). In this seminal work, 10 mL microdialysis samples were preconcentrated and desalted on a microscale solid-phase extraction column coupled on-line to a triple quadrupole mass spectrometer by a microelectrospray interface. Using this approach, endogenous met-enkephalin and neurotensin were detected in microdialysate. This method had several key features that enabled neuropeptide detection. Samples were injected onto a miniaturized solid-phase extraction column, which consisted of a 50 mm inner diameter (i.d.) capillary packed with a short length of reversed-phase HPLC particles. With this device, peptides injected onto the column in dialysate were concentrated based on their adsorption to the short packed bed of particles. The column was then rinsed with low ionic strength solution to ‘‘desalt’’ the dialysate samples. This was signiﬁcant because high levels of salt, such as found in physiological ﬂuids, can disrupt ionization processes necessary for MS analysis. The analytes were eluted from the column and then directly interfaced to the mass spectrometer by microelectrospray ionization. Electrospray ionization is a commonly used approach for ionization and vaporizing peptides; however, as traditionally practiced at the time, it was not useful for detection of low attomole quantities. This work, along with an initial in vitro study by the same group (Andren et al., 1994) illustrated that using electrospray from a small bore capillary at low ﬂow rates could enhance the sensitivity. Finally, the mass spectrometer was a triple quadrupole instrument that allowed selected reaction monitoring (SRM). In this mode of operation, ions created at the source (microelectrospray in this case) enter the ﬁrst quadrupole or stage of the instrument. This quadrupole acts as a mass ﬁlter allowing only ions of a particular mass to charge ratio (m/z) into the second quadrupole. Within this second quadrupole a collisionally induced dissociation (CID) reaction, initiated by injecting a low pressure of gas such as He into the quadrupole, is used to dissociate the peptide into characteristic fragments (Fig. 1). The third quadrupole acts as a mass ﬁlter to allow selected fragment ions (so-called ‘‘daughter ions’’) to reach the detector. The advantage of single reaction monitoring is that it is highly speciﬁc for the target

Fig. 1. Peptide fragmentation by collision-induced dissociation. Product ions, arising from cleavage of the precursor ion’s peptide backbone and charge migration, retain charge on either the N-terminus (a-, b-, or c-type ions) or the C-terminus (x-, y-, or z-type ions).

analyte because the only ions detected are those with parent ions of a given m/z that fragment by CID to a given m/z. Such speciﬁcity makes the measurement sequence speciﬁc. It also greatly reduces background chemicals from reaching the detector. Endogenous met-enkephalin and neurotensin were measured with this method, although in principle it could be applied to any peptide. A limitation of this method is that it could only be used for one peptide at a time. This is because the elution from the solid-phase extraction column was initiated by a step change in the carrier ﬂuid so that all compounds were stripped from the column at once. In addition, the SRM mode requires that the instrument monitor a single peptide at a time.

II.A.2. Improvements in endogenous peptide detection by LC– MS Improvements in sensitivity and separations for endogenous monitoring have been introduced since this seminal work. In one study, met-enkephalin and leu-enkephalin were simultaneously monitored in microdialysis samples (Haskins et al., 2001). This work reported detection limits of 1 amol in 3 mL samples. This improvement was achieved through several key modiﬁcations that included the use of: (1) further miniaturization of the solid-phase extraction columns (25 mm i.d.) and electrospray emitters (3 mm tips) for better sensitivity (see Fig. 2), (2) optimized mobile phase gradients rather than step changes in carrier ﬂuid composition to allow for LC peptide separations and multi-analyte monitoring, and (3) a quadrupole ion trap (QIT) mass spectrometer. The smaller LC columns and tips allowed use of lower

282

Fig. 2. Photomicrograph of a capillary LC columns with integrated electrospray emitter. The column consists of a 25 mm inner diameter fused silica capillary packed with HPLC particles and pulled to a ﬁne tip for electrospray ionization. The bright ﬁeld optical image shows the end of the column with an emitter tip showing the packing material, frit to retain the material, and electrospray tip.

ﬂow rates (20 nL/min) and improved the preconcentration and ionization efﬁciency for peptides. The use of a gradient allowed chromatographic separation and therefore detection of more than one peptide in the sample. Finally, the QIT allows tandem MS that generates more information on smaller quantities than a triple quadrupole instrument. In the QIT, ions with a preselected m/z are trapped by electric ﬁelds within the mass spectrometer. Introduction of He for CID reactions results in fragmentation of the trapped peptide. A spectrum of the resulting fragments can then be collected. The entire process requires o1 s. Unlike SRM on the triple quadrupole, this form of SRM allows all of the characteristic daughter ions to be detected thus providing greater conﬁdence in detection and lower background. Improvements by the MS manufacturer in the ion transmission characteristics also aided the sensitivity. In this method, to detect multiple ions, the SRM of the MS is operated in a time-segmented mode in which the ion of interest is monitored during the time that it should be eluting from the chromatography column. After elution of the target peak, the target ion can be changed to the next eluting peak. This feature was used to detection met- and leu-enkephalin in single samples. This approach offered excellent sensitivity and speciﬁcity for detection. Indeed, it was estimated that with the detection limits achieved, it would be possible to detect the enkephalins at 3–5 min intervals. However, this temporal resolution was

not realized because the system was operated on-line with microdialysis sampling. As a result, the temporal resolution was limited by the analysis time, which included sample loading and preconcentration, desalting, separation, and column reequilibration, to 30 min. A further enhancement of this approach to monitoring was reported recently in which higher order mass spectrometry, MS3, was used (Baseski et al., 2005). In MS3, a particular daughter ion is trapped and dissociated for further analysis. Much like the advantages of MS2, it is expected that with each stage of MS one gains speciﬁcity and improves the signal-to-noise ratio (S/N). In fact, for each stage of MS some loss of signal occurs along with the reduction of noise. Therefore, the S/N does improve with further stages until enough signal loss occurs to prevent detection of the ions. With MS3, the detection limit was improved for real samples and detection of enkephalins. This work also illustrated an improvement of temporal resolution to 15 min by further optimizing the gradient. Thus, the work for detecting preselected endogenous peptides by LC–MS in dialysates has demonstrated that it is possible to detect such peptides with good sensitivity and speciﬁcity. While the methods using the QIT illustrated better sensitivity, it is likely that newer triple quadrupole instruments, which have improved ion transmission characteristics, used with similarly miniaturized columns would offer sensitivity that approaches that achieved by the QIT. Furthermore, the triple quadrupole is more likely to yield better quantitative results because of better reproducibility achieved with this type of mass spectrometer. Preliminary work has suggested that in real samples, however, peptide quantiﬁcation at low concentrations may require an internal standard even on the triple quadrupole instrument (Sinnaeve et al., 2005). Therefore, for monitoring known endogenous compounds use of miniaturized LC with triple quadrupole MS in the SRM mode would offer the best analytical results; however, this has yet to be demonstrated.

II.A.3. Improving recovery of neuropeptides The initial reports of detecting endogenous neuropeptides by LC–MS emphasize the method of

283

analysis. Sampling and sample handling are also highly important in collecting samples. Several studies have discussed the importance of identifying the best dialysis membranes for a given peptide (Schutte et al., 2004; Hutchinson et al., 2005; Rosenbloom et al., 2005). Typically, the molecular weight cut-off and the material used for the dialysis membrane are both important in collecting peptides without signiﬁcant losses to adsorption. Storage of dialysates with trace level peptides can be problematic. Indeed, one study reported that overnight storage resulted in complete loss of signal for trace level enkephalins, presumably due to adsorptive losses (Haskins et al., 2001). While on-line analysis solves the storage problem, it is not always practical. An intriguing approach to solving some of the problems of low recovery and sample storage of peptides has been reported that uses solid supported enhanced microdialysis (Pettersson et al., 2001, 2004; Ao et al., 2004). This method builds on earlier work that demonstrated enhanced recovery for small molecules by including cyclodextrins in the dialysis media (Khramov and Stenken, 1999; Ao and Stenken, 2003). In this method, particles similar to HPLC supports or antibody-coated beads are included in the dialysis ﬂuid. When peptides cross the dialysis membrane they may adsorb to the particles. The microdialysis samples are collected in fractions as usual. The peptides are released from the solid particles by solvent treatment, dried, and resolved in a formic acid buffer to make them suitable for capillary LC–MS. As a result of adsorption to the solid particles during dialysis, the free peptide concentration remains low in the dialysis probe and the concentration gradient across the membrane remains high. Maintaining the concentration gradient improves the recovery. Using the HPLC supports, six endogenous neuropeptides were tested to investigate the feasibility of this enhanced microdialysis methodology. The improved relative recovery obtained from the solid supported enhanced microdialysis was varying from no effect to 10 times higher as compared with ordinary microdialysis. The most efﬁcient enrichment was obtained for luteinizing hormone releasing hormone, which was the largest but also the most hydrophilic of the

peptides tested. In contrast, no signiﬁcant difference in recovery was observed for leu-enkephalin being the smallest and the most hydrophobic peptide tested. For the antibody beads, several cytokines were tested and a more consistent 5–20 fold increase in recovery was observed (Ao et al., 2004). The differences in enhancement were attributed to differences in the diffusion and transport properties of the peptides and beads. An advantage of the antibody method is that the beads were directly suitable for ﬂow-cytometry type assays. A follow-up study illustrated that the temporal proﬁle of the monitoring is not disturbed by the presence of the beads (Ao and Stenken, 2006). The use of solid particles provides both an enhanced recovery, for some peptides, and possibly a method to avoid sample losses during peptide transport and storage. This method is yet to be utilized for actual in vivo sampling and analysis. Some problems will need to be overcome. The most signiﬁcant is developing a ﬂow system that is compatible with both the particles and in vivo sampling. Further investigation of this and similar approaches is warranted.

II.B. Detection of exogenous peptide and monitoring peptide metabolism by microdialysis and mass spectrometry The Caprioli method for monitoring endogenous peptides has also been extended to monitor metabolism of exogenously introduced peptides (Andren and Caprioli, 1995; Nydahl et al., 2003). In this experiment, a known peptide was infused through the dialysis probe so that it could be delivered to the extracellular space. The probe also recovered fragments of the peptide that were then analyzed by capillary LC with MS detection. To allow this analysis, several modiﬁcations in the analysis method were required. The solid-phase extraction column was lengthened and operated with a gradient of HPLC mobile phase so that peptides could be separated and enter the ionization source at different times. This allowed different metabolites to be detected in one sample. Another modiﬁcation was that the mass spectrometer was operated in ‘‘scanning’’ mode that

284

allowed new peptides to be detected and their fragments analyzed for identiﬁcation. Scanning modes on a triple quadrupole mass spectrometer are inherently less sensitive than SRM mode because of the low duty cycle for detection of a given ion. (In SRM, the ion is continuously detected; whereas in a scanning mode a given ion is only detected a small fraction of the time as the mass spectrometer scans across the entire m/z range.) For this experiment, however, the loss of sensitivity was not problematic because the exogenously added peptide was at a sufﬁciently high concentration for detection. A modiﬁcation of this method has been to use capillary electrophoresis and matrix-assisted laser desorption ionization (MALDI) MS for the analysis (Zhang et al., 1999). This approach seems to be a highly effective way to detect and identify the metabolism of peptides that are released into the extracellular space. In one intriguing example, the in vivo biotransformation of dynorphin A(1–17) (Dyn A) was studied in the striatum of hemiparkinsonian rats (Klintenberg and Andren, 2005). (For this experiment, the mass spectrometer was a time-of-ﬂight mass spectrometer.) The microdialysis probes were implanted into both hemispheres of unilaterally 6-hydroxydopamine (6-OHDA) lesioned rats. Dyn A was infused through the probes and the resulting metabolites monitored. It was observed that Dyn A metabolism was altered in the lesioned side. Dyn A metabolites 1–8, 1–16, 5–17, 10–17, 7–10, and 8–10 were detected in the lesioned side but not in the untreated striatum. Several other fragments were found in both sides; however, they were lower in the dopamine-depleted striatum. The results illustrate for the ﬁrst time that the extracellular in vivo processing of the dynorphin system is disturbed in this animal model of Parkinson’s disease. Several modiﬁcations to this general approach have been applied. In one case, MALDI-MS was used to monitor the metabolism of Dyn A (Reed et al., 2003). The peptides detected were consistent with enzymatic cleavage at the Arg7-Ile8 position of Dyn A), followed by terminal degradation of the resulting Dyn A (1–7) and Dyn A (8–17) peptides. Interestingly, novel post-translational modiﬁcations were found on C-terminal fragments of Dyn A. Using tandem MS, several other

modiﬁcations were detected including a covalent modiﬁcation of mass 172 Da on the tryptophan residue of C-terminal fragments. While the structure and role of these modiﬁcations was not determined, these results illustrate the power of using the exogenous peptide application combined with MS techniques to monitor peptide metabolism and discover unexpected metabolites. Similar approaches can be applied to peptidergic drug candidates (Prokai et al., 2001). In this case, the drug may be given intravenously and its transport into the brain monitored by microdialysis. Use of MS detection allows identiﬁcation and characterization of the metabolites. This work has used tandem MS on a QIT and nanoelectrospray on a Fourier transform ion cyclotron resonance mass spectrometer. These studies revealed that MS provides useful information about the extent or mechanism of transport and metabolic activation/ inactivation in early-phase discovery and development of CNS agents.

II.C. Discovery of novel peptides in dialysate II.C.1. Bottom-up proteomics The above results show that LC–MS is a useful approach for detecting known peptides in dialysis samples. This application however only touches on the power of MS because MS can also be used to identify novel compounds. The potential of using MS to identify neuropeptides became apparent with the advent of ‘‘shot-gun’’ or ‘‘bottom-up’’ proteomics (Hunt et al., 1986; Eng et al., 1994; Link et al., 1999; Washburn et al., 2001). In this form of proteomics, a protein mixture is treated with enzymes, such as trypsin, to digest the protein. The resulting peptides are separated by capillary LC with MS detection. The mass spectrometer used is capable of MS2, usually a QIT, quadrupoletime-of-ﬂight (Q-TOF), or similar instrument. For these experiments, the mass spectrometer is operated in a ‘‘data-dependent’’ mode. In this method, the MS acquires a spectrum of the ions eluting from the chromatography column. If an ion is present above a pre-determined threshold, the MS will then acquire a MS2 spectrum by applying CID to that ion. After collection of this spectrum, the

285

MS will begin full scans again and the process repeated throughout the elution window of the separation. The resulting data output of this experiment is a collection of MS2 spectra with parent ion m/z acquired at speciﬁc retention time. The next step in the analysis is to interpret the spectra and identify the peptides that were detected. While it is possible to interpret such spectra and attempt to identify the peptide manually, this is an exceedingly difﬁcult task, especially when hundreds or thousands of spectra are acquired in single analyses. Spectral interpretation and peptide identiﬁcation has been automated using software such as SEQUEST (Eng et al., 1994) and Mascot (Perkins et al., 1999; Creasy and Cottrell, 2002). These software packages use algorithms that correlate the MS2 spectra to the spectra predicted for peptides generated with a given enzyme speciﬁcity from proteins in a database. Statistical methods are used to score the correlation of experimental and predicted spectra and determine if the spectra actually matches one of the peptides. The resulting matches can then be used to identify the proteins in the mixture.

II.C.2. Application of peptidomic analysis to microdialysis samples Given that the neuropeptides in the brain arise from proteolytic cleavage of proteins, and that such peptides can be detected by tandem MS in dialysate, it is reasonable to expect that the LC–MS2 approach could be used to identify peptides detected in dialysate (see Fig. 3 for method ﬂow chart). Application of this method to endogenous neuropeptides in dialysate is complicated by several factors. A signiﬁcant issue is that the proteases used for producing the neuropeptides are not always known nor are they under experimental control. Therefore, the peptides are not necessarily tryptic, creating at least two difﬁculties. Peptides produced by other proteases may be less sensitively detected than tryptic peptides that have basic sites (because of the proteolytic speciﬁcity of trypsin, tryptic peptides are terminated by basic residues that ionize well in electrospray sources) resulting in good ionization efﬁciencies. The presence of basic residues also promotes efﬁcient

cleavage along the peptide backbone during CID resulting in high yield of the b- and y-type ions (see Fig. 1) used for peptide identiﬁcation. The second difﬁculty is that without knowledge of proteases involved in peptide production, a larger library of peptides must be searched thus increasing the probability of ﬁnding a random match. Another signiﬁcant issue is the effect of data-dependent scanning on sensitivity. Detection of known neuropeptides is more sensitive because the mass spectrometer can be used to target the speciﬁc ion of interest. In data-dependent scanning, if a peptide does not generate signal above the threshold, either because of low concentration or poor ionization efﬁciency, it will not be detected. More troublesome, if other peptides or ionizable substances with higher signal co-elute with a neuropeptide, then the neuropeptide will not be ‘‘selected’’ by the data-dependent scanning for MS2 analysis. These issues suggest that more care must be taken in using this approach for neuropeptides than for more conventional shotgun proteomics. Successful identiﬁcation of neuropeptides in dialysis samples has been reported using this approach (Haskins et al., 2001, 2004). Using chromatography conditions similar to that used for known neuropeptide and a QIT, neuropeptides were identiﬁed in dialysate collected from the striatum of anesthetized rats. In this project, dialysate was collected under both basal conditions and with high K+ perfused through the probe from 13 animals. In the basal conditions, no neuropeptides were successfully identiﬁed. (Subsequent work by this laboratory has identiﬁed several neuropeptides in the basal condition.) In the K+-stimulated case, 29 peptides, including 25 that had not previously been reported, were identiﬁed from 6 different protein precursors. Proteins identiﬁed include precursors to neuropeptides (proenkephalin A), synaptic proteins (neurogranin, brain acidic membrane protein), blood proteins (ﬁbrinogen), and transporters (excitatory amino acid transporter 1). Several factors were important in the success of this method. The use of K+ stimulation depolarized neurons and presumably increased the release of neuropeptides; although this may not always be

286 Neuron Sampled Region Release from Synaptic Vesicles Protein Precursors

Enzymatic Processing

Neuropeptides

Microdialysis-CLC-MS2

CLC-MS2 Chromatograms Data Reduction

MS2 Spectra Database Searching & Sequence Validation

Neuropeptides Compare ‘Basal’ & ‘Stimulated’

Protein Precursor Identification

Protein Precursors Fig. 3. Flow chart for multi-dimensional method of discovering neuropeptides by in vivo microdialysis-CLC-MS2. Neuropeptide synthesis in the brain by enzymatic processing of protein precursors released from synaptic vesicles is followed by collection of MS2 spectra at basal and stimulated levels, peptide sequencing and protein precursor identiﬁcation.

necessary. To improve the sensitivity sufﬁciently to allow multiple peptides to be identiﬁed, injection volumes onto the column were increased to 3.6 mL and the chromatography ﬂow rate reduced from 20 nL/min for monitoring to 10 nL/min for peptide discovery. For conﬁrmation of peptide identiﬁcation, two strategies were used. In some cases, the

identiﬁed peptides were synthesized and the synthetic version shown to match the retention time and MS2 spectra of the endogenous compounds. In other cases, it was found that using two different peptide identiﬁcation algorithms, SEQUEST and Mascot, provided higher conﬁdence in peptide identiﬁcation.

287

The most interesting of the 29 peptides identiﬁed in this study were the 6 fragments of proenkephalin A because this protein is a known source of neuroactive peptides. Fragments detected included met-enkephalin and leu-enkephalin, and several non-opioid fragments. Despite extensive studies of proenkephalin processing, six of the eight proenkephalin peptides detected had not been observed before or were only recently revealed by peptidomic methods. The heavy reliance of previous methods on antibodies that bound enkephalinergic epitopes contributed to the lack of previous observation of these peptides. Detection of these previously unknown peptides demonstrates the power of a peptidomic approach to study processing of endogenous neurotransmitter precursors. The biological signiﬁcance of the peptides from other proteins, such as neurogranin, that was detected remains to be determined. The presence of speciﬁc peptides in the dialysate samples raises the issue of proteases involved in generating these fragments. Four of the proenkephalin A peptides identiﬁed were encompassed by dibasic sites consistent with production through prohormone convertase activity followed by carboxypeptidase cleavage of C-terminal basic residues. Peptide fragment production from the other proteins is uncertain. A frequent observation was nesting of sequences. For neurogranin, seven peptides were found, but all were within the C-terminal region of 53–78 (out of 78) amino acids. A likely route to production of nested peptides is cleavage by an endopeptidase at a speciﬁc residue followed by carboxy- and aminopeptidase activity to create the other peptides.

II.C.3. Top-down proteomics: proteomic analysis dialysate Approaches to identify proteins in microdialysates have been reported. In one study, the cerebral protein expression in microdialysate from three stroke patients sampled from the hemisphere contralateral to the lesion was determined using a ‘‘top-down’’ proteomic method (Maurer et al., 2003). Top-down proteomics involves separation of whole proteins, usually by two-dimensional gel electrophoresis, rather than protein digests.

Resulting spots are extracted, digested with trypsin, and then analyzed by MS (commonly MALDI-MS). Protein identiﬁcation is achieved by correlating the individual masses to peptides expected to be found in a protein. In this study an average of 158 protein spots in the human cerebral microdialysate were found. A total of 27 individual proteins were identiﬁed. While most of proteins had previously been detected in human cerebrospinal ﬂuid, 10 additional proteins mainly of cerebral intracellular origin, were identiﬁed exclusively in the microdialysate. From this study it was concluded that the 10 proteins detected could be used to monitor progression of disease and stroke damage toward deterioration. Indeed, the correlation of protein composition in the human cerebral microdialysate with clinical condition may be a useful approach to future applications for neurological stroke diagnosis, prognosis, and treatment. While this method was applied to pathophysiological analysis, it is possible that with improvements in sensitivity, larger neuropeptides could be detected by this method. Several related studies on the proteome of cerebral spinal ﬂuid suggest that more work should be done in this area (Suzuyama et al., 2004; Yuan and Desiderio, 2005).

II.D. Peptidomic analysis of nervous tissue as a complement to in vivo analysis II.D.1. Peptidomic methods for CNS tissue Identiﬁcation of neuropeptides in microdialysate is still an emerging method. While some success has been achieved, the number of novel peptides detected remains relatively small. Another approach that may serve to complement the in vivo methods is peptidomic analysis of nervous tissue. Initial studies suggest that many more peptides can be detected and identiﬁed by this approach than the in vivo method. Peptides identiﬁed in tissue could serve as a guide for identiﬁcation of neuropeptides that are released and bioactive and should be present in dialysates. It may be easier to identify neuropeptides in tissue samples because the neuropeptides are more concentrated, there should be less chance for

288

extracellular enzymes to degrade the peptides resulting in less confusion to active peptides, and more sample is available. At the same time however, the resulting mixtures may contain more interfering substances because of the background from intracellular components. Therefore, for tissue methods a greater premium is placed on quality of the separation. As discussed below, it has also been found that sample preparation is a critical factor in determining the success of the method. Early work in this area utilized capillary LC with Q-TOF MS to analyze brain sections from rats (Skold et al., 2002; Svensson et al., 2003). Sample preparation included tissue homogenization and ﬁltration to remove larger proteins. The resulting peptide enriched fraction was then analyzed by capillary LC with data-dependent MS2. Initial studies, while revealing some neuropeptides, detected primarily peptides from blood such as fragments of hemoglobin. This effect can be attributed to post-mortem peptidase activity that not only reduces the levels of neuropeptides of interest, but also creates peptides from abundant proteins in the tissue samples such as hemoglobin. The high levels of such peptides can have the effect of masking the presence of lower levels of neuroactive compounds during data-dependent scanning operations. One interesting approach that avoids this problem has been to analyze tissue from mice with a mutation that inactivates carboxypeptidase E (Cpefat/fat mice) (Bures et al., 2001; Che et al., 2001, 2005c). This enzyme is one of the most important peptide degradation enzymes and often removes the basic residues from the C-terminus after prohormone convertase activity. Analysis of brain tissue from mice with this mutation allowed identiﬁcation of 100 peptides representing fragments of 16 known secretory pathway proteins including proenkephalin, proopiomelanocortin, protachykinins A and B, chromogranin A and B, and secretogranin II. Many of the identiﬁed peptides represent previously uncharacterized fragments of the precursors. For example, 12 of the 13 chromogranin B-derived peptides found in this study had not been previously reported. The ability to detect neuropeptides was likely due to the accumulation of peptides because of the lack of

degradation and the ability to selectively preconcentrate these peptides on an anhydrotrypsin afﬁnity resin. These factors could overcome the masking effect of abundant protein fragments and the post-mortem degradation of neuropeptides. While increasing the concentration of peptides related to signaling prior to the subsequent LC–MS2 analysis, this method is limited to intermediates with C-terminal basic residues (i.e., the peptides detected are not the mature peptide). The method is also limited by the availability of mice with the mutation. In a major advance for the ﬁeld of tissue neuropeptidomics, it was revealed that if the animal was sacriﬁced by focused microwave irradiation, then the yield of blood peptides was reduced and that of the neuropeptides was dramatically increased (Svensson et al., 2003). Focused microwave irradiation rapidly ﬁxes tissue and terminates enzyme activity by raising the brain temperature to 901C in o1 s. By rapidly stopping endopeptidase activity, neuropeptides are not degraded during post-mortem sample preparation resulting in higher detectable levels of the active peptides. Furthermore, abundant proteins from blood are not degraded by peptidases during sample preparation. Indeed, 550 pepidergic peaks were detected in this work and many were identiﬁed using database searching methods (see Fig. 4 for a sample data

Fig. 4. Two-dimensional graph of the neuropeptide content from 1 mg of rat hypothalamus. The neuropeptide map displays peptides in the m/z range 300–1,000 in 60-min gradient elution from a capillary LC column. Spot intensity is represented by color changes with black being the most intense reading and white the lowest.

289

trace from this method). Many of the classical neuropeptides, such as neurotensin, substance P, neurokinin A, corticotropinlike intermediate lobe peptide (CLIP), and beta-endorphin were identiﬁed in both rat and mouse brain tissue. From the proSAAS precursor, big PEN and little SAAS were identiﬁed. Relatively high-levels of the enkephalin variants leu-enkephalin, metenkephalin, met-enkephalinRF as well as the 4.44 kDa lipotropin gamma peptide from the proopiomelanocortin precursor were sequenced and identiﬁed. Furthermore, CLIP was identiﬁed containing Arg in the N-terminal. In addition to the neuropeptides, novel post-translational modiﬁcations of peptides were detected. In the rat hypothalamus, a novel peptide from the CART protein was identiﬁed carrying a phosphorylation that had not previously been described. Melanotropin alpha was modiﬁed with both amidation and acetylation at two sites. Besides these novel modiﬁcations, many expected modiﬁcations such as carboxyamidation were observed. The abrupt block of enzymatic activity also presumably accounts for the ability to detect the modiﬁed neuropeptides which might otherwise be modiﬁed by release of enzymatic activity during sample preparation. While the microwave ﬁxation method allows the use of rodents without mutations, the instrument used is expensive ($50,000 for a typical instrument) and as a result not widely available. An alternative approach that was also reported to reduce the background from degradation of abundant proteins was animal sacriﬁce by decapitation followed by rapid microwave using an ordinary microwave oven (Che et al., 2005b). While not completely eliminating background protein, the use of post-mortem microwave treatment dramatically reduced their presence suggesting that ﬁxation at this time is sufﬁcient to allow neuropeptidomic studies. In this study, 41 fragments of secretory pathway proteins were identiﬁed after microwave treatment. The peptides were derived from 15 proteins: proopiomelanocortin, proSAAS, proenkephalin, preprotachykinins A and B, provasopressin, prooxytocin, melanin-concentrating hormone, proneurotensin, chromogranins A and B, secretogranin II, prohormone convertases 1 and 2, and peptidyl amidating monooxygenase. Two of

the protein fragments corresponded to novel protein forms: VAP-33 with a 7-residue N-terminal extension and beta tubulin with a glutathione on the Cys near the N-terminus. The presence of these post-translationally processed peptides also demonstrates the effectiveness of the treatment in preserving such modiﬁcations. While prevention of post-mortem enzymatic activity appears to be a key step in allowing largescale peptidomic studies; several other signiﬁcant factors have also been found. Compared with previous work on dialysate, the tissue analysis methods utilized longer columns and shallower mobile phase gradients for higher resolution separations. The use of the Q-TOF instrument, which provides higher mass accuracy in the MS2 spectra, allowed more conﬁdent identiﬁcation of the peptides. Another study revealed that aggressive sample preparation such as ultrasonication for tissue disruption combined with surfactant and acid extraction can improve the yield of neuropeptides (Parkin et al., 2005). This study also revealed that the current bioinformatics for automated interpretation of mass spectra is not always adequate for identifying neuropeptides. For example, neurotensin is a known neuropeptide in striatum; however, its fragmentation pattern is atypical, due to a pair of Arg residues in the sequence. Such dibasic residues can sequester protons and prevent them from catalyzing the cleavage along the peptide backbone that is necessary to generate b- and ytype ions normally used for peptide identiﬁcation. In the case of known peptides, being aware of such problems does allow spectra for these compounds to be interpreted manually.

II.D.2. Quantification in tissue Quantiﬁcation of neuropeptides in tissue by peptidomic methods is not a simple task because the ion intensity in the scans is highly irreproducible. Borrowing from the shot-gun proteomics, the use of stable isotope labeling strategies has been reported (Che et al., 2005a; Che and Fricker, 2005). In this approach, a control and experimental samples are labeled with isotopically distinct tags. Both succinic anhydride with either four hydrogens or deuteriums and

290

[3-(2,5-dioxopyrrolidin-1-yloxycarbonyl)propyl]trimethylammonium chloride with either nine hydrogens or deuteriums were successfully used in analysis of mouse pituitary tissue. While both peptides could give useful results, performance differences were observed. All endogenous mouse pituitary peptides labeled with the light TMAB reagent eluted at the same time as the peptides labeled with the heavy reagent. Similar results were achieved with the succinyl labels. The greater mass difference with the TMAB label (9 Da per label compared with 4 Da per label for the succinyl label) allowed more accurate relative quantiﬁcation. Interestingly, with 82 total peptides detected, only 16 were detected with both labels. TMAB allowed detection of 41 peptides and the succinyl allowed detection of 25 peptides. These results suggest that the labeling strategies are complementary. This procedure has been used to quantify changes associated with the Cpefat/fat mutation in the prefrontal cortex (Lim et al., 2006). This method is limited to peptides with free amines.

II.D.3. Emerging methods for tissue peptidomics Other approaches are emerging for neuropeptidomic analysis of mammalian tissue. The use of capillary LC with electrospray ionization has allowed detection and identiﬁcation of impressive numbers of peptides; however, it is unlikely that all the neuropeptides in tissue are actually measured by this approach because many known peptides are not detected in CNS samples (Parkin et al., 2005). MALDI is a complementary ionization technique to electrospray; therefore, the combination of the two is expected to improve the yield of peptides that may be detected over use of either method alone. Indeed, MALDI-MS analysis has proven powerful for studies of invertebrate systems where speciﬁc neurons can be isolated and directly analyzed (Rubakhin et al., 2003). The use of LC to separate a peptide mixture prior to MS analysis is expected to improve the coverage of the peptidome by reducing ion suppression and resolving isobaric species. Indeed, off-line coupling of capillary LC to MALDI-MS has been applied to the analysis of invertebrate neuronal tissue homogenates (Hsieh et al., 1998; Floyd et al.,

1999). This method may prove useful for either tissue or dialysate analysis of neuropeptides.

II.D.4. Summary and comparison to in vivo methods Neuropeptidomic analysis of brain tissue has rapidly advanced over the past few years. Analysis of 1 mg tissue quantities can yield identiﬁcation of a 30–100 peptides. Often these peptides have not been previously reported. Quantiﬁcation is possible with stable isotope labeling strategies. These techniques will allow many new studies to be performed. The ability to quantify multiple peptides will allow determination of how treatments and pathophysiology alter peptide content and expression. The quantiﬁcation of the effect of the Cpe mutation is only the initial study in this regard. Novel peptides that are identiﬁed are also candidates as neuromodulators or neurotransmitters, perhaps providing the ligand for the many orphan receptors. While the tissue methods are invaluable, they cannot replace the in vivo measurements. Neuropeptidomic measurements made with in vivo sampling allow determination of the peptides that are actually released, their lifetime, and post-release processing. The extracellular neuropeptidome is more dynamic than the intracellular neuropeptidome and can change on rapid time scale with behavior, physiological state, pharmacological treatment, and external stimulus. Thus, the in vivo measurements may allow discovery of peptide patterns correlated with different behaviors, pathophyisological outcomes (Maurer et al., 2003), or consciousness. An intriguing example of this latter possibility was reported in a preliminary form (Kennedy et al., 2002). In this report, it was shown that the detected peptide proﬁle varied with the sleep–wake cycle. Unfortunately, the peptides were present at too low level to yield MS2 spectra with sufﬁcient quality for peptide sequence assignment. Nonetheless, this result suggests the exciting possibility of identifying signaling molecules associated with a given behavioral or conscious state. Making these applications method a reality will require several technical improvements. The most signiﬁcant is improving the S/N of the low-level

291

peptides during data-dependent scanning. This may be achieved through better sampling and sample preparation, such as use of sampling probes with higher peptide recovery and more extensive preconcentration. Newer MS instrumentation may also provide a route to improve the sensitivity. Previous work on dialysate neuropeptidomics has utilized QIT; however, the Q-TOF instrument and the new linear ion traps may offer better sensitivity.

IV. Screening and identiﬁcation of function of novel neuropeptides IV.A. Identifying active peptides The advent of neuropeptidomic methods has greatly improved our ability to detect and identify peptides in dialysate and tissue. The knowledge of these peptides however leads to the problem of determining if the detected peptides are neuroactive, that is, do they play a role in nervous system function by acting as neurotransmitters, neuromodulators, or regulating another process? Historically, active neuropeptides have typically been identiﬁed by using bioassay directed puriﬁcation. In this approach, fractions from separation of crude tissue or cultured cell extracts are tested for activity in a bioassay. Active samples are further fractionated until a pure bioactive compound is isolated (Jorpes and Mutt, 1966; Hughes et al., 1975). With neuropeptidomic methods, peptides are identiﬁed ﬁrst without regard to activity. More akin to this approach are methods that seek peptides with speciﬁc sequences, modiﬁcations, or distributions that are expected of neuropeptides. For example, in one approach cDNA clones were identiﬁed that had distribution consistent with neuropeptide distributions and corresponding amino acid sequences that had signal peptide at the N-terminus and prohormone cleavage sites. The predicted peptides could then be synthesized, used to raise antibodies, and the antibody used to extract the putative endogenous peptides (Douglass et al., 1995; De Lecea et al., 1998). It has also been possible to screen mixtures for

peptide modiﬁcations that are common in neuropeptides. A classic example is screening peptide fractions for carboxyamidation. This approach was used to discover peptides such as NPY and galanin (Tatemoto and Mutt, 1980). The biological activity of these peptides was later identiﬁed. The neuropeptidomic methods tend to be more unbiased than these approaches in the peptides that are identiﬁed. Furthermore, their application has resulted in an explosion of peptide sequences known to be expressed in the CNS. With the large number of peptides identiﬁed by new neuropeptidomic methods, it becomes advantageous to develop new approaches to screen for active peptides. The Cpefat/fat mice offer one opportunity for selecting possibly active peptides from the peptides detected (Che et al., 2005a). The rationale for selecting possible active peptides relies on the fact that Cpe is normally a late processing step for active peptide production. To capitalize on this observation, peptide levels were compared in wild type and Cpefat/fat mice (Lim et al., 2006). Peptides that decreased signiﬁcantly in the Cpefat/fat were presumed to be peptides that were formed by this enzyme and therefore potentially bioactive. Indeed, many of the peptides that decreased were known active peptides. This procedure also identiﬁed several previously unknown peptides such as novel fragments of VGF, procholecystokinin (proCCK), and prohormone convertase 2. Besides acting as neurotransmitters or neuromodulators, the peptides could have other activities. Sequence homology for some of the peptides found suggested that they could be involved in regulating prohormone covertase 1 (PC1). For example, proCCK fragments that contained LGALLA, which is similar to the LGALL sequence in peptide PEN that is known to be important in inhibiting PC1 and helping to regulate PEN-processing. The strategy discussed above allows peptides with likely activities to be identiﬁed based on homology with other known systems. However, this approach relies on knowledge of all peptide processing steps, which may be incomplete. In addition, it does not necessarily help resolve the function. Therefore, other methods are also needed.

292

IV.B. Determination of peptide function The in vivo measurements of peptides provide another strategy to screen for biological activity of peptides (Haskins et al., 2004). By identifying peptides that are released by depolarizing stimuli, the method provides the ﬁrst evidence for biological activity. Presumably released peptides are likely to be stored in neuronal vesicles. A caveat to this presumption is that once released, a peptide may be degraded; therefore, the actual active and/ or released form is not necessarily obvious. For example, both of the proenkephalinergic peptides SPQLEDEAKE (Peptide I 1–10) and SPQLEDEAKELQ (Peptide I 1–12) were observed in dialysate samples. It is unclear from this observation if both peptides were released or if the longer peptide was released and extracellular carboxypeptidase activity yielded the shorter sequence. Similar overlaps were found for many of the peptides detected in vivo. Observations from tissue analysis may help to resolve this issue, that is, previous reports detected the longer peptide is tissue. Another caveat of the release measurement is that the depolarizing stimuli increased the level of peptides (at least made them detectable in dialysate) that would not ordinarily be considered potential transmitters such as fragments of amino acid transporters or blood peptides. It is possible the high levels of K+ used for the stimulation resulted in alterations other than neuronal depolarization that caused the formation of these peptides. It is also possible that such fragments serve some signaling function apart from neurotransmission. The strategy of investigating released peptides in the brain has only barely been investigated. Future work would require experimental manipulations, such as electrical stimulation of speciﬁc tracts and assessment of tetrodotoxin (TTX)-sensitivity or Ca2+ dependency to better understand the conditions under which these peptides are released. A second in vivo strategy for screening for biological activity has also been demonstrated (Haskins et al., 2004). In this case, peptides identiﬁed by neuropeptidomic analysis of dialysate were synthesized and infused by ‘‘reverse dialysis’’ into the brain region where they were originally

discovered to be released by depolarizing stimuli. While the peptide was infused, amino acid neurotransmitters were monitored using on-line capillary electrophoresis to determine if the peptide altered their levels. The underlying rationale of this strategy is that if a peptide is a neurotransmitter or neuromodulator, it is likely to exert its effects by altering release of other transmitters. Using this approach, six novel peptides derived from proenkephalin were tested and three were found to increase the levels of at least one of the ﬁve neuroactive amino acids monitored (glutamate, aspartate, GABA, glycine, and taurine). Two of the peptides (Peptide I 1–10 and Peptide I 1–12) had similar effects by raising excitatory amino acids. The effects did not appear to require neuronal ﬁring because they were insensitive to TTX; however, alteration in excitatory amino acid level evoked by these peptides could alter excitability of neighboring neurons and therefore modulate neuronal signaling. Another peptide, BAM 8–18, increased the levels of GABA and taurine in a TTX-sensitive fashion suggesting it effects neuronal signaling. It was concluded from these results that proenkephalin A processing leads to peptide products that are released into the extracellular space with neurochemical activity not associated with the opioid receptor. The differential effects of these peptides illustrate the diverse signaling possible by a single precursor protein. While helping to ﬁnd possible active peptides, this method has limitations. The lack of an effect by a peptide on amino acid levels does not eliminate its consideration as a bioactive molecule. Clearly an active compound could exert its effects on other neurotransmitters such as acetylcholine, dopamine, or substance P. In addition, peptide effects on release of neurotransmitters may not be apparent except under certain conditions. A peptide may, for example, inhibit release when the target neuron is ﬁring at a certain rate. Therefore, this method could be improved by assessing the effect of peptides on other neurotransmitters and under different conditions. Nonetheless, this method seems to offer a novel general strategy for identifying active peptides.

293

Other more established strategies may also be used to identify peptides. In one case, an orphan G protein coupled receptor was screened against putative fragments of proenkephalin that had been synthesized (Lembo et al., 2002). In this study, a slight extension of BAM 8–18 was found to activate the receptor. It seems likely that a combination of methods would be productive. That is, use of neuropeptidomic methods may be used to detect and identify peptides. Potentially active peptides from this group could be selected for further study based on their possible processing sources and their release in the brain (assessed by in vivo methods). These peptides could then be tested with a variety of traditional receptor or other screens and newer strategies such as detection of the effect on other transmitters discussed above. This combination of results, along with knowledge of the location and conditions for secretion would help in identifying a novel peptide as a transmitter.

References Andren, P.E. and Caprioli, R.M. (1995) In-vivo metabolism of substance-P in rat striatum utilizing microdialysis liquidchromatography micro-electrospray mass-spectrometry. J. Mass Spectrom., 30(6): 817–824. Andren, P.E. and Caprioli, R.M. (1999) Determination of extracellular release of neurotensin in discrete rat brain regions utilizing in vivo microdialysis/electrospray mass spectrometry. Brain Res., 845(2): 123–129. Andren, P.E., Emmett, M.R. and Caprioli, R.M. (1994) Microelectrospray – zeptomole-attomole per microliter sensitivity for peptides. J. Am. Soc. Mass Spectrom., 5(9): 867–869. Andren, P.E., Emmett, M.R. and Caprioli, R.M. (1996) Speciﬁc molecular mass detection of endogenously released neuropeptides using in vivo microdialysis mass spectrometry. J. Neurochem., 66: S66. Ao, X.P. and Stenken, J.A. (2003) Water-soluble cyclodextrin polymers for enhanced relative recovery of hydrophobic analytes during microdialysis sampling. Analyst, 128(9): 1143–1149. Ao, X.P. and Stenken, J.A. (2006) No delayed temporal response to sample concentration changes during enhanced microdialysis sampling using cyclodextrins and antibodyimmobilized microspheres. Analyst, 131(1): 62–67. Ao, X.P., Sellati, T.J. and Stenken, J.A. (2004) Enhanced microdialysis relative recovery of inﬂammatory cytokines using antibody-coated microspheres analyzed by ﬂow cytometry. Anal. Chem., 76(13): 3777–3784.

Baseski, H.M., Watson, C.J., Cellar, N.A., Shackman, J.G. and Kennedy, R.T. (2005) Capillary liquid chromatography with MS3 for the determination of enkephalins in microdialysis samples from the striatum of anesthetized and freely-moving rats. J. Mass Spectrom., 40(2): 146–153. Bures, E.J., Courchesne, P.L., Douglass, J., Chen, K., Davis, M.T., Jones, M.D., McGinley, M.D., Robinson, J.H., Spahr, C.S., Sun, J., Wahl, R.C. and Patterson, S.D. (2001) Identiﬁcation of incompletely processed potential carboxypeptidase E substrates from CpEfat/CpEfat mice. Proteomics, 1(1): 79–92. Che, F.Y. and Fricker, L.D. (2005) Quantitative peptidomics of mouse pituitary: comparison of different stable isotopic tags. J. Mass Spectrom., 40(2): 238–249. Che, F.Y., Biswas, R. and Fricker, L.D. (2005a) Relative quantitation of peptides in wild-type and Cpe(fat/fat) mouse pituitary using stable isotopic tags and mass spectrometry. J. Mass Spectrom., 40(2): 227–237. Che, F.Y., Lim, J., Pan, H., Biswas, R. and Fricker, L.D. (2005b) Quantitative neuropeptidomics of microwaveirradiated mouse brain and pituitary. Mol. Cell. Proteomics, 4(9): 1391–1405. Che, F.Y., Yan, L., Li, H., Mzhavia, N., Devi, L.A. and Fricker, L.D. (2001) Identiﬁcation of peptides from brain and pituitary of Cpe(fat)/Cpe(fat) mice. Proc. Natl. Acad. Sci. U.S.A., 98(17): 9971–9976. Che, F.Y., Yuan, Q., Kalinina, E. and Fricker, L.D. (2005c) Peptidomics of Cpefat/fat mouse hypothalamus: effect of food deprivation and exercise on peptide levels. J. Biol. Chem., 280: 4451–4461. Consolo, S., Baldi, G., Russi, G., Civenni, G., Bartfai, T. and Vezzani, A. (1994) Impulse ﬂow dependency of galanin release in-vivo in the rat ventral hippocampus. Proc. Natl. Acad. Sci. U.S.A., 91(17): 8047–8051. Creasy, D.M. and Cottrell, J.S. (2002) Error tolerant searching of uninterpreted tandem mass spectrometry data. Proteomics, 2(10): 1426–1434. De Lecea, L., Kilduff, T.S., Peyron, C., Gao, X.B., Foye, P.E., Danielson, P.E., Fukuhara, C., Battenberg, E.L.F., Gautvik, V.T., Bartlett, F.S., Frankel, W.N., van den Pol, A.N., Bloom, F.E., Gautvik, K.M. and Sutcliffe, J.G. (1998) The hypocretins: hypothalamus-speciﬁc peptides with neuroexcitatory activity. Proc. Natl. Acad. Sci. U.S.A., 95(1): 322–327. Douglass, J., McKinzie, A.A. and Couceyro, P. (1995) PCR differential display identiﬁes a rat brain mRNA that is transcriptionally regulated by cocaine and amphetamine. J. Neurosci., 15: 2471–2481. Emmett, M.R., Andren, P.E. and Caprioli, R.M. (1995) Speciﬁc molecular mass detection of endogenously released neuropeptides using in vivo microdialysis mass spectrometry. J. Neurosci. Methods, 62(1–2): 141–147. Eng, J.K., McCormack, A.L. and Yates, J.R. (1994) An approach to correlate tandem mass-spectral data of peptides with amino-acid-sequences in a protein database. J. Am. Soc. Mass Spectrom., 5(11): 976–989. Floyd, P.D., Li, L., Moroz, T.P. and Sweedler, J.V. (1999) Characterization of peptides from aplysia using microbore

294 liquid chromatography with matrix-assisted laser desorption/ ionization time-of-ﬂight mass spectrometry guided puriﬁcation. J. Chromatogr. A, 830(1): 105–113. Haskins, W.E., Wang, Z.Q., Watson, C.J., Rostand, R.R., Witowski, S.R., Powell, D.H. and Kennedy, R.T. (2001) Capillary LC-MS2 at the attomole level for monitoring and discovering endogenous peptides in microdialysis samples collected in vivo. Anal. Chem., 73(21): 5005–5014. Haskins, W.E., Watson, C.J., Cellar, N.A., Powell, D.H. and Kennedy, R.T. (2004) Discovery and neurochemical screening of peptides in brain extracellular ﬂuid by chemical analysis of in vivo microdialysis samples. Anal. Chem., 76(18): 5523–5533. Hsieh, S., Dreisewerd, K., van der Schors, R.C., Jimenez, C.R., Stahl-Zeng, J., Hillenkamp, F., Jorgenson, J.W., Geraerts, W.P. and Li, K.W. (1998) Separation and identiﬁcation of peptides in single neurons by microcolumn liquid chromatography-matrix-assisted laser desorption/ionization time-ofﬂight mass spectrometry and postsource decay analysis. Anal. Chem., 70(9): 1847–1852. Hughes, J., Smith, T.W., Kosterlitz, H.W., Fothergill, L.A., Morgan, B.A. and Morris, H.R. (1975) Identiﬁcation of two related pentapeptides from the brain with potent opiate agonist activity. Nature, 258: 577–580. Hunt, D.F., Yates, J.R., Shabanowitz, J., Winston, S. and Hauer, C.R. (1986) Protein sequencing by tandem massspectrometry. Proc. Natl. Acad. Sci. U.S.A., 83(17): 6233–6237. Hutchinson, P.J., O’Connell, M.T., Nortje, J., Smith, P., AlRawi, P.G., Gupta, A.K., Menon, D.K. and Pickard, J.D. (2005) Cerebral microdialysis methodology – evaluation of 20 kDa and 100 kDa catheters. Physiol. Meas., 26(4): 423–428. Jorpes, E. and Mutt, V. (1966) Cholecystokinin and pancreozymin, one single hormone? Acta Physiol. Scand., 66: 196–202. Katugampola, S. and Davenport, A. (2003) Emerging roles for orphan G-protein-coupled receptors in the cardiovascular system. Trends Pharmacol. Sci., 24(1): 30–35. Kennedy, R.T., Watson, C.J., Haskins, W.E., Powell, D.H. and Strecker, R.E. (2002) In vivo neurochemical monitoring by microdialysis and capillary separations. Curr. Opin. Chem. Biol., 6(5): 659–665. Khramov, A.N. and Stenken, J.A. (1999) Enhanced microdialysis recovery of some tricyclic antidepressants and structurally related drugs by cyclodextrin-mediated transport. Analyst, 124(7): 1027–1033. Klintenberg, R. and Andren, P.E. (2005) Altered extracellular striatal in vivo biotransformation of the opioid neuropeptide dynorphin A(1–17) in the unilateral 6-OHDA rat model of Parkinson’s disease. J. Mass Spectrom., 40(2): 261–270. Lembo, P.M.C., Grazzini, E., Groblewski, T., O’Donnell, D., Roy, M.O., Zhang, J., Hoffert, C., Cao, J., Schmidt, R., Pelletier, M., Labarre, M., Gosselin, M., Fortin, Y., Banville, D., Shen, S.H., Strom, P., Payza, K., Dray, A., Walker, P. and Ahmad, S. (2002) Proenkephalin A gene products activate a new family of sensory neuron-speciﬁc GPCRs. Nat. Neurosci., 5(3): 201–209.

Lim, J., Berezniuk, I., Che, F.Y., Parikh, R., Biswas, R., Pan, H. and Fricker, L.D. (2006) Altered neuropeptide processing in prefrontal cortex of Cpe(fat/fat) mice: implications for neuropeptide discovery. J. Neurochem., 96(4): 1169–1181. Link, A.J., Eng, J., Schieltz, D.M., Carmack, E., Mize, G.J., Morris, D.R., Garvik, B.M. and Yates, J.R. (1999) Direct analysis of protein complexes using mass spectrometry. Nat. Biotechnol., 17(7): 676–682. Maidment, N.T., Brumbaugh, D.R., Rudolph, V.D., Erdelyi, E. and Evans, C.J. (1989) Microdialysis of extracellular endogenous opioid-peptides from rat-brain in vivo. Neuroscience, 33(3): 549–557. Maidment, N.T. and Evans, C.J. (1991) Measurement of extracellular neuropeptides in the brain: microdialysis linked to solid-phase radioimmunoassays with subfemtomole limits of detection. In: Robinson, T.E. and Justice, J.B. (Eds.), Microdialysis in the Neurosciences. Elsevier, New York, pp. 275–304. Maidment, N.T., Siddall, B.J., Rudolph, V.R., Erdelyi, E. and Evans, C.J. (1991) Dual determination of extracellular cholecystokinin and neurotensin fragments in rat forebrain – microdialysis combined with a sequential multiple antigen radioimmunoassay. Neuroscience, 45(1): 81–93. Maurer, M., Berger, C., Wolf, M., Futterer, C., Feldmann, R., Schwab, S. and Kuschinsky, W. (2003) The proteome of human brain microdialysate. Proteome Sci., 14: 7. Nydahl, K.S., Pierson, J., Nyberg, F., Caprioli, R.M. and Andren, P.E. (2003) In vivo processing of LVV-hemorphin-7 in rat brain and blood utilizing microdialysis combined with electrospray mass spectrometry. Rapid. Commun. Mass Spectrom., 17(8): 838–844. Parkin, M.C., Wei, H., O’Callaghan, J.P. and Kennedy, R.T. (2005) Sample-dependent effects on the neuropeptidome detected in rat brain tissue preparations by capillary liquid chromatography with tandem mass spectrometry. Anal. Chem., 77(19): 6331–6338. Perkins, D.N., Pappin, D.J.C., Creasy, D.M. and Cottrell, J.S. (1999) Probability-based protein identiﬁcation by searching sequence databases using mass spectrometry data. Electrophoresis, 20(18): 3551–3567. Pettersson, A., Amirkhani, A., Arvidsson, B., Markides, K. and Bergquist, J. (2004) A feasibility study of solid supported enhanced microdialysis. Anal. Chem., 76(6): 1678–1682. Pettersson, A., Markides, K. and Bergquist, J. (2001) Enhanced microdialysis of neuropeptides. Acta Biochim. Pol., 48(4): 1117–1120. Prokai, L., Zharikova, A., Janaky, T., Li, X.X., Braddy, A.C., Perjesi, P., Matveeva, L., Powell, D.H. and Prokai-Tatrai, K. (2001) Integration of mass spectrometry into early-phase discovery and development of central nervous system agents. J. Mass Spectrom., 36(11): 1211–1219. Reed, B., Zhang, Y., Chait, B.T. and Kreek, M.J. (2003) Dynorphin A(1–17) biotransformation in striatum of freely moving rats using microdialysis and matrix-assisted laser desorption/ionization mass spectrometry. J. Neurochem., 86(4): 815–823.

295 Rosenbloom, A.J., Sipe, D.M. and Weedn, V.W. (2005) Microdialysis of proteins: performance of the CMA/20 probe. J. Neurosci. Methods, 148(2): 147–153. Rubakhin, S.S., Greenough, W.T. and Sweedler, J.V. (2003) Spatial proﬁling with MALDI MS: distribution of neuropeptides within single neurons. Anal. Chem., 75(20): 5374–5380. Saito, Y. and Civelli, O. (2005) G-protein-coupled receptor deorphanizations. Int. Rev. Neurobiol., 65: 179–209. Schutte, R.J., Oshodi, S.A. and Reichert, W.M. (2004) In vitro characterization of microdialysis sampling of macromolecules. Anal. Chem., 76(20): 6058–6063. Sinnaeve, B.A., Storme, M.L. and Van Bocxlaer, J.F. (2005) Capillary liquid chromatography and tandem mass spectrometry for the quantiﬁcation of enkephalins in cerebrospinal ﬂuid. J. Sep. Sci., 28(14): 1779–1784. Skold, K., Svensson, M., Kaplan, A., Bjorkesten, L., Astrom, J. and Andren, P.E. (2002) A neuroproteomic approach to targeting neuropeptides in the brain. Proteomics, 2(4): 447–454. Strecker, R.E., McCarley, R.W., Dauphin, L., Wilkemeyer, M., Ramesh, V. and Thakkar, M.M. (2001) Effect of diurnal variation and behavioral state changes on extracellular orexin-A levels in the rat brain. Sleep, 24: A25. Suzuyama, K., Shiraishi, T., Oishi, T., Ueda, S., Okamoto, H., Furuta, M., Mineta, T. and Tabuchi, K. (2004) Combined proteomic approach with SELDI-TOF-MS and peptide mass

ﬁngerprinting identiﬁed the rapid increase of monomeric transthyretin in rat cerebrospinal ﬂuid after transient focal cerebral ischemia. Mol. Brain Res., 129(1–2): 44–53. Svensson, M., Skold, K., Svenningsson, P. and Andren, P.E. (2003) Peptidomics-based discovery of novel neuropeptides. J. Proteome Res., 2(2): 213–219. Tatemoto, K. and Mutt, V. (1980) Isolation of two novel candidate hormones using a chemical method for ﬁnding naturally occurring polypeptides. Nature, 285: 417–418. Washburn, M.P., Wolters, D. and Yates, J.R. (2001) Largescale analysis of the yeast proteome by multidimensional protein identiﬁcation technology. Nat. Biotechnol., 19(3): 242–247. Wise, A., Jupe, S.C. and Rees, S. (2004) The identiﬁcation of ligands at orphan G-protein coupled receptors. Annu. Rev. Pharmacol. Toxicol., 44: 43–66. Yuan, X.L. and Desiderio, D.M. (2005) Human cerebrospinal ﬂuid peptidomics. J. Mass Spectrom., 40(2): 176–181. Zhang, H., Stoeckli, M., Andren, P.E. and Caprioli, R.M. (1999) Combining solid-phase preconcentration, capillary electrophoresis and off-line matrix-assisted laser desorption/ ionization mass spectrometry: intracerebral metabolic processing of peptide E in vivo. J. Mass Spectrom., 34(4): 377–383.

CHAPTER 4.1

Microdialysis to study the effects of stress on serotonin, corticosterone and behaviour Astrid C.E. Linthorst Henry Wellcome Laboratories for Integrative Neuroscience and Endocrinology (LINE), University of Bristol, Bristol, UK

Abstract: During the past years, various reﬁnements have made in vivo microdialysis an extremely powerful method to study the mechanisms involved in stress physiology and behaviour in rats and (mutant) mice. In this chapter, microdialysis studies on the neurotransmitter serotonin during the sleep/wake cycle and various forms of stress are reviewed. Extracellular levels of serotonin and its metabolite 5-hydroxyindoleacetic acid correlate positively with the amount of behavioural activity of rats and mice in the home cage in many (but not all) forebrain regions studied so far. Moreover, there is now ample evidence that extracellular levels of serotonin in the terminal regions of serotonergic projections to the forebrain depend on the vigilance state of the animal. Serotonin levels decrease going from a wake state to non-rapid eye movement (non-REM) sleep and are lowest during rapid eye movement sleep. Importantly, the serotonin system has been described to respond to a stressful challenge in a stressor- and brain region-speciﬁc manner. Of interest with regard to the putative pathophysiology of stress-related psychiatric diseases are the observations that long-term alterations in key systems regulating the stress response, such as corticotropin-releasing factor and glucocorticoid receptors, result in an altered responsiveness of hippocampal serotonin to stress. Finally, the power of microdialysis is further underscored in the last section of this chapter, which describes the assessment of free corticosterone levels in dialysates as an index of the activity of the hypothalamic-pituitary-adrenocortical (HPA) axis. have been markedly reﬁned. These reﬁnements have all signiﬁcantly contributed to the suitability of microdialysis to study physiological processes and behaviour. The use of gas anaesthesia (e.g. halothane and isoﬂurane) and the implantation of guide cannulas, which allow longer recovery periods after surgery (as compared with ﬁxed implantation of the dialysis probe), minimise stress levels and thereby contribute to the expression of normal behavioural patterns in animals equipped for microdialysis. Importantly, the ﬁeld of mutant mouse research has been opened up for the dialysis approach ﬁrst by the introduction of thinner microdialysis membranes (diameter 0.24 mm or smaller as compared with 0.5–0.6 mm for standard dialysis membranes) and recently also by the production of

I. Introduction In vivo microdialysis is, as has been comprehensively discussed in the ﬁrst part of this book, a powerful method to study changes in extracellular levels of neurotransmitters and their metabolites in the brain and periphery of humans and experimental animals. Apart from being an important tool in neuropharmacological research, microdialysis is now increasingly being used to examine the role of speciﬁc neurotransmitters in stress physiology and animal behaviour. During the past years surgical, experimental and analytical methods Corresponding author: E-mail: Astrid.Linthorst@bristol. ac.uk

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

301

DOI: 10.1016/S1569-7339(06)16017-8 Copyright 2007 Elsevier B.V. All rights reserved

302

smaller guide cannulas and probes by various commercial microdialysis suppliers. A further important advantage of these thinner probes is that they allow the performance of microdialysis in small brain structures such as the paraventricular nucleus of the hypothalamus, the amygdala and the raphe nuclei; three brain structures intimately involved in the coordination of various aspects of the stress response. The study of mouse behaviour while performing microdialysis has also been made possible by the introduction of low-torque liquid swivels and light-weight ﬂexible counterbalance arms. The power of microdialysis studies for physiological and behavioural research is further enhanced by the reﬁnement of analytical methods. Until recently, microdialysis sample duration commonly was between 15 and 30 min, which clearly limited its use for the study of the more rapid changes in neurotransmitter levels probably underlying physiological and behavioural responses. At present, many research groups, therefore, invest considerable time and effort in optimising the detection limits and sample volumes needed for the analysis of neurotransmitters and other constituents of the extracellular ﬂuid. These efforts will most certainly contribute to enhancing the further potential of microdialysis in behavioural research. Some examples of rapid microdialysis sampling during behavioural paradigms are given later in this chapter. This chapter focusses on the study of the neurotransmitter serotonin by in vivo microdialysis in rats and (mutant) mice. Serotonin plays an important role in the modulation of a wide range of physiological and behavioural processes, among which sleep, food intake, the activity of the hypothalamic-pituitary-adrenocortical (HPA) axis, reproduction and (anxiety-related) behaviour (Jacobs and Azmitia, 1992; Lucki, 1998; Van de Kar and Blair, 1999; Jouvet, 1999; Lowry, 2002; Ursin, 2002; Linthorst, 2005). Disturbances at the level of serotonergic neurotransmission have been implicated in various psychiatric disorders such as major depression, generalised anxiety disorder and panic disorder (Maes and Meltzer, 1995; Ressler and Nemeroff, 2000; Nutt, 2002). Interestingly, aberrant coping with stress (De Kloet et al., 2005)

and disturbed sleep (Lauer et al., 1991; Adrien, 2002; Argyropoulos and Wilson, 2005) have both been found in patients with major depression. The ﬁrst section of this chapter, therefore, discusses microdialysis studies on the role of serotonin in the regulation of normal ongoing (home cage) behaviour and in sleep–wake behaviour. The subsequent section describes the responses of the serotonin system to stressful and fearful challenges. The third part discusses the potential of the microdialysis method to study the activity of the HPA axis by (co-)analysis of free corticosterone in brain dialysates.

II. Home cage and sleep/wake behaviour Serotonergic neurons are located in the raphe nuclei of the brainstem (Dahlstro¨m and Fuxe, 1964; Steinbusch, 1981) and can be divided in a rostral and caudal serotonin system (To¨rk, 1990; Jacobs and Azmitia, 1992). The caudal serotonin system, which mainly projects to the spinal cord, consists of the raphe pallidus nucleus, the raphe obscurus nucleus, the raphe magnus nucleus and serotonergic cell bodies in the ventral lateral medulla. The rostral serotonin system encompasses the caudal linear nucleus, the dorsal (DRN) and median (MRN) raphe nucleus and the supralemniscal region (B9 group). From these rostral nuclei, a dense network of serotonin ﬁbres innervates the whole forebrain, including brain structures with a pivotal role in the modulation of behaviour (such as the hippocampus, prefrontal cortex and amygdala) and in the regulation of the HPA axis (Chaouloff, 2000; Lowry, 2002). The ﬁring rate of the rostral raphe nuclei has been extensively characterised. A large proportion of serotonin neurons ﬁres in a highly regular discharge pattern consisting of solitary spikes (0.3–5 Hz; Aghajanian et al., 1968; Aghajanian and Vandermaelen, 1982). The activity of these neurons is highest during active waking behaviour. However, their discharge rate decreases during quiet waking and further reduces during non-rapid eye movement (non-REM) sleep. During rapid eye movement (REM) sleep, these serotonin neurons are (almost) silent (Jacobs and Azmitia, 1992). Interestingly, in the cat, a subset of serotonin

303

cell bodies has been found to be tonically active during oral–buccal movements such as grooming, licking and chewing (Fornal et al., 1996). These neurons decrease their ﬁring rate during locomotion and orienting behaviours. Apart from the above-mentioned serotonin neurons, a small subpopulation has been identiﬁed between the medial longitudinal fasciculi at the caudal interface of the DRN and the MRN, of which ﬁring frequency is not related to vigilance state (Rasmussen et al., 1984). Furthermore, there is now evidence that serotonin neurons in the DRN (and to a lesser extent in the MRN) can ﬁre in a burst-like manner (Hajo´s et al., 1995; Morzorati and Johnson, 1999). In an elegant study combining electrical stimulation and microdialysis, Gartside et al. (2000) could show that mimicking this burst-ﬁring pattern by electrical stimulation of the DRN results in a more pronounced increase in extracellular serotonin levels in the medial prefrontal cortex than observed after single-pulse stimulation. The striking correlation between vigilance state and ﬁring rate of the majority of serotonin neurons triggered our interest to ﬁnd out to which extent this relationship translates to the extracellular levels of serotonin in projection areas. During the past years, we have, therefore, extensively studied the relationship between extracellular levels of serotonin, as assessed by in vivo microdialysis, and spontaneous behavioural activity in the home cage in rats and mice. For each microdialysis experiment baseline and/or diurnal behaviours are scored live or from videotape. Every 30 s, the experimentater records the behaviour displayed by the animal, normally categorised as follows: resting (sleeping, lying and sitting), grooming and scratching, drinking or eating food pellets, exploration (locomotion, rearing, digging and snifﬁng), chewing, and other behaviours when displayed. For the duration of each microdialysis sample, the scores for each separate behaviour are accumulated to allow direct comparisons between the amount and the composition of behavioural activity on the one hand and the concentrations of dialysate constituents on the other hand. Our results demonstrate that increases in the behavioural activity of rodents are, under normal home cage conditions, correlated with enhanced extracellular

levels of serotonin and its metabolite 5-hydroxyindoleacetic acid (5-HIAA) in a wide variety of forebrain structures. The strongest correlation up to now has been found in the hippocampus. In this brain structure, extracellular serotonin levels are between 200 and 300% higher when the animal is behaviourally active during sample collection as compared with when the animal is at rest (Linthorst et al., 1994, 2000; Oshima et al., 2003; Beekman et al., 2005). In a recent study, we compared baseline levels of serotonin with behavioural activity in the hippocampus, prefrontal cortex, lateral septum and caudate-putamen of C57BL/6N mice and found that baseline levels varied most in the hippocampus but hardly in the caudateputamen (Beekman et al., 2005). A correlation between behavioural activity and extracellular serotonin levels is also underscored by the higher levels of serotonin and 5-HIAA during the dark phase as compared with the light phase of the light–dark cycle in the rodent hippocampus (Rueter and Jacobs, 1996b; Pen˜alva et al., 2002; Oshima et al., 2003) and other brain regions (Mendlin et al., 1996; Rueter and Jacobs, 1996b; Dudley et al., 1998). The different magnitude of the correlation between serotonin levels and activity in the brain structures studied may be explained by differences in the innervation pattern from the raphe nuclei (DRN vs. MRN) and in the extent of afferent stimulation/inhibition of these cell body regions. Thus, it is important to be aware of this strong correlation between behavioural activity and extracellular serotonin levels when studying the effects of (pharmacological) manipulations on serotonergic neurotransmission and we, therefore, recommend routine scoring of behaviour during experiments. Although not the topic of this chapter, a similar (but less pronounced) correlation between behavioural activity and extracellular concentrations has also been found for noradrenaline in several brain regions in rats and mice, such as the preoptic area and hippocampus (Linthorst et al., 1995a, 1996; Beekman and Linthorst, unpublished observations). The above-described ﬁndings are in line with the observations that the majority of raphe serotonin cell bodies displays a rise in ﬁring rate during a phase of active waking as compared with quiet

304

waking. However, the question arises whether levels of serotonin would actually decrease going from wakefulness to sleep, as has been found for the discharge rate of serotonergic cell bodies. This seems indeed to be the case. Using a rapid sampling microdialysis method (sampling rate 3 min) in combination with EEG measurements, we found that the extracellular concentrations of serotonin in the hippocampus are positively correlated with the amount of wakefulness during sample collection (Pen˜alva et al., 2003; see Fig. 1B). The lowest levels of serotonin were observed during REM sleep (Fig. 1A, C). Levels of serotonin in samples during which animals are in REM sleep for more than 50% of the sample duration are 70% lower than those in samples collected during a wake state. Extracellular levels of serotonin also decreased during non-REM sleep (55% if the proportion of non-REM sleep is more than 80% of the sample duration; Pen˜alva et al., 2003; see Fig. 1A, C). A similar relationship, but using longer sample durations, between hippocampal extracellular levels of serotonin and vigilance state has been reported by Park et al. (1999). Moreover, other brain regions innervated by the midbrain raphe nuclei, such as the frontal cortex (Portas et al., 1998), preoptic area (Python et al., 2001), hypothalamus (Houdouin et al., 1991; Imeri et al., 1994), DRN (Portas and McCarley, 1994) and amygdala and locus coeruleus (Shouse et al., 2000) also show a clear correlation between vigilance and extracellular levels of serotonin. Based on these observations together with the positive correlation between sleep and wake state and ﬁring rate of raphe serotonin neurons (see above), it can be hypothesised that regulatory mechanisms at the level of the cell bodies, rather than at the level of the terminal regions, are responsible for the changes in serotonin levels over the sleep/wake cycle. There seems, however, to be some speciﬁcity in this system as Sakai and Crochet have demonstrated the presence of subtypes of DRN serotonergic neurons that display sustained tonic activity during REM sleep or a high rate of tonic discharge during non-REM sleep (Sakai and Crochet, 2001). The role of hippocampal serotonin in the regulation of sleep/wake behaviour is at present still unclear. It is very well possible that

changes in extracellular serotonin in this brain structure are involved in the modulation of hippocampal theta rhythm (see Leung, 1998) and/or that they pass on information about the behavioural state of the animal (Jacobs and Fornal, 1999). To further elucidate the relationship between hippocampal serotonin and sleep, we studied the effects of sleep deprivation and the subsequent recovery phase by combined in vivo microdialysis/ EEG in rats. Four hours of sleep deprivation, induced by the gentle handling paradigm (introduction or removal of small objects from cage; occasionally, gentle movement of the cage to prevent the animal from falling asleep; no physical contact with the animal) resulted in a sustained elevation of hippocampal extracellular levels of serotonin (Pen˜alva et al., 2003). Importantly, these levels drop immediately during rebound sleep and the normal relationship between vigilance state (non-REM and REM) and serotonin levels is reinstated (Pen˜alva et al., 2003). These results may provide an explanation for the beneﬁcial effects of sleep deprivation in depressed patients (for review see Ringel and Szuba, 2001; Adrien, 2002). We hypothesise that the temporary increase in serotonin levels during the sleep deprivation period may result in improved mood (see also Section III); an effect which is abolished after the ﬁrst nap probably because of an immediate normalisation of serotonin levels (and their relationship with sleep/vigilance state). Clearly this hypothesis awaits further study. Increased serotonergic neurotransmission during sleep deprivation has also been demonstrated in other brain regions, such as the frontal cortex, hypothalamus, and brainstem (tissue level studies in rats; Asikainen et al., 1997) and the suprachiasmatic nucleus (microdialysis in hamsters; Grossman et al., 2000; but also see Bjorvatn et al., 2002).

III. Stressful challenges and anxiety tests The amount of stress perceived by individuals in today’s society seems to increase steadily, which may explain the rise in the incidence of stress-related psychiatric diseases such as major depression

305 A

3.0 2.5 2.0

100 75 50 25

100 75 50 25 13:30

13:15

13:00

12:45

12:30

12:15

12:00

11:45

100 75 50 25

time

B

C #

3.0 mean 5-HT (fmol/3 minsample)

3.0 mean 5-HT (fmol/3 min sample)

11:30

11:15

11:00

10:45

10:30

0.5

REM (% of time)

1.0

nonREM (% of time)

1.5

wake (% of time)

5-HT (fmol/3 min sample)

3.5

2.5 2.0 1.5 1.0 0.5 0.0

2.5 2.0 1.5

* +

>80% nonREM

>50% REM

0.5 0.0

80 -1 00 % 60 -8 0% 40 -6 0% 20 -4 0% 020 %

*

1.0

100% wake

wakefulness (% of time) Fig. 1. Relationship between hippocampal extracellular levels of 5-hydroxytryptamine (5-HT) and vigilance state under baseline conditions in male Wistar rats as assessed by a combined in vivo microdialysis/EEG method. (A) Time course of 5-HT levels (closed circles; fmol/3 min sample) and vigilance state (bars) measured between 10:30 and 13:30 h in a representative animal. The vigilance stages – non-rapid eye movement sleep (non-REMS), rapid eye movement sleep (REMS) and WAKE – were assessed from the analysis of the 10-s epochs of EEG and EMG, and averaged (as percentage of time) in 3-min intervals corresponding to the microdialysis samples. Please note that the 5-HT value of the sample collected between 12:45–12:48 h is missing because of an HPLC problem. (B) Relationship between 5-HT level and amount of wakefulness (expressed as percentage of sample time and averaged over 20% intervals). (C) Mean extracellular levels of 5-HT for each vigilance state under basal conditions. Only samples corresponding to 3-min periods showing 100% of time wakefulness, 80–100% non-REMS or >50% REMS were used. Values for B and C represent mean7SEM (n ¼ 11). ], signiﬁcant effect of amount of wakefulness (ANOVA with repeated measures design); *, signiﬁcantly different from wakefulness; +, signiﬁcantly different from non-REM (post-hoc contrast). Reproduced from Pen˜alva et al. (2003) with the permission of the European Journal of Neuroscience.

306

and anxiety disorders. At present, general practitioners deal with psychiatric and psychosomatic complaints during a large proportion of their daily patient visits and prescribe antidepressants and anxiolytics on a regular basis. Emotional states are controlled by a multitude of neurotransmitters and hormones. Serotonin is one of the key neurotransmitters involved in depression and anxiety. Depressed patients show decreased CSF levels of the 5-hydroxytryptamine (5-HT) metabolite 5-HIAA and changes in 5-HT receptor expression. Moreover, drugs affecting 5-HT neurotransmission, such as selective serotonin reuptake inhibitors (SSRI’s, e.g. ﬂuoxetine and paroxetine) and the 5-HT1A receptor agonist buspirone are now widely used in the treatment of these diseases (for review see Maes and Meltzer, 1995; Mann, 1998; Ressler and Nemeroff, 2000). During the past two decades, it has become increasingly clear that in many cases depressive illness is accompanied by aberrant functioning of the HPA axis. Changes at the level of the corticotropin-releasing factor (CRF) system (most likely hyperactivity) and at the level of glucocorticoid physiology have been described in cases of depression (Arborelius et al., 1999; Reul et al., 2000; Reul and Holsboer, 2002; De Kloet et al., 2005). Although our knowledge of the mechanisms underlying mood and anxiety disorders is growing, the available therapeutical tools are still suboptimal mainly because of side effects, long delay before onset of improvement and therapy resistance. It is, therefore, essential to increase our understanding of the brain mechanisms underlying healthy and aberrant coping with stress. During the past 15 years, my laboratory has focussed on the elucidation of the effects of stress on serotonergic neurotransmission by in vivo microdialysis in rats and mice. In this section, the effects of different stressful challenges varying in their intensity and physical versus psychological character are discussed. Our results together with data published in the literature clearly demonstrate that the serotonin system is able to generate a differentiated and stressor- and brain region-speciﬁc response to stressful challenges. Moreover, evidence has accumulated showing that long-term changes in key systems regulating the neuroendocrine, autonomic and behavioural responses to

stress [e.g. the CRF system, glucocorticoid receptors (GRs)] precipitate in alterations in serotonergic neurotransmission under basal and/or stressful conditions.

III.A. Forced swimming Originally, a paradigm applying forced swimming in rats and mice was developed by R. Porsolt in 1977 as a test to reveal putative antidepressant properties of new drug compounds (Porsolt et al., 1977; Cryan et al., 2002). At present, however, forced swimming in rodents is also widely used to study various aspects of the stress response, including the effects of stress on brain neurotransmission and behaviour. When we started combining forced swim stress and in vivo microdialysis in rats, we made an unexpected ﬁnding, that is swimming for 15 min in water at a temperature of 251C results in a dramatic increase in hippocampal extracellular levels of serotonin (900% of baseline) during the swim period only (Linthorst et al., 2002). However, we noticed that this enormous rise in hippocampal serotonin is only present in animals that dive during the swim session. During these experiments, the rats are connected to the swivel system via a plastic collar around their neck. We hypothesise that the diving animals, because of the plastic collar, appreciate the stressful situation differently as compared with the nondiving animals, probably leading to a panic-like response (Linthorst et al., 2002). Support for this hypothesis may stem from observations that the panic-inducing substance m-chloro-phenylpiperazine (Kahn et al., 1988) is also able to increase hippocampal serotonin levels up to 1,400% of baseline (Eriksson et al., 1999). Furthermore, we observed that the extreme rise in serotonin is absent in diving rats who are attached to the swivel system via a wire connected to a peg on their head, and in mice, which normally do not dive during the swim session. Wire-connected rats and mice show, however, an increase in hippocampal serotonin levels up to 200–300% during swimming in water at 251C (Linthorst et al., 2001; Pen˜alva et al., 2002; Fujino et al., 2002; Oshima et al., 2003). Interestingly, the CRF receptor antagonist

307

D-Phe-CRF12–41 inhibited the dramatic serotonin response in diving rats, pointing to an involvement of CRF and/or its congeners in this effect. The effects of forced swim stress on serotonin levels have also been studied by other research groups. Together with the above-described ﬁndings these literature data clearly demonstrate the complex nature of the forced swim paradigm. For instance, it has been found that hippocampal levels of serotonin moderately increase, decrease or do not change during swimming in water at 30–351C for 30 min (during the dark phase; Rueter and Jacobs, 1996a), in water at 251C for 5 min (Adell et al., 1997) and in water at 21–221C for 30 min (Kirby et al., 1995, 1997), respectively. Although often regarded as a psychologically stressful event, forced swimming is also accompanied by changes in the physiology of the animal; among others a signiﬁcant drop in body temperature is observed when lower water temperatures are used (Stone, 1970). Data from our laboratory shows that the exact response of hippocampal serotonin and 5-HIAA indeed depends on the water temperature and the concomitant decrease in body temperature (Linthorst et al., 2001). It may, therefore, be reasoned that the divergent data published up to now result from differences in the experimental protocols, especially with respect to the water temperature used. We conclude that while forced swim stress is a powerful tool to study the effects of stress on brain neurotransmission, it is of paramount importance to clearly deﬁne the experimental conditions and to interpret the data obtained also in the light of the combined psychological and physical aspects of this form of stress.

III.B. Predator stress Given the putative role of serotonin in stressrelated psychiatric disorders, we decided to assess in detail the effects of psychological stress on serotonergic neurotransmission in different brain regions in a behavioural animal model. For this purpose, we designed a series of microdialysis experiments in mice using predator stress as the stressful challenge (Beekman et al., 2005). Predator stress is an extremely valuable tool as it relates

to innate fear and is not based on physical stress, thereby possibly more closely resembling psychological stress in humans. Male C57BL/6N mice, equipped for microdialysis, are exposed for 30 min to a male rat, which is placed into a separate compartment of their home cage. The two compartments are separated by a clear Plexiglas wall containing two rows of small holes. No physical contact between the mouse and the rat is possible. To study the effects of predator stress on mouse behaviour other forms of arena’s are also being used (Blanchard et al., 2003), but our setup is especially suitable for concomitant performance of in vivo microdialysis. During the total duration of the experiment, mouse behaviour is monitored by cameras and recorded on videotape. To fully appreciate the time course of the effects of predator stress on hippocampal serotonin and 5-HIAA, dialysate samples are collected in short time intervals, that is 10-min intervals pre- and post-stress, and 5-min intervals during the rat exposure period and the ﬁrst 15 min after removal of the rat. Rat exposure causes a profound behavioural activation in C57BL/6N mice (Beekman et al., 2005). Mice wake up and/or become alert immediately at the start of the rat exposure. Often their ﬁrst response is to move to the corner of the cage from where risk-assessment behaviours, such as snifﬁng in the air, stretching towards the rat compartment and snifﬁng at the separation wall are started. These behaviours are most intense during the ﬁrst 5–15 min of the rat exposure period and decrease during the second part of the exposure. During this latter part, mice are still extensively snifﬁng in the bedding. Rat exposure causes a rise in extracellular levels of both serotonin and 5-HIAA in the hippocampus, the prefrontal cortex and the lateral septum of C57BL/6N mice (Beekman et al., 2005; see Fig. 2A–C). These effects are most pronounced in the hippocampus and, interestingly, no signiﬁcant effects are observed in the caudate-putamen (Fig. 2D). A second exposure (to a different rat) on the next day results in almost identical serotonin responses in the brain areas under study, whereas the rise in 5-HIAA is slightly reduced in the hippocampus. Importantly, the behavioural response to a second period of rat exposure is also remarkably similar to the response to the ﬁrst

308

p=0.011

+

150

+

*

350 120

300

250

250

90

200 60

150

30

20 0

60

150

C: Lateral septum

0

D: Caudate putamen 150

+

*

350

30

20 0

0

400

400

150

350 120

300 250

90

200

120

300 250

90

200 60

150

100

5-min samples

15-min samples

15-min samples

15-min samples

13.30 14.00

13.00

12.30

12.00

11.30

11.10

11.00

time

0 10.50

0 10.40

13.30 14.00

13.00

12.30

12.00

11.30

11.10

11.00

10.50

10.40

10.30

9.30

15-min samples

10.00

0 8.30 9.00

0

30

20 10.30

30

20

9.30

100

60

150

10.00

5-HT (% of baseline)

90

200

100

100

time

120

300

5-HIAA (% of baseline) time active (% of sample time)

350

150

+

*

8.30 9.00

5-HT (% of baseline)

400

400

5-HIAA (% of baseline) time active (% of sample time)

B: Prefrontal cortex

A: Hippocampus

15-min samples

Fig. 2. Effects of exposure to a male rat (rat is present during 30 min in a separate compartment of the home cage of the mouse) on serotonergic neurotransmission and behavioural activity in male C57Bl6/N mice as assessed by in vivo microdialysis. Extracellular levels of (Serotonin 5-HT) (squares, left y-axis), 5-hydroxyindoleacetic acid (5-HIAA; triangles, right y-axis) in percentage of baseline (+SEM), and time behaviourally active (bars, right y-axis) in percentage of sample time (+SEM), over the course of the experiment in hippocampus (n ¼ 9, panel A), prefrontal cortex (n ¼ 8, panel B), lateral septum (n ¼ 7, panel C) and caudate-putamen (n ¼ 7, panel D). Shaded area indicates the period of exposure to a rat. The sample duration was 10 min except for the rat exposure period and the ﬁrst 15 min thereafter during which 5-min samples were collected. ANOVA indicated effects of ‘time’ in the hippocampus, the prefrontal cortex and lateral septum, but not in the caudate-putamen for both 5-HT and 5-HIAA. Results from post-hoc tests are denoted by *, which indicates signiﬁcant difference from baseline 5-HT, and by +, which indicates signiﬁcant difference from baseline 5-HIAA (Bonferroni corrected for multiple comparisons, po0.008). Reproduced from Beekman et al. (2005) with the permission of the European Journal of Neuroscience.

exposure, although a small increase in coping behaviours such as grooming and nest building is found during the second exposure (Beekman et al., 2005). From these data, it may be hypothesised that an ethologically salient stimulus such as exposure to a predator results in a full neurochemical and behavioural response also on repeated stimulation. Recently, we found that predator exposure also increases extracellular levels of serotonin at the level of the hypothalamus (i.e. in the anterior

hypothalamus and in the paraventricular nucleus) in Balb/c mice (Beekman and Linthorst, unpublished observations). Unfortunately, there is only one microdialysis study from another research group available describing the effects of predator stress on serotonergic neurotransmission. Rueter and Jacobs (1996a) found that exposure of a rat to a cat results in elevated extracellular levels of 5-HT in the hippocampus, frontal cortex, amygdala and striatum. However, several research groups have

309

found evidence for an activation of the serotonin system induced by exposure to a predator or predator odour by assessing serotonin turnover in post-mortem brain tissue of hippocampus, frontal cortex and amygdala (Belzung et al., 2001; Hayley et al., 2001). In an earlier study, we investigated the neuroendocrine, behavioural and serotonergic responses to predator exposure in a mutant mouse model with life-long impaired GR function (GR-i mice; Pepin et al., 1992) using in vivo microdialysis, but with 30-min sampling intervals (Linthorst et al., 2000). We found that, whereas control B6C3F1 mice clearly show an anxious response during the presence of the rat (i.e. freezing and sitting in nest with eyes ﬁxed at the rat, subsequently followed by risk-assessment behaviours), GR-i mice show enhanced behavioural activity, especially investigating the separation wall. Control animals show a clear activation of the HPA axis due to the stressful challenge, as evidenced by increased dialysate levels of free corticosterone (see below), but this is not found in GR-i mice. Interestingly, the different behavioural and neuroendocrine responses in GR-i mice are accompanied by an exaggerated increase in hippocampal extracellular levels of serotonin, which cannot be explained by the increased level of behavioural activity alone (Linthorst et al., 2000). From these data, it may be concluded that life-long impairment of GR results in misjudgement of a potentially dangerous situation as demonstrated by the altered, putatively inappropriate, behavioural response and the absence of a rise in stress hormone levels. The exact role of the hyper-response of hippocampal serotonergic neurotransmission in these phenomena is not known at present.

III.C. Immune stress The organism responds to an infectious challenge with a variety of defence mechanisms, which apart from stimulation of the immune system also include activation of the HPA axis and induction of fever. Because infection and inﬂammation represent immediate physiologic threats, immune stress is often regarded as a systemic stressor without a

psychological component. However, this is a very simpliﬁed interpretation of the immune response. Immune stress also causes clear changes in the behaviour of humans and animals, such as increased or fragmented sleep, loss of interest in social contact, loss of appetite and immobility. These symptoms are collectively summarised as ‘sickness behaviour’ (Hart, 1988; Dantzer et al., 1991). It is, therefore, not surprising, as described below, that even higher brain regions including the hippocampus are activated during immune stress. Using in vivo microdialysis in rats, we have studied serotonin (and noradrenaline) responses during an inﬂammatory challenge (i.e. intraperitoneal (i.p.) administration of bacterial endotoxin) in different brain regions. Intraperitoneal injection of endotoxin results in a marked and prolonged dose-dependent increase in the extracellular levels of serotonin and 5-HIAA in the hippocampus (Linthorst et al., 1995b). Administration of speciﬁc cytokines showed that the effect of endotoxin can be mimicked by i.p. (Merali et al., 1997) as well as intracerebroventricular (i.c.v.) injection of interleukin-1b (Linthorst et al., 1995b) and interleukin-2 (Pauli et al., 1998). However, tumour necrosis factor alpha did not have an effect on hippocampal serotonergic neurotransmission (Pauli et al., 1998). Studies which have assessed serotonin turnover in post-mortem tissue of the hippocampus underscore these in vivo microdialysis ﬁndings (Kabiersch et al., 1988; Zalcman et al., 1994; Lacosta et al., 1999). Interestingly, it seems that inﬂammation may activate hippocampal serotonin via direct mechanisms within this brain region. Using reversed dialysis (also called retrodialysis; another extremely powerful feature of the microdialysis method) of interleukin-1b, we were able to demonstrate that intrahippocampal administration of this cytokine increases serotonin and 5-HIAA levels most likely via direct local stimulation of interleukin-1 receptors in the hippocampus (Linthorst et al., 1994). Post-mortem and dialysis studies also indicate that serotonergic neurotransmission in other higher brain structures, such as the prefrontal cortex, is enhanced during immune stress (Dunn and Welch, 1991; Zalcman et al., 1994; Lavicky and Dunn, 1995). Interestingly, we found that i.p.

310

administration of endotoxin had no effect on extracellular serotonin levels in the preoptic area of the hypothalamus, a key area in the control of body temperature (although 5-HIAA levels were slightly increased; Linthorst et al., 1995a). Wilkinson et al. (1991) observed similar effects on serotonergic neurotransmission in the hypothalamus of the cat using another pyrogen, that is muramyl dipeptide.

III.D. CRF and serotonin in the hippocampus During the past years, it has become increasingly clear that the CRF system plays a critical role in stress-related psychiatric diseases. CRF (and its congeners, urocortin (ucn) 1, 2 and 3) mediate various aspects of the stress response, such as activation of the HPA axis and the sympathetic nervous system, and the coordination of behavioural strategies to cope with the stressor. The activity of the CRF system may be increased in depressed patients, as elevated cerebrospinal ﬂuid levels of CRF (Nemeroff et al., 1984) and a reduced number of CRF receptors in the frontal cortex (Nemeroff et al., 1988) have been found in this disorder. Moreover, CRF mRNA expression and the number of CRF-expressing neurons are elevated in the paraventricular nucleus of the hypothalamus of depressed patients (Raadsheer et al., 1994, 1995). The reported aberrant functioning of both the CRF and serotonin system in major depression triggered our interest to ﬁnd out whether there exists a link between these two disturbances. In a series of microdialysis studies, we assessed the effects of acute i.c.v. administration of CRF and the urocortins and of chronic infusion of CRF on hippocampal extracellular levels of 5-HT and 5-HIAA. Both CRF and ucn 1 cause a dose-dependent (0.03– 10 mg/kg) increase in hippocampal 5-HT and 5HIAA after acute i.c.v. administration (Linthorst et al., 2002). Intracerebroventricular injections of the corticotropin-releasing factor receptor type 2 (CRF2) speciﬁc ligands ucn 2 and 3 also result in a rise in hippocampal 5-HT and 5-HIAA but these responses are less pronounced than those observed after administration of CRF or ucn 1 (De Groote et al., 2005). Rats in which brain levels of CRF are

elevated during 7 days by a chronic infusion of the peptide via an i.c.v. cannula attached to a miniosmotic pump show no alterations in baseline hippocampal extracellular levels of serotonin and 5-HIAA. Importantly, however, a signiﬁcantly diminished hippocampal serotonin response to immune stress is found in these animals (Linthorst et al., 1997). Our data were the ﬁrst to indicate that long-term aberrations at the level of the CRF system indeed result in aberrant responses at the level of hippocampal serotonergic neurotransmission. This conclusion is underscored by recent in vivo microdialysis studies in corticotropin-releasing factor receptor type 1 (CRF1) knockout mice and mice chronically treated with the CRF1 antagonist NBI 30775, demonstrating that longterm alterations at the level of CRF1 evolve in altered responsiveness of hippocampal serotonergic neurotransmission to (swim) stress (Pen˜alva et al., 2002; Oshima et al., 2003).

IV. Combination of neurotransmitter and corticosterone dialysis In vivo microdialysis is an extremely useful method for neuropharmacological research, that is for the assessment of the effects of drugs on brain neurotransmission. However, as can also be taken from the previous sections, microdialysis has become an increasingly important tool in the search for the mechanisms underlying physiological processes and behaviour. To study the role of neurotransmitter systems in the stress response, it is of great importance to assess simultaneously neurotransmission and the activity of the HPA axis. In the early 1990s, we decided therefore, to measure corticosterone levels in dialysates using a highly sensitive radioimmunoassay method. Corticosterone in dialysates represents the free, unbound, fraction of corticosterone because the brain extracellular ﬂuid is devoid of binding proteins such as corticosterone binding globulin. Dialysate corticosterone concentrations are, therefore, much lower than the concentrations normally measured in plasmaas the latter levels constitute the total (bound+ free) concentration of corticosterone. Measuring free corticosterone levels in the brain is also of

311

prolonged elevation of free extracellular corticosterone levels (for more than 6 h after a dose of 100 mg/kg body weight; Linthorst et al., 1995b). The measurement of free corticosterone levels in the brain resulted also in interesting unexpected ﬁndings in several of the studies performed. For instance, we found that whereas free corticosterone levels are below detection limit in CRF1 knockout mice (Pen˜alva et al., 2002), these levels are not affected in C57BL6/N mice treated for 2 weeks with the CRF1 antagonist NBI 30775 (for a detailed discussion on this apparent discrepancy see Oshima et al., 2003). Moreover, whereas i.c.v. administration of CRF, ucn 1 and the CRF2 ligand ucn 3 cause a rise in free corticosterone levels, another CRF2 ligand, ucn 2, was without effect (De Groote et al., 2005; see Fig. 3). Finally, the abovementioned studies in GR-i mice showed that the aberrant behavioural response to predator stress is not only accompanied by an exaggerated 5-HT

particular physiological signiﬁcance as these will be the levels ‘seen’ by the neurons expressing mineralocorticoid receptors and GRs (Reul et al., 2000). We are routinely using this method (not only in the hippocampus but also in other brain regions) to study the activity of the HPA axis during basal as well as stressful conditions in rats and mice. We observed clear diurnal rhythms of free corticosterone levels (with low levels during the early morning, a rise at the beginning of the afternoon and a peak in levels around the start of the dark phase) in rats and mice (Linthorst et al., 1994, 2000; Reul et al., 2000; Pen˜alva et al., 2002, 2003; Oshima et al., 2003). Stressful challenges such as forced swimming (Oshima et al., 2003), novelty (Droste et al., unpublished observations) and predator exposure (Linthorst et al., 2000) result in increased levels of free corticosterone with a peak 30–60 min after stress onset. In contrast, immune stress, induced by the injection of endotoxin, causes a

0.7

*

corticosterone (µg/dl)

0.6 0.5 0.4 0.3 0.2 0.1 0.0 09:00

10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 time (h) saline ( n=5)

CRF (n=5)

uc n 2 (n=6)

uc n 1 (n=6)

uc n 3 (n=7)

Fig. 3. Effects of intracerebroventricular (i.c.v.) administration of 1.0 mg corticotropin-releasing factor (CRF) or urocortin (ucn) 1, ucn 2 and ucn 3 on free corticosterone levels as assessed by in vivo microdialysis in the hippocampus of male Wistar rats. Saline-injected animals served as controls. Time points on the x-axis correspond to the time of the day at which collection of the samples was started. The arrow indicates the time point of injection. Values represent mean levels of corticosterone (mg/dL). ANOVA indicated a signiﬁcant effect of neuropeptide. *CRF, ucn 1 and ucn 3 signiﬁcantly different from saline (Dunnett; po0.05). Reproduced from De Groote et al. (2005) with the permission of the Journal of Neurochemistry.

312

response but also by the absence of an HPA axis response (Linthorst et al., 2000). These examples together with other reports published in the literatures (Azzi et al., 1998; Kitchener et al., 2004) clearly demonstrate the usefulness of analyses of free corticosterone levels in dialysates. The simultaneous assessment of both neurotransmission and HPA axis activity (and behaviour) in the same animal greatly enhances the power of the conclusions derived from the data (in contrast to studies in which these parameters are assessed in separate groups of animals). Importantly, this combination also contributes signiﬁcantly to the reduction in the number of animals used, one of the main goals of the improvement of animal welfare in experimental research.

CRF2 DRN GR HPA i.p. i.c.v. MRN Non-REM REM SSRI

corticotropin-releasing factor receptor type 1 corticotropin-releasing factor receptor type 2 dorsal raphe nucleus glucocorticoid receptor hypothalamic-pituitary-adrenocortical intraperitoneal intracerebroventricular median raphe nucleus non-rapid eye movement sleep rapid eye movement sleep selective serotonin reuptake inhibitor

Acknowledgements

V. Conclusions The studies discussed in this chapter demonstrate that in vivo microdialysis is an extremely powerful method to study serotonergic neurotransmission in behaviourally active rats and (mutant) mice. By combining microdialysis with detailed scoring of behaviour or EEG measurements, information can be gathered on the role of different neurotransmitters in the regulation of behaviour under basal (home cage and sleep–wake behaviour) and stressful conditions. Importantly, in vivo microdialysis can also be used to monitor the activity of the HPA axis over extended time periods and under different stressful conditions by assessing the extracellular concentrations of free corticosterone (the biologically active fraction) directly in the brain. It is exactly this combination of simultaneous monitoring of neurotransmitters together with behaviour and HPA axis activity that makes in vivo microdialysis an invaluable method in (behavioural) neuroscience and stress physiology.

List of abbreviations 5-HT 5-HIAA CRF

CRF1

serotonin (5-hydroxytryptamine) 5-hydroxyindoleacetic acid corticotropin-releasing factor

I would like to thank Dr Marjolein Beekman, Dr Lotte de Groote, Dr Akihiko Oshima and Dr Rosana G. Pen˜alva and Miss Cornelia Flachskamm for their invaluable contribution to the elucidation of the role of serotonin in the stress response. Moreover, I am grateful to my collaborator Professor Johannes M.H.M. Reul for his continuous support. The work of the author described here has been supported by the Max Planck Society and the Volkswagen Foundation (Germany).

References Adell, A., Casanovas, J.M. and Artigas, F. (1997) Comparative study in the rat of the actions of different types of stress on the release of 5-HT in raphe nuclei and forebrain areas. Neuropharmacology, 36: 735–741. Adrien, J. (2002) Neurobiological bases for the relation between sleep and depression. Sleep Med. Rev., 6: 341–351. Aghajanian, G.K. and Vandermaelen, C.P. (1982) Intracellular identiﬁcation of central noradrenergic and serotonergic neurons by a new double labeling procedure. J. Neurosci., 2: 1786–1792. Aghajanian, G.K., Foote, W.E. and Sheard, M.H. (1968) Lysergic acid diethylamide: sensitive neuronal units in the midbrain raphe. Science, 161: 706–708. Arborelius, L., Owens, M.J., Plotsky, P.M. and Nemeroff, C.B. (1999) The role of corticotropin-releasing factor in depression and anxiety disorders. J. Endocrinol., 160: 1–12.

313 Argyropoulos, S.V. and Wilson, S.J. (2005) Sleep disturbances in depression and the effects of antidepressants. Int. Rev. Psychiatry, 17: 237–245. Asikainen, M., Toppila, J., Alanko, L., Ward, D.J., Stenberg, D. and Porkka-Heiskanen, T. (1997) Sleep deprivation increases brain serotonin turnover in the rat. Neuroreport, 8: 1577–1582. Azzi, M., Betancur, C., Sillaber, I., Spangel, R., Rostene, W. and Berod, A. (1998) Repeated administration of the neurotensin receptor antagonist SR 48692 differentially regulates mesocortical and mesolimbic dopaminergic systems. J. Neurochem., 71: 1158–1167. Beekman, M., Flachskamm, C. and Linthorst, A.C.E. (2005) Effects of exposure to a predator on behaviour and serotonergic neurotransmission in different brain regions of C57bl/6N mice. Eur. J. Neurosci., 21: 2825–2836. Belzung, C., El Hage, W., Moindrot, N. and Griebel, G. (2001) Behavioral and neurochemical changes following predatory stress in mice. Neuropharmacology, 41: 400–408. Bjorvatn, B., Gronli, J., Hamre, F., Sorensen, E., Fiske, E., Bjorkum, A.A., Portas, C.M. and Ursin, R. (2002) Effects of sleep deprivation on extracellular serotonin in hippocampus and frontal cortex of the rat. Neuroscience, 113: 323–330. Blanchard, D.C., Griebel, G. and Blanchard, R.J. (2003) The mouse defense test battery: pharmacological and behavioral assays for anxiety and panic. Eur. J. Pharmacol., 463: 97–116. Chaouloff, F. (2000) Serotonin, stress and corticoids. J. Psychopharmacol., 14: 139–151. Cryan, J.F., Markou, A. and Lucki, I. (2002) Assessing antidepressant activity in rodents: recent developments and future needs. Trends Pharmacol. Sci., 23: 238–245. Dahlstro¨m, A. and Fuxe, K. (1964) Evidence for the existence of monoamine-containing neurons in the central nervous system. Acta Physiol. Scand., 62: 2–55. Dantzer, R., Bluthe´, R.M., Kent, S. and Kelley, K.W. (1991) Behavioural effects of cytokines. In: Rothwell, N. and Dantzer, R. (Eds.), Interleukin-1 in the Brain. Pergamon Press, Oxford, pp. 135–150. De Groote, L., Penalva, R.G., Flachskamm, C., Reul, J.M.H.M. and Linthorst, A.C.E. (2005) Differential monoaminergic, neuroendocrine and behavioural responses after central administration of corticotropin-releasing factor receptor type 1 and type 2 agonists. J. Neurochem., 94: 45–56. De Kloet, E.R., Joe¨ls, M. and Holsboer, F. (2005) Stress and the brain: from adaptation to disease. Nat. Rev. Neurosci., 6: 463–475. Dudley, T.E., Dinardo, L.A. and Glass, J.D. (1998) Endogenous regulation of serotonin release in the hamster suprachiasmatic nucleus. J. Neurosci., 18: 5045–5052. Dunn, A.J. and Welch, J. (1991) Stress- and endotoxin-induced increases in brain tryptophan and serotonin metabolism depend on sympathetic nervous system activity. J. Neurochem., 57: 1615–1622. Eriksson, E., Engberg, G., Bing, O. and Nissbrandt, H. (1999) Effects of mCPP on the extracellular concentrations of

serotonin and dopamine in rat brain. Neuropsychopharmacology, 20: 287–296. Fornal, C.A., Metzler, C.W., Marrosu, F., Ribiero-do-Valle, L.E. and Jacobs, B.L. (1996) A subgroup of dorsal raphe serotonergic neurons in the cat is strongly activated during oral-buccal movements. Brain Res., 716: 123–133. Fujino, K., Yoshitake, T., Inoue, O., Ibii, N., Kehr, J., Ishida, J., Nohta, H. and Yamaguchi, M. (2002) Increased serotonin release in mice frontal cortex and hippocampus induced by acute physiological stressors. Neurosci. Lett., 320: 91–95. Gartside, S.E., Hajo´s-Korcsok, E., Bagdy, E., Harsing, L.G., Sharp, T. and Hajo´s, M. (2000) Neurochemical and electrophysiological studies on the functional signiﬁcance of burst ﬁring in serotonergic neurons. Neuroscience, 98: 295–300. Grossman, G.H., Mistlberger, R.E., Antle, M.C., Ehlen, J.C. and Glass, J.D. (2000) Sleep deprivation stimulates serotonin release in the suprachiasmatic nucleus. Neuroreport, 11: 1929–1932. Hajo´s, M., Gartside, S.E., Villa, A.E.P. and Sharp, T. (1995) Evidence for a repetitive (burst) ﬁring pattern in a sub-population of 5-hydroxytryptamine neurons in the dorsal and median raphe nuclei of the rat. Neuroscience, 69: 189–197. Hart, B.L. (1988) Biological basis of the behavior of sick animals. Neurosci. Biobehav. Rev., 12: 123–137. Hayley, S., Borowski, T., Merali, Z. and Anisman, H. (2001) Central monoamine activity in genetically distinct strains of mice following a psychogenic stressor: effects of predator exposure. Brain Res, 892: 293–300. Houdouin, F., Cespuglio, R., Gharib, A., Sarda, N. and Jouvet, M. (1991) Detection of the release of 5-hydroxyindole compounds in the hypothalamus and the n. raphe dorsalis throughout the sleep–waking cycle and during stressful situations in the rat: a polygraphic and voltammetric approach. Exp. Brain Res., 85: 153–162. Imeri, L., De Simoni, M.G., Giglio, R., Clavenna, A. and Mancia, M. (1994) Changes in the serotonergic system during the sleep–wake cycle – simultaneous polygraphic and voltammetric recordings in hypothalamus using a telemetry system. Neuroscience, 58: 353–358. Jacobs, B.L. and Azmitia, E.C. (1992) Structure and function of the brain serotonin system. Physiol. Rev., 72: 165–229. Jacobs, B.L. and Fornal, C.A. (1999) Activity of serotonergic neurons in behaving animals. Neuropsychopharmacology, 21: S9–S15. Jouvet, M. (1999) Sleep and serotonin: an unﬁnished story. Neuropsychopharmacology, 21: 24S–27S. Kabiersch, A., del Rey, A., Honegger, C.G. and Besedovsky, H.O. (1988) Interleukin-1 induces changes in norepinephrine metabolism in the rat brain. Brain Behav. Immun., 2: 267–274. Kahn, R.S., Wetzler, S., van Praag, H.M., Asnis, G.M. and Strauman, T. (1988) Behavioral indications for serotonin receptor hypersensitivity in panic disorder. Psychiatry Res, 25: 101–104. Kirby, L.G., Allen, A.R. and Lucki, I. (1995) Regional differences in the effects of forced swimming on extracellular levels of 5-hydroxytryptamine and 5-hydroxyindoleacetic acid. Brain Res, 682: 189–196.

314 Kirby, L.G., Chougreen, J.M., Davis, K. and Lucki, I. (1997) The effects of different stressors on extracellular 5-hydroxytryptamine and 5-hydroxyindoleacetic acid. Brain Res, 760: 218–230. Kitchener, P., Di Blasi, F., Borrelli, E. and Piazza, P.V. (2004) Differences between brain structures in nuclear translocation and DNA binding of the glucocorticoid receptor during stress and the circadian cycle. Eur. J. Neurosci., 19: 1837–1846. Lacosta, S., Merali, Z. and Anisman, H. (1999) Behavioral and neurochemical consequences of lipopolysaccharide in mice: anxiogenic-like effects. Brain Res., 818: 291–303. Lauer, C.J., Riemann, D., Wiegand, M. and Berger, M. (1991) From early to late adulthood. Changes in EEG sleep of depressed patients and healthy volunteers. Biol. Psychiatry, 29: 979–993. Lavicky, J. and Dunn, A.J. (1995) Endotoxin administration stimulates cerebral catecholamine release in freely moving rats as assessed by microdialysis. J. Neurosci. Res., 40: 407–413. Leung, L.S. (1998) Generation of theta and gamma rhythms in the hippocampus. Neurosci. Biobehav. Rev., 22: 275–290. Linthorst, A.C.E. (2005) Stress, corticotropin-releasing factor and serotonergic neurotransmission. In: Steckler, T., Kalin, N.H. and Reul, J.M.H.M. (Eds.) Handbook of Stress and the Brain; Part 1: The Neurobiology of Stress, Vol. 15. Elsevier, Amsterdam, pp. 503–524. Linthorst, A.C.E., Flachskamm, C., Barden, N., Holsboer, F. and Reul, J.M.H.M. (2000) Glucocorticoid receptor impairment alters CNS responses to a psychological stressor: an in vivo microdialysis study in transgenic mice. Eur. J. Neurosci., 12: 283–291. Linthorst, A.C.E., Flachskamm, C., Holsboer, F. and Reul, J.M.H.M. (1994) Local administration of recombinant human interleukin-1 beta in the rat hippocampus increases serotonergic neurotransmission, hypothalamic-pituitary-adrenocortical axis activity, and body temperature. Endocrinology, 135: 520–532. Linthorst, A.C.E., Flachskamm, C., Holsboer, F. and Reul, J.M.H.M. (1995a) Intraperitoneal administration of bacterial endotoxin enhances noradrenergic neurotransmission in the rat preoptic area: relationship with body temperature and hypothalamic-pituitary-adrenocortical axis activity. Eur. J. Neurosci., 7: 2418–2430. Linthorst, A.C.E., Flachskamm, C., Holsboer, F. and Reul, J.M.H.M. (1996) Activation of serotonergic and noradrenergic neurotransmission in the rat hippocampus after peripheral administration of bacterial endotoxin: involvement of the cyclo-oxygenase pathway. Neuroscience, 72: 989–997. Linthorst, A.C.E., Flachskamm, C., Hopkins, S.J., Hoadley, M.E., Labeur, M.S., Holsboer, F. and Reul, J.M.H.M. (1997) Long-term intracerebroventricular infusion of corticotropin-releasing hormone alters neuroendocrine, neurochemical, autonomic, behavioral, and cytokine responses to a systemic inﬂammatory challenge. J. Neurosci., 17: 4448–4460.

Linthorst, A.C.E., Flachskamm, C., Mu¨ller-Preuss, P., Holsboer, F. and Reul, J.M.H.M. (1995b) Effect of bacterial endotoxin and interleukin-1 beta on hippocampal serotonergic neurotransmission, behavioral activity, and free corticosterone levels: an in vivo microdialysis study. J. Neurosci., 15: 2920–2934. Linthorst, A.C.E., Flachskamm, C., and Reul, J.M.H.M., (2001) Hippocampal serotonin responses to forced swim stress in rats: inﬂuence of water temperature. In: Proceedings of the 31st Annual Meeting of the Society for Neuroscience, San Diego, CA, U.S.A. Linthorst, A.C.E., Pen˜alva, R.G., Flachskamm, C., Holsboer, F. and Reul, J.M.H.M. (2002) Forced swim stress activates rat hippocampal serotonergic neurotransmission involving a corticotropin-releasing hormone receptor-dependent mechanism. Eur. J. Neurosci., 16: 2441–2452. Lowry, C.A. (2002) Functional subsets of serotonergic neurones: implications for control of the hypothalamic-pituitaryadrenal axis. J. Neuroendocrinol., 14: 911–923. Lucki, I. (1998) The spectrum of behaviors inﬂuenced by serotonin. Biol. Psychiatry, 44: 151–162. Maes, M. and Meltzer, H.Y. (1995) The serotonin hypothesis of major depression. In: Bloom, F.E. and Kupfer, D.J. (Eds.), Psychopharmacology: The Fourth Generation of Progress. Raven Press, New York, pp. 933–944. Mann, J.J. (1998) The neurobiology of suicide. Nat. Med., 4: 25–30. Mendlin, A., Martin, F.J., Rueter, L.E. and Jacobs, B.L. (1996) Neuronal release of serotonin in the cerebellum of behaving rats: an in vivo microdialysis study. J. Neurochem., 67: 617–622. Merali, Z., Lacosta, S. and Anisman, H. (1997) Effects of interleukin-1-beta and mild stress on alterations of norepinephrine, dopamine and serotonin neurotransmission – a regional microdialysis study. Brain Res, 761: 225–235. Morzorati, S.L. and Johnson, T.B. (1999) Serotonergic neuronal activity in the dorsal raphe nucleus of selectively bred alcohol-preferring and alcohol-nonpreferring rats and unselected wistar rats. Alcohol. Clin. Exp. Res., 23: 1362–1367. Nemeroff, C.B., Owens, M.J., Bissette, G., Andorn, A.C. and Stanley, M. (1988) Reduced corticotropin releasing factor binding sites in the frontal cortex of suicide victims. Arch. Gen. Psychiat., 45: 577–579. Nemeroff, C.B., Widerlov, E., Bissette, G., Walleus, H., Karlsson, I., Eklund, K., Kilts, C.D., Loosen, P.T. and Vale, W. (1984) Elevated concentrations of CSF corticotropin-releasing factor-like immunoreactivity in depressed patients. Science, 226: 1342–1344. Nutt, D.J. (2002) The neuropharmacology of serotonin and noradrenaline in depression. Int. Clin. Psychopharmacol., 17(Suppl. 1): S1–S12. Oshima, A., Flachskamm, C., Reul, J.M.H.M., Holsboer, F. and Linthorst, A.C.E. (2003) Altered serotonergic neurotransmission but normal hypothalamic-pituitary-adrenocortical axis activity in mice chronically treated with the corticotropin-releasing hormone receptor type 1 antagonist NBI 30775. Neuropsychopharmacology, 28: 2148–2159.

315 Park, S.P., Lopez-Rodriguez, F., Wilson, C.L., Maidment, N., Matsumoto, Y. and Engel, J. (1999) In vivo microdialysis measures of extracellular serotonin in the rat hippocampus during sleep–wakefulness. Brain Res, 833: 291–296. Pauli, S., Linthorst, A.C.E. and Reul, J.M.H.M. (1998) Tumour necrosis factor-alpha and interleukin-2 differentially affect hippocampal serotonergic neurotransmission, behavioural activity, body temperature and hypothalamic-pituitary-adrenocortical axis activity in the rat. Eur. J. Neurosci., 10: 868–878. Pen˜alva, R.G., Flachskamm, C., Zimmermann, S., Wurst, W., Holsboer, F., Reul, J.M.H.M. and Linthorst, A.C.E. (2002) Corticotropin-releasing hormone receptor type 1-deﬁciency enhances hippocampal serotonergic neurotransmission: an in vivo microdialysis study in mutant mice. Neuroscience, 109: 253–266. Pen˜alva, R.G., Lancel, M., Flachskamm, C., Reul, J.M.H.M., Holsboer, F. and Linthorst, A.C.E. (2003) Effect of sleep and sleep depriviation on serotonergic neurotransmission in the hippocampus: a combined in vivo microdialysis/EEG study in rats. Eur. J. Neurosci., 17: 1896–1906. Pepin, M.C., Pothier, F. and Barden, N. (1992) Impaired type II glucocorticoid-receptor function in mice bearing antisense RNA transgene. Nature, 355: 725–728. Porsolt, R.D., Le Pichon, M. and Jalfre, M. (1977) Depression: a new animal model sensitive to antidepressant treatments. Nature, 266: 730–732. Portas, C.M., Bjorvatn, B., Fagerland, S., Gronli, J., Mundal, V., Sorensen, E. and Ursin, R. (1998) On-line detection of extracellular levels of serotonin in dorsal raphe nucleus and frontal cortex over the sleep/wake cycle in the freely moving rat. Neuroscience, 83: 807–814. Portas, C.M. and McCarley, R.W. (1994) Behavioral state-related changes of extracellular serotonin concentration in the dorsal raphe nucleus: a microdialysis study in the freely moving cat. Brain Res., 648: 306–312. Python, A., Steimer, T., de Saint, H.Z., Mikolajewski, R. and Nicolaidis, S. (2001) Extracellular serotonin variations during vigilance states in the preoptic area of rats: a microdialysis study. Brain Res., 910: 49–54. Raadsheer, F.C., Hoogendijk, W.J.G., Stam, F.C., Tilders, F.J.H. and Swaab, D.F. (1994) Increased numbers of corticotropin-releasing hormone expressing neurons in the hypothalamic paraventricular nucleus of depressed patients. Neuroendocrinology, 60: 436–444. Raadsheer, F.C., Van Heerikhuize, J.J., Lucassen, P.J., Hoogendijk, W.J.G., Tilders, F.J.H. and Swaab, D.F. (1995) Corticotropin-releasing hormone mRNA levels in the paraventricular nucleus of patients with Alzheimer’s disease and depression. Am. J. Psychiatry, 152: 1372–1376.

Rasmussen, K., Heym, J. and Jacobs, B.L. (1984) Activity of serotonin-containing neurons in nucleus centralis superior of freely moving cats. Exp. Neurol., 83: 302–317. Ressler, K.J. and Nemeroff, C.B. (2000) Role of serotonergic and noradrenergic systems in the pathophysiology of depression and anxiety disorders. Depress. Anxiety, 12: 2–19. Reul, J.M.H.M., Gesing, A., Droste, S., Stec, I.S.M., Weber, A., Bachmann, C., Bilang-Bleuel, A., Holsboer, F. and Linthorst, A.C.E. (2000) The brain mineralocorticoid receptor: greedy for ligand, mysterious in function. Eur. J. Pharmacol., 405: 235–249. Reul, J.M.H.M. and Holsboer, F. (2002) Corticotropin-releasing hormone receptors 1 and 2 in anxiety and depression. Curr. Opin. Pharmacol., 2: 23–33. Ringel, B.L. and Szuba, M.P. (2001) Potential mechanisms of the sleep therapies for depression. Depress. Anxiety, 14: 29–36. Rueter, L.E. and Jacobs, B.L. (1996a) A microdialysis examination of serotonin release in the rat forebrain induced by behavioral environmental manipulations. Brain Res., 739: 57–69. Rueter, L.E. and Jacobs, B.L. (1996b) Changes in forebrain serotonin at the light–dark transition: correlation with behaviour. Neuroreport, 7: 1107–1111. Sakai, K. and Crochet, S. (2001) Differentiation of presumed serotonergic dorsal raphe neurons in relation to behavior and wake-sleep states. Neuroscience, 104: 1141–1155. Shouse, M.N., Staba, R.J., Saquib, S.F. and Farber, P.R. (2000) Monoamines and sleep: microdialysis ﬁndings in pons and amygdala. Brain Res., 860: 181–189. Steinbusch, H.W.M. (1981) Distribution of serotonin-immunoreactivity in the central nervous system of the rat-cell bodies and terminals. Neuroscience, 6: 557–618. Stone, E.A. (1970) Behavioral and neurochemical effects of acute swim stress are due to hypothermia. Life Sciences – Part 1. Physiol. Pharmacol., 9: 877–888. To¨rk, I. (1990) Anatomy of the serotonergic system. Ann. N.Y. Acad. Sci., 600: 9–35. Ursin, R. (2002) Serotonin and sleep. Sleep Med. Rev., 6: 55–69. Van de Kar, L.D. and Blair, M.L. (1999) Forebrain pathways mediating stress-induced hormone secretion. Front. Neuroendocrinol., 20: 1–48. Wilkinson, L.O., Auerbach, S.B. and Jacobs, B.L. (1991) Extracellular serotonin levels change with behavioral state but not with pyrogen-induced hyperthermia. J. Neurosci., 11: 2732–2741. Zalcman, S., Green-Johnson, J.M., Murray, L., Nance, D.M., Dyck, D., Anisman, H. and Greenberg, A.H. (1994) Cytokine-speciﬁc central monoamine alterations induced by interleukin-1, -2 and -6. Brain Res, 643: 40–49.

CHAPTER 4.2

Microdialysis of dopamine and norepinephrine during conditioning and operant behaviour Matthijs G.P. Feenstra Netherlands Institute for Neuroscience, Amsterdam, The Netherlands

Abstract: The catecholamines dopamine (DA) and norepinephrine (noradrenaline, NA) exert a strong modulatory control over the ‘‘motive circuit’’. This neuronal circuit comprises cortical, striatal and thalamic areas and underlies the behavioural expression of motivated behaviour, including classical and operant conditioning. Conditioned and unconditioned stimuli with a strong motivational salience have rapid access to the catecholaminergic systems and activation by such stimuli may result in the modulation of a wide array of neuronal processes in the brain. Microdialysis is the universally applicable method to determine the transmitter efﬂux resulting from this activation. In this chapter, the evidence that DA and NA efﬂux is increased by conditioned approach and avoidance behaviour and during instrumental actions is reviewed and data obtained using microdialysis are compared with results from in vivo electrophysiological and electrochemical studies. Special attention is paid to the suggestions that the catecholamines support ﬂexibility in operant responding and may have a unique role in reward-related learning. concentrations when the animals go from the sleeping to the waking state and when they are aroused, either by exposure to novelty, to a desirable reward or to a stressor like prolonged handling and restraint (see Table 2). The original, unitary concept of arousal has been critically

I. Introduction The catecholamines dopamine (DA) and norepinephrine (noradrenaline, NA) have been grouped together with the other monoamines, serotonin and histamine, and with acetylcholine as transmitters of the reticular activating system or ‘‘arousal transmitters’’ (see Marrocco et al., 1994; Panksepp, 1998; Marrocco and Field, 2002; Jones, 2003) (for abbreviations, see Table 1). The activity of these transmitters is strongly related to the waking state (Jones, 2005). In the context of this volume on microdialysis, it is a relevant concept: any scientist who has tried to measure the efﬂux of one of these transmitters in freely moving animals will have noticed the importance of the behavioural state of the animal. These transmitters show apparently similar increases in extracellular

Table 1. List of abbreviations CP CS DA DOPAC FR HVA LC NA NAC PFC TTX US VR VTA

Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

317

Caudate-putamen Conditioned stimulus Dopamine Dihydroxyphenylacetic acid Fixed ratio Homovanillic acid Locus coeruleus Noradrenaline, norepinephrine Nucleus accumbens Prefrontal cortex Tetrodotoxin Unconditioned stimulus Variable ratio Ventral tegmental area

DOI: 10.1016/S1569-7339(06)16018-X Copyright 2007 Elsevier B.V. All rights reserved

318 Table 2. Similar effects of arousing stimuli on extracellular brain levels of monoamines and acetylcholine measured by microdialysis in rats and mice (m, increased; ¼ , no change) Sleep or light-wake or dark

Novelty

Handling stress

NA

mKale´n et al. (1989); Stanley et al. (1989); Feenstra et al. (2000)

mDalley and Stanford (1995); Ihalainen et al. (1999); Feenstra et al. (2000)

mKale´n et al. (1989); Cenci et al. (1992); Feenstra et al. (1998, 2000); Ihalainen et al. (1999); Kawahara et al. (1999)

DA

mSmith et al. (1992b); Paulson and Robinson (1994); Feenstra et al. (2000); Lena et al. (2005)

mFeenstra et al. (1995, 2000); Feenstra and Botterblom (1996); Ladurelle et al. (1995); Saigusa et al. (1999); Ihalainen et al. (1999); Legault et al. (2001); ¼ Damsma et al. (1992); Fulford and Marsden (1998)

mCenci et al. (1992); Feenstra et al. (1995, 1998, 2000); Feenstra and Botterblom (1996); Enrico et al. (1998); Ihalainen et al. (1999); Kawahara et al. (1999)

5-HT

mKale´n et al. (1989); Rueter and Jacobs (1996); Portas et al. (2000); Park et al. (1999)

mBickerdike et al. (1993)

mKale´n et al. (1989); Adell et al. (1997); Cenci and Kale´n (2000); Fujino et al. (2002)

His

mYamatodani et al. (1996); Strecker et al. (2002)

ACh

mJime´nez-Capdeville and Dykes (1996); Mitsushima et al. (1996)

mWesterink et al. (2002) mAcquas et al. (1996); Aloisi et al., (1997); Thiel et al. (1998a); Miranda et al. (2000)

reviewed (Robbins, 1997) and the importance of individual evaluation and characterization of the activity and function of arousal transmitters has been stressed. Indeed, a closer look at the reactivity of these ‘‘arousal’’ systems to stimulus presentations and other changes in speciﬁc, welldeﬁned behavioural conditions has resulted in the discovery of sometimes clear, but often subtle differences (see, e.g. Dalley et al., 2001; Mingote et al., 2004; Rossetti and Carboni, 2005 for differential activation of acetylcholine and NA, or DA and NA, respectively, during behavioural tasks). Among the ‘‘arousal’’ transmitters the catecholamines DA and NA are most intimately related: not only do they share the same biosynthetic mechanism, but they also have similar afﬁnities for many of the key-regulatory proteins such as receptors and uptake carriers. Indeed, DA even has a higher afﬁnity for the NA transporter than NA itself (Hoffman, 1994). Moreover, in particular the cortical efﬂux of DA and NA shows similar reactivity to simple, unconditioned stimuli (Feenstra et al., 2000; Feenstra, 2000) and to some drugs (Westerink et al., 1998; Devoto et al., 2003). These similarities make a comparison of the

mNilsson et al. (1990); Kale´n et al. (1990); Moor et al. (1998); Thiel et al. (1998b)

activity of DA and NA systems in behaviour even more interesting. Therefore, in this chapter, the activity of the DA and NA systems as measured with in vivo microdialysis in two behavioural conditions, classical (Pavlovian) conditioning and operant conditioning, will be reviewed. The scope will be limited to ‘‘normal’’ rewards and punishments and microdialysis studies dealing with drugs of abuse will not be mentioned. The data will be presented and discussed in relation to other measurements of DA and NA activity, such as electrophysiological measurement of neuronal activity, electrochemical measurement of transmitter efﬂux and ex vivo measurements of transmitter turnover.

II. Methods II.A. Conditioning In the environment, an animal is exposed to many stimuli. The discovery of reliable and consistent associations between some of these stimuli may help to make reliable predictions about the

319

environment. Stimuli that by themselves seem to be neutral and not very important may predict the presence of a motivational salient stimulus, such as food or a predator. An animal is said to be conditioned when it shows a reaction to the originally neutral stimulus that is similar to its reaction to the motivationally salient stimulus itself. In classical or Pavlovian conditioning, the animal’s reaction cannot control the appearance of this salient stimulus, whereas in operant conditioning, it can. In a controlled set-up for a test of Pavlovian behaviour, conditions are programmed so that a certain sound or light (or environment) is always (or most of the time) automatically followed by the delivery of food, or the delivery of a foot shock. The animal will react to the sound or light as it would to the food or the foot shock itself. The sound and light are called conditioned stimulus (CS) and the food and foot-shock unconditioned stimulus (US). The Pavlovian association that is learned is the CS–US association and the important factor here is the contingency of the presentation of these stimuli (Rescorla, 1988). In operant conditioning, the animal reacts to a stimulus with a response to achieve something and reach a goal: be it the presentation of a reward or the escape from danger. In theory, this is much more relevant to daily life than Pavlovian conditioning – we (or any other animal) tend to act on or interact with the environment, not just react to it. The association is not a simple one in this case, as the stimulus (CS), the action and the outcome may form three different associations (Robbins and Everitt, 1996). The contingency between action and outcome determines the operant character of this behaviour (Balleine and Dickinson, 1998). However, Pavlovian mechanisms are strongly involved in operant conditioning also, as the stimulus is associated to both the outcome and the action. The ﬁrst association may resemble the CS–US relation in classical conditioning, whereas the second one may lead to instrumental habit formation. Thus, in practice, it is difﬁcult to study operant behaviour apart from Pavlovian behaviour. A neuronal circuit in the brain that underlies the behavioural expression of classical or operant conditioning is the ‘‘motive circuit’’ (Robbins and Everitt, 1996; Kalivas and Nakamura, 1999; Cardinal

et al., 2002). This limbic cortico-striatal thalamic circuit receives strong modulatory inputs from the ‘‘arousal’’ transmitters. As both conditioned and unconditioned stimuli have an attention-grabbing function and should have immediate access to important circuits on all levels of brain function, a broad spread of this salient information through modulatory monoaminergic systems seems a relevant mechanism. Moreover, one would expect that higher (cortical) areas might exert a supervisory control over these systems. Indeed, the prefrontal cortex (PFC), an important component of the motive circuit and the hypothesized seat of cognitive control in the brain (Miller and Cohen, 2001), sends efferents to all nuclei of origin of these modulatory transmitters (Uylings and van Eden, 1990; Groenewegen and Uylings, 2000) and has been shown to modulate their activity (e.g. Jodo et al., 1998; Sesack and Carrr, 2002; Puig et al., 2003).

II.B. Microdialysis and neurochemistry Neuroscientists have always tried to ﬁnd methods to study the ﬁnal and essential result of activation of a transmitter system, that is, the release of the transmitter itself. The aim was to use speciﬁc and sensitive techniques that could measure the concentration of the transmitter in the extracellular space (Ungerstedt, 1984). The introduction of microdialysis 20–25 years ago allowed researchers to measure extracellular DA and NA concentrations in striatal and cortical brain structures of the rat (Hernandez et al., 1983; Zetterstro¨m et al., 1983; L’Heureux et al., 1986; Sharp et al., 1986). This extracellular concentration is for monoamine transmitters several orders of magnitude lower than the total tissue concentration, which predominantly represents the intracellular stores of transmitter (Westerink et al., 1987; Di Chiara, 1991). An essential prerequisite for the acceptation of microdialysis as a technique has been the possibility to prove that the extracellular concentrations of the transmitter directly reﬂect neuronal release, that is, they are Ca- and TTX-sensitive (Imperato and Di Chiara, 1984; Westerink et al., 1987). A rapid decrease after TTX perfusion is regarded as the best indication that basal extracellular levels

320

that are determined directly depend on neuronal activity and impulse ﬂow (Westerink, 1995). A word on terminology is needed here: neuronal release refers to a well-deﬁned and speciﬁc process in which neurotransmitters are set free from the presynaptic terminal and enter the extracellular compartment. Release is an extremely rapid process that cannot be followed by a method relying on minute-to-minute sampling like microdialysis. What microdialysis can give us is a smoothed-out measure of the appearance, persistence and disappearance of the released transmitter in the extracellular ﬂuid. This indirect measure of release proper may be called transmitter overﬂow or efflux, the term that will be used in this chapter. Behaviour-related transmitter release or efﬂux cannot be determined in post mortem brain tissue samples. When repeated in vivo determinations are not possible, measurements of transmitter turnover are still used instead. Turnover of a transmitter is the amount that is being removed from the available metabolic pool by transport or metabolism. The turnover in a certain time period is the turnover rate, which is an indication of the activity of that transmitter system. The turnover rate can be determined after blocking synthesis and measuring the disappearance (utilization, as disappearance will depend on the use, that is, activity of the transmitter system), after blocking metabolism or transport and measuring the accumulation or by measuring the incorporation or disappearance of radioactive label in the transmitter pool (Pycock and Taberner, 1981). A rough measure of the transmitter activity is provided by the transmitter metabolite concentration. These techniques were employed in many previous studies, but have been replaced by the newer in vivo techniques, such as microdialysis. The main advantage is the possibility to study many brain areas in one animal. Microdialysis is the ﬁrst and still the only technique that is generally applicable to give an indication of transmitter release in the brain of an animal and relate it to the behaviour of that animal. The potential of microdialysis in the study of behaviour has been reviewed (see general reviews by Young, 1993; Westerink, 1995; Bruno et al., 1999; Fillenz, 2005). Whereas the progress has

been substantial, the challenge is still to fulﬁl the full potential of describing the transmitter activity during advanced behavioural settings.

II.C. Neurochemistry and behaviour The combination of in vivo neurochemical sampling with behavioural measurements imposes important demands on both the neurochemical and the behavioural methods. First, the methods need temporal tuning. Microdialysis samples need to be deﬁned with respect to the start or the end of a speciﬁc behavioural condition. Here, the restricted temporal resolution of microdialysis is frustrating. One would like to relate each separate behavioural action to a neurochemical measurement, but even fast microdialysis methods (e.g. 1- or 2-min sampling; Young, 2004; Lena et al., 2005) cannot be expected to reach this level of resolution. The only way this can be achieved is by using electrochemical in situ techniques such as fast-scan cyclic voltammetry that allow subsecond time resolution (although at the moment this is only possible for DA in striatal areas; Wightman and Robinson, 2002; Phillips et al., 2003b). However, such methods are less suitable to follow extracellular concentrations over longer periods of time – for that the use of microdialysis is indispensable. One way to obtain temporal tuning is to strive for a precise convergence of a dialysis sample with a deﬁned number of behavioural trials. Thus, in Cheng et al. (2003), each 7.5-min sample represented two trials in a classical conditioning paradigm and in Cheng and Feenstra (2006a), each 3.75-min sample also represented two trials, now in an instrumental learning task. It is important that the time taken by the dialysate to travel from the dialysis tubing to the collector (or injector when samples are on-line injected) is known exactly as it must be used to adjust the start and stop times of sample collection. Second, the fluid connection of the animal to the dialysis set-up should be possible and practical. This requires a relatively small behavioural apparatus, such as a Skinner box (Figs. 1–3) or a small open arena. Larger set-ups, such as a water maze,

321

Fig. 1. Set-up in the author’s laboratory for on-line microdialysis measurements during behavioural testing in a Skinner box. The box is enclosed in a sound-attenuating outer box with a hole in the top to accommodate the ﬂuid lines from microdialysis pump, via the two-channel swivel, to the injection port of the HPLC system and electrochemical detector (on the shelf).

can only be used when very long ﬂuid lines are used. These might lead to the build-up of backpressure and may increase problems with ﬂow-rate and occurrence of leaking and blocked probes. The ﬂuid connection must always include a ﬂuid swivel, as behavioural (and neurochemical) measurements are only useful when the animal can have freedom of movement in the box. Third, the behavioural set-up should not require frequent handling of the animal. Set-ups in which frequent handling of the animals cannot be avoided are, in principle, more sensitive to differences between experimenters and may include trial-to-trial variations. In addition, handling may lead to neurochemical effects that do not readily habituate. Placing rats into a Skinner box led each time to similar, signiﬁcant increases in cortical DA and NA efﬂux (Feenstra et al., 1999). Therefore, it is important to account for all possible experimenter–animal interactions. Only when animals are in full sound- and light-proof boxes no precautions are required. However, when only partial measures are taken, the collection of samples, the starting of behavioural tasks or just the entry into the behavioural testing room will be noticed

by the animal. Any arousal of the animal might be reﬂected in an increase of monoamine efﬂux. Therefore, incidental noise should be masked by background noise, for example, a radio playing constantly. Fourth, it seems logical to study behaviour in the active period of the animal. For rats and mice, this is the dark period. Many behavioural studies are still performed in the light period, when these animals are less active. It might be reasoned that even when they can perform the task well, the difference between their activity during the task and the resting periods around and in between task sessions might be exaggerated and task-related activations might be larger than they would be when behaviour is assessed during the active period. This question was addressed by Feenstra et al. (2000), who exposed rats to novelty and handling stress in the light and the dark periods and determined DA and NA efﬂux in PFC and nucleus accumbens (NAC). In the cortex, basal efﬂux was higher in the dark but relative stress increases were very similar in both periods. Absolute increases were often higher in the dark period. In the accumbens, no state-dependent

322

Fig. 2. The same set-up as in Fig. 1, now with the doors of the outer box opened. The inside of the Skinner box is visible. A water bottle is present for the rats that stay for several hours in this box during microdialysis experiments.

difference was observed in DA efﬂux, but handling induced increases only in the light, not in the dark. Thus, results obtained in the inactive period might not always reﬂect similar effects in the active period. Fifth, the recovery time between surgical implantation of microdialysis probes or guide cannulae and the behavioural test should be long enough for the animal to behave normally again. Normal circadian rhythms will reappear only after 1 week or more (Drijfhout et al., 1995; Westerink, 1995). Operant behaviour will need a longer recovery time than the less demanding Pavlovian conditioning (van der Stelt, de Bruin and Feenstra, unpublished observations).

III. DA and NA: basic considerations A major ﬁnding of microdialysis studies was the constant presence of low nanomolar concentrations of both NA and DA in the dialysates obtained from cortical areas and higher concentrations of DA in striatal areas. The rapid and strong decrease in extracellular concentrations when TTX is perfused shows that these basal levels relate directly to a constant neuronal release of the catecholamines (Westerink et al., 1987). One would expect that the extracellular concentrations of ‘‘arousal’’ transmitters would show a strong dependence on circadian rhythms. Classical electrophysiological studies indeed showed marked state-dependent activity of

323

Fig. 3. The operational panel of the Skinner box: (A) food tray, equipped with infrared sensors; (B) retractable levers; (C) stimulus lights; (D) movement sensor; (E) dispenser for the reward pellets and (F) grid ﬂoor.

locus coeruleus (LC) neurons (high during active waking and low during sleeping, depending on sleep stadia) but much less of DAergic neurons in the ventral tegmental area (VTA) and substantia nigra, with no dependence on sleep stadia (Jacobs, 1987). Microdialysis studies of striatal and accumbens efﬂux supported a moderate state-dependency of striatal efﬂux of DA (Smith et al., 1992b; Paulson and Robinson, 1994), but studies of cortical efﬂux showed a marked decrease of 50% in both NA and DA concentrations in the cortex of rats in the light period compared with the dark period (Feenstra et al., 2000; Lena et al., 2005). Cortical DA seems to share this property with NA and the other ‘‘arousal’’ transmitters (Table 2). A further outcome of the microdialysis studies was a similar reactivity of cortical DA and NA efﬂux to stressors and food rewards (Feenstra, 2000). The increase in efﬂux after novel, appetitive or aversive stimuli is instantaneous for both catecholamines. DA and NA in other brain areas (amygdala, hippocampus and hypothalamus) show responses to unconditioned stimuli, similar to the other ‘‘arousal’’ transmitters (Table 2). It is important to realize, however, that responses to such stimuli are not unitary and constant. DA responses to reward may depend on the

history of the animal, for example, responses may habituate, as demonstrated for DA in NAC shell (Bassareo and Di Chiara, 1997). Responses may also be devaluated and show negative contrast (they are lower in animals that were used to more preferred rewards; Genn et al., 2004) or sensoryspeciﬁc satiety (they are lower in animals that had previous unlimited access to the reward; Ahn and Phillips, 1999). On learning about predictive stimuli, responses to reward may shift to the predictor, so that the reward itself only evokes a response when presentation is unexpected (Schultz, 1998). Responses to punishing stimuli, such as foot shocks, may also vary: normal increases in cortical DA after foot shocks (see Feenstra, 2000) were absent in animals that pressed a lever for reward and obtained both the reward and the punishment and later received the foot shocks non-contingently (Beaufour et al., 2001). Novelty is not unitary, either: exposure to a novel environment that is very similar to the home cage increased NA and DA efﬂux by 50% (Feenstra et al., 2000), whereas a novel environment that is very different (Skinner box) induced an increase of more than 100% (Feenstra et al., 1999). Additional foot shocks also did not add to the cortical DA and NA responses to unfamiliar novelty,

324

whereas reward presentation attenuated this novelty-induced response (Feenstra et al., 1999). Many other examples of similar effects can be given. Both the motivational state of the animal and its familiarity to the situation, that is, the salience of the stimuli in the prevailing conditions, are important determinants of the DA and NA responses to any stimulus presented. Thus, already on the basis of this restricted evidence, we might conclude that salience is a very important factor in the activation of the catecholamine systems. An on-going discussion is whether the similar reactivity of DA and NA efﬂux in the cortex to appetitive and aversive stimuli extends to DA efﬂux in striatal areas such as NAC and caudateputamen (CP) (see Feenstra, 2000). These projections have been shown to have a lower reactivity to stressful stimuli than DA afferents in the PFC (Roth et al., 1988). Some groups have not been able to replicate effects of foot shock or aversive stimuli in general on DA efﬂux in striatal areas (see Feenstra, 2000; and this chapter, Section III.B). Moreover, DA neurons are not very responsive to aversive stimuli in the conscious primate (Mirenowicz and Schultz, 1996). In contrast, human imaging studies showed that during stress, D2-ligand binding to striatal receptors may be displaced, suggesting increased DA release (Pruessner et al., 2004). To clarify the issue of DA neuronal responses to aversive events, Ungless et al. (2004) studied neuronal responses in the VTA after giving foot pinches to the anaesthetized rat. Similar stimuli (e.g. tail pinches) have been shown to increase DA efﬂux in the conscious and freely moving animal using voltammetric and microdialysis techniques (e.g. Louillot et al., 1986; Pei et al., 1990). However, VTA neurons that were activated were not stained by an antibody to tyrosine hydroxylase. Thus, whereas most electrophysiological criteria for a classiﬁcation as DA neurons were fulﬁlled, these neurons turned out to be non-DAergic. In contrast, identiﬁed DAergic neurons were inhibited by foot pinch (Ungless et al., 2004). Subsequently, Ungless (2004) proposed a new theory to explain these apparently contradictory results and combine the microdialysis and electrophysiological data into one model: short-lasting aversive events might result in a longer lasting opponent process, that is, a

short decrease in DAergic activity might be followed by a longer lasting increase. It will be interesting to see whether future studies can support this hypothesis, as present microdialysis studies do not seem to do this (e.g. even using 1-min microdialysis measurements, Young (2004) could only detect increases in DA efﬂux following foot shocks). Some evidences for decreases in DA efﬂux after an aversive stimulus have been presented (see Ungless, 2004). However, these decreases were present exclusively in the shell of the NAC (after tail pinch, Di Chiara et al., 1999; or oral quinine administration, Bassareo et al., 2002), whereas rapid increases were obtained in other areas, such as the NAC core and PFC. The tail-pinch-induced decreases were not rapid (signiﬁcant effects were present only after the pinching was stopped) and were not followed by an increase (Di Chiara et al., 1999). An ‘‘opponent’’ (slow) increase was sometimes observed after quinine administration, but in these cases no rapid decrease was observed (Bassareo et al., 2002). Retarded increases in DA efﬂux from NAC have been observed more often (Feenstra et al., 1998, 2000), but cortical increases were always immediate. It is difﬁcult to explain these results by a general inhibitory effect of aversive events on DA activity. The opposite, a short-lasting increase in activity that is followed by a longer lasting decrease, has not been reported, either. These discussions have a more general importance: the apparent discrepancy in responsiveness of DA release and of VTA neuronal unit activity to aversive or appetitive events might reﬂect general differential temporal response properties of monoamine neurons. Maybe the time scales of current methods of release measurement are so different from those of electrophysiological methods that the two are reﬂecting different processes. Two distinct modes of activity could be represented by the phasic increases in activity, release or transmission following speciﬁc, short-lasting stimuli and by tonic activity, which is less well deﬁned (but includes the regular, pacemaker-like activity that has been described; Jacobs, 1987). DA cells were shown to be active in a single spike mode or in a bursting mode, with the latter being related to increased transmitter release (Bunney et al., 1991; Chergui et al., 1994). The bursting mode may be induced by speciﬁc

325

stimuli and represents a typical phasic response and may lead to accumulation of extracellular DA and to activation of receptors in larger volumes of brain tissue. Whereas empirical evidence for the existence of two modes of activity is available, it is, however, unknown how this is translated into behaviour-related ﬂuctuations in release as approximated by microdialysis of voltammetry measurements. Thus, it is not known if phasic and tonic modes of neuronal activity are related to different modes of release in the freely moving animal; it is not known either if methods like microdialysis (or human PETimaging of DA-receptor ligand displacement) are able to detect changes in release that are related to phasic changes in DA neuronal activity; it is not known if voltammetric methods can be used to follow extracellular concentrations over long periods of time; and it is not known if these modes are differentially regulated, that is, if tonic activity is increased by other internal or external states or events than phasic activity, although recent data reported by Floresco et al. (2003) suggest that this could be the case. The NA system has been less discussed in this respect, but similar issues may be relevant: NA may be active in different modes as well – LC neurons can ﬁre in a tonic or a bursting mode and the latter will lead to a higher terminal release, similar to that shown for DA (Florin-Lechner et al., 1996). Previous studies suggested clear phasic responses to stressful or noxious stimuli and cues that predicted those, but not to appetitive stimuli and predictive CS’s (Rasmussen and Jacobs, 1986). Contrary to DA systems, later studies, however, showed phasic responses to appetitive primary and conditioned stimuli, as well (e.g. Sara and Segal, 1991; Aston-Jones et al., 1997; Bouret and Sara, 2004). Therefore, LC neuronal activity compares well to NA efﬂux in these circumstances.

IV. Classical conditioning IV.A. General considerations Re-exposure to a stimulus that has been associated to an emotionally salient event, be it negative

(aversive or punishing) or positive (appetitive or rewarding), will lead to a conditioned response, that is, the subject reacts as if the US will be delivered. In the case of an aversive US, this is often called fear conditioning, and in the case of an appetitive US, reward prediction. The CS may be the context in which the US was presented: the environment, for example, the box in which the foot shocks or the rewards are delivered. This is called a contextual stimulus. However, the stimulus can also be explicit: a discrete, short-lasting sensory (e.g. auditory or visual) stimulus that will immediately precede the shocks or the rewards in an otherwise neutral environment. How can one show that monoamine release is activated by a CS? In a typical trial, the CS is immediately followed by the US. Whereas electrophysiological measurements of neuronal activity can easily discriminate these events, microdialysis measurements of monoamine efﬂux have restricted time resolution and are similar to the classic measurements of transmitter turnover or metabolism in which separate measurements are only possible when the CS and the US are presented in isolation. During normal conditioning trials, the response to the CS is measured in the same sample as the response to the US. As DA and NA (and all monoamines; see Table 2) may be reactive to the unconditioned stimuli that were employed in such studies (food reward and foot shock), this is often not a useful approach. Re-exposure to a stimulus (CS) that was learned to predict foot-shock or reward presentation should lead to conditioned release of the neurotransmitter, measured in the absence of effects of the US. Most studies take this approach. However, such an extinction trial leads to a prediction error and, if repeated, to new learning (Schultz and Dickinson, 2000), that is, now the CS does not predict the US anymore. Measurement of conditioned responses during extinction trials tacitly assumes that effects of prediction errors or extinction learning are less than the effect of the CS itself. A reliable way to study this is to measure the responses with high time resolution, allowing trialspeciﬁc responses or, even better, stimulus-speciﬁc responses. The latter is only open when subsecond measurements are carried out, using, for example,

326

cyclic voltammetry. Up to now, however, only studies using cocaine self-administration have been reported (Phillips et al., 2003b; Stuber et al., 2005). In extinction, DA responses after the stimulus were signiﬁcantly decreased only in the 10th trial (Stuber et al., 2005). Measuring effects of trialspeciﬁc responses is within the range of possibilities for the microdialysis methodology. Cheng et al. (2003) studied appetitive conditioning and showed that DA efﬂux was increased in each twotrial sample during two sessions of six extinction trials. Conditioned behaviour was, however, decreased in both sessions from trial 3 (i.e. sample 2) on. It might be that one longer extinction session of 10 or more trials would have resulted in decreased DA efﬂux, as well. Young (2004) also observed an increase in DA efﬂux upon CS alone presentation. These ﬁndings suggest that the ﬁrst extinction trials lead to increases in DA efﬂux, indicating that, indeed, CS-induced responses are stronger than possible prediction errors. A further prerequisite to show conditioned transmitter release is the use of proper control groups. A simple approach is the presentation of the CS-alone in a second group of animals: this should show that the CS has no (or weak) activating properties by itself. More elaborate approaches use control groups in which both CS and US have been presented to the animal but in an unrelated manner, that is, a non-paired or non-contingent stimulus presentation. This is sometimes also called ‘‘pseudoconditioning’’. The history of the animals is very similar: they were exposed to the same stimuli, but did not learn the association between the two. This procedure is most popular and has been applied often. But even with a truly random presentation (Rescorla, 1967) of CS and US, some behavioural conditioning does occur (Rescorla, 2000). However, on subsequent CS-alone presentation, in general, no evidence for conditioned behaviour is obtained (Cheng et al., 2003). The other approach is to use two different CS’s, CS+ and CS. When CS+ is always paired to the US and never to the CS, a good control situation might be obtained. A problem is that in this case also learning might occur, even if it is only negative learning (CS predicts absence of reward in a context where reward can be obtained).

Moreover, stimulus generalization may also develop, in which case the CS partially adopts response-inducing properties of the CS+.

IV.B. Aversive conditioning Re-exposure to a CS that predicts foot shocks was shown to increase central NA and DA turnover and metabolism (Tilson et al., 1975; Cassens et al., 1980; Herman et al., 1982; Tsuda et al., 1986; Goldstein et al., 1996; Morrow et al., 1999), as well as DA neuronal unit activity (Trulson and Preussler, 1984; Guarraci and Kapp, 1999) and NA neuronal activity in the LC (c-fos expression: Smith et al., 1992a; unit activity: Rasmussen and Jacobs, 1986). For more details see Feenstra (2000). Microdialysis measurements in fear conditioning showed increased DA and NA in several brain areas (Yokoo et al., 1990; Young et al., 1993; Yoshioka et al., 1996; for more details see Feenstra, 2000). McQuade and Stanford (2000) demonstrated increased NA efﬂux in the frontal cortex when rats were exposed to a sound that predicted transfer from a zone with low illumination to a zone that was brightly illuminated. This suggests that other inescapable aversive events can be used as a CS in aversive conditioning, in addition to foot shocks. All these studies were aimed at either DA or NA. A direct comparison of both catecholamines using combined measurements in the medial PFC showed parallel and similar increases for DA and NA (Feenstra et al., 2001). Whereas effects of conditioned fear on DA and/ or NA efﬂux in most brain areas have been reported (recently DA in amygdala; Yokoyama et al., 2005) or conﬁrmed and are not disputed, the responsivity of accumbens DA to conditioned aversive stimuli continues to be discussed. Previous ex vivo studies had suggested that fear-conditioned increase in DA metabolism in NAC did not occur (Herman et al., 1982; Deutch et al., 1985), but later an increase was reported (Goldstein et al., 1994, 1996; Morrow et al., 1999, 2000). These were conﬁrmed by microdialysis studies using an explicit sensory stimulus as CS (Young et al., 1993, 1998; Young, 2004) or a combination of sensory and contextual stimuli as CS (Saulskaya and

327

Marsden, 1995; Fulford and Marsden, 1998; Wilkinson et al., 1998). Conditioned activation of DA efﬂux in NAC subareas was again reported by Murphy et al. (2000) and Pezze et al. (2001, 2002), but Levita et al. (2002) reported negative ﬁndings. Guarraci and Kapp (1999) also observed that not all of their reactive DA neurons showed increased activity on CS presentation and that 22% had decreased activity. A possible solution might be that DA in the NAC shell apparently reacts to different stimuli than DA in the NAC core: Pezze et al. (2001) showed that a contextual CS increased DA in the core, not in the shell, whereas for an explicit (tone) stimulus, this was the other way around. Levita et al. (2002) indeed determined DA responses in the NAC core to an explicit stimulus. In the previous studies, either no distinction was made between NAC subareas or precise microdialysis probe locations were not presented. Therefore, it is difﬁcult to compare these studies in retrospect. A remarkable result was obtained by Besson and Louilot (1995): in a conditioned taste aversion procedure, a taste associated with an aversive LiCl injection induced a decrease in DA efﬂux in NAC core compared with control (conﬁrming previous results of Mark et al., 1991), but they observed an increase in NAC shell (conﬁrmed by Jeanblanc et al., 2002, who moreover showed that the increase compared with a control was present in the dorsal, not in the ventral, shell). In an interesting study, Young (2004) took 1-min dialysis samples and again observed that re-exposure to a CS predicting foot shocks led to a shortlasting, but clear increase in DA efﬂux. As the probes in this study were directed at the NAC shell, the results (as well as those of Besson and Louilot, 1995) are compatible with the ﬁndings of Pezze et al. (2001) (see also Pezze and Feldon, 2004). Moreover, DA efﬂux was increased more in the second and following tone–shock pairings, as compared with the ﬁrst pairing. This is a conﬁrmation of the ﬁndings by Wilkinson et al. (1998) that already after one pairing, the animals associated the tone with a prediction of a salient (in this case, punishing) event and that this led to an activation of DA efﬂux. These results suggest that voltammetric studies could shed more light on these matters. However, only one study has been presented

in which voltammetric measurements were carried out during appetitive and aversive conditioning (see abstract: Roitman et al., 2005). A CS associated to intraoral delivery of a quinine solution did not increase DA efﬂux, whereas a CS associated to a sucrose solution did. Clearly, these results must be conﬁrmed with other aversive stimuli. The ﬁnding of decreased DA measurements after the intraoral delivery of quinine is, however, remarkable. Quinine solutions have not often been used in conditioning studies, but a previous (microdialysis) study suggests that their use may result in complex and area-selective effects on DA efﬂux in NAC: increased DA efﬂux was normally observed, except in the NAC shell, where occasionally decreases were detected (Bassareo et al., 2002). Pre-exposure to the CS leads to latent inhibition, a retardation of a subsequent association between that CS and a US. Microdialysis measurements showed an involvement of NAC DA, in which preexposure leads to lower DA increases on re-exposure (Young et al., 1993). Later studies (Murphy et al., 2000; Jeanblanc et al., 2002) conﬁrmed this basic result, but their anatomically more detailed studies led to conﬂicting results regarding selective involvement of NAC subareas. For a more extensive discussion of DA and latent inhibition and the implications for DA’s role in representing saliency, see Young et al. (2005). Another explanation of the variability of results for NAC DA in fear conditioning is that the developmental history of the rats has been shown to exert a strong inﬂuence on the DA responses. Rearing rats in social isolation from weaning leads to strongly increased responses to foot shocks and CS’s associated with the shocks (Fulford and Marsden, 1998; Lapiz et al., 2003). It could well be that more subtle differences in developmental history also lead to differences in DA reactivity. In conclusion, microdialysis studies suggest that both NA and DA systems may be activated by a CS that was associated to foot shocks or other punishing events and some support is available from voltammetric studies. Thus, these systems are involved in the conditioned response, which is expressed in the expectancy of the US to come. Evidence is available that the reaction of DA systems may depend on the kind of CS that is used.

328

DA systems might react differently to either (complex) contextual stimuli or (simple) sensory stimuli. A similar differential reactivity was suggested for the NA system on the basis of lesion studies (Selden et al., 1990). It is not clear if different US’s lead to similar conditioned responses, as a CS associated to LiCl-induced sickness led to a decrease, instead of an increase, as did CS’s associated to foot shocks or a brightly lit arena. Results with a wide variety of US are required to make more general statements about catecholamine systems in aversive conditioning. The presently available evidence suggests that catecholamine activity to a CS in Pavlovian aversive conditioning is usually increased, but depends on the particular identity of both the CS and the US. IV.C. Appetitive conditioning Fear conditioning is a very direct and strong process: often a few or even a single exposure sufﬁces to produce long-lasting conditioning. A stimulus predicting danger has a high arousal potency. When the opposite, appetitive conditioning, is compared with aversive (fear) conditioning, often considerably more CS–US pairings are required to obtain conditioned responses: 2 tone–shock pairings were used by Young et al. (1993), whereas the same group used 15 tone–pellet pairings (Datla et al., 2002). Similarly, Feenstra et al. (2001) used 9 tone–shock pairings, whereas the same group used 42 (Mingote et al., 2004) and 60 (Cheng et al., 2003) tone–pellet pairings. An important practical difference with aversive conditioning is that most experimental set-ups for aversive conditioning use a US that cannot be ignored or avoided. The appetitive set-ups often only deliver an appetitive US and leave it to the subject to take it or not. Only invasive methods for delivery (e.g. intraoral or intravenous delivery) allow a direct comparison (e.g. Roitman et al., 2005). Such methods, however, have not been used in microdialysis studies of Pavlovian conditioning. IV.C.1. NA Appetitive classical conditioning was studied less often for NA than for DA and only few

microdialysis studies have been performed. However, the results show that NA is involved in the response to appetitive CS’s. Increased NA efﬂux after appetitive conditioning using olfactory stimuli was observed in the olfactory bulb of foodrestricted mice presented with sugar rewards hidden underneath odour-sprinkled wood shavings (Brennan et al., 1998). Mingote et al. (2004) studied NA and DA efﬂux in rats that showed strong conditioned approach behaviour to the food tray after training for 2 days. No conditioned NA (or DA) efﬂux was observed in the rats that were fed ad libitum, but after increasing the salience of the stimulus, by placing the rats on food restriction and increasing the reward from two to four pellets, a conditioned NA (but not DA) efﬂux on CS-alone presentation was observed in rats that were trained with paired presentation of CS and US, but not in rats that were used to unpaired presentations. The acquisition of the conditioned behaviour was not different from the ad libitum fed group, suggesting that in this paradigm, activation of NA is strongly dependent on the emotional signiﬁcance that the stimulus obtained. These results compare well to the results of two studies in which LC unit activity was determined upon presentation of a CS that signalled food reward: an increase in foodrestricted rats (Sara and Segal, 1991), but no effect in freely fed cats (Rasmussen and Jacobs, 1986).

IV.C.2. DA Studies directed at DA have been carried out more often (see also Section V). Previous studies are discussed by Blackburn et al. (1992) (see also Feenstra, 2000). Reward-associated stimuli (the sight and smell of food or a receptive female rat, ‘‘intrinsic’’ CS’s) increased DA efﬂux in NAC (Damsma et al., 1992; Wilson et al., 1995; Ahn and Phillips, 1999; Bassareo and Di Chiara, 1999b) and the PFC (Merali et al., 2004). In vivo voltammetric (D’Angio and Scatton, 1989; Louilot et al., 1991) and amperometric (Phillips et al., 1993) studies also indicated that DA-related parameters were increased in NAC upon presentation of such stimuli and this was recently conﬁrmed using more sophisticated fast-scan cyclic voltammetric methods (Robinson

329

et al., 2002; see Section V.D). Food-associated stimuli, however, apparently increase NAC DA efﬂux only in food-restricted rats (Wilson et al., 1995) and not in freely fed animals (Bassareo and Di Chiara, 1997). This was shown to apply to a particular subarea of the NAC, that is, the core, as DA in the shell did not respond to food-related stimuli either in restricted or in freely fed rats (Bassareo and Di Chiara, 1999a, b). Using other, ‘‘extrinsic’’ CS’s, conditioning of DA efﬂux has been shown by a number of groups: Mark et al. (1994) reported that an appetitively conditioned taste increased DA efﬂux in NAC, but not CP. More usual sensory stimuli (auditory and visual) were used by Datla et al. (2002) and Cheng et al. (2003) and were shown to increase NAC DA when presented as CS-alone. In the latter paper, the increase was observed in rats that were trained with paired presentation of CS and US, but not in rats that were used to unpaired presentations. NAC shell and core were studied separately and no differences were observed. A somewhat different approach was taken by Harmer and Phillips (1999) who showed that DA in the amygdala was responsive to paired presentations of a tone or light CS and reward, but not to unpaired presentation of a different CS and reward. No extinction experiments were carried out (CS-alone presentations). The fact that no increases were observed in the unpaired group is in contrast to the results of Cheng et al. (2003). This suggests that either the amygdala DA is only responsive to predictive stimuli and not to food presentation (not likely, given the results of Hajnal and Le´na´rd, 1997) or the unpaired CS acquired inhibitory properties: this always indicates the absence of reward, whereas the paired CS in the same animal always indicates the presence of rewards. The Cheng et al.’s (2003) study used two different groups of animals for the paired and unpaired presentations, so that the unpaired CS had intermediate properties, neither excitatory (shown by the absence of a conditioned DA response when a CS-alone session was given) nor inhibitory (shown by the presence of a DA response when rewards are presented). The full course of Pavlovian appetitive conditioning has not been described using microdialysis or other in vivo techniques. Phillips et al. (2003a)

studied ex vivo DA immunoreactivity, which depends more on DA synthesis than on release. In the ﬁrst acquisition session, increased levels of DA immunoreactivity were demonstrated in almost all areas (NAC shell, CP, amygdala and medial PFC), whereas in fully conditioned animals (20th session), no region showed any increase. In an intermediate fourth session, increases were observed in amygdala, PFC and, most strongly, NAC shell, but not in CP. NAC core showed a decrease. These time- and region-selective changes do not indicate changes in release, but nevertheless suggest a possible differential activation of DA neurons in the course of conditioning. A number of studies were directed at appetitive conditioning of DA efﬂux in the PFC following the paper by Bassareo and Di Chiara (1997). These authors placed a plastic box in the home cage of the rats and showed that when placement was followed by presentation of palatable food (Fonzies), DA efﬂux increased in rats that were not food-restricted. When the box did not predict Fonzies’ presentation, no increase was observed. Prediction of a familiar palatable food (presentation behind a screen) also increased DA efﬂux (Ahn and Phillips, 1999). As referred to above, Mingote et al. (2004) could not demonstrate conditioned DA efﬂux when an auditory cue was used as CS in a Skinner box, neither in rats fed ad libitum nor in food-restricted rats. The profound differences in procedures make it difﬁcult to compare these results to those of Bassareo and Di Chiara (1997) and Ahn and Phillips (1999), but a comparison with the results of the aversive conditioning (Feenstra et al., 2001) and with the results with NA (Mingote et al., 2004) suggests that in this case the cue did not acquire enough emotional signiﬁcance for the rat to lead to increases in DA efﬂux. A possible explanation is the difference in stimuli and context used in the test. The box used by Bassareo and Di Chiara (1997) was placed in the position where Fonzies were later presented. Thus, both the place and the placement were common between CS and US. It is well known that conditioned stimuli that have a higher intensity and a stronger similarity to the US (an implicit cue is more effective than an explicit one) and are presented in the same location as the US are associated more quickly.

330

It can therefore be concluded that appetitive conditioning of DA efﬂux is strongly dependent on the speciﬁc conditions and the motivational state of the animal, as was also found to be the case with aversive conditioning. IV.C.3. Comparison with voltammetric and electrophysiological data Conditioned responses of DA and NA systems have been demonstrated in a wide variety of appetitive conditioning paradigms, using a variety of neurochemical and physiological methods. The microdialysis and voltammetric studies mentioned above show that increased transmitter release is a likely event after CS presentation. Electrophysiological data in rodents (Sara and Segal, 1991; Pan et al., 2005) and primates (Schultz, 1998) indicate that neuronal activity is rapidly and phasically increased when a stimulus that predicts reward delivery is presented and this has led to important theories regarding DA (Schultz and Dickinson, 2000; see Section VI.A) and NA (Aston-Jones and Cohen, 2005) actions. These ﬁndings suggest that the CS-induced efﬂux is caused by increased impulse ﬂow from the cell body to the nerve terminal, not (or not exclusively) by local presynaptic actions. It also suggests that microdialysis measurements (even when relatively long sample periods are used) can be used to demonstrate eventrelated, phasic increases in catecholamine release.

motivational state of the animals. It should not be surprising to anyone that the wide variety in stimuli that have been used as US or even CS gives rise to variations in the responses of the DA and NA systems. A hypothetical representation of the relation between stimulus salience and cortical catecholamine responses is presented in Fig. 4, based on microdialysis results (Feenstra et al., 2001; Mingote et al., 2004, unpublished). Increasing salience will lead to increased chances of elevated catecholamine efﬂux. The actual form of the parabolic function

IV.D. Conclusion: classical conditioning Microdialysis studies show that DA and NA efﬂux can be increased upon presentation of a stimulus (CS) that predicts an aversive or an appetitive stimulus (US). Thus, activation of both DA and NA transmission may become part of the conditioned response after associative learning about appetitive or aversive events. Four aspects of these results need to be discussed: ﬁrst, the variability in responses to both appetitive and aversive stimuli and to cues predicting these primary stimuli. It will be very difﬁcult to explain all different results without referring to the differences either in stimuli, both conditioned and unconditioned, or in experimental set-ups and in

Fig. 4. Hypothetical relation of cortical DA and NA efﬂux to stimulus saliency in Pavlovian conditioning. Data reﬂect responses to CS-alone presentations and are taken from Feenstra et al. (2001), Mingote et al. (2004) and Mingote et al. (unpublished). CS is always a 10-s presentation of white noise. Controls were presented with the noise only (white bar). Appetitively conditioned animals (ad libitum food, light grey bars and food-restricted, dark grey bars) had learned that the noise predicts the presentation of two or four pellets. Aversively conditioned animals (black bars) had learned that the noise predicts the presentation of a foot shock.

331

may be different for each DA and NA and may depend on additional factors such as CS sensory modality and intensity, novelty, etc. Second, these results show that the thesis that reward-related responses of DA systems are standard and generally accepted ﬁnding (e.g. Ungless, 2004), whereas stress-related responses are not, ignores the fact that DA responses to reward are strongly dependent on the motivational state. To observe such responses, animals are often ‘‘motivated’’ by using food- or water-restriction regimes. Without this, responses are often not observed (see Section IV.C.2) or responses are different. The same applies for NA responses (Mingote et al., 2004). Third, it is becoming more and more clear that catecholamine systems do not respond in a unitary way. This is explicitly clear for the DA responses in both appetitive and aversive conditioning paradigms. The shell of the NAC, in particular, often shows responses that differ from, for example, the core of the NAC or the PFC (Bassareo and Di Chiara, 1997, 1999a, Di Chiara et al., 1999; Bassareo et al., 2002). It is not clear if such differential effects are related to differences in neuronal activity in the DA neurons (e.g. Guarraci and Kapp, 1999) or depend on local interactions at the nerve terminals. The fourth aspect is the possibility that even when all variable factors are taken into account, a fundamental difference is present between responses to rewarding and punishing stimuli. Such a difference may be related to innate differences in cellular reactivity or to time-related differences in response patterns. Combined studies of phasic and tonic release, determining transmitter activity over different time-windows at the same time, are eagerly awaited. The question what the functional signiﬁcance of the conditioned increase might be for behaviour is not the topic of this chapter. However, it may be relevant to say a few words about this. Most recent reviews indicate that DA and NA may have two functions, allow (a) the subject to take direct and relevant action in the given circumstances and (b) learning for a better behavioural reaction in future, similar conditions (Redgrave et al., 1999; Horvitz, 2000; Joseph et al., 2003; Salamone et al., 2003; Pezze and Feldon, 2004; Aston-Jones

and Cohen, 2005; Bouret and Sara, 2005). Consequently, it has been proposed for both NA and DA neurons that they may facilitate to switch responses (Redgrave et al., 1999; Bouret and Sara, 2005). The conditions in which this is required are typically those leading to (cognitive) arousal and the importance of novel or unexpected alterations in behavioural contexts is a recurrent aspect of all major theories regarding DA and NA functions. What is unresolved is which speciﬁc conditions lead to involvement of either catecholamine. Thus, Oades (1985) originally proposed a role for DA in switching and for NA in tuning, whereas Sara (1998) later suggested a role for NA in facilitating shifts in attention and information processing and for DA in maintaining behavioural responses. Data presented by various groups now indicate that both catecholamines may facilitate behavioural switching, but not that either of them is more involved than the other.

V. Operant conditioning V.A. Background In operant behaviour, an animal is not just reacting to stimuli, but is interacting with the environment. In fact, an action of the animal is required to get the reinforcement (a reward, an escape from punishment). The action is called goal-directed if the animal knows the contingency (action A leads to outcome X) and if the outcome is indeed a wanted outcome (Balleine and Dickinson, 1998). Actions can sometimes become habits, which are performed irrespective of the outcome. The action can be manifold and varies from pressing a lever (making a nose poke or pulling a chain) in a box to running or swimming in a maze. However, the character of a response like running may be different from that of pressing a lever. Running in a maze is basically searching for food and may not (only) reﬂect instrumental but also (possibly Pavlovian) approach responses. Behaviour in an operant box can be combined well with neurochemical or electrophysiological measurements as it may be controlled and measured in a very precise way (time-stamping on a millisecond

332

basis) and the only human interference is when the animal is placed in the box and taken out again (Figs. 1–3). This will be the focus of this section as this behaviour has been used more often for such combined studies than behaviour in mazes. However, a number of studies combined neurochemical measurements with maze behaviour and these will be discussed in a separate section. The role of catecholamines in operant behaviour has been discussed and studied for 40 years now. Although the ﬁrst studies were directed at aversively (or negatively) motivated behaviour, such as conditioned avoidance of a foot shock, later research was aimed almost exclusively at appetitively (or positively) motivated behaviour, such as lever pressing for water or food. Conditioned avoidance behaviour is discussed is a separate section. Many studies employing post mortem measurements of catecholamine utilization, turnover or metabolism indicated that both NA and DA may be activated during lever pressing for water or food (Schoenfeld and Seiden, 1967, 1969; Lewy and Seiden, 1972; Albert et al., 1977; EmmettOglesby et al., 1978; Heffner and Seiden, 1980; Heffner et al., 1981). From these early studies indicating the importance of catecholamines for free operant behaviour, the discussion has been what speciﬁc aspect of operant behaviour is related to activation of NA and DA. Indeed, Lewy and Seiden (1972), when discussing the activation of NA metabolism during lever pressing for water, nicely summed up the different possibilities: ‘‘Lever-pressing and water consumption are the two most obvious behaviors in this experiment; in addition, one must consider the contingency relationship between lever-pressing and water reward. y Certain stimuli, such as the click of the lever, the periodic click of the water dipper, or the presentation of the water, might contribute to the effect’’. The way this question was approached in these early post mortem studies was by varying the work (number of lever presses) and the outcome (number of water rewards; Albert et al., 1977; Emmett-Oglesby et al., 1978) and by using yoked control animals, which received rewards without having to press and in a rate determined by another animal (Seiden et al., 1975). It was concluded that NA metabolism in brainstem-diencephalon

and DA metabolism in the nucleus caudatus were activated by the stimuli previously paired with reinforcement (Emmett-Oglesby et al., 1978). This would point at the Pavlovian component of instrumental behaviour. The time resolution in these studies was of course not enough to measure separate effects of conditioned stimuli, operant actions and reward consumption. Moreover, the parameters studied were related to the turnover and metabolism of the transmitter, but were not direct indicators of release and may not have been sensitive enough to detect changes in all areas studied (e.g. Heffner et al. (1981) observed effects of operant behaviour on DA turnover in CP, amygdala and anterior hypothalamus, but not in 12 other areas, including NAC). The neurochemical methods often also included the administration of drugs which were required to block synthesis or transport of transmitters or metabolites, but which might have interfered with the behaviour itself. The ﬁrst studies of in vivo release or efﬂux of DA, NA or their metabolites used push–pull perfusion in combination with radioactive tracers in the ventricular ﬂuids (Sparber, 1975). Perfusion of the brain parenchym to obtain the transmitters themselves has been used in combination with behaviour, but has never been widely accepted, because of the danger of damaging the tissue by the direct contact between perfusion ﬂuid and brain tissue in the push–pull process (Westerink and Justice, 1991). Reliable methods to study in vivo release were therefore eagerly awaited. However, once microdialysis techniques were widely available, considerably more attention was paid to DA than to NA. An important factor was that theories regarding brain mechanisms of reward became dominated by the DA hypothesis of reward (Wise, 1978; Wise and Rompre, 1989). Consequently, most studies were directed at DA in the striatal areas, NAC and CP.

V.B. Striatal DA In early studies push–pull perfusion (Martin and Myers, 1976), microdialysis (Hernandez and Hoebel, 1988; Salamone et al., 1989) and voltammetry (Joseph et al., 1989) were applied to

333

measure extracellular concentrations of DA itself or the DA metabolites dihydroxyphenylacetic acid (DOPAC) or homovanillic acid (HVA) in NAC or CP during operant behaviour. Metabolite concentrations were increased in animals lever pressing for food or water, but generally after the operant session. DA itself was not only increased during lever pressing, but also remained increased for some time after lever pressing had completely stopped (Hernandez and Hoebel, 1988). In a comparison of striatal areas, Joseph and co-workers (Joseph et al., 1989; Joseph and Hodges, 1990) observed similar HVA increases in NAC and CP, whereas Hernandez and Hoebel (1988) could only detect increases in DA and metabolite in NAC. Later studies, however, did report DA increases in ventrolateral CP (Cousins and Salamone, 1996), although lower than those observed in NAC (Cousins et al., 1999). A completely different and unexplained result (a decrease of striatal DA efﬂux during a discrimination lever press task) was reported by Iwano et al. (1997). The question of what the increases in DA activity were related to led to a number of control studies. The importance of positive reinforcement was shown, as non-reinforced lever presses (i.e. extinction learning) had no effect (Hernandez and Hoebel, 1988) and intermittent pellet deliveries in the Skinner box increased HVA to the same extent as reinforced lever pressing (Salamone et al., 1989). Intermittent pellet deliveries also increased DA (McCullough and Salamone, 1992). In a set-up in which lever presses were separated in time from obtaining a sucrose solution, Doyon et al. (2004) showed increased DA efﬂux in NAC during drinking only (although this was not conﬁrmed in a second paper, Doyon et al., 2005). These results suggest that reinforcement is a prerequisite for increases in DA efﬂux after lever pressing. In contrast, it was also shown that striatal HVA increases were correlated with the number of lever presses in continuous reinforcement schedules (Joseph et al., 1989; Salamone et al., 1989). A similar relationship was observed for accumbens DA (McCullough et al., 1993a). This was suggested on the basis of a relation between response rate (number of presses) and DA, but not between reinforcement magnitude (pellets obtained) and DA (Sokolowski et al.,

1998). Thus, these results led to the conclusion that (in these experiments) the response costs of the operant behaviour determine DA increases more than the beneﬁts. A similar conclusion was obtained in a study in which post mortem DA metabolism was measured in rats pressing for intracranial self-stimulation (Neill et al., 2002). On the basis of these and other studies using lesions and selective DA-antagonist drugs, Salamone et al. (2003) concluded that DA does not mediate the reinforcing effects of food rewards, but is required for activational aspects of motivation. V.C. Other areas Lever pressing for food rewards also increased extrastriatal (in this case, prefrontal) DA (Hernandez and Hoebel, 1990). In a series of studies, a group of Japanese researchers studied DA efﬂux in various brain areas during discriminative operant tasks. Hori et al. (1993) tested rats in a one-lever task where a cue light indicated trials in which a reward could be obtained. Lever presses were not reinforced when the cue light was off. Measurements in the amygdala indicated that DA efﬂux was strongly increased, whereas it was not increased in rats that were presented the same lever–reward contingency, but without the discriminative stimulus (i.e. the light was always on). Similar measurements in the PFC did not result in increased DA levels, although the metabolites DOPAC and HVA were increased at the end of the behavioural session and later (Yamamuro et al., 1994). Interestingly the same research group showed that prefrontal DA did increase when rats learned to press a lever for food rewards (instrumental learning) in a one-lever continuous reinforcement schedule (Izaki et al., 1998). In a second group of rats that were trained on this task for some days, no increase was observed, although they pressed more and obtained more rewards. Finally, Nomura et al. (2004) reported that amygdalar DA efﬂux was increased during a session in young, but not in old, rats. DA increases correlated with the number of reinforced responses in young rats, but in old rats a negative correlation with the total number of responses was present.

334

Although not all these results are easily explained, in general, DA responses did not seem to reﬂect food intake or lever pressing by themselves, but rather the attention or effort that was related to correct performance of a reinforced operant task. Simple tasks in which the animals were not required to make a choice only led to increases in DA efﬂux during the acquisition, possibly because of the increased effort (e.g. attention) required for acquisition. However, other reports did not conﬁrm the absence of a DA activation after acquisition of the instrumental contingency or the absence of a discrimination. When Beaufour et al. (2001) tested rats in a simple lever press schedule that was previously acquired, DA efﬂux was increased. Feenstra et al. (2002) applied different ﬁxed ratio (FR) schedules (Fig. 5) and reported similar DA increases in PFC in all conditions. In this study, NA efﬂux was also determined. Similar, but lower increases were observed for NA as for DA (Fig. 6).

Fig. 5. Time schedule for an operant conditioning experiment combined with microdialysis measurements. After learning that a lever press will be followed by reward delivery (shaping), the animals learn that a number of lever presses are required for reward delivery (FR: ﬁxed ratio). The microdialysis test is carried out following a period of recovery from surgery and a test for behavioural performance.

The conclusion was that neither reward density nor work load was important factor for the induction of catecholamine activation in a well-trained task. In these tasks, the catecholamine increases seem to be related more with the general level of attention required to perform the task than with speciﬁc elements of the task. A similar conclusion was reached by van der Meulen et al. (2006) in a twolever task in which rats had to make a (spatial) discrimination between the levers. Again, both DA and NA efﬂux were activated, but DA increases were higher than those for NA (Fig. 7). Interesting results were obtained when contingency alterations were introduced: when a reversal of the two-lever contingencies was introduced, DA (but not NA!) was increased more and longer than during the control task (van der Meulen et al., 2006), but when the lever presses were not

Fig. 6. Microdialysis measurements of NA efﬂux in the medial PFC during behaviour as outlined in the schedule in Fig. 5. Separate groups of food-restricted rats (90% of free-feeding weight) were tested on schedules in which 1 (FR1), 10 (FR10) or 10 (FR20) lever presses were required or the delivery of one pellet. For comparison, one group of rats was tested on a schedule in which 20 lever presses resulted in the delivery of four pellets (FR104pel) and another group of rats was tested on FR10, but was not food-restricted (FR10adlib). The operant session had 12 trials and lasted 12 min. Sample time was 16 min. Results are means7SEM.

335

Fig. 7. Microdialysis measurements of maximal DA and NA efﬂux in the medial PFC during behaviour in a two-lever spatial discrimination task. In the control task, rats pressed the rewarded lever on a FR3 schedule to obtain a reward pellet. The session consisted of 64 trials. In the reversal task, the lever–reward contingency was reversed, so that rats had to press the other lever to obtain a reward. In the extinction task, no lever was rewarded anymore. Results are means7SEM of the maximal increase during the session. During the reversal and the extinction, the value for DA, but not that of NA, was signiﬁcantly different from that during the control task (*po0.05, ANOVA with post-hoc Student-Newman-Keuls test).

reinforced anymore (extinction learning), no signiﬁcant increases of either DA or NA were observed. Extra strong activations of both DA and NA were seen when the lever did not come out and the same number of pellets were presented in the same time period and also when the lever came out but the pellet delivery was random with respect to the lever press (Feenstra et al., 2002; Cheng and Feenstra, 2006b). Rewards came essentially unexpected for the rats in these paradigms. A similar approach was taken by Dalley et al. (2001) using the ﬁve-choice serial reaction time task. Rats make nose pokes on discriminative light presentation and can obtain a food reward on correct performance. Normal performance only induced a small NA increase, but loss of contingency induced a stronger and longer lasting increase. In all these studies, the abnormal increases were normalized when the loss of contingency was experienced for a few sessions. Thus, these studies indicate that unexpected rewards strongly increase not only DA, but also NA, efﬂux. Unexpected rewards are predicted to lead to strong event-related phasic activations of

DA neuronal activity in the temporal difference learning or prediction error theory of Schultz (1998) (see also Section VI.A). Thus, these results apparently provide a further support for the idea that microdialysis measurements can reﬂect phasic alterations in catecholamine release. Whether increases compared with basal levels are induced by normal performance of a trained task is not yet decided, although a majority of the evidence is in favour of an activation: all studies of DA efﬂux in the NAC and a number of studies of DA in the PFC suggest this to be the case. But it could be that the various operant schedules that were used make a difference. It might be of importance whether a continuous reinforcement schedule is maintained (where the lever is always available for pressing and rewards are delivered immediately) or a discrete trial set-up (where the animal has to wait until the next trial begins). A comparison to electrophysiological studies can be drawn: a monkey pressing for a reward on a FR30 schedule showed increased DA activity after presentation of the stimulus and during the ﬁrst 10 lever presses, whereas activity was normal during the ﬁnal lever presses and the

336

reward presentation (Nishino et al., 1987). It may be that at a certain stage, responding becomes automated and does not involve DAergic activity anymore. Separating the trials and introducing intertrial waiting periods may produce an activational effect for each trial, a process for which DA is required (Salamone et al., 2003), and may also lead to a longer period of expectancy for the next reward or uncertainty about the delivery for that reward. This may be related to DA increases, as well (Fiorillo et al., 2003). A common outcome seems to be that NA is less activated than DA. NA activity has been suggested to be strongly related to the attention required in a speciﬁc condition (Aston-Jones et al., 1999). A well-learned task or a relatively simple two-lever discrimination or reversal apparently does not require a high attentional level, as the basic task (lever press means direct reward) remains the same. DA may be more selectively related to speciﬁc action–outcome contingencies.

V.D. Comparison with voltammetric and electrophysiological data In vivo electrochemical studies would, in principle, be well suited to obtain more detailed information. A problem, however, is the identiﬁcation of the substances that are detected at the electrodes. Amperometry only measures the current without giving any additional information. Voltammetry provides more options but generally lacks sufﬁcient sensitivity (O’Neill, 1994; Wightman and Robinson, 2002). After the studies of Joseph and co-workers (see above), new developments were claimed to result in amperometric measurements of DA in NAC (Doherty and Gratton, 1992). Owing to the fact that convincing evidence of DA (or a DA metabolite) being the detected substance was not provided, the results of these studies are controversial (Wightman and Robinson, 2002). However, a short summary of the results may be discussed here. Kiyatkin and Gratton (1994) report measurements during a lever-pressing task that is not described in detail. The most signiﬁcant ﬁnding was that the current increased from detecting the stimulus light to peak at the moment of

the lever press and decreased during feeding. Tonic measurements resulted in a steady increase, which ﬁrst stabilized and later decreased when animals became sated and stopped eating. Richardson and Gratton (1996, 1998) conﬁrmed this (using measurements in NAC and PFC) in animals pressing for milk reward. The variations in current decreases that they reported were later shown not to reﬂect DA signals, but possibly pH changes (Roitman et al., 2004). Further developments in electrochemical measurements led to a more sophisticated and validated fast-scan cyclic voltammetric method, which could be used to detect rapid increases in extracellular DA (‘‘DA transients’’) (Wightman and Robinson, 2002; Phillips et al., 2003b; Roitman et al., 2004). Applied in rats lever pressing for a sugar solution, it was shown that DA increased after stimulus presentation and reached a peak at the lever press. Two separate peaks were observed when the animal decided to press later. Thus, phasic DA release is increased (1) after presentation of the stimulus predicting that reward may be available and (2) in the time immediately preceding the action (lever press) that results in reward presentation. Both the early ﬁndings of Seiden and co-workers and the theories of Salamone and co-workers are therefore conﬁrmed. These results compare also well with the electrophysiological responses obtained in monkeys (Schultz, 1998). The preoccupation of most researchers with DA effects led to a relative sparseness of information regarding NA activity during similar tasks. Electrophysiological studies in the monkey and, recently, the rat have provided evidence, however, that NA is involved in operant behaviour in a way that does not differ very much from DA. AstonJones et al. (1997) studied primate LC responses in an operant vigilance task. LC neurons responded with a phasic activation to a target stimulus and this activation switched to the new target on reversal of stimulus–reward contingencies. The LC response preceded the behavioural response with hundreds of trials. Interestingly, tonic activity was also increased for at least 10 min after reversal. Bouret and Sara (2004) studied NA neurons during a Go–No Go task in which an odour CS signalled the type of response required. A nose

337

poke led to the delivery of a reward. NA activity was increased on predictive stimuli (trial onset; CS Go presentation). NA activity was strongly related to the onset of the Go nose poke response. A similar outcome was reported by Clayton et al. (2004) in monkeys: NA neurons were activated preceding behavioural responses and would have a facilitating inﬂuence on behaviour. When new odours were presented by Bouret and Sara (2004), responses were strong to reward presentation, but rapidly shifted to the predictive stimuli (similar to what has been described for DA neurons; Schultz, 1998; Pan et al., 2005). In extinction NA, responses rapidly disappeared. In a reversal of odour–reward contingencies, NA responses extinguished to the old CS+ and shifted to the new CS+ (cf. Aston-Jones et al., 1997; Bouret and Sara, 2004). These electrophysiological results strongly suggest similar responses of NA and DA to the CS and before the action in reward-related behaviour. Further studies will need to uncover activational and functional differences between DA and NA activity.

V.E. Maze studies Despite the obvious problems associated with microdialysis studies during performance of maze tasks, a number of interesting studies were performed. As in all behavioural studies, special attention should be paid to control experiments to ensure that the condition that is under study is not confounded by other conditions. Men et al. (1999) studied spontaneous alternation performance during a 12-min session in a four-armed cross-maze. In principle, this is a test for working memory: the animal should not re-enter the arms that were already visited. NA efﬂux in the hippocampus was increased compared with a control group kept in a holding cage. It is not clear what the contribution is of the working memory to the NA increase as no controls were carried out in the maze. Using an eight-armed radial maze, Phillips et al. (2004) tested rats in a rewarded, delayed spatial win-shift task. The rat uses spatial information to ﬁnd arms in the maze that have a reward. During training, four arms were closed and four other arms were

baited. After a delay of 30 min, 1 h or 6 h, rats were placed in the maze again in which all arms were open: they should visit only the arms not visited in the training. Prefrontal DA efﬂux always increased not only during the training, but also during the test after 30 min or 1 h, but not 6 h. The DA increase during the test was inversely correlated to the number of errors. Omitting the availability of reward in the test phase (i.e. presenting extinction trials) at 30 min did not affect the DA increase. The authors suggest a relation between DA efﬂux and the retrieval of trial unique memories. Rossetti and Carboni (2005) studied spontaneous alternation in a T-maze: rats were trained to alternate between the two arms to receive a reward. These rats showed a strong increase in NA efﬂux during the test. In contrast, rats that could ﬁnd rewards in both arms and did not need to alternate had a much lower NA increase. DA was increased in a similar way in both groups and was, unlike NA, also increased while the rats were in the waiting cage, that is, during anticipation of the test. The authors conclude that NA is related to the level of selective attention required to achieve the goal, whereas DA is predominantly related to reward expectation. Using an odour box, where a rat learns in a few trials which odour predicts reward presentation, Tronel et al. (2004) determined NA in the period after acquisition and observed an increase around 2 h after learning. Pseudotrained rats that were exposed to the same, but now unrelated stimuli, did not show a similar increase. Previous pharmacological studies had indicated this time to be critical for a NA-dependent consolidation of memory. Thus, by focusing on particular aspects of the behavioural task at hand, these groups succeeded in demonstrating selective involvement of DA or NA in the underlying processes.

V.F. Conditioned avoidance behaviour Animals can learn to actively avoid a punishment by performing an action on the guidance of a CS. The action can be pressing a lever or moving to another side of a box. Catecholamine involvement has been considered for a long time and Fuxe and

338

Hanson (1967) already showed that (probably) NA neurons were activated during conditioned avoidance behaviour. The ﬁrst microdialysis studies were carried out by McCullough et al. (1993b), who reported a strong increase in DA efﬂux in the NAC during and after the session. The increases showed a correlation with the number of lever presses the animals made to avoid the shocks. Presenting periodic foot shocks did not result in a signiﬁcant increase in DA efﬂux, although the authors considered the sampling time (45 min) a restricting factor and mentioned that most animals had increased DA efﬂux. A series of studies were carried out by Stark and co-workers. Auditory stimuli signalled the possibility to escape from the side of a shuttle box where a shock would be presented to the other, safe side. The ﬁrst study (Stark and Scheich, 1997) was aimed at the auditory cortex where HVA was determined. A strong increase in HVA was observed during the ﬁrst session in the trained animals (where learning progress was the strongest), as compared with the pseudoconditioned and the auditory control animals. A later study reported on DA efﬂux in the medial PFC (Stark et al., 1999). The remarkable ﬁnding here was that animals that showed good learning in the session had a stronger increase in DA than the animals that had a good performance from the start or the animals that were bad learners. The increase in DA efﬂux was shown to be related to the acquisition of a behavioural strategy and in particular to a stage in which signal evaluation took place (Stark et al., 2000). Further evidence was presented in an experiment in which two cues that initially signalled a Go response now signalled either a Go or a No Go response (Stark et al., 2004). Animals that learned this discrimination rapidly had a strong increase in DA efﬂux. The authors conclude that DA is involved in the formation of new behavioural strategies. Also in these studies, animals that received many foot shocks (the bad learners) did not show strongly increased DA efﬂux. McIntyre et al. (2002) determined NA efﬂux in the amygdala during ﬁrst exposure to an avoidance procedure. Rats entering a dark arm of a runway received one foot shock and were returned to the holding cage. This increased NA during and

after the shock period in comparison with rats that did not receive a shock in the runway or rats that received a shock in the holding cage. The height of the NA increase showed a correlation with the latency to enter the same dark compartment in a retention test. These various studies show clearly that DA and NA responses to foot shocks are not absolute, but depend on the history of the animal and the context of shock presentation. All studies indicate a relation between catecholamine activation and the avoidance behaviour, more than the direct aversive experience of the shock. This relates to earlier suggestions that the (cortical DA) increase is ‘‘not connected to the emotional reaction caused by the aversive nature of the stressor but may rather reﬂect a heightened attention of the animal or activation of cognitive processes in an attempt to cope with the stressor’’ (D’Angio et al., 1988).

V.G. Conclusion: operant conditioning Given the importance of Pavlovian mechanisms for operant behaviour, it comes as no surprise that the relations between DA and NA responses and stimulus presentation observed during appetitive and aversive classical conditioning are reﬂected in the responses during operant behaviour. Although the various operant behavioural paradigms are very different, common ﬁndings include increased efﬂux or activity after presentation of a predictive cue and just before the action that leads to the outcome. In addition, there is a more general increase that is related to the amount of effort required for task performance. This may be the level of attention, the demand for working memory or the activational investment that is required in the given conditions. As discussed above (Section IV.D), many factors may lead to variable results in DA or NA efﬂux during Pavlovian behaviour. An extra factor that is speciﬁc for operant behaviour is the habit formation that occurs when performance reached a stable level and may even be more or less automated. Habitual performance is not determined anymore by the evaluation of outcomes (Balleine and Dickinson, 1998). A number of examples were

339

presented in which behaviour may have been habitual and may not have involved the same level of attention and reward expectancy as during acquisition or more effortful processing. Simple repeated actions like lever pressing may be more susceptible to habit formation than, for example, food-searching behaviour in a maze. It has indeed been demonstrated that conditioned responses become independent of DA, D1-receptor-mediated neurotransmission on extended habit training (Choi et al., 2005).

VI. Dopamine, noradrenaline and learning VI.A. Rodent studies Most of the above-mentioned studies were directed at measuring neurochemical correlates of performance of a previously learned operant behaviour. However, on the basis of data indicating that DA might be involved in the acquisition or adaptation of speciﬁc behaviours, it has been suggested that DA might have a role in this learning process (Beninger, 1983). These theories were extended when measurements of DA cellular activity in monkeys resulted in the proposal of a formal role for DA in learning and experiments showed DA activity measurements to support the theory (Schultz and Dickinson, 2000; Schultz, 2002). This theory proposes that DA acts as a reward prediction error, being activated when rewards are higher than expected and inhibited when rewards are lower than expected. Unexpected rewards strongly activate DA activity, but on prediction of reward presentation by stimuli, DA is activated by this stimulus. In rats, this is a gradual process in which responses to reward slowly decrease and responses to the stimulus increase at the same time (Pan et al., 2005). The highest total activity was observed at the beginning of learning (unexpected reward), the lowest when learning was over (fully expected reward). A major question was, of course, whether studies of DA release would conﬁrm this theory. Some microdialysis studies are available that are relevant to this question, although they cannot fully answer it, given the restricted time resolution. Stark et al. (1999, 2004)

also reported higher DA increases in gerbils that learned a conditioned avoidance response faster than animals that were slow learners. Izaki et al. (1998) studied PFC DA of rats in their ﬁrst and in a later session of lever pressing. Only in the ﬁrst session, DA efﬂux was increased. Although the stronger increase at the start of the learning is according to the reward prediction error theory, the complete absence of activity when reward was predicted was unexpected. It is also in contrast with previous data (Hernandez and Hoebel, 1990) and with our experience that prefrontal DA is activated each time a rat is performing an operant, rewarded task (Feenstra et al., 2002, in preparation). Cheng and Feenstra (2006a) performed a similar study on instrumental learning, measuring DA in the NAC subareas shell and core (Fig. 8). Part of the rats had learned the lever-pressing procedure after one session, whereas others did not improve their performance. In all sessions, DA was increased, but only in the ﬁrst session was there a difference between the learning group and the non-learning group, the learning group having a higher DA efﬂux (Fig. 9). In the second session,

Fig. 8. Schedule for a trial in the instrumental learning paradigm used by Cheng and Feenstra (2006a). Two trials were programmed in the period in which one microdialysis sample was taken (3.75 min). The session had 30 trials.

340

Fig. 9. Mean changes in extracellular dopamine (DA) levels in the nucleus accumbens shell in the learning (squares) and non-learning (down triangles) groups during two sessions of instrumental learning (upper vertical axis). All concentrations are expressed as percentage of basal7SEM. Filled symbols indicate signiﬁcant differences compared with all four baseline samples (po0.05). Mean numbers of lever presses ( ¼ food pellets obtained) in two trials (one dialysis sample time) are given as bar graphs (middle vertical axis: grey bars ¼ learning group; black bars ¼ non-learning group). Mean numbers of nose pokes in two trials are given in the lower graphs. Filled symbols in the nose poke graph indicate signiﬁcant differences compared with the ﬁrst sample (po0.01). * Signiﬁcant differences between learning (n ¼ 8) and non-learning (n ¼ 6) groups (po0.05). Reproduced from Cheng and Feenstra (2006a) r Cold Spring Harbor Laboratory Press.

no differences were detected. Although in the learning group only in the shell DA efﬂux was higher in the ﬁrst than in the second session, these data seem to provide a better ﬁt to the learning theory. van der Meulen et al. (2006) determined DA and NA efﬂux in PFC dialysates during adaptation of a discriminated lever-pressing response, that is, reversal learning: the rats had to switch responses from one (previously rewarded) lever to the other (previously unrewarded) lever. DA efﬂux was always activated when rewards were obtained, but the activation was higher during the ﬁrst reversal (Fig. 7). During later reversals, activation was ‘‘normal’’ again. When reward was withheld (extinction), no DA activation was detectable. NA behaved quite differently: it was increased in every session in which rats were lever pressing, but no differences in activation were observed during reversal or extinction learning (Fig. 7). These data suggest no selective activation of NA during adaptation of instrumental behaviour (see, however, the results of Bouret and

Sara, 2004, in an earlier section). Still, NA is reproducibly activated during this behaviour.

VI.B. Primate studies VI.B.1. Reward processing DA release in the human brain may be visualized as displacement of radioactive ligand binding in PET studies (no equivalent method for detecting NA release is available). Koepp et al. (1998) showed that binding to D2-receptors was decreased during performance of a video game. Reaching a higher level of performance (rewarded with a sum of money) was related to a stronger displacement of binding in NAC and, to a lesser extent, in CP. These studies lacked a proper control group and Zald et al. (2004) performed experiments that were modelled on the rat operant schedules: one group had to turn cards and was rewarded on a variable ratio 4 (VR4, reward presentation on average every four lever presses)

341

schedule, another group on a ﬁxed ratio 4 (FR4, reward presentation exactly after four lever presses) schedule and a third control group performed all the acts but was not rewarded. The VR schedule produced increases (medial CP), but also decreases (other CP subareas), compared with the control task, but no effects in NAC. The FR schedule produced modest increases, no decreases. The important difference between the VR and the FR schedules is the unpredictability of reward presentation in the VR condition. DA efﬂux was also studied in a discrimination task which was manipulated so that blocks with mostly (unpredictable) rewarded trials were alternated with blocks with mostly (unpredictable) punished trials (Pappata et al., 2002). Only the rewarded blocks led to increased DA. Unpredictability and expectancy were involved in another study, as well (de la Fuente-Ferna`ndez et al., 2002). Parkinsonian patients expecting a therapeutic treatment responded to a placebo injection with a DA increase. These studies have in common that the context leads one to expect rewards, but the presentation of a reward in a single trial is unpredictable. Human DA efﬂux was not solely related to reward processing: Pruessner et al. (2004) studied effects of stress and observed stress-induced displacement of D2-ligand binding in susceptible persons. Thus, these PET studies show an effect on a measure indicating DA efﬂux that is comparable to those obtained with microdialysis studies in rats. Results from both PET studies and microdialysis studies suggest higher DA efﬂux after delivery of unpredicted rewards, similar to what was described for phasic DA activity (Schultz, 1998).

VI.B.2. Learning and effortful processing Microdialysis studies in the human brain are scarce, but Fried et al. (2001) determined DA efﬂux in the amygdala of patients that had intractable epilepsy and required electrode implantations. In a paired-associates task, clear increases were shown by slow learners, whereas fast learners only had a short DA elevation. In a working memory task, an increase was only seen when the task was performed for the ﬁrst time, after previous experience with a control reading task. Thus, in both

tasks, the increases may reﬂect increased attention or motivation during learning (Fried et al., 2001). The results of the working memory task may be compared with those obtained in a monkey microdialysis study (Watanabe et al., 1997). Prefrontal DA was increased during the working memory task, when compared with a control task. Next to working memory, attentional demand and task difﬁculty were considered to underlie this activation. Displacement of a D2-receptor ligand from non-striatal binding sites during working memory and sustained attention was reported by Aalto et al. (2005). Various prefrontal and medial temporal cortical areas were involved and increased DA efﬂux was related to better performance. The authors suggest an attention- or arousal-related increase in DA activity. It is interesting that these two tasks, which have been associated with DAergic and NAergic functions, respectively (AstonJones et al., 1999; Goldman-Rakic et al., 2000), are now both shown to involve DAergic activity. Further evidence for DA involvement in effortful cognitive processing was presented by Tettamanti et al. (2005), who reported D2-ligand displacement in the basal ganglia during phonological processing. These studies suggest an important role for cognitive effort in inducing DA release. Could this be the counterpart of DA’s role in the activational aspects of motivation (cf. Salamone et al., 2003)?

VII. Conclusions DA is still regarded as an arousal transmitter and recent examples include DA’s importance for a general behavioural characterization in lower animals (Andretic et al., 2005) and for the involvement in attention-requiring, effortful cognitive processing in primates, including man (see previous section). The typical inverted U-shape relationship of arousal and cognitive functions (Yerkes and Dodson, 1908) is still of current interest and has been described for prefrontal DA (e.g. Goldman-Rakic et al., 2000), for NA (e.g. Aston-Jones et al., 1999) and for acetylcholine (e.g. Newhouse et al., 2004). Such descriptions of DA function are similar to conclusions reached 10–20 years ago on the basis

342

of microdialysis and electrophysiological studies (e.g. D’Angio et al., 1988: ‘‘We propose that the enhanced cortical DA metabolism associated with stress may rather reﬂect activation of attentional and/or cognitive processes in an attempt to deal with the stressor’’; Schultz et al., 1993: ‘‘dopamine neurons respond phasically to alerting external stimuli with behavioral signiﬁcance whose detection is crucial for learning and performing delayed response tasks.’’). At the same time, more detailed electrophysiological studies indicate a crucial and very speciﬁc role for DA in learning theories (Schultz, 2002). No studies of DA release are yet available to support these theories, but the reported alterations of DA efﬂux, measured using microdialysis during behavioural tasks, are compatible with the predictions based on this theory. Microdialysis studies can indeed contribute to the evaluation of such theories, as they provide unique measurements of DA efﬂux (reﬂecting DA release) in all possible brain areas and many mammalian species, during tasks that are not readily used or applicable in primates. Moreover, it will be of considerable importance to describe processes related to transmitter release in the different activity modes for monoamines, reﬂecting tonic and phasic responses. A basic requirement for such a role of DA and possibly other monoamines in more speciﬁc behavioural theories is a reactivity to motivational salient stimuli and the acquisition of incentive motivation by stimuli. This is strongly dependent on speciﬁc external conditions and individual differences in motivational state, which together determine the salience of the stimuli that may activate DA. Variations in these will give rise to varying results in apparently similar test conditions. A better control of experimental conditions will lead to more comparable study outcomes. Before DA’s rise to fame, NA was considered to be an important signal for reward and learning (e.g. Anlezark et al., 1973). Since then, NA has primarily been associated with arousal and attention. It is signiﬁcant that some recent data support a role in goal-directed behaviour and decision making (see van der Meulen et al., 2002; Bouret and Sara, 2004; Clayton et al., 2004), functions that were predominantly associated with DA activity.

Further comparative studies using microdialysis, similar to those by Dalley et al. (2001), Mingote et al. (2004) and Rossetti and Carboni (2005), will teach us more about the different involvement of the catecholamine arousal transmitters in cognitive processes.

Acknowledgment I am indebted to Geoffrey van der Plasse for making the photographs and for helpful comments.

References Aalto, S., Bruck, A., Laine, M., Nagren, K. and Rinne, J.O. (2005) Frontal and temporal dopamine release during working memory and attention tasks in healthy humans: a positron emission tomography study using the high-afﬁnity dopamine D2 receptor ligand [11C]FLB 457. J. Neurosci., 25: 2471–2477. Acquas, E., Wilson, C. and Fibiger, H.C. (1996) Conditioned and unconditioned stimuli increase frontal cortical and hippocampal acetylcholine release: effects of novelty, habituation, and fear. J. Neurosci., 16: 3089–3096. Adell, A., Casanovas, J.M. and Artigas, F. (1997) Comparative study in the rat of the actions of different types of stress on the release of 5-HT in raphe nuclei and forebrain areas. Neuropharmacology, 36: 735–741. Ahn, S. and Phillips, A.G. (1999) Dopaminergic correlates of sensory-speciﬁc satiety in the medial prefrontal cortex and nucleus accumbens of the rat. J. Neurosci., 19: RC29. Albert, L.H., Emmett-Oglesby, M. and Seiden, L.S. (1977) Effects of schedules of reinforcement on brain catecholamine metabolism in the rat. Pharmacol. Biochem. Behav., 6: 481–486. Aloisi, A.M., Casamenti, F., Scali, C., Pepeu, G. and Carli, G. (1997) Effects of novelty, pain and stress on hippocampal extracellular acetylcholine levels in male rats. Brain Res., 748: 219–226. Andretic, R., van Swinderen, B. and Greenspan, R.J. (2005) Dopaminergic modulation of arousal in Drosophila. Curr. Biol., 15: 1165–1175. Anlezark, G.M., Crow, T.J. and Greenway, A.P. (1973) Impaired learning and decreased cortical norepinephrine after bilateral locus coeruleus lesions. Science, 181: 682–684. Aston-Jones, G. and Cohen, J.D. (2005) An integrative theory of locus coeruleus-norepinephrine function: adaptive gain and optimal performance. Annu. Rev. Neurosci., 28: 403–450. Aston-Jones, G., Rajkowski, J. and Cohen, J. (1999) Role of locus coeruleus in attention and behavioral ﬂexibility. Biol. Psychiatry, 46: 1309–1320. Aston-Jones, G., Rajkowski, J. and Kubiak, P. (1997) Conditioned responses of monkey locus coeruleus neurons

343 anticipate acquisition of discriminate behavior in a vigilance task. Neuroscience, 80: 697–715. Balleine, B.W. and Dickinson, A. (1998) Goal-directed instrumental action: contingency and incentive learning and their cortical substrates. Neuropharmacology, 37: 407–419. Bassareo, V. and Di Chiara, G. (1997) Differential inﬂuence of associative and non-associative learning mechanisms on the responsiveness of prefrontal and accumbal dopamine transmission to food stimuli in rats fed ad libitum. J. Neurosci., 17: 851–861. Bassareo, V. and Di Chiara, G. (1999a) Differential responsiveness of dopamine transmission to food-stimuli in nucleus accumbens shell/core compartments. Neuroscience, 89: 637–641. Bassareo, V. and Di Chiara, G. (1999b) Modulation of feedinginduced activation of mesolimbic dopamine transmission by appetitive stimuli and its relation to motivational state. Eur. J. Neurosci., 11: 4389–4397. Bassareo, V., De Luca, M.A. and Di Chiara, G. (2002) Differential expression of motivational stimulus properties by dopamine in nucleus accumbens shell versus core and prefrontal cortex. J. Neurosci., 22: 4709–4719. Beaufour, C.C., Le Bihan, C., Hamon, M. and Thie´bot, M.-H. (2001) Extracellular dopamine in the rat prefrontal cortex during reward-, punishment- and novelty-associated behaviour. Effects of diazepam. Pharmacol. Biochem. Behav., 69: 133–142. Beninger, R.J. (1983) The role of dopamine in locomotor activity and learning. Brain Res. Rev., 6: 173–196. Besson, C. and Louilot, A. (1995) Asymmetrical involvement of mesolimbic dopaminergic neurons in affective perception. Neuroscience, 68: 963–968. Bickerdike, M.J., Wright, I.K. and Marsden, C.A. (1993) Social isolation attenuates rat forebrain 5-HT release induced by KCI stimulation and exposure to a novel environment. Behav. Pharmacol., 4: 231–236. Blackburn, J.R., Pfaus, J.G. and Phillips, A.G. (1992) Dopamine functions in appetitive and defensive behaviours. Prog. Neurobiol., 39: 247–279. Bouret, S. and Sara, S.J. (2004) Reward expectation, orientation of attention and locus coeruleus-medial frontal cortex interplay during learning. Eur. J. Neurosci., 20: 791–802. Bouret, S. and Sara, S.J. (2005) Network reset: a simpliﬁed overarching theory of locus coeruleus noradrenaline function. Trends Neurosci., 28: 574–582. Brennan, P.A., Schellinck, H.M., De La Riva, C., Kendrick, K.M. and Keverne, E.B. (1998) Changes in neurotransmitter release in the main olfactory bulb following an olfactory conditioning procedure in mice. Neuroscience, 87: 583–590. Bruno, J.P., Sarter, M., Moore Arnold, H. and Himmelhebr, A.M. (1999) In vivo neurochemical correlates of cognitive processes: methodological and conceptual challenges. Rev. Neurosci., 10: 25–48. Bunney, B.S., Chiodo, L.A. and Grace, A.A. (1991) Midbrain dopamine system electrophysiological functioning: a review and new hypothesis. Synapse, 9: 79–94.

Cardinal, R.N., Parkinson, J.A., Hall, J. and Everitt, B.J. (2002) Emotion and motivation: the role of the amygdala, ventral striatum, and prefrontal cortex. Neurosci. Biobehav. Rev., 26: 321–352. Cassens, G., Roffman, M., Kuruc, A., Orsaluk, P.J. and Schildkraut, J.J. (1980) Alterations in brain norepinephrine metabolism induced by environmental stimuli previously paired with inescapable shock. Science, 209: 1138–1140. Cenci, M.A. and Kale´n, P. (2000) Serotonin release from mesencephalic raphe neurons grafted to the 5,7-dihydroxytryptamine-lesioned rat hippocampus: effects of behavioral activation and stress. Exp. Neurol., 164: 351–361. Cenci, M.A., Kale´n, P., Mandel, R.J. and Bjo¨rklund, A. (1992) Regional differences in the regulation of dopamine and noradrenaline release in medial frontal cortex, nucleus accumbens and caudate-putamen: a microdialysis study in the rat. Brain Res., 581: 217–228. Cheng, J.J., de Bruin, J.P.C. and Feenstra, M.G.P. (2003) Dopamine efﬂux in nucleus accumbens shell and core in response to appetitive classical conditioning. Eur. J. Neurosci., 18: 1306–1314. Cheng, J. and Feenstra, M.G.P. (2006a) Individual differences in dopamine efﬂux in nucleus accumbens shell and core during instrumental learning. Learn. Mem., 13: 168–177. Cheng, J. and Feenstra, M.G.P. (2006b) Stimulus contingency in operant task performance: effects on prefrontal DA and NA efﬂux. In: Di Chiara, G., Carboni, E., Valentini, V., Acquas, E., Bassareo, V. and Cadoni, C. (Eds.), Monitoring Molecules in Neuroscience. Univ. of Cagliari, Cagliari, pp. 525–526. Chergui, K., Suaud-Chagny, M.F. and Gonon, F. (1994) Nonlinear relationship between impulse ﬂow, dopamine release and dopamine elimination in the rat brain in vivo. Neuroscience, 62: 641–645. Choi, W.Y., Balsam, P.D. and Horvitz, J.C. (2005) Extended habit training reduces dopamine mediation of appetitive response expression. J. Neurosci., 25: 6729–6733. Clayton, E.C., Rajkowski, J., Cohen, J.D. and Aston-Jones, G. (2004) Phasic activation of monkey locus ceruleus neurons by simple decisions in a forced-choice task. J. Neurosci., 24: 9914–9920. Cousins, M.S. and Salamone, J.D. (1996) Involvement of ventrolateral striatal dopamine in movement initiation andexecution: a microdialysis and behavioral investigation. Neuroscience, 70: 849–859. Cousins, M.S., Trevitt, J., Atherton, A. and Salamone, J.D. (1999) Different behavioral functions of dopamine in the nucleus accumbens and ventrolateral striatum: a microdialysis and behavioral investigation. Neuroscience, 91: 925–934. D’Angio, M. and Scatton, B. (1989) Feeding or exposure to food odors increases extracellular DOPAC levels (as measured by in vivo voltammetry) in the prefrontal cortex of fooddeprived rats. Neurosci. Lett., 96: 223–228. D’Angio, M., Serrano, A., Driscoll, P. and Scatton, B. (1988) Stressful environmental stimuli increase extracellular DOPAC levels in the prefrontal cortex of hypoemotional (Roman high-avoidance) but not hyperemotional (Roman

344 low-avoidance) rats. An in vivo voltammetric study. Brain Res., 451: 237–247. Dalley, J.W. and Stanford, S.C. (1995) Incremental changes in extracellular noradrenaline availability in the frontal cortex induced by naturalistic environmental stimuli: a microdialysis study in the freely moving rat. J. Neurochem., 65: 2644–2651. Dalley, J.W., McGaughy, J., O’Connell, M.T., Cardinal, R.N., Levita, L. and Robbins, T.W. (2001) Distinct changes in cortical acetylcholine and noradrenaline efﬂux during contingent and noncontingent performance of a visual attentional task. J. Neurosci., 21: 4908–4914. Damsma, G., Pfaus, J.G., Wenkstern, D., Phillips, A.G. and Fibiger, H.C. (1992) Sexual behavior increases dopamine transmission in the nucleus accumbens and striatum of male rats: comparison with novelty and locomotion. Behav. Neurosci., 106: 181–191. Datla, K.P., Ahier, R.G., Young, A.M.J., Gray, J.A. and Joseph, M.H. (2002) Conditioned appetitive stimulus increases extracellular dopamine in the nucleus accumbens of the rat. Eur. J. Neurosci., 16: 1987–1993. de la Fuente-Ferna`ndez, R., Phillips, A.G., Zamburlini, M., Sossi, V., Calne, D.B., Ruth, T.J. and Stoessl, A.J. (2002) Dopamine release in human ventral striatum and expectation of reward. Behav. Brain Res., 136: 359–363. Deutch, A.Y., Tam, S.-Y. and Roth, R.H. (1985) Footshock and conditioned stress increase 3,4-dihydroxy-phenylacetic acid (DOPAC) in the ventral tegmental area but not substantia nigra. Brain Res., 333: 143–146. Devoto, P., Flore, G., Vacca, G., Pira, L., Arca, A., Casu, M.A., Pani, L. and Gessa, G.L. (2003) Co-release of noradrenaline and dopamine from noradrenergic neurons in the cerebral cortex induced by clozapine, the prototype atypical antipsychotic. Psychopharmacology, 167: 79–84. Di Chiara, G. (1991) Brain dialysis of monoamines. In: Robinson, T.E. and Justice, Jr., J.B. (Eds.), Microdialysis in the Neurosciences. Elsevier Science, Amsterdam, pp. 175–187. Di Chiara, G., Loddo, P. and Tanda, G. (1999) Reciprocal changes in prefrontal and limbic dopamine responsiveness to aversive and rewarding stimuli after chronic mild stress: implications for the psychobiology of depression. Biol. Psychiatry, 46: 1624–1633. Doherty, M.D. and Gratton, A. (1992) High-speed chronoamperometric measurements of mesolimbic and nigrostriatal dopamine release associated with repeated daily stress. Brain Res., 586: 295–302. Doyon, W.M., Anders, S.K., Ramachandra, V.S., Czachowski, C.L. and Gonzales, R.A. (2005) Effect of operant selfadministration of 10% ethanol plus 10% sucrose on dopamine and ethanol concentrations in the nucleus accumbens. J. Neurochem., 93: 1469–1481. Doyon, W.M., Ramachandra, V., Samson, H.H., Czachowski, C.L. and Gonzales, R.A. (2004) Accumbal dopamine concentration during operant self-administration of a sucrose or a novel sucrose with ethanol solution. Alcohol, 34: 261–271. Drijfhout, W.J., Kemper, R.H.A., Meerlo, P., Koolhaas, J.M., Grol, C.J. and Westerink, B.H.C. (1995) A telemetry study on the chronic effects of microdialysis probe implantation on

the activity pattern and temperature rhythm of the rat. J. Neurosci. Methods, 61: 191–196. Emmett-Oglesby, M.W., Lewy, A.J., Albert, L.H. and Seiden, L.S. (1978) Role of lever responding and water presentation in altering rat brain catecholamine metabolism. J. Pharmacol. Exp. Ther., 204: 406–415. Enrico, P., Bouma, M., de Vries, J.B. and Westerink, B.H.C. (1998) The role of afferents to the ventral tegmental area in the handling stress-induced increase in the release of dopamine in the medial prefrontal cortex: a dual-probe microdialysis study in the rat brain. Brain Res., 779: 205–213. Feenstra, M.G.P. (2000) Dopamine and noradrenaline release in the prefrontal cortex in relation to conditioned and unconditioned stress and reward. In: Uylings, H.B.M., van Eden, C.G., de Bruin, J.P.C., Feenstra, M.G.P. and Pennartz, C.M.A. (Eds.), Cognition, Emotion and Autonomic Responses: The Integrative Role of the Prefrontal Cortex and Limbic Structures. Elsevier, Amsterdam. Prog. Brain Res., 126: 133–163. Feenstra, M.G.P. and Botterblom, M.H.A. (1996) Rapid sampling of extracellular dopamine in the rat prefrontal cortex during food consumption, handling and exposure to novelty. Brain Res., 742: 17–24. Feenstra, M.G.P., Botterblom, M.H.A. and Mastenbroek, S. (2000) Dopamine and noradrenaline efﬂux in the prefrontal cortex in the light and dark period: effects of novelty and handling and comparison to the nucleus accumbens. Neuroscience, 100: 741–748. Feenstra, M.G.P., Botterblom, M.H.A. and van Uum, J.F.M. (1995) Novelty-induced increase in dopamine release in the rat prefrontal cortex in vivo: inhibition by diazepam. Neurosci. Lett., 189: 81–84. Feenstra, M.G.P., Botterblom, M.H.A. and van Uum, J.F.M. (1998) Local activation of metabotropic glutamate receptors inhibits the handling-induced increased release of dopamine in the nucleus accumbens but not that of dopamine or noradrenaline in the prefrontal cortex: comparison with inhibition of ionotropic receptors. J. Neurochem., 70: 1104–1113. Feenstra, M., IJff, B., Lesman, A., de Weijer, B., Joosten, R. and de Bruin, J. (2002) Effects of lever pressing and expected rewards on dopamine (DA) and noradrenaline (NA) efﬂux measured by microdialysis in the prefrontal cortex during operant responding. FENS Abstr., 1: A210.5. Feenstra, M.G.P., Teske, G., Botterblom, M.H.A. and de Bruin, J.P.C. (1999) Dopamine and noradrenaline release in the prefrontal cortex of rats during classical aversive and appetitive conditioning to a contextual stimulus: interference by novelty effects. Neurosci. Lett., 272: 179–182. Feenstra, M.G.P., Vogel, M., Botterblom, M.H.A., Joosten, R.N.J.M.A. and de Bruin, J.P.C. (2001) Dopamine and noradrenaline efﬂux in the rat prefrontal cortex after classical aversive conditioning to an auditory cue. Eur. J. Neurosci., 13: 1051–1054. Fillenz, M. (2005) In vivo neurochemical monitoring and the study of behaviour. Neurosci. Biobehav. Rev., 29: 949–962. Fiorillo, C.D., Tobler, P.N. and Schultz, W. (2003) Discrete coding of reward probability and uncertainty by dopamine neurons. Science, 299: 1898–1902.

345 Floresco, S.B., West, A.R., Ash, B., Moore, H. and Grace, A.A. (2003) Afferent modulation of dopamine neuron ﬁring differentially regulates tonic and phasic dopamine transmission. Nat. Neurosci., 6: 968–973. Florin-Lechner, S.M., Druhan, J.P., Aston-Jones, G. and Valentino, R.J. (1996) Enhanced norepinephrine release in prefrontal cortex with burst stimulation of the locus coeruleus. Brain Res., 742: 89–97. Fried, I., Wilson, C.L., Morrow, J.W., Cameron, K.A., Behnke, E.D., Ackerson, L.C. and Maidment, N.T. (2001) Increased dopamine release in the human amygdala during performance of cognitive tasks. Nat. Neurosci., 4: 201–206. Fujino, K., Yoshitake, T., Inoue, O., Ibii, N., Kehr, J., Ishida, J., Nohta, H. and Yamaguchi, M. (2002) Increased serotonin release in mice frontal cortex and hippocampus induced by acute physiological stressors. Neurosci. Lett., 320: 91–95. Fulford, A.J. and Marsden, C.A. (1998) Effect of isolationrearing on conditioned dopamine release in vivo in the nucleus accumbens of the rat. J. Neurochem., 70: 384–390. Fuxe, K. and Hanson, C.F. (1967) Central catecholamine neurons and conditioned avoidance behaviour. Psychopharmacology, 11: 439–447. Genn, R.F., Ahn, S. and Phillips, A.G. (2004) Attenuated dopamine efﬂux in the rat nucleus accumbens during successive negative contrast. Behav. Neurosci., 118: 869–873. Goldman-Rakic, P.S., Muly, E.C. III. and Williams, G.V. (2000) D(1) receptors in prefrontal cells and circuits. Brain Res. Rev., 31: 295–301. Goldstein, L.E., Rasmusson, A.M., Bunney, B.S. and Roth, R.H. (1994) The NMDA glycine site antagonist (+)-HA-966 selectively regulates conditioned stress-induced metabolis activation of the mesoprefrontal cortical dopamine but not serotonin systems: a behavioral, neuroendocrine, and neurochemical study in the rat. J. Neurosci., 14: 4937–4950. Goldstein, L.E., Rasmusson, A.M., Bunney, B.S. and Roth, R.H. (1996) Role of amygdala in the coordination of behavioral, neuroendocrine, and prefrontal cortical monoamine responses to psychological stress in the rat. J. Neurosci., 16: 4787–4798. Groenewegen, H.J. and Uylings, H.B. (2000) The prefrontal cortex and the integration of sensory, limbic and autonomic information. Prog. Brain Res., 126: 3–28. Guarraci, F.A. and Kapp, B.S. (1999) An electrophysiological characterization of ventral tegmental area dopaminergic neurons during differential Pavlovian fear conditioning in the awake rabbit. Behav. Brain Res., 99: 169–179. Hajnal, A. and Le´na´rd, L. (1997) Feeding-related dopamine in the amygdala of freely moving rats. Neuroreport, 8: 2817–2820. Harmer, C.J. and Phillips, G.D. (1999) Enhanced dopamine efﬂux in the amygdala by a predictive, but not a non-predictive, stimulus: facilitation by prior repeated D-amphetamine. Neuroscience, 90: 119–130. Heffner, T.G., Luttinger, D., Hartman, J.A. and Seiden, L.S. (1981) Regional changes in brain catecholamine turnover in the rat during performance on ﬁxed ratio and variable interval schedules of reinforcement. Brain Res., 214: 215–218.

Heffner, T.G. and Seiden, L.S. (1980) Synthesis of catecholamines from [3 H]tyrosine in brain during the performance of operant behavior. Brain Res., 183: 403–419. Herman, J.P., Guillonneau, D., Dantzer, R., Scatton, B., Semerdjian-Rouquier, L. and Le Moal, M. (1982) Differential effects of inescapable footshocks and of stimuli previously paired with inescapable footshocks on dopamine turnover in cortical and limbic areas of the rat. Life Sci., 30: 2207–2214. Hernandez, L. and Hoebel, B.G. (1988) Feeding and hypothalamic stimulation increase dopamine turnover in the accumbens. Physiol. Behav., 44: 599–606. Hernandez, L. and Hoebel, B.G. (1990) Feeding can enhance dopamine turnover in the prefrontal cortex. Brain Res. Bull., 25: 975–979. Hernandez, L., Paez, X. and Hamlin, C. (1983) Neurotransmitters extraction by local intracerebral dialysis in anesthetized rats. Pharmacol. Biochem. Behav., 18: 159–162. Hoffman, B.J. (1994) Molecular biology of dopamine transporters. In: Niznik, H.B. (Ed.), DA Receptors and Transporters. Marcel Dekker, New York, pp. 645–668. Hori, K., Tanaka, J. and Nomura, M. (1993) Effects of discrimination learning on the rat amygdala dopamine release: a microdialysis study. Brain Res., 621: 296–300. Horvitz, J.C. (2000) Mesolimbocortical and nigrostriatal dopamine responses to salient non-reward events. Neuroscience, 96: 651–656. Ihalainen, J.A., Riekkinen, P. Jr. and Feenstra, M.G.P. (1999) Comparison of dopamine and noradrenaline release in mouse prefrontal cortex, striatum and hippocampus using microdialysis. Neurosci. Lett., 277: 71–74. Imperato, A. and Di Chiara, G. (1984) Trans-striatal dialysis coupled to reverse phase high performance liquid chromatography with electrochemical detection: a new method for the study of the in vivo release of endogenous dopamine and metabolites. J. Neurosci., 4: 966–977. Iwano, H., Yamamuro, Y., Hori, K., Yamauchi, T. and Nomura, M. (1997) Effects of discrimination learning on the rat striatal dopaminergic activity: a microdialysis study. Neuroreport, 8: 1095–1098. Izaki, Y., Hori, K. and Nomura, M. (1998) Dopamine and acetylcholine elevation on lever-press acquisition in rat prefrontal cortex. Neurosci. Lett., 258: 33–36. Jacobs, B.L. (1987) Central monoaminergic neurons: singleunit studies in behaving animals. In: Meltzer, H.Y. (Ed.), Psychopharmacology: The Third Generation of Progress. Raven Press, New York, pp. 159–169. Jeanblanc, J., Hoeltzel, A. and Louilot, A. (2002) Dissociation in the involvement of dopaminergic neurons innervating the core and shell subregions of the nucleus accumbens in latent inhibition and affective perception. Neuroscience, 111: 315–323. Jime´nez-Capdeville, M.E. and Dykes, R.W. (1996) Changes in cortical acetylcholine release in the rat during day and night: differences between motor and sensory areas. Neuroscience, 71: 567–579. Jodo, E., Chiang, C. and Aston-Jones, G. (1998) Potent excitatory inﬂuence of prefrontal cortex activity on noradrenergic locus coeruleus neurons. Neuroscience, 83: 63–79.

346 Jones, B.E. (2003) Arousal systems. Front. Biosci., 8: s438–s451. Jones, B.E. (2005) From waking to sleeping: neuronal and chemical substrates. Trends Pharmacol. Sci., 26: 578–586. Joseph, M.H. and Hodges, H. (1990) Lever pressing for food reward and changes in dopamine turnover and uric acid in rat caudate and nucleus accumbens studied chronically by in vivo voltammetry. J. Neurosci. Methods, 34: 143–149. Joseph, M.H., Hodges, H. and Gray, J.A. (1989) Lever pressing for food reward and in vivo voltammetry: evidence for increases in extracellular homovanillic acid, the dopamine metabolite, and uric acid in the rat caudate nucleus. Neuroscience, 32: 195–201. Joseph, M.H., Datla, K. and Young, A.M.J. (2003) The interpretation of the measurement of nucleus accumbens dopamine by in vivo dialysis: the kick, the craving or the cognition? Neurosci. Biobehav. Rev., 27: 527–541. Kale´n, P., Nilsson, O.G., Cenci, M.A., Rosengren, E., Lindvall, O. and Bjo¨rklund, A. (1990) Intracerebral microdialysis as a tool to monitor transmitter release from grafted cholinergic and monoaminergic neurons. J. Neurosci. Methods, 34: 107–115. Kale´n, P., Rosegren, E., Lindvall, O. and Bjo¨rklund, A. (1989) Hippocampal noradrenaline and serotonin release over 24 hours as measured by the dialysis technique in freely moving rats: correlation to behavioural activity state, effect of handling and tail-pinch. Eur. J. Neurosci., 1: 181–188. Kalivas, P.W. and Nakamura, M. (1999) Neural systems for behavioral activation and reward. Curr. Opin. Neurobiol., 9: 223–227. Kawahara, Y., Kawahara, H. and Westerink, B.H.C. (1999) Comparison of effects of hypotension and handling stress on the release of noradrenaline and dopamine in the locus coeruleus and medial prefrontal cortex of the rat. Naunyn Schmiedeberg’s Arch. Pharmacol., 360: 42–49. Kiyatkin, E. and Gratton, A. (1994) Electrochemical monitoring of extracellular dopamine in nucleus accumbens of rats lever-pressing for food. Brain Res., 652: 225–234. Koepp, M.J., Gunn, R.N., Lawrence, A.D., Cunningham, V.J., Dagher, A., Jones, T., Brooks, D.J., Bench, C.J. and Grasby, P.M. (1998) Evidence for striatal dopamine release during a video game. Nature, 393: 266–268. Ladurelle, N., Roques, B.P. and Dauge´, V. (1995) The transfer of rats from a familiar to a novel environment prolongs the increase of extracellular dopamine efﬂux induced by CCK8 in the posterior nucleus accumbens. J. Neurosci., 15: 3118–3127. Lapiz, M.D., Fulford, A., Muchimapura, S., Mason, R., Parker, T. and Marsden, C.A. (2003) Inﬂuence of postweaning social isolation in the rat on brain development, conditioned behavior, and neurotransmission. Neurosci. Behav. Physiol., 33: 13–29. Legault, M. and Wise, R.A. (2001) Novelty-evoked elevations of nucleus accumbens dopamine: dependence on impulse ﬂow from the ventral subiculum and glutamatergic neurotransmission in the ventral tegmental area. Eur. J. Neurosci., 13: 819–828.

Lena, I., Parrot, S., Deschaux, O., Muffat-Joly, S., Sauvinet, V., Renaud, B., Suaud-Chagny, M.-F. and Gottesmann, C. (2005) Variations in extracellular levels of dopamine, noradrenaline, glutamate, and aspartate across the sleep-wake cycle in the medial prefrontal cortex and nucleus accumbens of freely moving rats. J. Neurosci. Res., 81: 891–899. Levita, L., Dalley, J.W. and Robbins, T.W. (2002) Nucleus accumbens dopamine and learned fear revisited: a review and some new ﬁndings. Behav. Brain Res., 137: 115–127. Lewy, A.J. and Seiden, L.S. (1972) Operant behavior changes norepinephrine metabolism in rat brain. Science, 175: 454–456. L’Heureux, R., Dennis, T., Curet, O. and Scatton, B. (1986) Measurement of endogenous noradrenaline release in the rat cerebral cortex in vivo by transcortical dialysis: effects of drugs affecting noradrenergic transmission. J. Neurochem., 46: 1794–1801. Louilot, A., Gonzalez-Mora, J.L., Guadalupe, T. and Mas, M. (1991) Sex-related olfactory stimuli induce a selective increase in dopamine release in the nucleus accumbens of male rats. A voltammetric study. Brain Res., 553: 313–317. Mark, G.P., Blander, D.S. and Hoebel, B.G. (1991) A conditioned stimulus decreases extracellular dopamine in the nucleus accumbens after the development of a learned taste aversion. Brain Res., 551: 308–310. Mark, G.P., Smith, S.E., Rada, P.V. and Hoebel, B.G. (1994) An appetitively conditioned taste elicits a preferential increase in mesolimbic dopamine release. Pharmacol. Biochem. Behav., 48: 651–660. Marrocco, R.T. and Field, B.A. (2002) Arousal. In: Ramachandran, V.S. (Ed.), Encyclopedia of the Human Brain. Academic Press, New York, pp. 223–236. Marrocco, R.T., Witte, E.A. and Davidson, M.C. (1994) Arousal systems. Curr. Opin. Neurobiol., 4: 166–170. Martin, G.E. and Myers, R.D. (1976) Dopamine efﬂux from the brain stem of the rat during feeding, drinking and leverpressing for food. Pharmacol. Biochem. Behav., 4: 551–560. McCullough, L.D. and Salamone, J.D. (1992) Involvement of nucleus accumbens dopamine in the motor activity induced by periodic food presentation: a microdialysis and behavioral study. Brain Res., 592: 29–36. McCullough, L.D., Cousins, M.S. and Salamone, J.D. (1993a) The role of nucleus accumbens dopamine in responding on a continuous reinforcement operant schedule: a neurochemical and behavioral study. Pharm. Biochem. Behav., 46: 581–586. McCullough, L.D., Sokolowski, J.D. and Salamone, J.D. (1993b) A neurochemical and behavioral investigation of the involvement of nucleus accumbens dopamine in instrumental avoidance. Neuroscience, 52: 919–925. McIntyre, C.K., Hatﬁeld, T. and McGaugh, J.L. (2002) Amygdala norepinephrine levels after training predict inhibitory avoidance retention performance in rats. Eur. J. Neurosci., 16: 1223–1226. McQuade, R. and Stanford, S.C. (2000) A microdialysis study of the noradrenergic response in rat frontal cortex and hypothalamus to a conditioned cue for aversive, naturalistic environmental stimuli. Psychopharmacology, 148: 201–208.

347 Men, D., McCarty, R. and Gold, P.E. (1999) Enhanced release of norepinephrine in rat hippocampus during spontaneous alternation tests. Neurobiol. Learn. Mem., 71: 2890300. Merali, Z., McIntosh, J. and Anisman, H. (2004) Anticipatory cues differentially provoke in vivo peptidergic and monoaminergic release at the medial prefrontal cortex. Neuropsychopharmacology, 29: 1409–1418. Miller, E.K. and Cohen, J.D. (2001) An integrative theory of prefrontal cortex function. Annu. Rev. Neurosci., 24: 167–202. Mingote, S., de Bruin, J.P.C. and Feenstra, M.G.P. (2004) Noradrenaline and dopamine efﬂux in the prefrontal cortex in relation to appetitive classical conditioning. J. Neurosci., 24: 2475–2480. Miranda, M.I., Ramirez-Lugo, L. and Bermudez-Rattoni, F. (2000) Cortical cholinergic activity is related to the novelty of the stimulus. Brain Res., 882: 230–235. Mirenowicz, J. and Schultz, W. (1996) Preferential activation of midbrain dopamine neurons by appetitive rather than aversive stimuli. Nature, 379: 449–451. Mitsushima, D., Mizuno, T. and Kimura, F. (1996) Age–related changes in diurnal acetylcholine release in the prefrontal cortex of male rats as measured by microdialysis. Neuroscience, 72: 429–434. Moor, E., Schirm, E., Jacso, J. and Westerink, B.H.C. (1998) Effects of neostigmine and atropine on basal and handlinginduced acetylcholine output from ventral hippocampus. Neuroscience, 82: 819–825. Morrow, B.A., Elsworth, J.D., Rasmusson, A.M. and Roth, R.H. (1999) The role of mesoprefrontal dopamine neurons in the acquisition and expression of conditioned fear in the rat. Neuroscience, 92: 553–564. Morrow, B.A., Redmond, A.J., Roth, R.H. and Elsworth, J.D. (2000) The predator odor, TMT, displays a unique, stresslike pattern of dopaminergic and endocrinological activation in the rat. Brain Res., 864: 146–151. Murphy, C.A., Pezze, M.-A., Feldon, J. and Heidbreder, C. (2000) Differential involvement of dopamine in the shell and core of the nucleus accumbens in the expression of latent inhibition to an aversive conditioned stimulus. Neuroscience, 97: 469–477. Neill, D.B., Fenton, H. and Justice, J.B. (2002) Increase in accumbal dopaminergic transmission correlates with response cost not reward of hypothalamic stimulation. Behav. Brain Res., 137: 129–138. Newhouse, P.A., Potter, A. and Singh, A. (2004) Effects of nicotinic stimulation on cognitive performance. Curr. Opin. Pharmacol., 4: 36–46. Nilsson, O.G., Kale´n, P., Rosengren, E. and Bjo¨rklund, A. (1990) Acetylcholine release in the rat hippocampus as studied by microdialysis is dependent on axonal impulse ﬂow and increases during behavioural activation. Neuroscience, 36: 325–338. Nishino, H., Ono, T., Muramoto, K., Fukuda, M. and Sasaki, K. (1987) Neuronal activity in the ventral tegmental area (VTA) during motivated bar press feeding in the monkey. Brain Res., 413: 302–313.

Nomura, M., Izaki, Y., Takita, M., Tanaka, J. and Hori, K. (2004) Extracellular level of basolateral amygdalar dopamine responding to reversal of appetitive-conditioned discrimination in young and old rats. Brain Res., 1018: 241–246. Oades, R.D. (1985) The role of noradrenaline in tuning and dopamine in switching between signals in the CNS. Neurosci. Biobehav. Rev., 9: 261–282. O’Neill, R.D. (1994) Microvoltammetric techniques and sensors for monitoring neurochemical dynamics in vivo. Analyst, 119: 767–779. Pan, W.-X., Schmidt, R., Wickens, J.R. and Hyland, B.I. (2005) Dopamine cells respond to predicted events during classical conditioning: evidence for eligibility traces in the rewardearning network. J. Neurosci., 25: 6235–6242. Panksepp, J. (1998) Affective Neuroscience. Oxford University Press, Oxford. Pappata, S., Dehaene, S., Poline, J.B., Gregoire, M.C., Jobert, A., Delforge, J., Frouin, V., Bottlaender, M., Dolle, F., Di Giamberardino, L. and Syrota, A. (2002) In vivo detection of striatal dopamine release during reward: a PET study with [11C]raclopride and a single dynamic scan approach. Neuroimage, 26: 1015–1121. Park, S.P., Lopez-Rodriguez, F., Wilzon, C.L., Maidment, N., Matsumotot, Y. and Engel, J. Jr. (1999) In vivo microdialysis measures of extracellular serotonin in the rat hippocampus during sleep-wakefulness. Brain Res., 833: 291–296. Paulson, P.E. and Robinson, T.E. (1994) Relationship between circadian changes in spontaneous motor activity and dorsal versus ventral striatal dopamine neurotransmission assessed with on-line microdialysis. Behav. Neurosci., 108: 624–635. Pei, Q., Zetterstro¨m, T. and Fillenz, M. (1990) Tail pinchinduced changes in the turnover and release of dopamine and 5-hydroxytryptamine in different brain regions of the rat. Neuroscience, 35: 133–138. Pezze, M.A. and Feldon, J. (2004) Mesolimbic dopaminergic pathways in fear conditioning. Prog. Neurobiol., 74: 301–320. Pezze, M.A., Feldon, J. and Murphy, C.A. (2002) Increased conditioned fear response and altered balance of dopamine in the shell and core of the nucleus accumbens during amphetamine withdrawal. Neuropharmacology, 42: 633–643. Pezze, M.A., Heidbreder, C.A., Feldon, J. and Murphy, C.A. (2001) Selective responding of nucleus accumbens core and shell dopamine to aversively conditioned contextual and discrete stimuli. Neuroscience, 108: 91–102. Phillips, A.G., Ahn, S. and Floresco, S.B. (2004) Magnitude of dopamine release in medial prefrontal cortex predicts accuracy of memory on a delayed response task. J. Neurosci., 24: 547–553. Phillips, A.G., Atkinson, L.J., Blackburn, J.R. and Blaha, C.D. (1993) Increased extracellular dopamine in the nucleus accumbens of the rat elicited by a conditional stimulus for food: an electrochemical study. Can. J. Physiol. Pharmacol., 71: 387–393. Phillips, G.D., Setzu, E., Vugler, A. and Hitchcott, P.K. (2003a) Immunohistochemical assessment of mesotelencephalic dopamine activity during the acquisition and expression of

348 Pavlovian versus instrumental behaviours. Neuroscience, 117: 755–767. Phillips, P.E.M., Stuber, G.D., Heien, M.L.A.V., Wightman, R.M. and Carelli, R.M. (2003b) Subsecond dopamine release promotes cocaine seeking. Nature, 422: 614–618. Portas, C.M., Bjorvatn, B. and Ursin, R. (2000) Serotonin and the sleep/wake cycle: special emphasis on microdialysis studies. Prog. Neurobiol., 60: 13–35. Pruessner, J.C., Champagne, F., Meaney, M.J. and Dagher, A. (2004) Dopamine release in response to a psychological stress in humans and its relationship to early life maternal care: a positron emission tomography study using [11C]raclopride. J. Neurosci., 24: 2825–2831. Puig, M.V., Celada, P., Diaz-Mataix, L. and Artigas, F. (2003) In vivo modulation of the activity of pyramidal neurons in the rat medial prefrontal cortex by 5-HT2A receptors: relationship to thalamocortical afferents. Cereb. Cortex, 13: 870–882. Pycock, C.J. and Taberner, P.V. (1981) Central Neurotransmitter Turnover. Croom Helm, London. Rasmussen, K. and Jacobs, B.L. (1986) Single unit activity of locus coeruleus neurons in the freely moving cat. II. Conditioning and pharmacologic studies. Brain Res., 371: 335–344. Redgrave, P., Prescott, T.J. and Gurney, K. (1999) Is the shortlatency dopamine response too short to signal reward error? Trends Neurosci., 22: 146–151. Rescorla, R.A. (1967) Pavlovian conditioning and its proper control procedures. Psychol. Rev., 74: 71–80. Rescorla, R.A. (1988) Pavlovian conditioning. It’s not what you think it is. Am. Psychol., 43: 151–160. Rescorla, R.A. (2000) Associative changes with a random CS–US relationship. Q. J. Exp. Psychol. B., 53: 325–340. Richardson, N.R. and Gratton, A. (1996) Behavior-relevant changes in nucleus accumbens dopamine transmission elicited by food reinforcement: an electrochemical study in rat. J. Neurosci., 16: 8160–8269. Richardson, N.R. and Gratton, A. (1998) Changes in medial prefrontal cortical dopamine levels associated with responsecontingent food reward: an electrochemical study in rat. J. Neurosci., 18: 9130–9138. Robbins, T.W. (1997) Arousal systems and attentional processes. Biol. Psychol., 45: 57–71. Robbins, T.W. and Everitt, B.J. (1996) Neurobehavioural mechanisms of reward and motivation. Curr. Opin. Neurobiol., 6: 228–236. Robinson, D.L., Heien, M.L. and Wightman, R.M. (2002) Frequency of dopamine concentration transients increases in dorsal and ventral striatum of male rats during introduction of conspecifics. J. Neurosci., 22: 10477–10486. Roitman, M.F., Stuber, G.D., Heien, M.L.A., Wightman, R.M. and Carelli, R.M. (2005) Subsecond dopamine release is selectively evoked by rewarding and not aversive taste stimuli. Program No. 657.1. Abstract Viewer/Itinerary Planner. Washington, DC: Society for Neuroscience, 2005. Online. Roitman, M.F., Stuber, G.D., Phillips, P.E.M., Wightman, R.M. and Carelli, R.M. (2004) Dopamine operates as a subsecond modulator of food seeking. J. Neurosci., 24: 1265–1271.

Rossetti, Z.L. and Carboni, S. (2005) Noradrenaline and dopamine elevations in the rat prefrontal cortex in spatial working memory. J. Neurosci., 25: 2322–2329. Roth, R.H., Tam, S.-Y., Ida, Y., Yang, J.-X. and Deutch, A.Y. (1988) Stress and the mesocorticolimbic dopamine systems. Ann. N. Y. Acad. Sci., 537: 138–147. Rueter, L.E. and Jacobs, B.L. (1996) Changes in forebrain serotonin at the light–dark transition: correlation with behaviour. Neuroreport, 7: 1107–1111. Saigusa, T., Tuinstra, T., Koshikawa, N. and Cools, A.R. (1999) High and low responders to novelty: effects of a catecholamine synthesis inhibitor on novelty-induced changes in behaviour and release of accumbal dopamine. Neuroscience, 88: 1153–1163. Salamone, J.D., Correa, M., Mingote, S. and Weber, S.M. (2003) Nucleus accumbens dopamine and the regulation of effort in food-seeking behavior: implications for studies of natural motivation, psychiatry, and drug abuse. J. Pharmacol. Exp. Ther., 305: 1–8. Salamone, J.D., Keller, R.W., Zigmond, M.J. and Stricker, E.M. (1989) Behavioural activation in rats increases striatal dopamine metabolism measured by dialysis perfusion. Brain Res., 487: 215–224. Sara, S.J. (1998) Learning by neurones: role of attention, reinforcement and behaviour. C. R. Acad. Sci. Life Sci., 321: 193–198. Sara, S.J. and Segal, M. (1991) Plasticity of sensory responses of locus coeruleus neurons in the behaving rat: implications for cognition. [In: Barnes, C.D. and Pompeiano, O. (Eds.).] Prog. Brain Res., 88: 571–585. Saulskaya, N. and Marsden, C.A. (1995) Conditioned dopamine release: dependence upon N-methyl-D-aspartate receptors. Neuroscience, 67: 57–63. Schoenfeld, R. and Seiden, L.S. (1967) Alpha-methyltyrosine: effects on ﬁxed ratio schedules of reinforcement. J. Pharm. Pharmacol., 19: 771–772. Schoenfeld, R.I. and Seiden, L.S. (1969) Effect of alpha-methyltyrosine on operant behavior and brain catecholamine levels. J. Pharmacol. Exp. Ther., 167: 319–327. Schultz, W. (1998) Predictive reward signal of dopamine neurons. J. Neurophysiol., 80: 1–27. Schultz, W. (2002) Getting formal with dopamine and reward. Neuron, 36: 241–263. Schultz, W., Apicella, P. and Ljungberg, T. (1993) Responses of monkey dopamine neurons to reward and conditioned stimuli during successive steps of learning a delayed response task. J. Neurosci., 13: 900–913. Schultz, W. and Dickinson, A. (2000) Neuronal coding of prediction errors. Annu. Rev. Neurosci., 23: 473–500. Seiden, L.S., MacPhail, R.C. and Oglesby, M.W. (1975) Catecholamines and drug–behavior interactions. Fed. Proc., 34: 1823–1831. Selden, N.R., Robbins, T.W. and Everitt, B.J. (1990) Enhanced behavioral conditioning to context and impaired behavioral and neuroendocrine responses to conditioned stimuli following ceruleocortical noradrenergic lesions: support for an attentional hypothesis of central noradrenergic function. J. Neurosci., 10: 531–539.

349 Sesack, S.R. and Carr, D.B. (2002) Selective prefrontal cortex inputs to dopamine cells: implications for schizophrenia. Physiol. Behav., 77: 513–517. Sharp, T., Zetterstrom, T. and Ungerstedt, U. (1986) An in vivo study of dopamine release and metabolism in rat brain regions using intracerebral dialysis. J. Neurochem., 47: 113–122. Smith, M.A., Banerjee, S., Gold, P.W. and Glowa, J. (1992a) Induction of c-fos mRNA in rat brain by conditioned and unconditioned stressors. Brain Res., 578: 135–141. Smith, A.D., Olson, R.J. and Justice, J.B. Jr. (1992b) Quantitative microdialysis of dopamine in the striatum: effect of circadian variation. J. Neurosci. Methods, 44: 33–41. Sokolowski, J.D., Conlan, A.N. and Salamone, J.D. (1998) A microdialysis study of nucleus accumbens core and shell dopamine during operant responding in the rat. Neuroscience, 86: 1001–1009. Sparber, S.B. (1975) Neurochemical changes associated with schedule-controlled behavior. Fed. Proc., 34: 1802–1812. Stanley, B.G., Schwartz, D.H., Hernandez, L., Hoebel, B.G. and Leibowitz, S.F. (1989) Patterns of extracellular norepinephrine in the paraventricular hypothalamus: relationship to circadian rhythm and deprivation-induced eating behavior. Life Sci., 45: 275–282. Stark, H., Bischof, A. and Scheich, H. (1999) Increase of extracellular dopamine in prefrontal cortex of gerbils during acquisition of the avoidance strategy in the shuttle-box. Neurosci. Lett., 264: 77–80. Stark, H., Bischof, A., Wagner, T. and Scheich, H. (2000) Stages of avoidance strategy formation in gerbils are correlated with dopaminergic transmission activity. Eur. J. Pharmacol., 405: 263–275. Stark, H., Rothe, T., Wagner, T. and Scheich, H. (2004) Learning a new behavioral strategy in the shuttle-box increases prefrontal dopamine. Neuroscience, 126: 21–29. Stark, H. and Scheich, H. (1997) Dopaminergic and serotonergic neurotransmission systems are differentially involved in auditory cortex learning: a long-term microdialysis study of metabolites. J. Neurochem., 68: 691–697. Strecker, R.E., Nalwalk, J., Dauphin, L.J., Thakkar, M.M., Chen, Y., Ramesh, V., Hough, L.B. and Mccarley, R.W. (2002) Extracellular histamine levels in the feline preoptic/ anterior hypothalamic area during natural sleep-wakefulness and prolonged wakefulness: an in vivo microdialysis study. Neuroscience, 113: 663–670. Stuber, G.D., Wightman, R.M. and Carelli, R.M. (2005) Extinction of cocaine self-administration reveals functionally and temporally distinct dopaminergic signals in the nucleus accumbens. Neuron, 46: 661–669. Tettamanti, M., Moro, A., Messa, C., Moresco, R.M., Rizzo, G., Carpinelli, A., Matarrese, M., Fazio, F. and Perani, D. (2005) Basal ganglia and language: phonology modulates dopaminergic release. Neuroreport, 16: 397–401. Thiel, C.M., Huston, J.P. and Schwarting, R.K. (1998a) Hippocampal acetylcholine and habituation learning. Neuroscience, 85: 1253–1262. Thiel, C.M., Huston, J.P. and Schwarting, R.K. (1998b) Cholinergic activation in frontal cortex and nucleus

accumbens related to basic behavioral manipulations: handling, and the role of post-handling experience. Brain Res., 812: 121–132. Tilson, H.A., Rech, R.H. and Sparber, S.B. (1975) Release of 14 C-norepinephrine into the lateral cerebroventricle of rats by exposure to a conditioned aversive stimulus. Pharmacol. Biochem. Behav., 3: 385–392. Tronel, S., Feenstra, M.G.P. and Sara, S.J. (2004) Noradrenergic action in prefrontal cortex in the late stage of memory consolidation. Learn. Mem., 11: 453–458 Erratum 658. Trulson, M.E. and Preussler, D.W. (1984) Dopamine-containing ventral tegmental area neurons in freely moving cats: activity during the sleep–waking cycle and effects of stress. Exp. Neurol., 83: 367–377. Tsuda, A., Tanaka, M., Ida, Y., Tsujimaru, S., Ushijima, I. and Nagasaki, N. (1986) Effects of preshock experience on enhancement of rat brain noradrenaline turnover by psychological stress. Pharmacol. Biochem. Behav., 24: 115–119. Ungerstedt, U. (1984) Measurement of neurotransmitter release by intracranial dialysis. In: Marsden, C.A. (Ed.), Measurement of Neurotransmitter Release In Vivo. John Wiley & Sons, Chichester, pp. 81–105. Ungless, M.A. (2004) Dopamine: the salient issue. Trends Neurosci., 27: 702–706. Ungless, M.A., Magill, P.J. and Bolam, J.P. (2004) Uniform inhibition of dopamine neurons in the ventral tegmental area by aversive stimuli. Science, 303: 2040–2042. Uylings, H.B. and van Eden, C.G. (1990) Qualitative and quantitative comparison of the prefrontal cortex in rat and in primates, including humans. Prog. Brain Res., 85: 31–62. van der Meulen, J.A.J., Joosten, R.N.J.M.A., de Bruin, J.P.C. and Feenstra, M.G.P. (2006, in press) Dopamine and noradrenaline efﬂux in the medial prefrontal cortex during serial reversals and extinction of instrumental goal-directed behavior. Cerebral Cortex 2006; doi: 10.1093/cercor/bhl057. Watanabe, M., Kodama, T. and Hikosaka, K. (1997) Increase of extracellular dopamine in primate prefrontal cortex during a working memory task. J. Neurophysiol., 78: 2795–2798. Westerink, B.H.C. (1995) Brain microdialysis and its application for the study of animal behaviour. Behav. Brain Res., 70: 103–124. Westerink, B.H.C., Cremers, T.I., de Vries, J.B., Liefers, H., Tran, N. and De Boer, P. (2002) Evidence for activation of histamine H3 autoreceptors during handling stress in the prefrontal cortex of the rat. Synapse, 43: 238–243. Westerink, B.H.C., Damsma, G., Rollema, H., de Vries, J.B. and Horn, A.S. (1987) Scope and limitations of in vivo brain dialysis: a comparison of its application to various neurotransmitter systems. Life Sci., 4: 1763–1776. Westerink, B.H.C., de Boer, P., de Vries, J.B., Kruse, C.G. and Long, S.K. (1998) Antipsychotic drugs induce similar effects on the release of dopamine and noradrenaline in the medial prefrontal cortex of the rat brain. Eur. J. Pharmacol., 361: 27–33. Westerink, B.H.C. and Justice, J.B. Jr. (1991) Microdialysis compared with other in vivo release models. In: Robinson,

350 T.E. and Justice, Jr., J.B. (Eds.), Microdialysis in the Neurosciences. Elsevier Science, Amsterdam, pp. 23–43. Wightman, R.M. and Robinson, D.L. (2002) Transient changes in mesolimbic dopamine and their association with ‘reward’. J. Neurochem., 82: 721–735. Wilkinson, L.S., Humby, T., Killcross, A.S., Torres, E.M., Everitt, B.J. and Robbins, T.W. (1998) Dissociations in dopamine release in medial prefrontal cortex and ventral striatum during the acquisition and extinction of classical aversive conditioning in the rat. Eur. J. Neurosci., 10: 1019–1026. Wilson, C., Nomikos, G.G., Collu, M. and Fibiger, H.C. (1995) Dopaminergic correlates of motivated behavior: importance of drive. J. Neurosci., 15: 5169–5178. Wise, R.A. (1978) Catecholamine theories of reward: a critical review. Brain Res., 152: 215–247. Wise, R.A. and Rompre, P.P. (1989) Brain dopamine and reward. Annu. Rev. Psychol., 40: 191–225. Yamamuro, Y., Hori, K., Iwano, H. and Nomura, M. (1994) The relationship between learning performance and dopamine in the prefrontal cortex of the rat. Neurosci. Lett., 177: 83–86. Yamatodani, A., Mochizuki, T. and Mammoto, T. (1996) New vistas on histamine arousal hypothesis: microdialysis studies. Methods Find. Exp. Clin. Pharmacol., 18(Suppl. A): 113–117. Yerkes, R.M. and Dodson, J.D. (1908) The relation of strength of stimulus to rapidity of habit-formation. J. Comp. Neurol. Psychol., 18: 459–482. Yokoo, H., Tanaka, M., Yoshida, M., Tsuda, A., Tanaka, T. and Mizoguchi, K. (1990) Direct evidence of conditioned fear-elicited enhancement of noradrenaline release in the rat hypothalamus assessed by intracranial microdialysis. Brain Res., 536: 305–308.

Yokoyama, M., Suzuki, E., Sato, T., Maruta, S., Watanabe, S. and Miyaoka, H. (2005) Amygdalic levels of dopamine and serotonin rise upon exposure to conditioned fear stress without elevation of glutamate. Neurosci. Lett., 379: 37–41. Yoshioka, M., Matsumoto, M., Togashi, H. and Saito, H. (1996) Effect of conditioned fear stress on dopamine release in the rat prefrontal cortex. Neurosci. Lett., 209: 201–203. Young, A.M.J. (1993) Intracerebral microdialysis in the study of physiology and behaviour. Rev. Neurosci., 4: 373–395. Young, A.M.J. (2004) Increased extracellular dopamine in nucleus accumbens in response to unconditioned and conditioned aversive stimuli: studies using 1 min microdialysis in rats. J. Neurosci. Methods, 138: 57–63. Young, A.M.J., Ahier, R.G., Upton, R.L., Joseph, M.H. and Gray, J.A. (1998) Increased extracellular dopamine in the nucleus accumbens of the rat during associative learning of neutral stimuli. Neuroscience, 83: 1175–1183. Young, A.M.J., Joseph, M.H. and Gray, J.A. (1993) Latent inhibition of conditioned dopamine release in rat nucleus accumbens. Neuroscience, 54: 5–9. Young, A.M., Moran, P.M. and Joseph, M.H. (2005) The role of dopamine in conditioning and latent inhibition: what, when, where and how? Neurosci. Biobehav. Rev., 29: 963–976. Zald, D.H., Boileau, I., El-Dearedy, W., Gunn, R., McGlone, F., Dichter, G.S. and Dagher, A. (2004) Dopamine transmission in the human striatum during monetary reward tasks. J. Neurosci., 24: 4105–4112. Zetterstro¨m, T., Sharp, T., Marsden, C.A. and Ungerstedt, U. (1983) In vivo measurement of dopamine and its metabolites by intracerebral dialysis: changes after D-amphetamine. J. Neurochem., 41: 1769–1773.

CHAPTER 4.3

Microdialysis in the study of behavior reinforcement and inhibition Pedro V. Rada1,2,, Ximena Pa´ez1, Luis F. Herna´ndez1, Nicole M. Avena2 and Bartley G. Hoebel2 1

Laboratory of Behavioral Physiology, School of Medicine, Universidad de los Andes, Me´rida, Venezuela 2 Department of Psychology, Princeton University, Princeton, NJ, USA

Abstract: Brain microdialysis has been a valuable technique in the neuroscience ﬁeld for more than 20 years. In vivo microdialysis in freely-moving rats allows measurement of neurotransmitter release in response to ongoing behaviors. In this chapter we review ﬁndings using microdialysis in the study of behavior reinforcement and inhibition. This literature leads to the development of the dopamine hypothesis of reward and the cholinergic hypothesis of aversion and their underlying neural circuitry. Within the context of natural rewards, we discuss many of the key ﬁndings using microdialysis in the nucleus accumbens, ventral tegmental area and hypothalamus to study feeding, water intake, and mating. Artiﬁcial rewards, such as intracranial self-stimulation and drug reward, are also reviewed. Finally, data are summarized that suggest a natural reward, sugar, may take on behavioral and neurochemical properties of an artiﬁcial reward, such as a drug of abuse, under certain conditions

following chemical gradients. The collected amount of the analytes in the dialysate is proportional to what is released near the probe (Segovia et al., 1986; Schwartz et al., 1990). Since the dialysates are pure protein-free ultraﬁltrates, they do not suffer enzymatic degradation and can be chemically analyzed without pretreatments. Thus, it is possible to continuously sample from the extracellular environment in virtually all living tissues and organs of sufﬁcient size. The analytes are typically analyzed by high-performance liquid chromatography (HPLC). Coupling microdialysis with improved analytical techniques such as capillary electrophoresis or mass spectrometry allows for the study of chemical changes in very short periods and in very small sample volumes (see Chapter 3.4). The main application of microdialysis has traditionally been the study of chemicals in the brain. Before the invention of microdialysis,

I. Introduction I.A. Brain microdialysis The microdialysis technique is a valuable tool in basic and clinical research. It has been extensively and successfully used, particularly in the neuroscience ﬁeld, for almost three decades (Mark et al., 1991; Ungerstedt, 1991; Westerink, 1995; Muller, 2002; Bourne, 2003; Plock and Kloft, 2005). The popularity of this technique is due to its advantages with respect to other in vivo sampling techniques. It permits recovery of endogenous and exogenous substances from the brain or body, or infusion of drugs through the microdialysis probe (i.e., reverse microdialysis). Dialysis occurs by solute exchange through the porous membrane of the probe Corresponding author: E-mail: [email protected] or [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

351

DOI: 10.1016/S1569-7339(06)16019-1 Copyright 2007 Elsevier B.V. All rights reserved

352

neurotransmission studies were based on techniques such as push–pull perfusion and tissue homogenates (Korf, 1986; Myers et al., 1998; Kottegoda et al., 2002; Rui and Lebaron, 2005). Microdialysis initially studied chemical changes in discrete brain areas of anesthetized rats (Hernandez et al., 1983; Segovia et al., 1986), but rapid improvements in the technique soon made it possible to sample from freely-behaving animals (Zetterstrom and Ungerstedt, 1984; Hernandez et al., 1986; Carboni et al., 1989; Westerink, 1995; Fillenz, 2005). Microdialysis in freely behaving animals is ideal for studying the relationship between particular chemical messengers and changes in ongoing behaviors, such as feeding (Hoebel et al., 1989; Hernandez et al., 1991; Meguid et al., 1996; Bassareo and Di Chiara, 1999a, b; Rouch et al., 1999; Smith, 2004; Rada et al., 2005), drinking (Yoshida et al., 1992; Tanaka et al., 2004; Molander et al., 2005), mating (Pfaus et al., 1990; Dominguez and Hull, 2005), exercising (Meeusen et al., 2001), cognitive processes (Pepeu and Giovannini, 2004), pain (Stiller et al., 2003), intracranial self-stimulation (ICSS) (Hernandez and Hoebel, 1988a; You et al., 2001), drug self-administration (Hurd et al., 1999; Ranaldi et al., 1999), and in neuropsychiatric animal models (Hernandez et al., 1990, 1991; Hoebel et al., 1992; Joseph et al., 2003; Invernizzi and Garattini, 2004; Wilcox et al., 2005). A Medline search indicates there have been more than 10,000 microdialysis publications in the last 25 years, with no indication of a decline in its use. So far, 71% of microdialysis studies are in rats, and of those 74% are in the brain. Behavioral studies represent 18% of all microdialysis studies, and reward studies correspond to about 10% of those, almost all of which have been conducted in rats.

I.B. The concept of neural reward and aversion processes This chapter will review some of the ﬁndings that the microdialysis technique has offered in the ﬁeld of behavior reinforcement and its inhibition. Pavlov (1927) used the term ‘‘reinforcement’’ to

refer to the strengthening of the association between an unconditioned stimulus and a conditioned stimulus that results when the two are paired. Skinner (1938) deﬁned a ‘‘reinforcer’’ as a stimulus administered following a correct, arbitrarily chosen, response that increases the probability of occurrence of the response. The terms reinforcement and reward will be used interchangeably in this chapter and will include both positive and negative reinforcement (i.e., behavior to get, or get rid of, the reinforcer). Primary reinforcers, such as food, water, and sex, have an inherited role in fostering the survival of the animal or species, while secondary reinforcers, such as associated environmental stimuli, are learned. Aversion, or behavior inhibition, will be considered as the opposite of reward in that it is a process that suppresses, instead of increases, a behavior. This type of response can also be lifesaving. For example, a conditioned taste aversion prevents an animal from eating a food that has previously been associated with sickness. Normally, once the consummatory phase of a natural behavior has been satisﬁed, an inhibition of the behavior occurs. This behavior inhibition, with regard to feeding, is deﬁned as the process leading to satiety, and although the animal is averse to further eating, the state is generally described as pleasant. In this chapter, we will review results that suggest the dopamine (DA) system in the nucleus accumbens (NAc) is part of a reinforcement system. DA release can activate an animal and contribute to a desired state. Cholinergic interneurons may counteract DA and play a role in behavior inhibition. Thus depending on the balance of neurotransmitter functions and the circuits in which they are embedded, an animal will either increase or decrease its response rate or response force. We hypothesize that high extracellular DA reinforces behavior, be it approach or escape behavior. If acetylcholine (ACh) also rises the approach behavior becomes inhibited and satiation ensues. Escape behavior lowers ACh (Rada and Hoebel, 2001). Low extracellular DA coupled with chronically increased ACh release depresses an animal and may lead to a condition variously described as immobility, helplessness, or despair.

353

I.C. Anatomical substrates of reward The neural circuitry involved in reward and aversion is part of brain systems engaged in emotions. The limbic system was originally conceived as a series of interconnected brain areas that played an important role in the acquisition, storage, and expression of emotion, and was described as a circuit between hypothalamus and the cerebral cortex (Papez, 1995), and later expanded to include other structures such as the amygdala (AMYG), hippocampus (HIPP), and NAc. From early studies by Mogenson and collaborators, it was concluded that the NAc was a structure that linked ‘‘emotions to action’’ (Mogenson et al., 1980). In the 1950s, Olds and Milner (1954) serendipitously discovered that rats would self-administer electrical pulses directly into many of the sites that coincided with areas of the limbic system, suggesting that the brain had

specialized ‘‘centers’’ for reward. Many researchers contributed to the effort of tracing the ‘‘reward pathway’’ starting with the circuit from the lateral hypothalamus (LH) to the ventral tegmental area (VTA) and the NAc as proposed by Wise and Hoffman (1992). As can be seen in Fig. 1, the neural circuitry of reward comprised at least two loops. The ﬁrst loop includes the LH and connects to structures in the hindbrain, returning to the NAc and then back to the hypothalamus directly and indirectly through the ventral pallidum (VP) (Leibowitz and Hoebel, 2004; Kelley et al., 2005). Others have described links from midbrain cholinergic cells to the VTA (Yeomans et al., 2001), and a glutamatergic/orexin path directly from the LH to the VTA and to the NAc (Harris et al., 2005). A second ‘‘loop’’ is composed of cortical glutamatergic inputs to the NAc with connections from the NAc to the VP, then to the mediodorsal

Fig. 1. Diagram of the reward circuitry showing the convergence in the NAc of the cortical and subcortical loops (from Leibowitz and Hoebel, 2004).

354

thalamus projecting back to the prefrontal cortex (PFC). This second loop, which is actually a series of concentric loops or a spiral, receives important afferents from the AMYG and HIPP (McGinty, 1999; Napier and Mitrovic, 1999; McFarland et al., 2004). This cortical–subcortical loop involving the basal ganglia motor system also branches off in the thalamus to connect with the primary motor system. It is evident from the perspective of Fig. 1, following Mogenson’s ‘‘reward circuitry’’ theory, that the NAc is at one of the intersections between limbic and motor systems. Another classic view has received renewed attention with modern evidence that the LH connects reciprocally with the cortex (Rolls, 1984; Oomura, 1988). Neuroanatomists later described the extended amygdala as interconnecting structures that shared strong histological homologies, including the AMYG, stria terminalis, substantia innominata, and the NAc (de Olmos and Heimer, 1999). This designation clearly differentiated the NAc shell, as part of the extended AMYG, from the NAc core and dorsal striatum (STR) (Zahm et al., 1996; Cadoni and Di Chiara, 1999; Weiner, 2003). Efforts to differentiate the speciﬁc functions of the NAc shell vs. core have led to very interesting hypotheses of Pavlovian-instrumental transfer, learning, motivation, and habit formation, which are reviewed elsewhere (Robbins et al., 1989; Kalivas and Nakamura, 1999; Di Chiara, 2002).

I.D. Dopamine hypothesis of reward Initial studies in the early 1970s showed that speciﬁc lesions of the DA projections from the midbrain to the STR produced deﬁcits in feeding and drinking, resembling the ‘‘lateral hypothalamic syndrome’’ characterized by aphagia and adipsia (Teitelbaum and Epstein, 1962; Ungerstedt, 1970, 1971). The deﬁcit in feeding and drinking occurred without somatosensory impairment but rather as a form of sensory neglect (Zigmond and Stricker, 1972; Marshall and Teitelbaum, 1974; Lindholm et al., 1975). It was suggested that the dopaminergic system in the medial forebrain bundle was probably damaged as the ascending ﬁbers ran through the far-lateral LH. Studies by Wise and

colleagues and others have demonstrated that neuroleptics (DA receptor antagonists) attenuate the expression of various behaviors such as feeding, drinking, ICSS, and drug self-administration (Wise, 1978; Xenakis and Sclafani, 1981; Bailey et al., 1986; Ettenberg and Camp, 1986; Ettenberg, 1989; Wise and Rompre, 1989; Horvitz et al., 1993; Samson and Chappell, 2004; Xi et al., 2005, 2006). Microdialysis studies have given further support to these behavioral ﬁndings by showing that natural reinforcers, as well as artiﬁcial reinforcers such as ICSS and drugs of abuse increase extracellular DA in the NAc (Hernandez and Hoebel, 1988b; Nakahara et al., 1989a, b; Wise et al., 1995a; Ranaldi et al., 1999). Electrophysiological studies in monkeys have emphasized the role of dopaminergic neurons in the animal’s capacity to predict the occurrence of a novel event and reward learning (Schultz, 1997, 1998a, b; Tobler et al., 2005). Berridge and Robinson (1998) also suggest that DA may be involved in ‘‘incentive salience’’ and emphasize this role in motivation. Their conclusions are based on several experiments using ‘‘taste reactivity’’ to measure affective reactions suggesting that DA-depleted rats have normal hedonic reactions and associative learning (Berridge and Robinson, 1998). Moreover, DA blockers do not seem to block oral approach reﬂexes, nor do speciﬁc DA agonists potentiate such measures, but they do alter the incentive value of the reward (Pecina et al., 1997, 2003; Berridge and Robinson, 1998). Recently, in DA-deﬁcient mice, it was demonstrated that DA was not necessary for learning or liking, but it was necessary for seeking of the reward (Robinson et al., 2005), which could be reinstated by localized restoration of DA function with viral vectors (Szczypka et al., 2001). Most recently, brain imaging studies have been used to detect the active parts during various aspects of reinforcement, with attention drawn to changes in DA receptor function during the cognitive aspects of reinforcement (Kalivas and Volkow, 2005). Stressful events and aversive stimuli can also increase DA levels in the NAc (Abercrombie et al., 1989; Keefe et al., 1993; Salamone, 1994; Salamone et al., 1997; Rada et al., 1998b). On the surface this seems to argue against any theory that DA mediates reward, but on reﬂection, there is more to

355

reward than simple pleasure. DA could also be involved in the relief from stress and pain, both in the motivation to achieve relief and the reward of achieving it. This ﬁts with the theory that DA has a role in salience and in negative reinforcement as well as positive reinforcement. None of these discoveries rule out the possibility that DA activates a pleasure response at some point in the circuit, including mu-opioid receptor ‘‘hot spots’’ in the NAc for ‘‘liking’’ a taste (Pecina and Berridge, 2005). The opioid peptide enkephalin in the NAc has been related to reward and can activate both mu and delta receptors to increase the release of DA (Di Chiara and Imperato, 1988; Bals-Kubik et al., 1989). Opiates are involved in eliciting feeding in the NAc (Kelley et al., 2000) as well as in many other limbic system sites (Levine and Billington, 2004). Moreover, it was shown in 1980, in one of the ﬁrst local self-administration studies, that rats will self-administer morphine directly into the NAc (Olds, 1982). Microdialysis experiments have contributed to all of these interesting dimensions in the study of DA’s role in reward. DA increases in the NAc shell during an animal’s ﬁrst exposure to a novel food, and this response disappears at a second meal even though the animal consumes as much as the ﬁrst time (Bassareo and Di Chiara, 1999b; Bassareo et al., 2002; Rada et al., 2005). The DA response can be reinstated by food deprivation (Bassareo and Di Chiara, 1999b) or an animal can learn to restore it, and obtain a DA surge every day, by binge eating (Rada et al., 2005).

I.E. Cholinergic hypothesis of aversion Microdialysis has also contributed to a theory of ‘‘aversion’’ or behavior inhibition. The details are given in later sections of this review under the topics of feeding satiation, mating satiation, ICSS escape and drug withdrawal. In brief, we ﬁnd that ACh is released in the NAc in a variety of situations that all have behavior inhibition as a common feature. Extracellular ACh rises toward the end of a meal (Mark et al., 1992; Rada et al., 2005) and is also elevated after experiencing the forced swim test and thus may contribute to behavioral

depression (Chau et al., 2001). Acetylcholine also increases in the NAc during aversive hypothalamic stimulation, and most telling, extracellular ACh levels decrease when the animal performs stimulation–escape responses (Rada and Hoebel, 2001). Evidence will be presented for the general principle that accumbens ACh is relatively high compared with DA during withdrawal from addictive substances (Rada et al., 1996, 2004; Colantuoni et al., 2002; Rada and Hoebel, 2005). Since withdrawal is an aversive condition that results in many behavioral signs of distress, including anxiety and depression, one can surmise that relatively high levels of ACh in the NAc can enhance a circuit that can cause either behavior inhibition or aversion or both. II. Natural rewards Rewards can be classiﬁed as non-drug rewards or drug rewards (Di Chiara, 2002), or as natural and artiﬁcial rewards. We will classify natural rewards as those reinforcers studied in natural behaviors (feeding, thirst, and mating) and artiﬁcial rewards as drug reward and ICSS. Artiﬁcial rewards act via natural pathways but have magniﬁed responses. We will focus ﬁrst on natural rewards and later show how they too can become magniﬁed, thereby blurring the distinction between non-drug and drug rewards. II.A. Feeding Of all the natural rewards under investigation, feeding behavior is the most studied using the microdialysis technique. The study of ingestive behavior has progressed substantially during the last decade with the discovery of new peptides involved in feeding regulation and the development of new animal models of feeding disorders (see comprehensive reviews by Berthoud, 2004; Leibowitz and Hoebel, 2004). In the present chapter, we will focus speciﬁcally on ﬁndings using brain microdialysis that have impacted the ingestive behavior ﬁeld. Most research on feeding behavior using brain microdialysis has concentrated on the hypothalamus

356

and NAc. Fig. 1 shows a way in which these areas are linked (also see Rolls, 1994; Berthoud, 2000). The hypothalamus has maintained its early start in feeding research as an area clearly involved in feeding initiation and satiation, and a place where most feeding neuropeptides act (Leibowitz and Wortley, 2004). Circuits in the hypothalamus can foster foraging for one macronutrient or another (Leibowitz and Wortley, 2004) and are under the control of hormones that signal nutrient stores in the body (Leibowitz and Hoebel, 2004; Strader and Woods, 2005). The NAc is a terminal ﬁeld of the mesolimbic dopaminergic system involved in hedonic and motivational aspects of feeding as discussed above. Many other limbic areas are also involved, but these two sites capture the basic subcortical functions of sensory input, physiological modulation and motor output. Higher structures project to limbic sites with information from learning and memory stores (Rolls, 2000). One can start anywhere in these wonderful circuits. Given the topic of this chapter, we start with the NAc.

II.A.1. NAc microdialysis during feeding II.A.1.a. Dopamine in the NAc. Dopamine is the most studied neurotransmitter in relation to natural rewards. As mentioned brieﬂy in the Introduction, lesion of the LH is characterized by aphagia and adipsia (Anand and Brobeck, 1951a, b; Teitelbaum and Epstein, 1962), and this syndrome was due in part to damage incurred by dopaminergic projections to the STR (Ungerstedt, 1970, 1971). This spurred the study of a possible role for DA in an animal’s reaction to natural rewards. DA antagonists injected locally into the accumbens temporarily increase operant responding for food, which then gradually declines suggesting a loss of reward and not simply motor impairment (Wise, 1978). There is also a loss of reaction to the taste of sugar (Schneider et al., 1986). Microdialysis studies have shown that feeding releases DA in the NAc (Hernandez and Hoebel, 1988a, b; Radhakishun et al., 1988; Hoebel et al., 1989; Westerink et al., 1994). An initial use of microdialysis to understand if DA in the NAc is involved in conditioning was carried out by Mark et al. (1991), showing that oral infusion of a

palatable food, saccharin, increased DA release in the NAc, and the opposite response was observed if the saccharin had been associated with sickness in a conditioned taste aversion paradigm. Di Chiara and colleagues (Bassareo and Di Chiara, 1997, 1999b; Tanda and Di Chiara, 1998) ﬁnd that a novel, unconditioned, palatable food stimuli (Fonzies) releases DA only during the ﬁrst experience, and this effect habituates on the second exposure, suggesting that DA in the NAc is involved in acquisition rather than maintenance of incentive motivation (Bassareo and Di Chiara, 1997). The ﬁrst experience with a novel food raises DA selectively in the shell and not the core of the NAc (Tanda and Di Chiara, 1998; Bassareo and Di Chiara, 1999a), and this increase is dependent on stimulation of mu-opioid receptors located in the VTA (Tanda and Di Chiara, 1998). Similar results have been found using sucrose. Ingestion of a 10% sucrose solution releases DA in the NAc shell during the ﬁrst exposure, but less on a second exposure 24 h or 21 days later, again suggesting that DA release in the NAc shell depends on the novelty of the stimulus and deprivation state (Rada et al., 2005). DA release is proportional to sucrose concentration (Hajnal et al., 2004), and under certain feeding conditions DA can be released with a palatable food time after time, as discussed in Section III of this chapter (Rada et al., 2005). Postingestional factors can inﬂuence DA release in the NAc. In a conditioned taste preference paradigm, a neutral taste that was previously associated with infusion of a highly caloric solution into the stomach can increase DA levels in the NAc (Mark et al., 1994), suggesting that the DA increase observed when rats drink sucrose could be due, in part, to prior experience with its caloric content (Hajnal and Norgren, 2001; Rada et al., 2005). An alternative explanation could be that DA increases in the NAc shell as a consequence of the orosensory stimulation. In order to bypass most postingestional factors, a gastric ﬁstula can be implanted so that a liquid diet (e.g., sucrose) can be drained out (Mook et al., 1988; Smith, 1998). Sham-feeding conﬁrms that the taste of sucrose can release DA in the NAc (Hajnal et al., 2004; Avena et al., 2006), thereby corroborating

357

the effect of saccharin (Mark et al., 1991). Together, these results demonstrate that DA release in the NAc can vary with deprivation, food conditioned stimuli, orosensory stimulation, postingestive factors, and the presence of a novel palatable food. II.A.1.b. Acetylcholine in the nucleus accumbens (behavior inhibition). Acetylcholine in the STR and accumbens is released from interneurons that represent 5% of the cellular population (Bolam et al., 1984; Meredith et al., 1989). Some clues as to the effects of cholinergic interneurons on reward mechanisms can be deduced from Parkinson’s disease. These patients have destruction of DA neurons in the basal ganglia and present symptoms such as bradikinesia, rigidity, and abnormal movements (dorsal STR), and also an anhedonic state (ventral STR) (Fibiger, 1984; Isella et al., 2003; Lemke et al., 2005) including weight loss (Chen et al., 2003; Palhagen et al., 2005; Tuite et al., 2005). A general hypothesis suggests that normal functioning of the basal ganglia depends on a balance between DA/ACh in the STR (Grewaald et al., 1974). When the DA neurons degenerate, an imbalance occurs, with a relative increase in ACh release (Spehlmann and Stahl, 1976; RodriguezPuertas et al., 1994). This led us to the theory that ACh in the NAc could counteract DA’s rewarding signal and thus produce behavioral inhibition (Hoebel et al., 1999). Few studies have investigated the role of ACh in the NAc during feeding. Acetylcholine increases in the accumbens at the end of the meal and probably signals satiety (Mark et al., 1992). When neostigmine, an acetylcholinesterase inhibitor and indirect cholinergic agonist, is infused through a dialysis probe bilaterally in the NAc, a decrease in food intake is observed (Mark et al., 1992). Moreover, bilateral infusion of a muscarinic (M1) receptor agonist (arecoline) shortens the time interval to reach satiety (Rada et al., unpublished data). Lesioning the cholinergic interneurons with a speciﬁc neurotoxin makes rats eat more, although they lose body weight (Galosi et al., 1997; Hajnal et al., 2000). Further evidence that ACh could be involved in behavioral inhibition comes from conditioned taste aversion experiments, in

which an aversive taste increases ACh levels in the NAc (Mark et al., 1995) and local injection of a cholinergic agonist into the accumbens induces a conditioned taste aversion (Taylor et al., 1992). Finally, the hypothesis that ACh signals satiety was tested using the sham-feeding paradigm. If ACh signals satiety, then sham-fed rats should consume large amounts of food without any change in accumbens ACh. Indeed, rats that are sham-fed drink large amounts of a liquid diet (10% sucrose) and show no signiﬁcant change in ACh levels in the NAc during the meal. Evidently postingestional signals are needed to activate ACh interneurons during feeding behavior as can be seen in Fig. 2 (Avena et al., 2006). Several anomalous behavioral results question the idea that ACh serves as a behavior-inhibition signal in the NAc, but there are alternative explanations for these results. Carbachol, a nonspeciﬁc cholinergic agonist, is self-administered directly into the accumbens (Ikemoto et al., 1998). In a feeding paradigm, local injection of the nonspeciﬁc muscarinic antagonist, scopolamine, reduces lever pressing for food and sucrose consumption and increases locomotor activity (Pratt and Kelley, 2004; Kelley et al., 2005; Pratt and Kelley, 2005). In these experiments a nonspeciﬁc agonist and antagonist, were used, making the interpretation difﬁcult. For instance, brain microdialysis has shown that local infusion of scopolamine into the accumbens increases ACh release, probably by antagonizing M2 presynaptic autoreceptors (Chau et al., 1999, 2001), and local infusion of carbachol not only decreases ACh release but also increases DA release, possibly explaining why rats would self-administer this drug (Fig. 3). II.A.1.c. Glutamate and GABA in the nucleus accumbens. Evidence has been accumulating for roles of accumbens glutamate (GLU) in reward, motivation and novelty (Saulskaya and Mikhailova, 2004; Kalivas and Volkow, 2005). Local injection in the NAc of an AMPA/kainate antagonist is sufﬁcient to stimulate feeding in satiated rats (Maldonado-Irizarry et al., 1995). This was later corroborated by microdialysis, showing that free feeding in food-deprived rats signiﬁcantly decreased GLU release in the NAc (Rada et al.,

358

Fig. 2. Changes in extracellular levels of ACh in the NAc of sham-feeding and real-feeding rats. Acetylcholine only increased in realfeeding rats during sugar intake. Bar graphs indicate the amount of sugar consumed in the sham-feeding rats (black bars) compared with real-feeding rats (open bars). Asterisks indicate po0.05. Adapted from Avena et al., 2006.

Fig. 3. A 40-min infusion of carbachol (100 mM) in the accumbens by reverse microdialysis simultaneously decreased extracellular ACh and increased DA levels in the NAc. Asterisks indicate po0.05.

1997). Intraaccumbal injection of raclopride, a D2 antagonist, prevents the GLU decrease (Saulskaya and Mikhailova, 2002). Glutamate is released in the NAc following the presentation of an inedible object when the rat expects food (Saulskaya and Mikhailova, 2002) and also during the presentation of a conditioned aversive stimuli (Saulskaya

and Mikhailova, 2004), suggesting that GLU is released in the NAc when the motivational value of the reward is changed or aversive. Several pharmacological studies show that local injections of GABA agonists into the NAc increase food intake (Reynolds and Berridge, 2002; Hanlon et al., 2004); however, presently there are no

359

known microdialysis studies that have measured the effect of feeding on GABA release. II.A.2.a. Ventral tegmental area microdialysis during feeding. This brain area is of utmost importance since DA neurons, forming the mesolimbic system, originate in the VTA (Fallon and Moore, 1978). These DA neurons are known to be under the inﬂuence of other neurotransmitter systems including GLU (Taber and Fibiger, 1997; Westerink et al., 1997; Floresco et al., 2001; Harris and Aston-Jones, 2003), GABA (Cruz et al., 2004; Ye et al., 2004), and the opioids (Cowen and Lawrence, 1999). II.A.2.b. Dopamine in the ventral tegmental area. Dopamine is not only released in its terminal region, the NAc, but also in its dendritic area. While the majority of research has focused on the NAc, studies have shown that extracellular DA increases in the VTA in response to opioids (Klitenick et al., 1992; Yoshida et al., 1993). With regard to feeding, extracellular DA increases in the VTA while an animal eats (Yoshida et al., 1992). II.A.2.c. Acetylcholine in the ventral tegmental area. The VTA receives cholinergic inputs from the pedunculopontine nuclei (Woolf, 1991). Infusion of nicotine directly into the VTA by reverse microdialysis stimulates DA release in the NAc (Nisell et al., 1994). Muscarinic receptors in the VTA have also been shown to modulate feeding and drinking behavior. For instance, local injection of a muscarinic antagonist suppresses feeding and ICSS (Rada et al., 2000; Sharf and Ranaldi, 2006). Conversely, ICSS stimulates ACh release in the VTA (Rada et al., 2000). In this brain area, ACh seems to facilitate behavior by directly stimulating DA neurons that release DA in the NAc (Forster and Blaha, 2000; Yeomans et al., 2001). II.A.3. Hypothalamic microdialysis during feeding In the following paragraphs we will discuss some of the neurochemical ﬁndings using brain microdialysis to measure amines and amino acids in the hypothalamic region that have important links to the reward circuitry.

II.A.3.a. Dopamine in the hypothalamus. Dopamine in the hypothalamus is thought to play a very different role in feeding than in the NAc. Early studies suggested that DA in the LH might be involved in the anorectic effect of amphetamine (Leibowitz, 1975). This was later conﬁrmed when a local LH injection of sulpiride, a relatively speciﬁc D2 antagonist, was sufﬁcient to induce feeding and drinking (Parada et al., 1988) and, with repeated injections, obesity (Baptista, 1999). DA in the LH may be essential in locomotion related to food and water seeking (Parada et al., 1990). Microdialysis studies have found that eating induces a signiﬁcant increase in DA levels in the LH, with no change observed in rats fed intragastrically, suggesting that oropharyngeal stimulation is important (Yang et al., 1996). However, in this report samples were taken every 20 min, so it is difﬁcult to know if the DA increase was signaling satiety. DA increase is directly correlated with the meal size and it has been suggested that obese Zucker rats may have an inherently higher DA ‘‘threshold’’ level for satiety in the LH (Yang and Meguid, 1995). It is difﬁcult to know if DA signals satiety and obese rats have a higher threshold or whether obese rats eat more because they release more DA in the LH. Behavioral experiments support the DA satiety explanation since DA agonists act as anorectics when injected into the LH (Leibowitz, 1975). Moreover, rats self-administer a DA antagonist directly into the LH and this releases DA in the NAc, thus linking LH DA to the inhibition of NAc DA reinforcement (Parada et al., 1995). The ventromedial hypothalamus (VMH) usually has an opposing response to that observed in the LH. The VMH sends GABAergic projections to the LH that may inhibit feeding (Beverly and Martin, 1989). The VMH also receives dopaminergic inputs, which respond in the opposite manner to the LH, with a decrease in extracellular DA concentration during a meal (Yang et al., 1997), which depends directly on the size of the meal (Meguid et al., 1997). The DA decrease in the VMH depends in part on oropharyngeal stimulation (Yang et al., 1997). In summary, DA release in the hypothalamus is involved in feeding behavior and the LH and VMH probably have opposing dopaminergic functions. The source of DA

360

in the hypothalamus could be local DA cell clusters as well as the mesolimbic system (Fuxe and Ungerstedt, 1968). II.A.3.b. Norepinephrine in the hypothalamus. Extracellular norepinephrine (NE) in the hypothalamus follows a circadian rhythm, suggesting that it plays an important role in the animal’s overall state of arousal (Margules et al., 1972; Jacobs and Chan, 1987; Stanley et al., 1989). The earliest studies found that NE in the paraventricular nucleus (PVN), can enhance food intake (Grossman, 1960; Leibowitz, 1970, 1972). Microdialysis studies found that NE is released in the PVN at the beginning of the active feeding period coinciding with the onset of the active cycle of the rat (dark onset) (Hoebel et al., 1989; Stanley et al., 1989; Mitome, 1994; Morien et al., 1995; Tachibana et al., 2000, 2001). It was later found that PVN NE increases in satiated rats during a large meal at dark onset, and it also increases at the start of the dark cycle in food-deprived rats given a carbohydrate meal (Paez et al., 1993). In rats maintained on a restricted schedule, NE levels rise just before the meal (Mitome et al., 1994). Further studies have looked at the effect of various drugs or peptides, which can modify ingestive behavior, on hypothalamic NE. For instance, galanin (GAL) and neuropeptide-Y (NPY) are both peptides that, if injected in the PVN, can induce feeding (Kyrkouli et al., 1990), and also increase NE levels in the PVN in rats if food is present (Kyrkouli et al., 1992). If food is not present, GAL still increases NE, but NPY decreases it (Kyrkouli et al., 1992). Intraventricular injection of NPY increases both food intake and NE in the PVN (Matos et al., 1996). Alpha-adrenergic antagonists are capable of blocking the GAL- but not the NPY feeding response (Kyrkouli et al., 1990). These results are consistent with previous behavioral studies showing that GAL-induced eating is probably mediated through the NE system. Anorectic drugs modify NE release in the hypothalamus. For instance, systemic injection of phenylpropanolamine, an alpha-1 adrenoceptor agonist, suppresses food intake in rats and simultaneously decreases extracellular levels of NE in the PVN (Davies et al., 1993). In contrast, the

alpha-2 blocker, idazoxan, also suppresses food intake in the rat, but instead of the expected decrease in NE levels in the PVN an increase occurs, which may be mediated through a presynaptic autoreceptor (Paez and Leibowitz, 1993). These results illustrate the importance of microdialysis in recognizing possible pre- and postsynaptic mechanisms of action for various peptides and drugs that modulate ingestive behavior. Few studies have looked at the effect of obesity on NE levels in the hypothalamus, although it is known that chronic infusions of NE into the VMH or the PVN induce hyperphagia and obesity (Leibowitz et al., 1984; Cincotta et al., 2000). One model of obesity uses male offspring of female rats that were undernourished during the ﬁrst two trimesters of pregnancy or had been injected with insulin during the third trimester (Jones and Friedman, 1982; Jones and Dayries, 1990). Microdialysis of the medial hypothalamus in the obese offspring showed a signiﬁcant elevation in extracellular NE levels compared with control rats (Jones et al., 1995). Thus, NE may play a role in feeding or body weight regulation in this model of gestation-linked obesity. II.A.3.c. Histamine in the hypothalamus. Pharmacological studies have demonstrated that local hypothalamic histamine modulates food intake (Ookuma et al., 1989). This was conﬁrmed by locally manipulating histamine levels in the PVN and VMH using an inhibitor of the synthetic enzyme histidine decarboxylase. Experimentally decreasing histamine levels induces feeding, but only at the start of the light cycle and only in the PVN or VMH, not in the LH or dorsomedial hypothalamus (Ookuma et al., 1993). Histamine may modulate ingestive behavior in the hypothalamus by interacting with the noradrenergic system. Local injection of a histamine H1 receptor antagonist increases feeding and extracellular NE, and this effect is blocked by a speciﬁc alpha-2 adrenoceptor antagonist (Kurose and Terashima, 1999). II.A.3.d. Serotonin in the hypothalamus. Medial hypothalamic (MH) injection of serotonin (5-HT) or its agonists inhibit feeding behavior (Leibowitz, 1986; Leibowitz et al., 1987, 1988). Initial

361

microdialysis studies of 5-HT during a meal demonstrated that 5-HT increases in the LH and also in the medial hypothalamus in anticipation of a meal when smelling food, as well as during the meal (Schwartz et al., 1990). This suggests that 5-HT may play a role in the response to foodrelated appetitive stimuli and then contribute to satiety when postingestional factors release CCK or related satiety signals. Later studies demonstrated that 5-HT increases if the rat is given a carbohydrate meal, and this increase is detected 15 min after the start of the meal, possibly contributing to satiety. In this same report, it was found that a protein or fat meal could decrease 5-HT levels (Rouch et al., 1999). This may relate in part to tryptophan uptake for 5-HT synthesis after a carbohydrate meal (Fernstrom and Wurtman, 1974). Other researchers have found changes in hypothalamic 5-HT levels in normal and obese rats consistent with the theory that 5-HT can act as a satiety signal (Mori et al., 1999; Fetissov et al., 2000; De Fanti et al., 2001). Several groups have investigated the effect of satiety peptides on hypothalamic 5-HT release. Enterostatin or leptin can increase extracellular 5-HT in the LH (Koizumi and Kimura, 2002; Telles et al., 2003). Conversely, Nicolaidis, Orosco and collaborators suggest that 5-HT causes the release of insulin in the PVNVMH region (Rouch et al., 1999; Orosco et al., 2000). These microdialysis studies point to a role for 5-HT in modulating circuits that control food intake, including a strong synergistic effect with postingestional satiety factors. II.A.3.e. Acetylcholine in the hypothalamus. In the LH several studies have demonstrated the potentiation of eating and drinking water following local injection of a muscarinic agonist (Grossman, 1960; De Parada et al., 2000). However, so far there are no known published microdialysis studies of ACh release in the LH during feeding behavior. II.A.3.f. Glutamate and GABA in the hypothalamus. Retrograde labeling shows that GLU inputs to the LH originate in the frontal cortex, AMYG, NAc, preoptic area, SN, VTA, parabraquial nuclei, and the nucleus of the solitary tract (Duva et al., 2005). Glutamate has been shown

to induce feeding when injected locally in the LH (Stanley et al., 1993a, b; Khan et al., 1999; Duva et al., 2002). Moreover, injection of NMDA into the LH induces eating without affecting locomotion (Duva et al., 2001, 2002). Extracellular levels of GLU in the LH increase at the beginning of a meal and then decrease by the end of the meal (Rada et al., 2003). It would seem that GLU initiates eating as a fast-acting neurotransmitter, and the maintenance of the behavior probably depends on other neurotransmitters. It has been suggested that there exists reciprocal connections between the medial and LH. In the medial hypothalamus GABA can induce eating (Beverly and Martin, 1989, 1990). Furthermore, following acute glucoprivation that would increase appetite, extracellular GABA increases in the VMH while an opposite response occurs in the LH (Beverly et al., 1995). Early studies injecting a GABA agonist into the LH showed an increase in feeding, while antagonists did the opposite (Kelly et al., 1977; Kelly and Grossman, 1979; Tsujii and Bray, 1991). However, later studies suggest a decrease in feeding when a GABA agonist is locally injected into the LH (Maldonado-Irizarry et al., 1995), but an antagonist does not initiate feeding (Stratford and Kelley, 1999). Monitoring GABA in the LH every 30 s during a meal revealed that GABA increased at the end of the meal, possibly signaling satiety (Rada et al., 2003). This could also explain the absence of response to induce feeding using antagonists in the LH, since GABA levels might be too low before the meal for an antagonist to show an effect at that time.

II.A.4. Hypothalamic accumbens connections There are both direct and indirect connections between the hypothalamus and the NAc (Kelley et al., 2005). Microdialysis has provided evidence that these sites may interact. Most, but not all, neurotransmitters or peptides that promote feeding when injected into the hypothalamus also stimulate DA and decrease ACh release in the NAc, while satiating peptides do the opposite (Leibowitz and Hoebel, 2004). Injection of the cholinergic agonist, carbachol or the DA antagonist, sulpiride, into the LH both increases food

362

intake and simultaneously increases DA release in the NAc (De Parada et al., 2000). Similarly, injection of NE or GAL into the PVN of the hypothalamus increases food consumption, which is correlated with a signiﬁcant increase in DA and decrease in ACh in the NAc (Hajnal et al., 1997; Rada et al., 1998a). Conversely, cholecystokinin (CCK) injected into the PVN decreases food ingestion while decreasing DA in the NAc (Helm et al., 2003). Cholecystokinin does not increase ACh unless it is injected simultaneously with 5-HT, which then produces a profound decrease in food intake along with increased ACh release (Helm et al., 2003). This suggests that under normal circumstances, postingestional factors would provide CCK release for interaction with 5-HT contributing to the decrease in DA and increase in ACh release in the NAc. In summary, as a general rule, the hypothalamus controls feeding behavior, in part, by modulating DA and ACh in the NAc (Hoebel et al., 1999). An exception to the hypothalamic accumbens rule is NPY. Injection of this peptide into the PVN induces eating in the rat, however, no changes in extracellular DA or ACh were detected (Rada et al., 1998a), suggesting that NPY can induce feeding by some other mechanism, or requires other cofactors that were not present.

II.B. Water intake Compared with the abundant data on food intake, there are relatively few brain microdialysis studies on water intake. It has been reported that drinking releases DA in the NAc and VTA (Yoshida et al., 1992; Young et al., 1992). Injection of a D2 receptor antagonist in perifornical LH can induce drinking and increase DA in the NAc (Parada et al., 1988, 1990). This activation of the DA mesolimbic system could mediate some of the rewarding aspects of drinking behavior. Similar to the effect seen with food intake, ACh increases in the VTA during drinking, and blockade with muscarinic receptors in the same area inhibits water intake (Rada et al., 2000). In the LH, extracellular ACh also increases during drinking, and the exogenous administration of cholinergic

drugs or D2 receptor blockers increases water intake (Puig de Parada et al., 1997). These microdialysis studies are in agreement with behavioral studies that show an increase in water intake following pharmacological manipulation of the cholinergic system with carbachol (Grossman, 1960). Angiotensin II injections into the subfornical organ induce drinking (McKinley et al., 2001) and release NE in the PVN and LH, even when water is not available (Gerstberger et al., 1992). However, this NE increase is attenuated if rats are allowed to drink (Ushigome et al., 2002; Tanaka et al., 2003). These studies indicate that a hypothalamic component of the noradrenergic system participates in drinking behavior. Angiotensin injection in the lateral ventricle releases DA in the NAc (Jones, 1986). This effect is enhanced if the rat is allowed to drink in response to the injection (Hoebel et al., 1994).

II.C. Mating The NAc participates in the control of sexual behavior in the same way that it does with other natural reinforcers. Microdialysis in the NAc of male and female rats shows an elevation of extracellular DA levels during sexual behavior (Damsma et al., 1992; Mas et al., 1995; Becker et al., 2001). In the NAc of sexually active male rats DA levels increase when a receptive female rat is presented, and increase even more during copulation (Pfaus et al., 1990; Pleim et al., 1990). Damsma and colleagues have examined the effects of locomotion, exposure to a novel chamber, sex odors, and sexual activity on DA transmission in the NAc and STR (Damsma et al., 1992). The DA increase seen during copulation is greater than following active locomotion (wheel running) or exposure to novel stimuli (mating chamber, fresh bedding, or soiled bedding). This increase was more intense in the NAc than the STR. Thus, neither novelty nor locomotion can account for the increase of DA in either area, suggesting that the anticipatory and consummatory aspects of sexual behavior are naturally occurring events in which reinforcement is likely mediated by DA release in the NAc.

363

Dopamine in the NAc and STR is also increased during sexual behavior in females, but only when the female controls the time of intromissions (Becker et al., 2001). The timing of copulatory stimuli is critical for the magnitude of the increase in accumbens DA. When intromissions are spaced 1–2 min apart and insemination would most likely result in pregnancy, DA signiﬁcantly increases in the NAc. This increase is not a passive response to coital stimuli or copulation-related motor activity, and possibly reﬂects other qualitative information about the copulatory stimuli. The medial preoptic area (MPOA) is located at the rostral end of the hypothalamus and is a critical integrative site for male sexual behavior in most vertebrate species. DA agonists in MPOA facilitate sexual behavior, while antagonists impair copulation, genital reﬂexes and sexual motivation (Mas et al., 1987; Bitran et al., 1988; Warner et al., 1991; Dominguez and Hull, 2005). Microdialysis in the MPOA has shown a DA increase during appetitive (noncontact exposure to sexual stimuli such as exposure to a receptive female) and consummatory (copulation) phases (Fumero et al., 1994; Dominguez et al., 2001; Triemstra et al., 2005). Recently, it was shown that GLU, by stimulating nitric oxide, is responsible for this increase in MPOA DA during copulation (Dominguez et al., 2004). Bilateral olfactory bulbectomy completely prevents mating in male rodents; however, unilateral bulbectomy does not. Microdialysis performed in the MPOA of animals with unilateral bulbectomy during mating shows that DA increases only in the contralateral and not in the ipsilateral side, suggesting that somatosensory cues alone are not sufﬁcient to release DA in the MPOA during sexual behavior in the absence of chemosensory input (Triemstra et al., 2005). Sexually excitatory olfactory stimuli activate the medial AMYG, which in turn projects to the bed nucleus of the stria terminalis and the MPOA leading to DA release and facilitation of copulation (Kostarczyk, 1986; Gomez and Newman, 1992; Dominguez et al., 2001). Microdialysis in the anterior LH area suggests that 5-HT inhibits behaviors during the postcopulatory phase of male sexual behavior (Lorrain et al., 1999). Serotonin injection in the LH also

inhibits basal and female-induced DA release in the NAc. This suggests that the neural circuit promoting sexual quiescence during the postejaculatory interval include serotonergic input to the LH, which in turn inhibits DA release in the NAc. This fact may have relevance for understanding the sexual side effects common to antidepressants medications (Rudkin et al., 2004).

III. Artiﬁcial rewards III.A. Intracranial self-stimulation Olds and Milner (1954) discovered ICSS in the mid 1950s to produce positive reinforcement. Research using intracerebral microdialysis has focused mainly on how ICSS modulates DA release in the NAc. Most studies show that LH or VTA ICSS releases DA in the NAc (Hernandez and Hoebel, 1988a; Nakahara et al., 1989a, b; Phillips et al., 1992; Fiorino et al., 1993; You et al., 1998, 2001). Similarly, electrical brain stimulation of the PFC signiﬁcantly increases DA release in the NAc (Taber and Fibiger, 1995; You et al., 1998, 2001). This rise in NAc DA is intensity dependent and can be blocked with the excitatory amino acid antagonist kynurenic acid injected either into the VTA or directly into NAc, suggesting that the neural signal engages the VTA (You et al., 1998). In addition, like food reward, self-stimulation of the LH also increases ACh release in the VTA and an infusion of atropine, through reverse dialysis, completely blocks self-stimulation conﬁrming that this cholinergic system in the hindbrain is involved in activating the DA neurons in the VTA (Rada et al., 2000; Sharf and Ranaldi, 2006). As discussed in previous paragraphs, an electrolytic lesion of the LH produces aphagia and adipsia. Conversely, electrical stimulation of the LH induces feeding in satiated animals (Anand and Brobeck, 1951a, b; Hoebel and Teitelbaum, 1962; Valenstein et al., 1968; Hoebel, 1976). Using microdialysis to monitor DA levels in the NAc, an increase in DA levels was demonstrated following electrical stimulation of the perifornical LH that induced eating (Hernandez and Hoebel, 1988a; Rada et al., 1998b). This suggests that ICSS and

364

feeding share common circuitry and that DA in the NAc could be part of the signal. Using hypothalamic sites where electrical stimulation was either positively or negatively reinforcing, or both, DA was released in the NAc by automatic stimulation, self-stimulation or stimulation–escape responding. DA increases even during stimulation escape using a MH site that was purely aversive (Rada et al., 1998b). These results conﬁrm that DA in the NAc is not only involved in positive reinforcement, but in negative reinforcement as well. As cited in the prior section on ACh in the NAc, the aversive LH stimulation causes release of accumbens ACh. Stimulation–escape responding signiﬁcantly decreases extracellular ACh levels (Rada and Hoebel, 2001). This supports the theory that elevated ACh in the accumbens is aversive, and reducing it is part of the reward earned by stimulation–escape responses.

III.B. Drug reward Almost all drugs abused by humans increase DA in the NAc (Di Chiara and Imperato, 1988; Hernandez and Hoebel, 1988b; Pothos et al., 1991; Rada et al., 1991a, 2001; Tanda et al., 1997; Di Chiara, 1998; Koob et al., 1998; Hoebel et al., 1999) with the exception of benzodiazepines and barbiturates (Masuzawa et al., 2003; Rada and Hoebel, 2005). Withdrawal in contrast decreases DA release in the NAc (Parsons et al., 1991; Pothos et al., 1991; Weiss et al., 1992; Diana et al., 1993, Hildebrand et al., 1998; Rada et al., 2004) and, in several cases, increases ACh release (Rada et al., 1991b, 1996, 2001, 2004). Although the benzodiazepine, diazepam (Valium), does not release DA, its withdrawal does release ACh. On the theory that relatively high extracellular ACh is aversive, this could contribute to the use of diazepam for self-medication (Rada and Hoebel, 2005). Note that with natural satiety, ACh increases, but DA is ‘‘normal’’ or elevated. However, during drug withdrawal DA and ACh often respond in opposite directions, with a decrease in DA and increase in ACh. It was hypothesized that this

imbalance probably gives rise to an aversive state (Hoebel et al., 1999).

IV. Natural and artiﬁcial rewards: do they share common reward mechanisms and circuitry? IV.A. Sugar addiction Natural reward mechanisms presumably were selected in the wild to promote behaviors that are necessary for the survival of the species. In normal animals these reinforcers have opposing mechanisms that inhibit the behavior once the speciﬁc need has been satisﬁed. However, drugs of abuse activate the reward circuits without necessarily activating the inhibitory components of the regulatory system. A large body of evidence suggests that natural reinforcers and drugs of abuse share common reward circuitry. For example, both food and drug reinforcers increase extracellular DA in the NAc (Di Chiara and Imperato, 1988; Hernandez and Hoebel, 1988b; Radhakishun et al., 1988; Pothos et al., 1991; Rada et al., 1991a; Salamone, 1994; Wise et al., 1995a, b; Tanda and Di Chiara, 1998; Cappendijk et al., 1999; Acquas et al., 2002). These reinforcers seem to also share behavioral responses. For instance, sweet taste or morphine prolongs a meal. This can be blocked with naloxone, an opiate antagonist (Sclafani et al., 1982; Nader et al., 1994; Gosnell et al., 1996). Consumption of sugar can act as an analgesic by releasing endogenous opioids (Kanarek et al., 1991). Weight loss increases opiate-induced eating and also drug self-administration (Hagan and Moss, 1991; Specker et al., 1994; Cabeza de Vaca and Carr, 1998). An animal model of binge eating has been developed by the Hoebel laboratory to systematically study whether excessive sugar intake can elicit behavioral and neurochemical changes similar to those of drugs of abuse (Avena et al., in press). Several diagnostic criteria used to study drug abuse reveal that binge eating of palatable foods, a behavioral component of obesity, may have some addictive-like properties. For example, rats maintained on a diet of intermittent access to a sugar solution and chow gradually escalate their intake

365

of sugar over the course of one month and ‘‘binge’’ on the sugar when it becomes available each day (Colantuoni et al., 2001). These animals also have increased D1 and mu-opioid receptor binding, and D3 receptor mRNA in the NAc (Colantuoni et al., 2001; Spangler et al., 2004). Drugs of abuse are known to repeatedly increase DA release in the NAc without habituation of the response as seen with palatable food (Bassareo and Di Chiara, 1997). Sugar bingeing also repeatedly releases DA in the NAc (Rada et al., 2005), similar to addictive drugs (Fig. 4). A similar result is obtained when sugar-bingeing rats sham-feed during the binge, suggesting that the taste of sugar is sufﬁcient to release DA repeatedly in the NAc (Avena et al., 2006). Signs of withdrawal such as teeth chattering, grooming, anxiety, depression, and distress vocalization are found in sugar-bingeing rats. This is most noticeable when withdrawal is precipitated by an opioid antagonist, suggesting that the endogenous opioid system is altered by the excessive bingeing. During both naloxone precipitated and even during simple withdrawal of the sugar microdialysis reveals decreases in DA and increases ACh in the NAc (Colantuoni et al., 2002; Avena, unpublished), indicative of a drug-like withdrawal state (Pothos et al., 1991; Rada et al., 1991a, 1996). Craving-like behavior is seen in sugar-binge animals as they manifest increased intake after 2 weeks of abstinence, known as a ‘‘deprivation

effect’’ (Avena et al., 2005). After a month of abstinence the animals respond more than before for cues previously associated with sugar (Grimm et al., 2005). Evidence shows cross-sensitization between sugar bingeing and amphetamine-induced locomotion (Avena and Hoebel, 2003), cocaine-induced locomotion (Gosnell, 2005), and enhanced alcohol intake (Avena et al., 2004), suggesting that a common neural pathway, presumably DA, mediates these behaviors. Thus, each of these similarities suggests that binge eating on sugar results in a state qualitatively similar to drug abuse, and that this state persists and can foster future intake of sugar or drugs of abuse. Apparently a natural reinforcer such as sugar can change from a substance of use to a substance of abuse under certain conditions. All of these results point to a natural function for addiction that is usurped by the more powerful drugs of abuse.

V. Conclusions Brain microdialysis was invented by Ungerstedt (1984) and continues to be an immensely valuable technique for monitoring biochemicals and their metabolites in vivo. The measurement of release is one of the criteria for proving that a substance is a neurotransmitter. Moreover, measurement of

Fig. 4. Dopamine increases the ﬁrst time rats have access to sugar; however, this response habituates and disappears in following trials if rats have ad libitum access to the sugar. In contrast, rats receiving intermittent sugar show the same DA response every time, similar to drugs of addiction. Adapted from Rada et al., 2005.

366

release during behavior provides a critical piece of information in determining when and where the neurotransmitter-coded system is active. As faster time-sampling techniques are invented to match the rate of nerve impulse ﬂow, microdialysis will continue to have a place for verifying measurement of the actual neurotransmitter in relation to behavior and underlying neural processes that depend on volume conduction through the extracellular and ventricular ﬂuids of the brain.

Acknowledgments We would like to thank Miriam Bocarsly and Caroline Lee for their assistance with the preparation of this chapter.

References Abercrombie, E.D., Keefe, K.A., DiFrischia, D.S. and Zigmond, M.J. (1989) Differential effect of stress on in vivo dopamine release in striatum, nucleus accumbens, and medial frontal cortex. J. Neurochem., 52: 1655–1658. Acquas, E., Tanda, G. and Di Chiara, G. (2002) Differential effects of caffeine on dopamine and acetylcholine transmission in brain areas of drug-naive and caffeine-pretreated rats. Neuropsychopharmacology, 27: 182–193. Anand, B.K. and Brobeck, J.R. (1951a) Hypothalamic control of food intake in rats and cats. Yale J. Biol. Med., 24: 123–140. Anand, B.K. and Brobeck, J.R. (1951b) Localization of a ‘‘feeding center’’ in the hypothalamus of the rat. Proc. Soc. Exp. Biol. Med., 77: 323–324. Avena, N.M., Carrillo, C.A., Needham, L., Leibowitz, S.F. and Hoebel, B.G. (2004) Sugar-dependent rats show enhanced intake of unsweetened ethanol. Alcohol, 34: 203–209. Avena, N.M. and Hoebel, B.G. (2003) A diet promoting sugar dependency causes behavioral cross-sensitization to a low dose of amphetamine. Neuroscience, 122: 17–20. Avena, N.M., Long, K.A. and Hoebel, B.G. (2005) Sugardependent rats show enhanced responding for sugar after abstinence: evidence of a sugar deprivation effect. Physiol. Behav., 84: 359–362. Avena, N.M., Rada, P. and Hoebel, B.G. (in press) Sugar bingeing in rats. In: Crawley, J.N. et al. (Eds.), Current Protocols in Neuroscience. Wiley, New York. Avena, N.M., Rada, P., Moise, N. and Hoebel, B.G. (2006) Sucrose sham-feeding on a binge schedule releases accumbens dopamine repeatedly and eliminates the acetlycholine satiety response. Neuroscience, 139: 813–820.

Bailey, C.S., Hsiao, S. and King, J.E. (1986) Hedonic reactivity to sucrose in rats: modiﬁcation by pimozide. Physiol. Behav., 38: 447–452. Bals-Kubik, R., Herz, A. and Shippenberg, T.S. (1989) Evidence that the aversive effects of opioid antagonists and kappa-agonists are centrally mediated. Psychopharmacology (Berl.), 98: 203–206. Baptista, T. (1999) Body weight gain induced by antipsychotic drugs: mechanisms and management. Acta Psychiatr. Scand., 100: 3–16. Bassareo, V. and Di Chiara, G. (1997) Differential inﬂuence of associative and nonassociative learning mechanisms on the responsiveness of prefrontal and accumbal dopamine transmission to food stimuli in rats fed ad libitum. J. Neurosci., 17: 851–861. Bassareo, V. and Di Chiara, G. (1999a) Differential responsiveness of dopamine transmission to food-stimuli in nucleus accumbens shell/core compartments. Neuroscience, 89: 637–641. Bassareo, V. and Di Chiara, G. (1999b) Modulation of feedinginduced activation of mesolimbic dopamine transmission by appetitive stimuli and its relation to motivational state. Eur. J. Neurosci., 11: 4389–4397. Bassareo, V., De Luca, M.A. and Di Chiara, G. (2002) Differential expression of motivational stimulus properties by dopamine in nucleus accumbens shell versus core and prefrontal cortex. J. Neurosci., 22: 4709–4719. Becker, J.B., Rudick, C.N. and Jenkins, W.J. (2001) The role of dopamine in the nucleus accumbens and striatum during sexual behavior in the female rat. J. Neurosci., 21: 3236–3241. Berridge, K.C. and Robinson, T.E. (1998) What is the role of dopamine in reward: hedonic impact, reward learning, or incentive salience? Brain Res. Brain Res. Rev., 28: 309–369. Berthoud, H.R. (2000) An overiew of neural pathways and networks involved in the control of food intake and selection. In: Berthoud, H.R. and Seeley, R.J. (Eds.), Neural and Metabolic Control of Macronutrient Intake. CRC Press, New York. Berthoud, H.R. (2004) Mind versus metabolism in the control of food intake and energy balance. Physiol. Behav., 81: 781–793. Beverly, J.L., Beverly, M.F. and Meguid, M.M. (1995) Alterations in extracellular GABA in the ventral hypothalamus of rats in response to acute glucoprivation. Am. J. Physiol., 269: R1174–R1178. Beverly, J.L. and Martin, R.J. (1989) Increased GABA shunt activity in VMN of three hyperphagic rat models. Am. J. Physiol., 256: R1225–R1231. Beverly, J.L. and Martin, R.J. (1990) Inﬂuence of serum glucose on glutamate decarboxylase activity in the ventromedial nucleus of rats. Am. J. Physiol., 258: R697–R703. Bitran, D., Hull, E.M., Holmes, G.M. and Lookingland, K.J. (1988) Regulation of male rat copulatory behavior by preoptic incertohypothalamic dopamine neurons. Brain Res. Bull., 20: 323–331. Bolam, J.P., Wainer, B.H. and Smith, A.D. (1984) Characterization of cholinergic neurons in the rat neostriatum.

367 A combination of choline acetyltransferase immunocytochemistry, Golgi-impregnation and electron microscopy. Neuroscience, 12: 711–718. Bourne, J.A. (2003) Intracerebral microdialysis: 30 years as a tool for the neuroscientist. Clin. Exp. Pharmacol. Physiol., 30: 16–24. Cabeza de Vaca, S. and Carr, K.D. (1998) Food restriction enhances the central rewarding effect of abused drugs. J. Neurosci., 18: 7502–7510. Cadoni, C. and Di Chiara, G. (1999) Reciprocal changes in dopamine responsiveness in the nucleus accumbens shell and core and in the dorsal caudate-putamen in rats sensitized to morphine. Neuroscience, 90: 447–455. Cappendijk, S.L., Hurd, Y.L., Nylander, I., van Ree, J.M. and Terenius, L. (1999) A heroin-, but not a cocaine-expecting, self-administration state preferentially alters endogenous brain peptides. Eur. J. Pharmacol., 365: 175–182. Carboni, E., Imperato, A., Perezzani, L. and Di Chiara, G. (1989) Amphetamine, cocaine, phencyclidine and nomifensine increase extracellular dopamine concentrations preferentially in the nucleus accumbens of freely moving rats. Neuroscience, 28: 653–661. Chau, D.T., Rada, P., Kosloff, R.A., Taylor, J.L. and Hoebel, B.G. (2001) Nucleus accumbens muscarinic receptors in the control of behavioral depression: antidepressant-like effects of local M1 antagonist in the Porsolt swim test. Neuroscience, 104: 791–798. Chau, D., Rada, P.V., Kosloff, R.A. and Hoebel, B.G. (1999) Cholinergic, M1 receptors in the nucleus accumbens mediate behavioral depression. A possible downstream target for ﬂuoxetine. Ann. N.Y. Acad. Sci., 877: 769–774. Chen, H., Zhang, S.M., Hernan, M.A., Willett, W.C. and Ascherio, A. (2003) Weight loss in Parkinson’s disease. Ann. Neurol., 53: 676–679. Cincotta, A.H., Luo, S., Zhang, Y., Liang, Y., Bina, K.G., Jetton, T.L. and Scislowski, P.W. (2000) Chronic infusion of norepinephrine into the VMH of normal rats induces the obese glucose-intolerant state. Am. J. Physiol. Regul. Integr. Comp. Physiol., 278: R435–R444. Colantuoni, C., Rada, P., McCarthy, J., Patten, C., Avena, N.M., Chadeayne, A. and Hoebel, B.G. (2002) Evidence that intermittent, excessive sugar intake causes endogenous opioid dependence. Obes. Res., 10: 478–488. Colantuoni, C., Schwenker, J., McCarthy, J., Rada, P., Ladenheim, B., Cadet, J.L., Schwartz, G.J., Moran, T.H. and Hoebel, B.G. (2001) Excessive sugar intake alters binding to dopamine and mu-opioid receptors in the brain. Neuroreport, 12: 3549–3552. Cowen, M.S. and Lawrence, A.J. (1999) The role of opioid– dopamine interactions in the induction and maintenance of ethanol consumption. Prog. Neuropsychopharmacol. Biol. Psychiatry, 23: 1171–1212. Cruz, H.G., Ivanova, T., Lunn, M.L., Stoffel, M., Slesinger, P.A. and Luscher, C. (2004) Bi-directional effects of GABA(B) receptor agonists on the mesolimbic dopamine system. Nat. Neurosci., 7: 153–159. Damsma, G., Pfaus, J.G., Wenkstern, D., Phillips, A.G. and Fibiger, H.C. (1992) Sexual behavior increases dopamine

transmission in the nucleus accumbens and striatum of male rats: comparison with novelty and locomotion. Behav. Neurosci., 106: 181–191. Davies, B.T., Wellman, P.J. and Morien, A. (1993) An assessment of the involvement of paraventricular hypothalamic alpha2-adrenoceptors in phenylpropanolamine anorexia. Physiol. Behav., 54: 121–128. De Fanti, B.A., Hamilton, J.S. and Horwitz, B.A. (2001) Mealinduced changes in extracellular 5-HT in medial hypothalamus of lean (Fa/Fa) and obese (fa/fa) Zucker rats. Brain Res., 902: 164–170. de Olmos, J.S. and Heimer, L. (1999) The concepts of the ventral striatopallidal system and extended amygdala. Ann. N.Y. Acad. Sci., 877: 1–32. De Parada, M.P., Parada, M.A., Rada, P., Hernandez, L. and Hoebel, B.G. (2000) Dopamine–acetylcholine interaction in the rat lateral hypothalamus in the control of locomotion. Pharmacol. Biochem. Behav., 66: 227–234. Di Chiara, G. (1998) A motivational learning hypothesis of the role of mesolimbic dopamine in compulsive drug use. J. Psychopharmacol., 12: 54–67. Di Chiara, G. (2002) Nucleus accumbens shell and core dopamine: differential role in behavior and addiction. Behav. Brain Res., 137: 75–114. Di Chiara, G. and Imperato, A. (1988) Drugs abused by humans preferentially increase synaptic dopamine concentrations in the mesolimbic system of freely moving rats. Proc. Natl. Acad. Sci. U.S.A., 85: 5274–5278. Diana, M., Pistis, M., Carboni, S., Gessa, G.L. and Rossetti, Z.L. (1993) Profound decrement of mesolimbic dopaminergic neuronal activity during ethanol withdrawal syndrome in rats: electrophysiological and biochemical evidence. Proc. Natl. Acad. Sci. U.S.A., 90: 7966–7969. Dominguez, J.M. and Hull, E.M. (2005) Dopamine, the medical preoptic area, and male sexual behavior. Physiol. Behav., 86: 356–368. Dominguez, J.M., Muschamp, J.W., Schmich, J.M. and Hull, E.M. (2004) Nitric oxide mediates glutamate-evoked dopamine release in the medial preoptic area. Neuroscience, 125: 203–210. Dominguez, J., Riolo, J.V., Xu, Z. and Hull, E.M. (2001) Regulation by the medial amygdala of copulation and medial preoptic dopamine release. J. Neurosci., 21: 349–355. Duva, M.A., Tomkins, E.M., Moranda, L.M., Kaplan, R., Sukhaseum, A. and Stanley, B.G. (2005) Origins of lateral hypothalamic afferents associated with N-methyl-D-aspartic acid-elicited eating studied using reverse microdialysis of NMDA and ﬂuorogold. Neurosci. Res., 52: 95–106. Duva, M.A., Tomkins, E.M., Moranda, L.M., Kaplan, R., Sukhaseum, A., Bernardo, J.P. and Stanley, B.G. (2002) Regional differences in feeding and other behaviors elicited by N-methyl-D-aspartic acid in the rodent hypothalamus: a reverse microdialysis mapping study. Brain Res., 925: 141–147. Duva, M.A., Tomkins, E.M., Moranda, L.M., Kaplan, R., Sukhaseum, A., Jimenez, A. and Stanley, B.G. (2001) Reverse microdialysis of N-methyl-D-aspartic acid into the

368 lateral hypothalamus of rats: effects on feeding and other behaviors. Brain Res., 921: 122–132. Ettenberg, A. (1989) Dopamine, neuroleptics and reinforced behavior. Neurosci. Biobehav. Rev., 13: 105–111. Ettenberg, A. and Camp, C.H. (1986) A partial reinforcement extinction effect in water-reinforced rats intermittently treated with haloperidol. Pharmacol. Biochem. Behav., 25: 1231–1235. Fallon, J.H. and Moore, R.Y. (1978) Catecholamine innervation of the basal forebrain. IV. Topography of the dopamine projection to the basal forebrain and neostriatum. J. Comp. Neurol., 180: 545–580. Fernstrom, J.D. and Wurtman, R.J. (1974) Control of serotonin levels by the diet. Adv. Biochem. Psychopharmacol., 11: 133–142. Fetissov, S.O., Meguid, M.M., Chen, C. and Miyata, G. (2000) Synchronized release of dopamine and serotonin in the medial and lateral hypothalamus of rats. Neuroscience, 101: 657–663. Fibiger, H.C. (1984) The neurobiological substrates of depression in Parkinson’s disease: a hypothesis. Can. J. Neurol. Sci., 11: 105–107. Fillenz, M. (2005) In vivo neurochemical monitoring and the study of behaviour. Neurosci. Biobehav. Rev., 29: 949–962. Fiorino, D.F., Coury, A., Fibiger, H.C. and Phillips, A.G. (1993) Electrical stimulation of reward sites in the ventral tegmental area increases dopamine transmission in the nucleus accumbens of the rat. Behav. Brain Res., 55: 131–141. Floresco, S.B., Todd, C.L. and Grace, A.A. (2001) Glutamatergic afferents from the hippocampus to the nucleus accumbens regulate activity of ventral tegmental area dopamine neurons. J. Neurosci., 21: 4915–4922. Forster, G.L. and Blaha, C.D. (2000) Laterodorsal tegmental stimulation elicits dopamine efﬂux in the rat nucleus accumbens by activation of acetylcholine and glutamate receptors in the ventral tegmental area. Eur. J. Neurosci., 12: 3596–3604. Fumero, B., Fernandez-Vera, J.R., Gonzalez-Mora, J.L. and Mas, M. (1994) Changes in monoamine turnover in forebrain areas associated with masculine sexual behavior: a microdialysis study. Brain Res., 662: 233–239. Fuxe, K. and Ungerstedt, U. (1968) Histochemical studies on the distribution of catecholamines and 5-hydroxytryptamine after intraventricular injections. Histochemie, 13: 16–28. Galosi, R., Hajnal, A. and Lenard, L. (1997) Nucleus accumbens and body weight regulation: neurochemical lesions. Neurobiology, 5: 67–70. Gerstberger, R., Muller, A.R. and Simon-Oppermann, C. (1992) Functional hypothalamic angiotensin II and catecholamine receptor systems inside and outside the bloodbrain barrier. Prog. Brain Res., 91: 423–433. Gomez, D.M. and Newman, S.W. (1992) Differential projections of the anterior and posterior regions of the medial amygdaloid nucleus in the Syrian hamster. J. Comp. Neurol., 317: 195–218. Gosnell, B.A. (2005) Sucrose intake enhances behavioral sensitization produced by cocaine. Brain Res., 1031: 194–201.

Gosnell, B.A., Yracheta, J.M., Bell, S.M. and Lane, K.E. (1996) Intravenous self-administration of cathinone by rats. Behav. Pharmacol., 7: 526–531. Grewaald, D.S., Fibiger, H.C. and McGeer, E.G. (1974) 6-Hydroxydopamine and striatal acetylcholine levels. Brain Res., 73: 372–375. Grimm, J.W., Fyall, A.M. and Osincup, D.P. (2005) Incubation of sucrose craving: effects of reduced training and sucrose pre-loading. Physiol. Behav., 84: 73–79. Grossman, S.P. (1960) Eating or drinking elicited by direct adrenergic or cholinergic stimulation of hypothalamus. Science, 132: 301–302. Hagan, M.M. and Moss, D.E. (1991) An animal model of bulimia nervosa: opioid sensitivity to fasting episodes. Pharmacol. Biochem. Behav., 39: 421–422. Hajnal, A. and Norgren, R. (2001) Accumbens dopamine mechanisms in sucrose intake. Brain Res., 904: 76–84. Hajnal, A., Mark, G.P., Rada, P.V., Lenard, L. and Hoebel, B.G. (1997) Norepinephrine microinjections in the hypothalamic paraventricular nucleus increase extracellular dopamine and decrease acetylcholine in the nucleus accumbens: relevance to feeding reinforcement. J. Neurochem., 68: 667–674. Hajnal, A., Smith, G.P. and Norgren, R. (2004) Oral sucrose stimulation increases accumbens dopamine in the rat. Am. J. Physiol. Regul. Integr. Comp. Physiol., 286: R31–R37. Hajnal, A., Szekely, M., Galosi, R. and Lenard, L. (2000) Accumbens cholinergic interneurons play a role in the regulation of body weight and metabolism. Physiol. Behav., 70: 95–103. Hanlon, E.C., Baldo, B.A., Sadeghian, K. and Kelley, A.E. (2004) Increases in food intake or food-seeking behavior induced by GABAergic, opioid, or dopaminergic stimulation of the nucleus accumbens: is it hunger? Psychopharmacology (Berl.), 172: 241–247. Harris, G.C. and Aston-Jones, G. (2003) Critical role for ventral tegmental glutamate in preference for a cocaineconditioned environment. Neuropsychopharmacology, 28: 73–76. Harris, G.C., Wimmer, M. and Aston-Jones, G. (2005) A role for lateral hypothalamic orexin neurons in reward seeking. Nature, 437: 556–559. Helm, K.A., Rada, P. and Hoebel, B.G. (2003) Cholecystokinin combined with serotonin in the hypothalamus limits accumbens dopamine release while increasing acetylcholine: a possible satiation mechanism. Brain Res., 963: 290–297. Hernandez, L. and Hoebel, B.G. (1988a) Feeding and hypothalamic stimulation increase dopamine turnover in the accumbens. Physiol. Behav., 44: 599–606. Hernandez, L. and Hoebel, B.G. (1988b) Food reward and cocaine increase extracellular dopamine in the nucleus accumbens as measured by microdialysis. Life Sci., 42: 1705–1712. Hernandez, L., Baptista, T. and Hoebel, B.G. (1990) Neurochemical effects of chronic haloperidol and lithium assessed by brain microdialysis in rats. Prog. Neuropsychopharmacol. Biol. Psychiatry, 14: S17–S35.

369 Hernandez, L., Paez, X. and Hamlin, C. (1983) Neurotransmitters extraction by local intracerebral dialysis in anesthetized rats. Pharmacol. Biochem. Behav., 18: 159–162. Hernandez, L., Parada, M., Baptista, T., Schwartz, D., West, H.L., Mark, G.P. and Hoebel, B.G. (1991) Hypothalamic serotonin in treatments for feeding disorders and depression as studied by brain microdialysis. J. Clin. Psychiatry, 52(Suppl.): 32–40. Hernandez, L., Stanley, B.G. and Hoebel, B.G. (1986) A small, removable microdialysis probe. Life Sci., 39: 2629–2637. Hildebrand, B.E., Nomikos, G.G., Hertel, P., Schilstrom, B. and Svensson, T.H. (1998) Reduced dopamine output in the nucleus accumbens but not in the medial prefrontal cortex in rats displaying a mecamylamine-precipitated nicotine withdrawal syndrome. Brain Res., 779: 214–225. Hoebel, B.G. (1976) Brain-stimulation reward and aversion in relation to behavior. In: Wauquier, A. and Rolls, E.T. (Eds.), Brain-Stimulation Reward. North Holland Publisher, Amsterdam, pp. 335–372. Hoebel, B.G. and Teitelbaum, P. (1962) Hypothalamic control of feeding and self-stimulation. Science, 135: 375–377. Hoebel, B.G., Hernandez, L., Schwartz, D.H., Mark, G.P. and Hunter, G.A. (1989) Microdialysis studies of brain norepinephrine, serotonin, and dopamine release during ingestive behavior. Theoretical and clinical implications. Ann. N.Y. Acad. Sci., 575: 171–191 discussion 192–173. Hoebel, B.G., Leibowitz, S.F. and Hernandez, L. (1992) Neurochemistry of anorexia and bulimia. In: Anderson, H. (Ed.), The Biology of Feast and Famine: Relevance to Eating Disorders. Academic Press, New York, pp. 21–45. Hoebel, B.G., Rada, P., Mark, G.P. and Hernandez, L. (1994) The power of integrative peptides to reinforce behavior by releasing dopamine. In: Strand, F.L., et al. (Eds.), Models of Neuropeptide Action. Annals of the New York Academy of Sciences, New York, pp. 36–41. Hoebel, B.G., Rada, P., Mark, G.P. and Pothos, E. (1999) Neural systems for reinforcement and inhibition of behavior: relevance to eating, addiction, and depression. In: Kahneman, D., et al. (Eds.), Well-Being: The Foundations of Hedonic Psychology. Russell Sage Foundation, New York, pp. 558–572. Horvitz, J.C., Richardson, W.B. and Ettenberg, A. (1993) Dopamine receptor blockade and reductions in thirst produce differential effects on drinking behavior. Pharmacol. Biochem. Behav., 45: 725–728. Hurd, Y.L., Svensson, P. and Ponten, M. (1999) The role of dopamine, dynorphin, and CART systems in the ventral striatum and amygdala in cocaine abuse. Ann. N.Y. Acad. Sci., 877: 499–506. Ikemoto, S., Glazier, B.S., Murphy, J.M. and McBride, W.J. (1998) Rats self-administer carbachol directly into the nucleus accumbens. Physiol. Behav., 63: 811–814. Invernizzi, R.W. and Garattini, S. (2004) Role of presynaptic alpha2-adrenoceptors in antidepressant action: recent ﬁndings from microdialysis studies. Prog. Neuropsychopharmacol. Biol. Psychiatry, 28: 819–827.

Isella, V., Iurlaro, S., Piolti, R., Ferrarese, C., Frattola, L., Appollonio, I., Melzi, P. and Grimaldi, M. (2003) Physical anhedonia in Parkinson’s disease. J. Neurol. Neurosurg. Psychiatry, 74: 1308–1311. Jacobs, W.R. and Chan, Y.L. (1987) Effect of norepinephrine on renal tubular Na-K-ATPase and oxygen consumption. Life Sci., 40: 1571–1578. Jones, A.P. and Dayries, M. (1990) Maternal hormone manipulations and the development of obesity in rats. Physiol. Behav., 47: 1107–1110. Jones, A.P. and Friedman, M.I. (1982) Obesity and adipocyte abnormalities in offspring of rats undernourished during pregnancy. Science, 215: 1518–1519. Jones, A.P., Pothos, E.N., Rada, P., Olster, D.H. and Hoebel, B.G. (1995) Maternal hormonal manipulations in rats cause obesity and increase medial hypothalamic norepinephrine release in male offspring. Brain Res. Dev. Brain Res., 88: 127–131. Jones, D.L. (1986) Central integration of cardiovascular and drinking responses elicited by central administration of angiotensin II: divergence of regulation by the ventral tegmental area and nucleus accumbens. Can. J. Physiol. Pharmacol., 64: 1011–1016. Joseph, M.H., Datla, K. and Young, A.M. (2003) The interpretation of the measurement of nucleus accumbens dopamine by in vivo dialysis: the kick, the craving or the cognition? Neurosci. Biobehav. Rev., 27: 527–541. Kalivas, P.W. and Nakamura, M. (1999) Neural systems for behavioral activation and reward. Curr. Opin. Neurobiol., 9: 223–227. Kalivas, P.W. and Volkow, N.D. (2005) The neural basis of addiction: a pathology of motivation and choice. Am. J. Psychiatry, 162: 1403–1413. Kanarek, R.B., White, E.S., Biegen, M.T. and Marks-Kaufman, R. (1991) Dietary inﬂuences on morphine-induced analgesia in rats. Pharmacol. Biochem. Behav., 38: 681–684. Keefe, K.A., Sved, A.F., Zigmond, M.J. and Abercrombie, E.D. (1993) Stress-induced dopamine release in the neostriatum: evaluation of the role of action potentials in nigrostriatal dopamine neurons or local initiation by endogenous excitatory amino acids. J. Neurochem., 61: 1943–1952. Kelly, J., Alheid, G.F., Newberg, A. and Grossman, S.P. (1977) GABA stimulation and blockade in the hypothalamus and midbrain: effects on feeding and locomotor activity. Pharmacol. Biochem. Behav., 7: 537–541. Kelley, A.E., Bakshi, V.P., Fleming, S. and Holahan, M.R. (2000) A pharmacological analysis of the substrates underlying conditioned feeding induced by repeated opioid stimulation of the nucleus accumbens. Neuropsychopharmacology, 23: 455–467. Kelley, A.E., Baldo, B.A., Pratt, W.E. and Will, M.J. (2005) Corticostriatal-hypothalamic circuitry and food motivation: integration of energy, action and reward. Physiol. Behav., 86: 773–795. Kelly, J. and Grossman, S.P. (1979) GABA and hypothalamic feeding systems. II. A comparison of GABA, glycine and

370 acetylcholine agonists and their antagonists. Pharmacol. Biochem. Behav., 11: 647–652. Khan, A.M., Curras, M.C., Dao, J., Jamal, F.A., Turkowski, C.A., Goel, R.K., Gillard, E.R., Wolfsohn, S.D. and Stanley, B.G. (1999) Lateral hypothalamic NMDA receptor subunits NR2A and/or NR2B mediate eating: immunochemical/ behavioral evidence. Am. J. Physiol., 276: R880–R891. Klitenick, M.A., DeWitte, P. and Kalivas, P.W. (1992) Regulation of somatodendritic dopamine release in the ventral tegmental area by opioids and GABA: an in vivo microdialysis study. J. Neurosci., 12: 2623–2632. Koizumi, M. and Kimura, S. (2002) Enterostatin increases extracellular serotonin and dopamine in the lateral hypothalamic area in rats measured by in vivo microdialysis. Neurosci. Lett., 320: 96–98. Koob, G.F., Sanna, P.P. and Bloom, F.E. (1998) Neuroscience of addiction. Neuron, 21: 467–476. Korf, J. (1986) Release of endogenous amino acids, dopamine, and cyclic AMP from the rat brain: methodological aspects and mutual interferences. Ann. N.Y. Acad. Sci., 473: 418–433. Kostarczyk, E.M. (1986) The amygdala and male reproductive functions: I. Anatomical and endocrine bases. Neurosci. Biobehav. Rev., 10: 67–77. Kottegoda, S., Shaik, I. and Shippy, S.A. (2002) Demonstration of low ﬂow push–pull perfusion. J. Neurosci. Methods, 121: 93–101. Kurose, Y. and Terashima, Y. (1999) Histamine regulates food intake through modulating noradrenaline release in the paraventricular nucleus. Brain Res., 828: 115–118. Kyrkouli, S.E., Stanley, B.G., Hutchinson, R., Seiraﬁ, R.D. and Leibowitz, S.F. (1990) Peptide–amine interactions in the hypothalamic paraventricular nucleus: analysis of galanin and neuropeptide Y in relation to feeding. Brain Res., 521: 185–191. Kyrkouli, S.E., Stanley, B.G. and Leibowitz, S.F. (1992) Differential effects of galanin and neuropeptide Y on extracellular norepinephrine levels in the paraventricular hypothalamic nucleus of the rat: a microdialysis study. Life Sci., 51: 203–210. Leibowitz, S.F. (1970) Hypothalamic beta-adrenergic ‘‘satiety’’ system antagonizes an alpha-adrenergic ‘‘hunger’’ system in the rat. Nature, 226: 963–964. Leibowitz, S.F. (1972) Central adrenergic receptors and the regulation of hunger and thirst. Res. Publ. Assoc. Res. Nerv. Ment. Dis., 50: 327–358. Leibowitz, S.F. (1975) Catecholaminergic mechanisms of the lateral hypothalamus: their role in the mediation of amphetamine anorexia. Brain Res., 98: 529–545. Leibowitz, S.F. (1986) Brain monoamines and peptides: role in the control of eating behavior. Fed. Proc., 45: 1396–1403. Leibowitz, S.F. and Hoebel, B.G. (2004) Behavioral neuroscience and obesity. In: Bray, G., et al. (Eds.), The Handbook of Obesity. Marcel Dekker, Inc, New York. Leibowitz, S.F., Roossin, P. and Rosenn, M. (1984) Chronic norepinephrine injection into the hypothalamic

paraventricular nucleus produces hyperphagia and increased body weight in the rat. Pharmacol. Biochem. Behav., 21: 801–808. Leibowitz, S.F., Weiss, G.F. and Shor-Posner, G. (1987) Medial hypothalamic serotonin in the control of eating behavior. Int. J. Obes., 11(Suppl. 3): 109–123. Leibowitz, S.F., Weiss, G.F. and Shor-Posner, G. (1988) Hypothalamic serotonin: pharmacological, biochemical, and behavioral analyses of its feeding-suppressive action. Clin. Neuropharmacol., 11(Suppl. 1): S51–S71. Leibowitz, S.F. and Wortley, K.E. (2004) Hypothalamic control of energy balance: different peptides, different functions. Peptides, 25: 473–504. Lemke, M.R., Brecht, H.M., Koester, J., Kraus, P.H. and Reichmann, H. (2005) Anhedonia, depression, and motor functioning in Parkinson’s disease during treatment with pramipexole. J. Neuropsychiatry Clin. Neurosci., 17: 214–220. Levine, A.S. and Billington, C.J. (2004) Opioids as agents of reward-related feeding: a consideration of the evidence. Physiol. Behav., 82: 57–61. Lindholm, E., Shumway, G.S., Grijalva, C.V., Schallert, T. and Ruppel, M. (1975) Gastric pathology produced by hypothalamic lesions in rats. Physiol. Behav., 14: 165–169. Lorrain, D.S., Riolo, J.V., Matuszewich, L. and Hull, E.M. (1999) Lateral hypothalamic serotonin inhibits nucleus accumbens dopamine: implications for sexual satiety. J. Neurosci., 19: 7648–7652. Maldonado-Irizarry, C.S., Swanson, C.J. and Kelley, A.E. (1995) Glutamate receptors in the nucleus accumbens shell control feeding behavior via the lateral hypothalamus. J. Neurosci., 15: 6779–6788. Margules, D.L., Lewis, M.J., Dragovich, J.A. and Margules, A.S. (1972) Hypothalamic norepinephrine: circadian rhythms and the control of feeding behavior. Science, 178: 640–643. Mark, G.P., Blander, D.S. and Hoebel, B.G. (1991) A conditioned stimulus decreases extracellular dopamine in the nucleus accumbens after the development of a learned taste aversion. Brain Res., 551: 308–310. Mark, G.P., Rada, P., Pothos, E. and Hoebel, B.G. (1992) Effects of feeding and drinking on acetylcholine release in the nucleus accumbens, striatum, and hippocampus of freely behaving rats. J. Neurochem., 58: 2269–2274. Mark, G.P., Smith, S.E., Rada, P.V. and Hoebel, B.G. (1994) An appetitively conditioned taste elicits a preferential increase in mesolimbic dopamine release. Pharmacol. Biochem. Behav., 48: 651–660. Mark, G.P., Weinberg, J.B., Rada, P.V. and Hoebel, B.G. (1995) Extracellular acetylcholine is increased in the nucleus accumbens following the presentation of an aversively conditioned taste stimulus. Brain Res., 688: 184–188. Marshall, J.F. and Teitelbaum, P. (1974) Further analysis of sensory inattention following lateral hypothalamic damage in rats. J. Comp. Physiol. Psychol., 86: 375–395. Mas, M., Fumero, B. and Gonzalez-Mora, J.L. (1995) Voltammetric and microdialysis monitoring of brain monoamine

371 neurotransmitter release during sociosexual interactions. Behav. Brain Res., 71: 69–79. Mas, M., Rodriguez del Castillo, A., Guerra, M., Davidson, J.M. and Battaner, E. (1987) Neurochemical correlates of male sexual behavior. Physiol. Behav., 41: 341–345. Masuzawa, M., Nakao, S., Miyamoto, E., Yamada, M., Murao, K., Nishi, K. and Shingu, K. (2003) Pentobarbital inhibits ketamine-induced dopamine release in the rat nucleus accumbens: a microdialysis study. Anesth. Analg., 96: 148–152. Matos, F.F., Guss, V. and Korpinen, C. (1996) Effects of neuropeptide Y (NPY) and [D-Trp32]NPY on monoamine and metabolite levels in dialysates from rat hypothalamus during feeding behavior. Neuropeptides, 30: 391–398. McFarland, K., Davidge, S.B., Lapish, C.C. and Kalivas, P.W. (2004) Limbic and motor circuitry underlying footshockinduced reinstatement of cocaine-seeking behavior. J. Neurosci., 24: 1551–1560. McGinty, J.F. (1999) Regulation of neurotransmitter interactions in the ventral striatum. In: McGinty, J.F., et al. (Eds.) Advancing from the Ventral Striatum to the Extended Amygdala: Implications for Neuropsychiatry and Drug Abuse, 877. The New York Academy of Sciences, New York, pp. 129–139. McKinley, M.J., Allen, A.M., Mathai, M.L., May, C., McAllen, R.M., Oldﬁeld, B.J. and Weisinger, R.S. (2001) Brain angiotensin and body ﬂuid homeostasis. Jpn. J. Physiol., 51: 281–289. Meeusen, R., Piacentini, M.F. and De Meirleir, K. (2001) Brain microdialysis in exercise research. Sports Med., 31: 965–983. Meguid, M.M., Yang, Z.J. and Gleason, J.R. (1996) The gutbrain brain-gut axis in anorexia: toward an understanding of food intake regulation. Nutrition, 12: S57–S62. Meguid, M.M., Yang, Z.J. and Laviano, A. (1997) Meal size and number: relationship to dopamine levels in the ventromedial hypothalamic nucleus. Am. J. Physiol., 272: R1925–R1930. Meredith, G.E., Blank, B. and Groenewegen, H.J. (1989) The distribution and compartmental organization of the cholinergic neurons in nucleus accumbens of the rat. Neuroscience, 31: 327–345. Mitome, M. (1994) The central mechanism of feeding-associated circadian corticosterone rhythm in rats: analyses of paraventricular noradrenaline by in vivo microdialysis. Hokkaido Igaku Zasshi, 69: 120–135. Mitome, M., Honma, S., Yoshihara, T. and Honma, K. (1994) Prefeeding increase in paraventricular NE release is regulated by a feeding-associated rhythm in rats. Am. J. Physiol., 266: E606–E611. Mogenson, G.J., Jones, D.L. and Yim, C.Y. (1980) From motivation to action: functional interface between the limbic system and the motor system. Prog. Neurobiol., 14: 69–97. Molander, A., Lof, E., Stomberg, R., Ericson, M. and Soderpalm, B. (2005) Involvement of accumbal glycine receptors in the regulation of voluntary ethanol intake in the rat. Alcohol. Clin. Exp. Res., 29: 38–45.

Mook, D.G., Wagner, S. and Schwartz, L.A. (1988) Glucose sham drinking in the rat: satiety for the sweet taste without the sweet taste. Appetite, 10: 89–102. Mori, R.C., Guimaraes, R.B., Nascimento, C.M. and Ribeiro, E.B. (1999) Lateral hypothalamic serotonergic responsiveness to food intake in rat obesity as measured by microdialysis. Can. J. Physiol. Pharmacol., 77: 286–292. Morien, A., Wellman, P.J. and Fojt, J. (1995) Diurnal rhythms of paraventricular hypothalamic norepinephrine and food intake in rats. Pharmacol. Biochem. Behav., 52: 169–174. Muller, M. (2002) Science, medicine, and the future: microdialysis. BMJ, 324: 588–591. Myers, R.D., Adell, A. and Lankford, M.F. (1998) Simultaneous comparison of cerebral dialysis and push–pull perfusion in the brain of rats: a critical review. Neurosci. Biobehav. Rev., 22: 371–387. Nader, K., Bechara, A., Roberts, D.C. and van der Kooy, D. (1994) Neuroleptics block high- but not low-dose heroin place preferences: further evidence for a two-system model of motivation. Behav. Neurosci., 108: 1128–1138. Nakahara, D., Ozaki, N., Kapoor, V. and Nagatsu, T. (1989a) The effect of uptake inhibition on dopamine release from the nucleus accumbens of rats during self- or forced stimulation of the medial forebrain bundle: a microdialysis study. Neurosci. Lett., 104: 136–140. Nakahara, D., Ozaki, N., Miura, Y., Miura, H. and Nagatsu, T. (1989b) Increased dopamine and serotonin metabolism in rat nucleus accumbens produced by intracranial self-stimulation of medial forebrain bundle as measured by in vivo microdialysis. Brain Res., 495: 178–181. Napier, T.C. and Mitrovic, I. (1999) Opioid modulation of ventral pallidal inputs. Ann. N.Y. Acad. Sci., 877: 176–201. Nisell, M., Nomikos, G.G. and Svensson, T.H. (1994) Infusion of nicotine in the ventral tegmental area or the nucleus accumbens of the rat differentially affects accumbal dopamine release. Pharmacol. Toxicol., 75: 348–352. Olds, J. and Milner, P. (1954) Positive reinforcement produced by electrical stimulation of septal area and other regions of rat brain. J. Comp. Physiol. Psychol., 47: 419–427. Olds, M.E. (1982) Reinforcing effects of morphine in the nucleus accumbens. Brain Res., 237: 429–440. Ookuma, K., Sakata, T., Fukagawa, K., Yoshimatsu, H., Kurokawa, M., Machidori, H. and Fujimoto, K. (1993) Neuronal histamine in the hypothalamus suppresses food intake in rats. Brain Res., 628: 235–242. Ookuma, K., Yoshimatsu, H., Sakata, T., Fujimoto, K. and Fukagawa, F. (1989) Hypothalamic sites of neuronal histamine action on food intake by rats. Brain Res., 490: 268–275. Oomura, Y. (1988) Chemical and neuronal control of feeding motivation. Physiol. Behav., 44: 555–560. Orosco, M., Rouch, C. and Gerozissis, K. (2000) Activation of hypothalamic insulin by serotonin is the primary event of the insulin–serotonin interaction involved in the control of feeding. Brain Res., 872: 64–70. Paez, X. and Leibowitz, S.F. (1993) Changes in extracellular PVN monoamines and macronutrient intake after idazoxan

372 or ﬂuoxetine injection. Pharmacol. Biochem. Behav., 46: 933–941. Paez, X., Stanley, B.G. and Leibowitz, S.F. (1993) Microdialysis analysis of norepinephrine levels in the paraventricular nucleus in association with food intake at dark onset. Brain Res., 606: 167–170. Palhagen, S., Lorefalt, B., Carlsson, M., Ganowiak, W., Toss, G., Unosson, M. and Granerus, A.K. (2005) Does L-dopa treatment contribute to reduction in body weight in elderly patients with Parkinson’s disease? Acta Neurol. Scand., 111: 12–20. Papez, J.W. (1995) A proposed mechanism of emotion. 1937. J. Neuropsychiatry Clin. Neurosci., 7: 103–112. Parada, M.A., Hernandez, L. and Hoebel, B.G. (1988) Sulpiride injections in the lateral hypothalamus induce feeding and drinking in rats. Pharmacol. Biochem. Behav., 30: 917–923. Parada, M.A., Hernandez, L., Puig de Parada, M., Paez, X. and Hoebel, B.G. (1990) Dopamine in the lateral hypothalamus may be involved in the inhibition of locomotion related to food and water seeking. Brain Res. Bull., 25: 961–968. Parada, M.A., Puig de Parada, M. and Hoebel, B.G. (1995) Rats self-inject a dopamine antagonist in the lateral hypothalamus where it acts to increase extracellular dopamine in the nucleus accumbens. Pharmacol. Biochem. Behav., 52: 179–187. Parsons, L.H., Smith, A.D. and Justice Jr, J.B. (1991) Basal extracellular dopamine is decreased in the rat nucleus accumbens during abstinence from chronic cocaine. Synapse, 9: 60–65. Pavlov, I.P. (1927) Conditioned Reﬂexes: An Investigation of the Physiological Activity of the Cerebral Cortex. Clarendon Press, London. Pecina, S. and Berridge, K.C. (2005) Hedonic hot spot in nucleus accumbens shell: where do mu-opioids cause increased hedonic impact of sweetness? J. Neurosci., 25: 11777–11786. Pecina, S., Berridge, K.C. and Parker, L.A. (1997) Pimozide does not shift palatability: separation of anhedonia from sensorimotor suppression by taste reactivity. Pharmacol. Biochem. Behav., 58: 801–811. Pecina, S., Cagniard, B., Berridge, K.C., Aldridge, J.W. and Zhuang, X. (2003) Hyperdopaminergic mutant mice have higher ‘‘wanting’’ but not ‘‘liking’’ for sweet rewards. J. Neurosci., 23: 9395–9402. Pepeu, G. and Giovannini, M.G. (2004) Changes in acetylcholine extracellular levels during cognitive processes. Learn. Mem., 11: 21–27. Pfaus, J.G., Damsma, G., Nomikos, G.G., Wenkstern, D.G., Blaha, C.D., Phillips, A.G. and Fibiger, H.C. (1990) Sexual behavior enhances central dopamine transmission in the male rat. Brain Res., 530: 345–348. Phillips, A.G., Coury, A., Fiorino, D., LePiane, F.G., Brown, E. and Fibiger, H.C. (1992) Self-stimulation of the ventral tegmental area enhances dopamine release in the nucleus accumbens: a microdialysis study. Ann. N.Y. Acad. Sci., 654: 199–206. Pleim, E.T., Matochik, J.A., Barﬁeld, R.J. and Auerbach, S.B. (1990) Correlation of dopamine release in the nucleus

accumbens with masculine sexual behavior in rats. Brain Res., 524: 160–163. Plock, N. and Kloft, C. (2005) Microdialysis: theoretical background and recent implementation in applied life-sciences. Eur. J. Pharm. Sci., 25: 1–24. Pothos, E., Rada, P., Mark, G.P. and Hoebel, B.G. (1991) Dopamine microdialysis in the nucleus accumbens during acute and chronic morphine, naloxone-precipitated withdrawal and clonidine treatment. Brain Res., 566: 348–350. Pratt, W.E. and Kelley, A.E. (2004) Nucleus accumbens acetylcholine regulates appetitive learning and motivation for food via activation of muscarinic receptors. Behav. Neurosci., 118: 730–739. Pratt, W.E. and Kelley, A.E. (2005) Striatal muscarinic receptor antagonism reduces 24-h food intake in association with decreased preproenkephalin gene expression. Eur. J. Neurosci., 22: 3229–3240. Puig de Parada, M., Paez, X., Parada, M.A., Hernandez, L., Molina, G., Murzi, E. and Contreras, Q. (1997) Changes in dopamine and acetylcholine release in the rat lateral hypothalamus during deprivation-induced drinking. Neurosci. Lett., 227: 153–156. Rada, P. and Hoebel, B.G. (2005) Acetylcholine in the accumbens is decreased by diazepam and increased by benzodiazepine withdrawal: a possible mechanism for dependency. Eur. J. Pharmacol., 508: 131–138. Rada, P., Avena, N.M. and Hoebel, B.G. (2005) Daily bingeing on sugar repeatedly releases dopamine in the accumbens shell. Neuroscience, 134: 737–744. Rada, P., Jensen, K. and Hoebel, B.G. (2001) Effects of nicotine and mecamylamine-induced withdrawal on extracellular dopamine and acetylcholine in the rat nucleus accumbens. Psychopharmacology (Berl.), 157: 105–110. Rada, P., Johnson, D.F., Lewis, M.J. and Hoebel, B.G. (2004) In alcohol-treated rats, naloxone decreases extracellular dopamine and increases acetylcholine in the nucleus accumbens: evidence of opioid withdrawal. Pharmacol. Biochem. Behav., 79: 599–605. Rada, P., Mark, G.P. and Hoebel, B.G. (1998a) Galanin in the hypothalamus raises dopamine and lowers acetylcholine release in the nucleus accumbens: a possible mechanism for hypothalamic initiation of feeding behavior. Brain Res., 798: 1–6. Rada, P., Mark, G.P., Pothos, E. and Hoebel, B.G. (1991a) Systemic morphine simultaneously decreases extracellular acetylcholine and increases dopamine in the nucleus accumbens of freely moving rats. Neuropharmacology, 30: 1133–1136. Rada, P., Mendialdua, A., Hernandez, L. and Hoebel, B.G. (2003) Extracellular glutamate increases in the lateral hypothalamus during meal initiation, and GABA peaks during satiation: microdialysis measurements every 30 s. Behav. Neurosci., 117: 222–227. Rada, P., Pothos, E., Mark, G.P. and Hoebel, B.G. (1991b) Microdialysis evidence that acetylcholine in the nucleus accumbens is involved in morphine withdrawal and its treatment with clonidine. Brain Res., 561: 354–356.

373 Rada, P., Tucci, S., Murzi, E. and Hernandez, L. (1997) Extracellular glutamate increases in the lateral hypothalamus and decreases in the nucleus accumbens during feeding. Brain Res., 768: 338–340. Rada, P.V. and Hoebel, B.G. (2001) Aversive hypothalamic stimulation releases acetylcholine in the nucleus accumbens, and stimulation-escape decreases it. Brain Res., 888: 60–65. Rada, P.V., Mark, G.P. and Hoebel, B.G. (1998b) Dopamine release in the nucleus accumbens by hypothalamic stimulation-escape behavior. Brain Res., 782: 228–234. Rada, P.V., Mark, G.P., Taylor, K.M. and Hoebel, B.G. (1996) Morphine and naloxone, i. p. or locally, affect extracellular acetylcholine in the accumbens and prefrontal cortex. Pharmacol. Biochem. Behav., 53: 809–816. Rada, P.V., Mark, G.P., Yeomans, J.J. and Hoebel, B.G. (2000) Acetylcholine release in ventral tegmental area by hypothalamic self-stimulation, eating, and drinking. Pharmacol. Biochem. Behav., 65: 375–379. Radhakishun, F.S., van Ree, J.M. and Westerink, B.H. (1988) Scheduled eating increases dopamine release in the nucleus accumbens of food-deprived rats as assessed with on-line brain dialysis. Neurosci. Lett., 85: 351–356. Ranaldi, R., Pocock, D., Zereik, R. and Wise, R.A. (1999) Dopamine ﬂuctuations in the nucleus accumbens during maintenance, extinction, and reinstatement of intravenous D-amphetamine self-administration. J. Neurosci., 19: 4102–4109. Reynolds, S.M. and Berridge, K.C. (2002) Positive and negative motivation in nucleus accumbens shell: bivalent rostrocaudal gradients for GABA-elicited eating, taste ‘‘liking’’/’’disliking’’ reactions, place preference/avoidance, and fear. J. Neurosci., 22: 7308–7320. Robbins, T.W., Cador, M., Taylor, J.R. and Everitt, B.J. (1989) Limbic–striatal interactions in reward-related processes. Neurosci. Biobehav. Rev., 13: 155–162. Robinson, S., Sandstrom, S.M., Denenberg, V.H. and Palmiter, R.D. (2005) Distinguishing whether dopamine regulates liking, wanting, and/or learning about rewards. Behav. Neurosci., 119: 5–15. Rodriguez-Puertas, R., Pazos, A. and Pascual, J. (1994) Cholinergic markers in degenerative parkinsonism: autoradiographic demonstration of high-afﬁnity choline uptake carrier hyperactivity. Brain Res., 636: 327–332. Rolls, E.T. (1984) The neurophysiology of feeding. Int. J. Obes., 8(Suppl. 1): 139–150. Rolls, E.T. (1994) Neurophysiology and cognitive functions of the striatum. Rev. Neurol. (Paris), 150: 648–660. Rolls, E.T. (2000) Memory systems in the brain. Annu. Rev. Psychol., 51: 599–630. Rouch, C., Nicolaidis, S. and Orosco, M. (1999) Determination, using microdialysis, of hypothalamic serotonin variations in response to different macronutrients. Physiol. Behav., 65: 653–657. Rudkin, L., Taylor, M.J. and Hawton, K. (2004) Strategies for managing sexual dysfunction induced by antidepressant medication. Cochrane Database Syst. Rev., CD003382.

Rui, H. and Lebaron, M.J. (2005) Creating tissue microarrays by cutting-edge matrix assembly. Expert. Rev. Med. Devices, 2: 673–680. Salamone, J.D. (1994) The involvement of nucleus accumbens dopamine in appetitive and aversive motivation. Behav. Brain Res., 61: 117–133. Salamone, J.D., Cousins, M.S. and Snyder, B.J. (1997) Behavioral functions of nucleus accumbens dopamine: empirical and conceptual problems with the anhedonia hypothesis. Neurosci. Biobehav. Rev., 21: 341–359. Samson, H.H. and Chappell, A.M. (2004) Effects of raclopride in the core of the nucleus accumbens on ethanol seeking and consumption: the use of extinction trials to measure seeking. Alcohol. Clin. Exp. Res., 28: 544–549. Saulskaya, N.B. and Mikhailova, M.O. (2002) Feeding-induced decrease in extracellular glutamate level in the rat nucleus accumbens: dependence on glutamate uptake. Neuroscience, 112: 791–801. Saulskaya, N.B. and Mikhailova, M.O. (2004) Increased glutamate release into the intercellular space of the nucleus accumbens (N. accumbens) during substitution of food reinforcement with aversive or neutral stimuli. Neurosci. Behav. Physiol., 34: 109–113. Schneider, L.H., Gibbs, J. and Smith, G.P. (1986) D-2 selective receptor antagonists suppress sucrose sham feeding in the rat. Brain Res. Bull., 17: 605–611. Schultz, W. (1997) Dopamine neurons and their role in reward mechanisms. Curr. Opin. Neurobiol., 7: 191–197. Schultz, W. (1998a) The phasic reward signal of primate dopamine neurons. Adv. Pharmacol., 42: 686–690. Schultz, W. (1998b) Predictive reward signal of dopamine neurons. J. Neurophysiol., 80: 1–27. Schwartz, D.H., Hernandez, L. and Hoebel, B.G. (1990) Serotonin release in lateral and medial hypothalamus during feeding and its anticipation. Brain Res. Bull., 25: 797–802. Sclafani, A., Aravich, P.F. and Xenakis, S. (1982) Dopaminergic and endorphinergic mediation of a sweet reward. In: Hoebel, B.G. and Novin, D. (Eds.), The Neural Basis of Feeding and Reward. Haer Institute for Electrophysiological Research, pp, Brunswick, pp. 507–516. Segovia, J., Tossman, U., Herrera-Marschitz, M., GarciaMunoz, M. and Ungerstedt, U. (1986) Gamma-aminobutyric acid release in the globus pallidus in vivo after a 6-hydroxydopamine lesion in the substantia nigra of the rat. Neurosci. Lett., 70: 364–368. Sharf, R. and Ranaldi, R. (2006) Blockade of muscarinic acetylcholine receptors in the ventral tegmental area disrupts food-related learning in rats. Psychopharmacology (Berl.), 184: 87–94. Skinner, B.F. (1938) The Behavior of Organisms: An Experimental Analysis. B.F. Skinner Foundation, New York. Smith, G.P. (1998) Sham feeding in rats with chronic, reversible gastric ﬁstulas. In: Crawley, J.N., et al. (Eds.), Current Protocols in Neuroscience. Wiley, New York, pp. 8.6:D.1–8.6:D.6.

374 Smith, G.P. (2004) Accumbens dopamine mediates the rewarding effect of orosensory stimulation by sucrose. Appetite, 43: 11–13. Spangler, R., Wittkowski, K.M., Goddard, N.L., Avena, N.M., Hoebel, B.G. and Leibowitz, S.F. (2004) Opiate-like effects of sugar on gene expression in reward areas of the rat brain. Brain Res. Mol. Brain Res., 124: 134–142. Specker, S.M., Lac, S.T. and Carroll, M.E. (1994) Food deprivation history and cocaine self-administration: an animal model of binge eating. Pharmacol. Biochem. Behav., 48: 1025–1029. Spehlmann, R. and Stahl, S.M. (1976) Dopamine acetylcholine imbalance in Parkinson’s disease. Possible regenerative overgrowth of cholinergic axon terminals. Lancet, 1: 724–726. Stanley, B.G., Ha, L.H., Spears, L.C. and Dee, M.G. (1993a) Lateral hypothalamic injections of glutamate, kainic acid, D,L-alpha-amino-3-hydroxy-5-methyl-isoxazole propionic acid or N-methyl-D-aspartic acid rapidly elicit intense transient eating in rats. Brain Res., 613: 88–95. Stanley, B.G., Schwartz, D.H., Hernandez, L., Hoebel, B.G. and Leibowitz, S.F. (1989) Patterns of extracellular norepinephrine in the paraventricular hypothalamus: relationship to circadian rhythm and deprivation-induced eating behavior. Life Sci., 45: 275–282. Stanley, B.G., Willett, V.L. III., Donias, H.W., Ha, L.H. and Spears, L.C. (1993b) The lateral hypothalamus: a primary site mediating excitatory amino acid-elicited eating. Brain Res., 630: 41–49. Stiller, C.O., Taylor, B.K., Linderoth, B., Gustafsson, H., Warsame Afrah, A. and Brodin, E. (2003) Microdialysis in pain research. Adv. Drug Deliv. Rev., 55: 1065–1079. Strader, A.D. and Woods, S.C. (2005) Gastrointestinal hormones and food intake. Gastroenterology, 128: 175–191. Stratford, T.R. and Kelley, A.E. (1999) Evidence of a functional relationship between the nucleus accumbens shell and lateral hypothalamus subserving the control of feeding behavior. J. Neurosci., 19: 11040–11048. Szczypka, M.S., Kwok, K., Brot, M.D., Marck, B.T., Matsumoto, A.M., Donahue, B.A. and Palmiter, R.D. (2001) Dopamine production in the caudate putamen restores feeding in dopamine-deﬁcient mice. Neuron, 30: 819–828. Taber, M.T. and Fibiger, H.C. (1995) Electrical stimulation of the prefrontal cortex increases dopamine release in the nucleus accumbens of the rat: modulation by metabotropic glutamate receptors. J. Neurosci., 15: 3896–3904. Taber, M.T. and Fibiger, H.C. (1997) Feeding-evoked dopamine release in the nucleus accumbens: regulation by glutamatergic mechanisms. Neuroscience, 76: 1105–1112. Tachibana, T., Tazawa, M. and Sugahara, K. (2001) Feeding increases 5-hydroxytryptamine and norepinephrine within the hypothalamus of chicks. Comp. Biochem. Physiol. A. Mol. Integr. Physiol., 130: 715–722. Tachibana, T., Utimura, D., Kato, H., Kubo, T. and Sugahara, K. (2000) Extracellular norepinephrine in the medial hypothalamus increases during feeding in chicks: a microdialysis study. Comp. Biochem. Physiol. A Mol. Integr. Physiol., 127: 331–338. Tanaka, J., Hayashi, Y., Yamato, K., Miyakubo, H. and Nomura, M. (2004) Involvement of serotonergic systems in the

lateral parabrachial nucleus in sodium and water intake: a microdialysis study in the rat. Neurosci. Lett., 357: 41–44. Tanaka, J., Kariya, K. and Nomura, M. (2003) Drinking attenuates the noradrenaline release in the lateral hypothalamic area induced by angiotensin II activation of the subfornical organ in rats. Behav. Brain Res., 140: 49–55. Tanda, G. and Di Chiara, G. (1998) A dopamine-mu1 opioid link in the rat ventral tegmentum shared by palatable food (Fonzies) and non-psychostimulant drugs of abuse. Eur. J. Neurosci., 10: 1179–1187. Tanda, G., Pontieri, F.E. and Di Chiara, G. (1997) Cannabinoid and heroin activation of mesolimbic dopamine transmission by a common mu1 opioid receptor mechanism. Science, 276: 2048–2050. Taylor, K.M., Davidson, K., Mark, G.P., Rada, P. and Hoebel, B.G. (1992) Conditioned taste aversion induced by increased acetylcholine in the nucleus accumbens. Soc. Neurosci. Abstr., 18: 1066. Teitelbaum, P. and Epstein, A.N. (1962) The lateral hypothalamic syndrome: recovery of feeding and drinking after lateral hypothalamic lesions. Psychol. Rev., 69: 74–90. Telles, M.M., Guimaraes, R.B. and Ribeiro, E.B. (2003) Effect of leptin on the acute feeding-induced hypothalamic serotonergic stimulation in normal rats. Regul. Pept., 115: 11–18. Tobler, P.N., Fiorillo, C.D. and Schultz, W. (2005) Adaptive coding of reward value by dopamine neurons. Science, 307: 1642–1645. Triemstra, J.L., Nagatani, S. and Wood, R.I. (2005) Chemosensory cues are essential for mating-induced dopamine release in MPOA of male Syrian hamsters. Neuropsychopharmacology, 30: 1436–1442. Tsujii, S. and Bray, G.A. (1991) GABA-related feeding control in genetically obese rats. Brain Res., 540: 48–54. Tuite, P.J., Maxwell, R.E., Ikramuddin, S., Kotzd, C.M., Billingtond, C.J., Laseski, M.A. and Thielen, S.D. (2005) Weight and body mass index in Parkinson’s disease patients after deep brain stimulation surgery. Parkinsonism Relat. Disord., 11: 247–252. Ungerstedt, U. (1970) Is interruption of the nigro-striatal dopamine system producing the ‘‘lateral hypothalamus syndrome’’? Acta Physiol. Scand., 80: 35A–36A. Ungerstedt, U. (1971) Adipsia and aphagia after 6-hydroxydopamine induced degeneration of the nigro-striatal dopamine system. Acta Physiol. Scand. Suppl., 367: 95–122. Ungerstedt, U. (1984) Measurement of neurotransmitter release by intracranial dialysis. Measurement of neurotransmitter release in vivo. In: Marsden, C.A. (Ed.), Measurement of Neurotransmitter Release. Wiley, pp, New York, pp. 81–105. Ungerstedt, U. (1991) Microdialysis – principles and applications for studies in animals and man. J. Intern. Med., 230: 365–373. Ushigome, A., Tanaka, J., Kariya, K. and Nomura, M. (2002) Paraventricular noradrenergic systems participate in angiotensin II-induced drinking. Peptides, 23: 2169–2175. Valenstein, E.S., Cox, V.C. and Kakolewski, J.W. (1968) Modiﬁcation of motivated behavior elicited by electrical stimulation of the hypothalamus. Science, 159: 1119–1121.

375 Warner, R.K., Thompson, J.T., Markowski, V.P., Loucks, J.A., Bazzett, T.J., Eaton, R.C. and Hull, E.M. (1991) Microinjection of the dopamine antagonist cis-ﬂupenthixol into the MPOA impairs copulation, penile reﬂexes and sexual motivation in male rats. Brain Res., 540: 177–182. Weiner, I. (2003) The ‘‘two-headed’’ latent inhibition model of schizophrenia: modeling positive and negative symptoms and their treatment. Psychopharmacology (Berl.), 169: 257–297. Weiss, F., Markou, A., Lorang, M.T. and Koob, G.F. (1992) Basal extracellular dopamine levels in the nucleus accumbens are decreased during cocaine withdrawal after unlimited-access self-administration. Brain Res., 593: 314–318. Westerink, B.H. (1995) Brain microdialysis and its application for the study of animal behaviour. Behav. Brain Res., 70: 103–124. Westerink, B.H., Kwint, H.F. and de Vries, J.B. (1997) Eatinginduced dopamine release from mesolimbic neurons is mediated by NMDA receptors in the ventral tegmenatl area: a dual-probe microdialysis study. J. Neurochem., 69: 662–668. Westerink, B.H., Teisman, A. and de Vries, J.B. (1994) Increase in dopamine release from the nucleus accumbens in response to feeding: a model to study interactions between drugs and naturally activated dopaminergic neurons in the rat brain. Naunyn Schmiedebergs Arch. Pharmacol., 349: 230–235. Wilcox, K.M., Kimmel, H.L., Lindsey, K.P., Votaw, J.R., Goodman, M.M. and Howell, L.L. (2005) In vivo comparison of the reinforcing and dopamine transporter effects of local anesthetics in rhesus monkeys. Synapse, 58: 220–228. Wise, R.A. (1978) Neuroleptic attenuation of intracranial selfstimulation: reward or performance deﬁcits? Life Sci., 22: 535–542. Wise, R.A. and Hoffman, D.C. (1992) Localization of drug reward mechanisms by intracranial injections. Synapse, 10: 247–263. Wise, R.A. and Rompre, P.P. (1989) Brain dopamine and reward. Annu. Rev. Psychol., 40: 191–225. Wise, R.A., Leone, P., Rivest, R. and Leeb, K. (1995a) Elevations of nucleus accumbens dopamine and DOPAC levels during intravenous heroin self-administration. Synapse, 21: 140–148. Wise, R.A., Newton, P., Leeb, K., Burnette, B., Pocock, D. and Justice, J.B. Jr. (1995b) Fluctuations in nucleus accumbens dopamine concentration during intravenous cocaine self-administration in rats. Psychopharmacology (Berl.), 120: 10–20. Woolf, N.J. (1991) Cholinergic systems in mammalian brain and spinal cord. Prog. Neurobiol., 37: 475–524. Xenakis, S. and Sclafani, A. (1981) The effects of pimozide on the consumption of a palatable saccharin-glucose solution in the rat. Pharmacol. Biochem. Behav., 15: 435–442. Xi, Z.X., Gilbert, J.G., Pak, A.C., Ashby, C.R. Jr., Heidbreder, C.A. and Gardner, E.L. (2005) Selective dopamine D3 receptor antagonism by SB-277011A attenuates cocaine reinforcement as assessed by progressive-ratio and variablecost-variable-payoff ﬁxed-ratio cocaine self-administration in rats. Eur. J. Neurosci., 21: 3427–3438. Xi, Z.X., Newman, A.H., Gilbert, J.G., Pak, A.C., Peng, X.Q., Ashby, C.R., Gitajn, L. and Gardner, E.L. (2006) The novel

dopamine D(3) receptor antagonist NGB 2904 inhibits cocaine’s rewarding effects and cocaine-induced reinstatement of drug-seeking behavior in rats. Neuropsychopharmacology, 31: 1393–1405. Yang, Z.J. and Meguid, M.M. (1995) LHA dopaminergic activity in obese and lean Zucker rats. Neuroreport, 6: 1191–1194. Yang, Z.J., Koseki, M., Meguid, M.M. and Laviano, A. (1996) Eating-related increase of dopamine concentration in the LHA with oronasal stimulation. Am. J. Physiol., 270: R315–R318. Yang, Z.J., Meguid, M.M. and Oler, A. (1997) Eating-associated VMN-dopamine levels of rats: comparison of oral and intragastric feeding. Neuroreport, 8: 1543–1547. Ye, J.H., Wang, F., Krnjevic, K., Wang, W., Xiong, Z.G. and Zhang, J. (2004) Presynaptic glycine receptors on GABAergic terminals facilitate discharge of dopaminergic neurons in ventral tegmental area. J. Neurosci., 24: 8961–8974. Yeomans, J., Forster, G. and Blaha, C. (2001) M5 muscarinic receptors are needed for slow activation of dopamine neurons and for rewarding brain stimulation. Life Sci., 68: 2449–2456. Yoshida, M., Yokoo, H., Mizoguchi, K., Kawahara, H., Tsuda, A., Nishikawa, T. and Tanaka, M. (1992) Eating and drinking cause increased dopamine release in the nucleus accumbens and ventral tegmental area in the rat: measurement by in vivo microdialysis. Neurosci. Lett., 139: 73–76. Yoshida, M., Yokoo, H., Tanaka, T., Mizoguchi, K., Emoto, H., Ishii, H. and Tanaka, M. (1993) Facilitatory modulation of mesolimbic dopamine neuronal activity by a mu-opioid agonist and nicotine as examined with in vivo microdialysis. Brain Res., 624: 277–280. You, Z.B., Chen, Y.Q. and Wise, R.A. (2001) Dopamine and glutamate release in the nucleus accumbens and ventral tegmental area of rat following lateral hypothalamic self-stimulation. Neuroscience, 107: 629–639. You, Z.B., Tzschentke, T.M., Brodin, E. and Wise, R.A. (1998) Electrical stimulation of the prefrontal cortex increases cholecystokinin, glutamate, and dopamine release in the nucleus accumbens: an in vivo microdialysis study in freely moving rats. J. Neurosci., 18: 6492–6500. Young, A.M., Joseph, M.H. and Gray, J.A. (1992) Increased dopamine release in vivo in nucleus accumbens and caudate nucleus of the rat during drinking: a microdialysis study. Neuroscience, 48: 871–876. Zahm, D.S., Williams, E. and Wohltmann, C. (1996) Ventral striatopallidothalamic projection. IV: Relative involvements of neurochemically distinct subterritories in the ventral pallidum and adjacent parts of the rostroventral forebrain. J. Comp. Neurol., 364: 340–362. Zetterstrom, T. and Ungerstedt, U. (1984) Effects of apomorphine on the in vivo release of dopamine and its metabolites, studied by brain dialysis. Eur. J. Pharmacol., 97: 29–36. Zigmond, M.J. and Stricker, E.M. (1972) Deﬁcits in feeding behavior after intraventricular injection of 6-hydroxydopamine in rats. Science, 177: 1211–1214.

CHAPTER 4.4

Changes in acetylcholine extracellular levels during cognitive processes Giancarlo Pepeu and Maria G. Giovannini Department of Pharmacology, University of Florence, Florence, Italy

Abstract: The microdialysis studies in which acetylcholine (ACh) extracellular levels were measured during the performance of spontaneous and acquired behaviors involving cognitive processes have been reviewed. If diffusion and inactivation are kept stable, the extracellular ACh levels are an indication of the release from the neurons and depend on neuronal ﬁring. The term ‘‘cognitive processes’’ comprises many brain functions including attention, information acquisition, and memory formation, which can be investigated in animals implanted with microdialysis probes. Novelty induces arousal and attention and a diffuse increase in ACh release from the cerebral cortex and hippocampus. Conversely, sensory stimulation induces diffuse activation of the cortical cholinergic system and a more intense ACh release from the speciﬁc receiving cortical areas. Habituation attenuates the cholinergic activation. The relationship between levels of attention and ACh release was also examined. Spatial memory formation is associated with an activation of the hippocampal cholinergic system. The acquisition of a conditioned stimulus is accompanied by an increase in ACh release occurring not only in the cortex and hippocampus but also in the nucleus accumbens, for some stimuli. Working memory and reference memory involve the cortical and hippocampal cholinergic systems, respectively, and the shifting from hippocampal to striatal cholinergic activation accompanies the changing of rat strategy in mastering a maze for food. From the papers reviewed, microdialysis appears a powerful tool for investigating the role of brain cholinergic system in speciﬁc areas and different cognitive processes. Using this tool, the extensive involvement of the brain cholinergic pathways in the ongoing cognitive activity has been conﬁrmed. Papers describing drug effect on release and behavior, investigated separately, are not included. The changes in ACh extracellular levels during cognitive processes have been the object of a shorter review (Pepeu and Giovannini, 2004).

I. Introduction I.A. Aims and limits of the chapter A PubMed search with the terms ‘‘acetylcholine and microdialysis’’ produced, in January 2006, 914 titles. In this chapter, we have reviewed only the papers in which acetylcholine (ACh) extracellular levels were measured during the performance of spontaneous and acquired behaviors involving cognitive processes with the aim to deﬁne when and where the cholinergic systems are involved.

I.B. Definitions The expression ‘‘cognitive processes’’ comprises many brain functions not only including attention, information acquisition, and memory formation but also the acquisition of motor skills, language, and emotions. With the obvious exception of language, these functions can be investigated in

Corresponding author: E-mail: giancarlo.pepeu@uniﬁ.it

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

377

DOI: 10.1016/S1569-7339(06)16020-8 Copyright 2007 Elsevier B.V. All rights reserved

378

laboratory animals in which they are present in simpler forms than in humans. Changes in ACh extracellular levels can be measured by microdialysis in discrete brain regions during these brain functions. The extracellular levels of ACh represent the equilibrium between the release of the neurotransmitter from nerve endings and its removal by diffusion and hydrolysis by cholinesterases (ChE). If diffusion and inactivation are kept stable, the extracellular levels are an indication of the release from the neurons impinging on the area where the neurotransmitter is measured and depend on neuronal ﬁring and are considered an indication of changes in activity of the cholinergic neurons. This is demonstrated by the observation that inhibition of propagated potentials by tetrodoxin (TTX) strongly reduces ACh extracellular levels (Pepeu, 1973; Westerink et al., 1987; Bianchi et al., 2003). For these reasons, the terms ‘‘release’’, which we have used throughout the text, ‘‘output’’, and ‘‘efﬂux’’ are frequently used in place of the more correct but longer expression ‘‘extracellular levels’’.

I.C. From the cortical cup and push– pull cannula to microdialysis Fig. 1 shows the techniques used, before the introduction of microdialysis, to investigate ACh release and correlate their change with modiﬁcations in neuronal activity in different brain areas. The initial studies began more than 40 years ago with the so-called cortical cup technique and the push–pull cannula with the aim to understand the role of the cholinergic system in brain functioning. For a review of the early literature, see Pepeu (1973) and DeFeudis (1974). With the cortical cup technique it was possible to quantify ACh diffusing from the cerebral cortex, into a Ringer solution placed in small cylinders applied on the exposed brain to form a cup, when a ChE inhibitor was added to the Ringer. An equilibrium between ACh in the cortex and the solution in the cup is reached in 10–15 min. Thus, the solution in the cup was substituted every 15–20 min and its ACh content quantiﬁed. The push–pull cannula (Gaddum, 1961), consisted of two concentric stainless-steel

tubes implanted in a discrete brain region. The oxygenated Ringer solution, containing a ChE inhibitor, perfused through the cannula removes the neurotransmitters released from the neurons around its tip. Unfortunately, the damage made by the cannula around the tip, and its frequent clogging, limited the use of this technique, which was never used in behavioral experiments. The push–pull cannula may be considered the precursor of the microdialysis technique. Fig. 1 also summarizes the main ﬁndings obtained with the cup technique. It was shown that the changes in ACh levels in the cortical cups depended on brain activity since arousal, demonstrated by electrocortical activation, was accompanied by increased ACh levels in the cup. Conversely, slow wave sleep and sedation, either spontaneous or drug induced, were associated with decreased ACh levels (Kanai and Szerb, 1965; Celesia and Jasper, 1966). Moreover, the stimulation and destruction of the cholinergic neurons of the NB result in an increase (Casamenti et al., 1986) and in a decrease of ACh diffusing in the cup (Lo Conte et al., 1982), respectively. Therefore, the changes in ACh levels in the cup express the changes in ACh released from the cortical cholinergic nerve endings under the cup. Most of the ﬁndings obtained with the cup technique have been reproduced and conﬁrmed with microdialysis. These data generated the background knowledge on the function of the cholinergic system in different states of cortical activity on which the progresses achieved with the microdialysis method have been built. However, even if the cup experiments could be done on un-anesthetized rats, these had to be kept in a conﬁned space and only observations on changes in ACh release during gross behaviors were possible. It should also be mentioned that in the cortical cup era, ACh was quantiﬁed by bioassay on leech dorsal muscle, frog rectus, or guinea pig ileum, the only methods sufﬁciently sensitive, at that time, for detecting ACh in nanogram amounts. The introduction of microdialysis was accompanied by the development of a reliable and rapid-HPLC method for quantifying ACh at femtomole level (Damsma et al., 1987). However, a radioenzymatic assay for quantiﬁcation of ACh (Consolo et al., 1987) gave identical results (Bertorelli et al., 1990).

379

a) Cortical cup

c) Microdialysis Transversal probe 400 µm Vertical I-shaped probes

5000 µm

b) Push-pull cannula

Guide cannula

3 mm 500 µm home-made

700 µm

CMA 12

Stainless steel Epoxy glue Dialysis membrane Fused silica

Fig. 1. Methods used for investigating the extracellular levels of acetylcholine (ACh). MacIntosh and Oboring (1955) were the ﬁrst to develop the cortical cup (a). Gaddum (1961) used the push–pull cannula (b). Fisone et al. (1987) were the ﬁrst to use the microdialysis technique for studying ACh release (c).

Recently, the use of chromatography/mass spectrometry has increased the sensitivity of ACh assay to 0.2 fmol (Uutela et al., 2005). The ﬁrst paper on the release of ACh using microdialysis was published by Fisone et al. (1987) and the initial studies were reviewed by Westerink et al. (1987). The aim of the early studies was to investigate the effect of drugs and the mechanisms regulating ACh release (Westerink et al., 1988), and to identify the neuronal pathways from which ACh was released (Kurosawa et al., 1989). Most of the studies were carried out in freely moving rats and correlations between ACh release and motor activity (Mizuno et al., 1991), wakefulness and sleep (Kametani and Kawamura, 1990), handling (Nilsson et al., 1990), feeding and drinking (Mark

et al., 1992) were made. All these studies were reviewed by Westerink (1995). The ﬁrst observation in which changes in ACh release were directly investigated during a behavior involving a cognitive process was reported by Inglis et al. (1994) and Inglis and Fibiger (1995) who demonstrated in the rat an increase in ACh release due to attention and food expectation.

II. Methodological issues II.A. General considerations In this section, the methodological aspects concerning the use of microdialysis for the study of

380

ACh release in animals performing tasks involving cognitive processes are presented. A microdialysis membrane acts similarly to a blood capillary, sampling the interstitial ﬂuid by a process of diffusion along a logarithmic concentration gradient toward and away from the probe. The principle of dialysis follows Fick’s ﬁrst law of diffusion, which describes the passive movement of molecules down a concentration gradient, therefore allowing not only collection of neurotransmitters from the outside to the inside of the membrane but also the delivery of exogenous compounds directly into the brain structure of interest. The recovery of neurotransmitters in microdialysis experiments depends on several variables, including temperature, pH, molecular weight, shape, and charge as well as on the surface area of the dialysis probe, the ﬂow rate of the perfusion liquid and its composition, and on the chemical composition of the membrane (see Benveniste, 1989; Benveniste and Huttemeier, 1990). In Table 1, the experimental parameters inﬂuencing the output and recovery of ACh in different brain areas are shown with examples of the extracellular levels found.

II.B. Membrane recovery The concentration of an endogenous neurotransmitter in the dialysate depends on membrane recovery. Recovery in vivo is a measure of the rate at which a substance ﬂows from the extracellular ﬂuid (ECF) to the dialysate. This in turn depends on diffusion barriers within the brain, on the neurotransmitter’s and the membrane’s chemical characteristics, the surface area of the dialysis membrane, and the perfusion ﬂow rate. A question raised frequently is whether the transmitter levels in the dialysate reﬂect ‘true’ variations of the molecule in the ECF, not to mention at the synapse. Several models have been experimentally designed to quantify the recovery both in vitro, simple to perform but of limited value, and in vivo, usually very cumbersome and time consuming to perform (for review see Khan and Shuaib, 2001). In the early microdialysis papers, the recovery of ACh through the dialysis probes were calculated in vitro and reported. For instance according to Wu et al.

(1988), the recovery was 49.271.8% in the striatum, 5072% in the frontal cortices, and 48.97 1.4% in the hippocampus. Slightly higher values were obtained by Giovannini et al. (1991), while lower recovery was reported by Day et al. (1991). However, in no paper the ACh values were corrected for recovery. In behavioral studies, the relative changes in the dialysate content of neurotransmitter are important as they indicate the relative changes in ACh extracellular level – averaged over time and space, while the absolute levels cannot be calculated.

II.C. Probes (transversal vs. vertical) Not all authors describe the microdialysis membrane used in their experiments. The most commonly used membranes in ACh microdialysis studies are the following: CMA type 10, 11, 12, Carnegie Medicine, Sweden; AN 69-HF Hospal Dasco, Italy; AN 69 Filtral 8, 16 Hospal Industries, Meyzieu, France; A-I-8-02, Eicom; BAS microdialysis probe; Spectrum Medical Industries, Los Angeles, CA. cellulose ﬁber. The external diameter ranges from 240 to 500 mm, the latter being the most commonly used. The molecular weight cutoff goes from 6 to 15 kDa. Microdialysis experiments can be performed with probes of different geometry, transversal or vertical, which have different and somehow complementary characteristics and are suited for different experiment purposes. Transversal microdialysis probes are based on the design ﬁrst described by Imperato and Di Chiara (1984) with later modiﬁcations by several investigators who adapted their design for the study of ACh release not only from the striatum (Damsma et al., 1988) but also from other structures such as dorsal hippocampus and frontal cortex (Giovannini et al., 1991) in awake rats. Transversal probes are simple and inexpensive to make, having a large dialysis surface area, which allows high recovery. Nonetheless, transversal probes are not suited for all applications. Indeed, sampling ACh from small and ventral brain areas such as the basal forebrain (Szerb et al., 1994; Miranda and Bermudez-Rattoni, 1999), the amygdala (Hajnal et al., 1998; Cangioli et al., 2002),

Table 1. Basal acetylcholine (ACh) extracellular levels, and parameters used in microdialysis experiments investigating the release during different behaviors Behavior

Brain region

Probe

Flow rate K+

Ca2+

Mg2+

Glu

AChEI

Min

Basal ACh levels

Circadian rhythm

Motor cortex

V-3 (dual probes)

1.5

2.4

–

–

Neo 5

60

Motor cortex 2.11 22 h pmol/mm/h

4

SS cortex Visual cortex Circadian rhythm

Prefrontal cortex V-1

2.5

4

2.3

–

–

Phy 10

20

20

Motor activity, Dorsal circadian rhythm hippocampus

V-1

2.5

4

2.3

–

–

Phy 10

Immobilization stress

Dorsal hippocampus

V-1

2.0

5.4

2.4

1.3

4

Neo 0.1 15

Inescapable stress

Nucleus accumbens

V-2

1

3.9

1.2

1

–

Neo 0.5 15

Dorsal hippocampus

V-2

Amygdala

V-1.5

mPFC

V-3 (dual probes) T-8

4

4

1.2

–

–

Phy 7

1.0

4

1.2

–

–

Neo 0.3 20

Novelty, habituation

Frontoparietal cortex

T-6 V-5

Reference JimenezCapdeville and Dykes (1996)

5

Mitsushima et al. (1996)

4

Mizuno et al. (1991)

2

Tajima et al. (1996)

Nucleus 7 accumbens 0.519 pmol/15 min Dorsal hippocampus 0.199 pmol/15 min Amygdala 0.291 pmol/15 min mPFC 0.521 pmol/ 15 min Frontoparietal 2 cortex 2.42 pmol/ 10 min Dorsal hippo 1.19 pmol/10 min 7

Mark et al. (1996)

Giovannini et al. (2001)

Puig et al. (1997)

381

Dorsal hippocampus DeprivationLateral induced drinking hypothalamus

10

SS cortex 2.18 pmol/mm/h Visual cortex 1.62 pmol/mm/h Light 5.72 pmol/ 20 min Dark 12.06 pmol/ 20 min Light 12.3 pmol/ 20 min Dark 20.9 pmol/ 20 min 51.7 nmol/L

Days

382

Table 1 (continued ) Behavior

Brain region

Probe

Flow rate K+

Ca2+

Mg2+

Glu

AChEI

Min

Basal ACh levels

Days

Reference

Conditioned taste aversion

NBM

V-3

2

4.7

2.5

–

–

Neo 10

15

2.0 pmol/30 min

10

Miranda and BermudezRattoni (1999)

Insular cortex Nucleus accumbens

V-3 V-3

1.8

3.9

1.2

1.0

–

Neo 0.5 10

Striatum

V-4

Ventral hippocampus Frontal cortex

V-4 –

4

2.25

–

–

Neo 0.3 10

V-2

2.0

2.9

1.2

0.8

5

Neo 0.1 6

Nucleus 7 (18 h accumbens 1.37 probe) pmol/25 mL Striatum 0.78 pmol/25 mL Ventral hippo 0.42 pmol/25 mL Frontal cortex 0.11 pmol/sample Nucleus accumbens core 0.84 pmol/sample Nucleus accumbens shell 0.23 pmol/sample 0.048 pmol/10 min 3

V-4

2.2

3

1.3

1

–

Neo 1.0 10

64.6 fmol/mL

7

mPFC

V-3

3

3

1.2

0.27

7.2

Neo 0.5 5

5

Ventral hippocampus

V-3 (dual probes)

Ventral hippocampus

Dual probes

2

2.5

1.3

2.1

1

Neo 0.1 5

Feeding and drinking

Handling

V-2

Nucleus V-2 accumbens (core and shell)

Attentional Frontoparietal demands cortex Radial arm maze Hippocampus Inhibitory avoidance

Spatial memory

Striatum

T

2.2

3

1.3

1

–

Neo 0.1 10

mPFC 2.72 fmol/ mL Ventral hippocampus 2.86 fmol/mL Ventral hippocampus 7.3 fmol/mL Striatum 37.3 3 fmol/mL 3.36 pmol/10 min

Visual discrimination Attentional task

mPFC

V-2

2

2.9

1.2

0.8

5

Neo 0.5 12

0.33 pmol/min

mPFC

V-2

2

3

1.3

1

–

Neo 0.05

244.9 fmol/10 min

Cross maze

Striatum

10

Mark et al. (1992)

Thiel et al. (1998a)

Himmelheber et al. (2001) Fadda et al. (2000) Giovannini et al. (2005)

7

Chang and Gold (2003)

2

Chang and Gold (2003) Himmelheber et al. (1997) Dalley et al. (2001)

2

Visual stimulation

mPFC

V-2

2

3

1.3

1.0

–

S.S. cortex

V-2

Visual cortex

V-2 (triple probes) V-2 2

3

1.3

1.0

–

2.3

–

–

20

0.96 pmol/ 40 mL

Neo 10 + Atrop 10

20

540 pmol/sample

80 min

Fournier et al. (2004)

Passetti et al. (2000) NailBoucherie et al. (2000)

172 pmol/sample 190 pmol/sample

Visual attentional task Contextual fear conditioning

mPFC Hippocampus (CA3)

V-4

1

4

Extinction of feared behavior

mPFC

V-1

2

Ringer’s Phy 10 solution

Neo 0.05 Neo 10

3

10

1.11 pmol/h

2

20

12.5 nM

5

Izaki et al. (2001)

Note: mPFC, medial prefront cortex; SS cortex, somatosensory cortex; NBM, nucleus basalis magnocellularis. Probe: V, I-shaped; T, transversal (length in mm). Flow rate: mL/min. K+, Ca2+, Mg2+, Glu: concentration in mM. AChEI: Neo, neostigmine; Phy, physostigmine (concentration in mM). Min: sample collection intervals in minutes. Days: days post-surgery (unless stated).

383

384

or the ventral hippocampus and medial prefrontal cortex (Giovannini et al., 2005) requires the use of vertical I-shaped probes. These probes can be made in the laboratory or purchased by commercial sources which are rather expensive, although reusable for a few experiments if properly handled. The recovery of ACh from vertical probes is usually low due to the small exposed area. As reported in Table 1, the length of the I-shaped membrane ranges between 0.5 and 4 mm. ‘‘Dual membrane’’ experiments can also be performed, to simultaneously measure the release of ACh from two different, related or not, structures during the performance of a behavioral task (Jimenez-Capdeville and Dykes, 1996; Mark et al., 1996; Giovannini et al., 2005; Laplante et al., 2005), thus allowing a direct comparison of ACh involvement in cognitive function in the two different brain areas (McIntyre et al., 2003a, b). Since microdialysis studies during behavior need active, fully recovered animals, they are usually performed several days after probe implantation (typically between 2 and 7 days). This methodological requirement poses a further problem since it has been shown that by the third day after probe insertion an astrocytic reaction occurs around the membrane and axonal anterograde and retrograde degeneration is evident. A glial reactive scar develops around the membrane, which inﬂuences the release of neurotransmitters (Benveniste and Diemer, 1987). Most of the laboratories that study the release of ACh during behavior use the vertical probe design, which allows the insertion of the guide cannula under anesthesia a few days before the experiment and the insertion of the probe tip after the animal has recovered (see Table 1), a few hours before the behavioral task is performed. The inlet and outlet tubings of the membrane must be long enough to allow free and unrestrained movements of the animal in the microdialysis cage during behavioral testing and the testing apparatus is adapted to their unimpeded passage. To make any correlation between neurotransmitter release and behavior possible, it is also necessary that the outlet tubing has a ﬁxed and known time lag between the sampling of dialysate and its collection into the test tube.

II.D. Dialysis medium: calcium and cholinesterase inhibitors A variety of perfusate ﬂow rates and compositions are used for ACh microdialysis (Table 1). The ﬂow rate affects the recovery; ﬂow rates between 0.5 and 5 mL/min are used, depending on the applications. It is advisable that the composition of the perfusion mediums should be as close as possible to cerebrospinal ﬂuid (CSF) or brain ECF electrolyte composition. We use an artiﬁcial CSF (aCSF), of the following composition (in mM): 140 NaCl, 1.2 CaCl2, 3.0 KCl, 0.27 MgCl2, 0.27 Na2HPO4, and 7.2 glucose, pH 7.4. The concentration of K+, Mg+, but especially of Ca2+ in the perfusion medium markedly affect ACh release. Indeed, perfusion ﬂuid devoid of Ca2+ depletes calcium around the probe and blocks synaptic release of ACh (Damsma et al., 1988), whereas variation of this ion concentration in the perfusion ﬂuid (1.2–3.4 mM) may affect the pharmacological responses in microdialysis experiments (de Boer et al., 1990). Furthermore, infusion of the Na+ channel blocker TTX decreases ACh release to undetectable values (Damsma et al., 1988; Giovannini et al., 1991, 2001; Westerink, 1995). The fulﬁllment of the two classical criteria of neurotransmitter release, Ca2+-dependency and TTX sensitivity, implies that the dialysate ACh content strictly mirrors synaptic release, which in turn is a reliable index of the functional activity of cholinergic neurons (Pepeu, 1973). The most controversial aspect of ACh microdialysis experiments is the need to include a ChE inhibitor in the perfusate to obtain levels of ACh easily detectable with the HPLC methods currently available (sensitivity limits 50–100 fmol/injection), although the use of mass spectrometry increases the sensitivity of ACh assay to 0.2 fmol (Uutela et al., 2005). Both physostigmine and neostigmine have been used through the years by different laboratories, at concentration ranging from 0.05 to 10 mM (Table 1). Recently, however, the use of neostigmine in a concentration range of 0.1–0.5 mM prevails in behavioral experiments, allowing detectable levels of ACh in 5–10 min sampling periods (Day et al., 2001; Giovannini et al., 2005). The presence of high levels of ACh in the extracellular

385

space brought about by ChE inhibition causes continuous occupation of muscarinic presynaptic inhibitory receptors, thereby maintaining an inhibitory tone which controls ACh release from the cholinergic terminal. In contrast, it has been demonstrated that concentrations of neostigmine up to 5 mM do not induce changes in sensoryevoked cortical activity, assessed by ﬁeld potential recording (Oldford and Castro-Alamancos, 2003) and that varying the concentration of neostigmine does not affect signiﬁcantly the magnitude and the duration of sensory-evoked ACh efﬂux (Himmelheber et al., 1998). If the addition of a ChE inhibitor is not physiological for the above reasons even the performance of microdialysis with no inhibitors has been seen as non-physiological, since depleting the ECF of ACh (Benveniste and Huttemeier, 1990) will artiﬁcially decrease receptor occupancy and hence the inhibitory feedback. The presence of ChE inhibitors in the perfusion ﬂuid allows to shorten sample collection time during performance of a behavioral task, thus enabling a correlation between ACh release and memory acquisition, encoding, or recall. For instance, a direct correlation between object recognition and increased activity of the cortical cholinergic network is not possible (Himmelheber et al., 1997; Giovannini et al., 1998). Object recognition, a simple spontaneous task involving exploratory activity and working memory, takes a few seconds only, while to detect changes in ACh release by microdialysis a collection time of at least 5 min is needed even with a sensitivity of the ACh assay of 100 fmol/injection (Giovannini et al., 1998). Present microdialysis time resolution (5 min, at the shortest, see Table 1) is insufﬁcient to detect transient events. More sensitive methods are therefore needed to collect samples for shorter time, with the temporal resolution needed to enable the visualization of transient changes of ACh release.

rabbit, cat, and monkey. Microdialysis in the hippocampus of rabbit has been used to study the involvement of the cholinergic system in the classical conditioned rabbit nictitating reﬂex (Meyer et al., 1996). The implication of ACh release from the basal forebrain and pontine reticular formation (Vazquez and Baghdoyan, 2001, 2004) in REM sleep was deﬁned by microdialysis experiments using the cat, while rhesus monkeys (Macaca mulatta) have been used to study the effects of aging on ACh release (Tsukada et al., 2000). Furthermore, Tang and Aigner (1996) studied ACh release from the inferior temporal cortex (IT), perirhinal cortex (PR), and dentate gyrus (DG) of the hippocampus of rhesus monkeys performing a visual recognition task and a memoryindependent task. The results provide biochemical evidence for cerebral cholinergic system activation during visually mediated behavior in non-human primates, possibly a prerequisite for visual recognition memory. Microdialysis has been adapted to the mouse to study the cholinergic system in different transgenic mouse strain (Erb et al., 2001; Bellucci et al., 2004). II.F. The cholinergic pathways involved in cognitive processes investigated by microdialysis According to the description of Mesulam et al. (1983), the basal forebrain contains four cholinergic cell groups:

II.E. Animals: rodents, monkeys, cats Although the animal of choice for the study of ACh involvement in cognitive processes by microdialysis is the rat, some studies have been performed on different animals such as the mouse,

The Ch1 and Ch2 groups of septal and vertical diagonal band nuclei provide the major cholinergic projections of the hippocampus. The activity of this pathway is investigated by implanting the probes in the ventral or dorsal hippocampus. The Ch3 cell group, located within the horizontal diagonal band nucleus, provides the major cholinergic innervation to the olfactory bulb. This pathway is not directly involved in cognitive processes and has not been the target of microdialysis studies. The Ch4 cell group, located within the nucleus basalis (NB), provides the major cholinergic innervation of the amygdala and cerebral cortex. Probes have been implanted

386

in different cortical areas and in the amygdala. The cholinergic projection from the basal forebrain to the neocortex has been described as either diffuse or highly speciﬁc. This discrepancy depends on the type of stimulation and the cortical areas investigated. Fournier et al. (2004) demonstrated in anesthetized rats that visual stimulation evoked a large increase in ACh release from the visual cortex, little change from the somatosensory cortex, and no change in the frontal cortex. Skin stimulation evoked a large increase from the somatosensory cortex. Conversely, stimulations inducing arousal and attention induce a similar increase in ACh levels in all cortical areas indicating that changes in the activity of the cholinergic cortical network are not area speciﬁc, as argued by Sarter and Bruno (2000). In the rat, the forebrain cholinergic neurons invade the innermost third part of the globus pallidus in a diffuse manner. In addition, large cholinergic neurons reside in the striatum and other less conspicuous cholinergic cell groups are found in the thalamus (midline, intralaminar). Probes have been implanted in the nucleus accumbens and in the striatum to study the release from the intrinsic cholinergic neurons, and in the thalamus. The probes in the thalamus have been used to study the activity of the cholinergic pathway originating in the mesopontine region during the sleep cycle (Williams et al., 1994). Cholinergic neurons have been detected also in the epithalamus (medial habenula), and hypothalamus (posterior region and arcuate nucleus). In the brainstem, the motor and autonomic preganglionic neurons of the cranial nerve are cholinergic, and cholinergic neurons form a diffuse reticular system and several nuclei including the peduncolopontine tegmental (Ch5), lateral dorsal tegmental (Ch6), and parabrachial (Cuello and Sofroniew, 1984), projecting to the thalamus and forebrain region. Probes have been implanted in the hypothalamus to study the effect of stress on neurotransmitter release (Gotoh et al., 1998) and in the brainstem ventral tegmental area to study the effect of self-stimulation, eating and drinking

on ACh release from the peduncolopontine tegmental, lateral dorsal tegmental nuclei (Rada et al., 2000).

II.G. Handling of the animals: stress A major problem in the study of ACh release during cognitive process is whether the susceptibility of un-anesthetized animals to stress due to handling and environmental changes introduces artefacts in neurotransmitter measurements. Indeed, several reports indicate that simply moving the animals from their home cage to the dialysis cage (Acquas et al., 1996; Giovannini et al., 1998; Thiel et al., 1998b), or handling the animal (Thiel et al., 1998a) causes a signiﬁcant increase in the release of ACh from different brain regions. This can be overcome by allowing the animal to habituate to the experimenter, the room, and the cage. It has been reported that, when the animals are handled the day before the experimental session, the simple manipulation for transferring them from their home cage to the arena does not modify signiﬁcantly the release of ACh from the cortex and hippocampus (Giovannini et al., 1998, 2001; Thiel et al., 1998a, b; Day et al., 2001; Bianchi et al., 2003).

III. Microdialysis studies in animals performing spontaneous behaviors involving congnitive processes III.A. Response to novelty Novelty can be a sensory stimulation or an unknown environment. The reaction to novelty is a behavioral arousal which can be associated with alert immobility (freezing) or exploratory activity. The electroencephalogram shows low-voltage fast activity in the cerebral cortex (Weinberger and Lindsley, 1964) and rhythmical slow activity or ‘‘theta’’ rhythm in the hippocampus (Whishaw and Vanderwolf, 1973). From a cognitive viewpoint, novelty elicits attention, followed by information acquisition and analysis of its relevance. Therefore, exposure to novelty can be considered

387

the simplest cognitive test, with the caveat that novelty also represents a stressful condition. III.A.1. Sensory stimulations Inglis and Fibiger (1995) demonstrated that the presentation of four different sensory stimuli, auditory, visual, olfactory, and tactile, evoked an immediate increase in ACh release in both hippocampus and frontal cortex lasting about 20 min. In the frontal cortex, the increase following a tactile stimulation was larger than that caused by other stimuli. Even if all the rats showed an activation of the cortical and hippocampal cholinergic systems, they presented different behavioral responses varying from freezing during the tactile stimulus to exploratory behavior after the visual stimulus, a ﬁnding demonstrating that the cholinergic system is a constant component but not the only one of the response to sensory stimulation. Indeed, there is evidence that the cortical cholinergic system is modulated by noradrenergic, GABAergic, and dopaminergic inputs, presumably at the level of its origin in the NB (Acquas et al., 1998). It should be noted that the cholinergic neurons of the NB ascending to the cortex and of the septo-hippocampal pathway respond as a single-nuclear group to sensory stimulation. A novel taste also induces an activation of the cortical cholinergic network (Miranda et al., 2000). However, the increase in ACh release was statistically signiﬁcant only in the insular cortex which is involved in the mnemonic gustatory representation but not in the parietal cortex, a ﬁnding demonstrating that taste activates a speciﬁc group of cholinergic neurons. Feeding, after food deprivation, which represents a mixture of familiar sensory stimulation and reward stimulation, did not increase ACh release in the hippocampus but only in the nucleus accumbens; while drinking, after water deprivation, caused an increase also in the striatum (Mark et al., 1992). III.A.2. Novel environment When the rats were moved from their home cage to the experimental cage, a twofold increase in ACh release was observed in the hippocampus and a fourfold in the cortex (Inglis and Fibiger, 1995). A 50% increase in ACh release was observed also

(Aloisi et al., 1997) when rats were introduced for 30 min in an open ﬁeld. The peak was reached within the ﬁrst 10 min of exposure to the novel environment and was accompanied by sustained exploratory activity whose rapid decrease was paralleled by the return of ACh release to basal. Giovannini et al. (1998) investigated the release of ACh from the frontal cortex during object recognition. The exploratory activity during object recognition was monitored by microwave sensors placed above the arena. Young rats were placed in the arena for 5 min and left to explore the same objects twice with an intertrial of 60 min and then a third time after 30 min with two differently shaped objects, one familiar and one new. The rats recognized the new object as demonstrated by the longer time spent in its exploration, in comparison with that spent with the familiar object. Every time that the rats were placed in the arena there was a sharp increase in ACh release ranging between 60 and 100%. Twenty-ﬁve-month-old rats, unable to discriminate between familiar and new object, showed no statistically signiﬁcant increase in ACh release, a ﬁnding indicating a relationship between the ability to discriminate the new object and the activity of the cortical cholinergic system. Giovannini et al. (2001) also observed that, in rats exposed to a novel environment for 30 min, there was a much larger increase in ACh release in the hippocampus than in the frontal cortex (Fig. 2). As seen after sensory stimulation, the cholinergic neurons projecting to the cerebral cortex and to the hippocampus responded to novelty as a single group. The increase in ACh was accompanied by an increase in motor activity, but no linear relationship was detected between the two events, as also observed by Bianchi et al. (2003).

III.A.3. Habituation Repeated exposure to a stimulus and repeated or prolonged exposure to an environment rapidly lead to habituation. Habituation is a non-associative form of implicit memory (Bailey et al., 2004) and indicates that the animal has acquired and stored information received in the ﬁrst exposure. Habituation is a gradual reduction of response to a stimulus and involves decline of existing response (Dudai, 1989).

388

A

B 250

ACh release(% of basal)

Cortex 200

150

100

C 350 0

Hippocampus

60 min

24 h

Time

250

D 750 AUC (arbitrary units)

ACh release (% of basal)

50

150

50

500 250

0 0

60 min

0 min

60 min

24 h

Time

Fig. 2. Novelty increases cortical and hippocampal ACh release. (A) Rat exploring a novel environment. (B) ACh extracellular levels in the cerebral cortex during: (1) 30 min exploratory period; (2) exploratory period 60 min later; and (3) exploratory period after 24 h. (C) ACh extracellular levels in the hippocampus during the ﬁrst and second exploratory periods. (D) Changes in ACh extracellular levels expressed as area under the curve: white columns ¼ cerebral cortex; black columns ¼ hippocampus. Note the habituation after 60 min and the loss of the memory of the environment after 24 h. Data taken from Giovannini et al. (1998, 2001). With permission from Elsevier.

The cholinergic forebrain systems undergo habituation. Acquas et al. (1996) demonstrated that after prolonged exposition to a tone and light stimulus during the training sessions, in the test session the rats showed no increase in ACh release in the frontal cortex and hippocampus and no increase in motor activity. Conversely, in naı¨ ve rats, the tone and light stimulus induced a sharp increase in ACh release in both regions and a marked fear-related behavior. According to Miranda et al. (2000), repeated taste stimulations evoked gradually smaller increase in ACh release

in the insular cortex indicating an inverse relation between taste familiarity and cortical ACh release. Giovannini et al. (1998) observed no habituation in the cholinergic and behavioral responses in the second exposure after an intertrial time of 60 min, if the ﬁrst exposure to a novel environment lasted only 5 min. Conversely, if the rats were left in the novel environment for 30 min, a second exposure 60 min later elicited a much smaller increase in ACh release in the cerebral cortex and hippocampus than the ﬁrst, as shown in panels B, C, and D of Fig. 2,

389

and much less exploratory activity (Giovannini et al., 2001), as conﬁrmed also by Bianchi et al. (2003). However, when the intertrial interval was 24 h, there was no decrement in the response of the cholinergic system and motor activity between the ﬁrst and second exposition. These experiments indicate that the onset of habituation to a novel environment, devoid of salient stimuli, requires an exposition longer than 5 min and that memory of the environment is lost within 24 h. Similarly, Thiel et al. (1998b) found no difference in the cholinergic activation between the ﬁrst 10 min exposure to an open ﬁeld and re-exposure 24 h later.

III.A.4. Relationship between ACh release and motor activity Contradictory results were obtained when attempting to correlate ACh release and motor activity, taken as a measure of exploratory activity. A correlation between spontaneous motility and ACh release from the striatum was observed by Watanabe et al. (1990), under conditions minimizing the effects of stress and novelty. A correlation between ACh release from the cerebral cortex, hippocampus, and striatum and spontaneous motility, considered as a measure of behavioral arousal, was observed in the rat also by Day et al. (1991). Mizuno et al. (1991) demonstrated a relationship between ACh release from the hippocampus and spontaneous motor activity, irrespective of the circadian rhythm of ACh release. Bianchi et al. (2003) observed a correlation between ACh release in the ventral hippocampus and exploratory activity in a novel environment. However, this correlation was not always found. Giovannini et al. (2001), placing the rats twice in the same novel environment, observed no correlation between ACh release from dorsal hippocampus and motor activity at the ﬁrst exposure, but only at the second exposure. Thiel et al. (1998b) observed no correlation between ACh release and rearing at the second exposure. These discrepancies may depend on the functional differences between the dorsal and ventral part of the hippocampus (Hock and Bunsey, 1998), and demonstrate that the increase in ACh release has several components besides motor activity, including

stress which is always present when a rat is facing a novel environment.

IV. Acetylcholine release during attention Novelty induces arousal and attention. Arousal has been deﬁned as ‘‘some level of non-speciﬁc neuronal excitability deriving from structure formerly known as the reticular formation’’ (Robbins and Everitt, 1995) while ‘‘attention is a fuzzy concept used in a variety of difference sense ranging from orienting to selection and to vigilance or sustained attention’’ (Treisman, 2004). Arousal is associated with an increase in extracellular ACh levels in the cerebral cortex (see Pepeu, 1973). Arousal is a prerequisite of attention and therefore activation of the cholinergic system underlies the attentional mechanisms. This is conﬁrmed by the observation that tasks requiring attention are impaired by administration of anticholinergic drugs (Jones and Higgins, 1995) and by selective lesions of the NB (Sarter and Bruno, 2000; Chudasama et al., 2004). Microdialysis studies have been aimed to deﬁne if there is a speciﬁc relationship between cortical cholinergic output and attentional performance. Passetti et al. (2000) demonstrated a 100% increase in ACh release in the prefrontal cortex during the performance of a ﬁve-choice serial reaction time task in which the rats were trained to detect the brief ﬂashes of light presented randomly in one of ﬁve spatial locations to obtain a food reward. The test obliged the animals to sustain and divide the attention between the ﬁve locations. If the demand on the rats was increased or decreased by modulating the stimulus duration, there was no evidence of changes in the increase of ACh release in the more or less demanding conditions. Himmelheber et al. (2001) observed a marked increase in ACh release from the frontoparietal cortex during both the performance of an operant task requiring sustained attention, and a low-demand operant task. Conversely, if ACh release in the frontoparietal cortex of rats performing a task requiring sustained attention is compared with that of rats performing two simpler tasks in the same operant chamber, an increase of nearly 140% was

390

Ach EFFLUX (% change from mean baseline)

150 Sus, Attn, Task Retrac, Lev, Task Fl-9 Task

125 100 75 50 25 0 -25

behavioral task

bs

fn 1 bs fn 2 bs fn 3 bs fn 4 bl oc k bl 1 oc k bl 2 oc k bl 3 oc k bl 4 oc k bl 5 oc k po 6 st ta s po k st 1 ta po sk st 2 t po ask st 3 t po ask st 4 ta sk 5

-50

Collection Intervals (6 min) Fig. 3. ACh release during tasks involving different levels of attention. U: Sustained attention; U: simple lever pressing; D: press a lever on a ﬁxed interval. Data taken from Arnold et al. (2002). With permission from Elsevier.

observed in the sustained attention task and an increase of 50% in the easier tasks (Arnold et al., 2002) as illustrated by Fig. 3. Himmelheber et al. (2000) detected a stricter relationship between attentional effort and ACh release. They found that transferring animals from the home cage to the operant chamber elicited an initial increase in cortical ACh efﬂux not related to novelty, since the rats had been trained in the chamber, but presumably due to food reward expectation (see Inglis and Fibiger, 1995). In the operant chamber, the performance of a sustained attention task was accompanied by a further increase in cortical ACh efﬂux and a further increase was caused by a visual distracter loading the attentional demand. Recently, it has been shown that in rats trained in a sustained attention task, associated with a 140% increase in prefrontal cortex ACh release, if the animal’s ability to detect signals was attenuated by bilateral basal forebrain infusion of a NMDA antagonist, a further increase in ACh efﬂux was observed. It appears that the rats make an effort to maintain the performance under challenging

conditions by straining the activation of the cholinergic system (Kozak et al., 2006). Similarly, a decline in attentional performance but an increase in cortical ACh efﬂux was observed during presentation of a distracter while the rats were performing a sustained attention task (Himmelheber et al., 2000). In conclusion, while all experiments demonstrate an increase in cortical ACh release during sustained attention tasks, it may be assumed that the type and difﬁculty of the task, the duration of training, the presence of distracters, and/or the efﬁciency of the basal forebrain cholinergic neurons are variables inﬂuencing the correlation between level of attention and ACh release. Sustained attention is a complex process controlled by ascending and descending inputs. The prefrontal cortex plays a pivotal role in the ‘‘top–down’’ control (see Sarter et al., 2001). Nelson et al. (2005) observed that ACh release from the posterior parietal cortex is regulated by the prefrontal cortex via glutamatergic and cholinergic mechanisms. Basal forebrain glutamate

391

(G. Pepeu).receptors contribute to the stimulation of cortical ACh efﬂux in response to behavioral stimuli (Fadel et al., 2001), and GABA receptors modulate attention and the activity of the basal forebrain cholinergic neurons (Sarter and Bruno, 1997). V. Acetylcholine release in learning, memory, and recall Attention is a crucial early stage in learning. It may be assumed that the increase in ACh release occurring during attention plays a role in information acquisition. In contrast, it is impossible to study information acquisition separately from attention.

inhibitory conditioned avoidance response in the rat, Giovannini et al. (2005) observed a threefold increase in ACh release from the prefrontal cortex and ventral hippocampus during training and a smaller but consistent increase during recall, 60 min later. However, the functional meaning of the two increases was different. Only ACh increase during training activated the ERK cascade, and only the acquisition of the conditioned response but not the recall was prevented by muscarinic receptor blockade and by an inhibitor of ERK activation. This ﬁnding demonstrated that the increase in ACh induced by the novel environment and the painful electric shock was important for the formation of the conditioned inhibitory response.

V.A. Conditioned responses

V.B. Other forms of memory

Conditioning is a simple form of associative learning. If sensory stimulation, which evokes arousal and attention, is associated with an aversive stimulation, for example, a foot shock, it acquires the meaning of a conditioned stimulus. Acquas et al. (1996), by monitoring simultaneously ACh release in the frontal cortex and hippocampus, demonstrated that both novel, unconditioned stimulation, and conditioned stimulation, eliciting a fear behavior, induced a marked increase in ACh release in both regions. A conditioned stimulus reminding the animal of an aversive event, such as nausea induced by lithium chloride paired with saccharin ﬂavored water, caused an increase in ACh release in the nucleus accumbens, whose intrinsic cholinergic neurons do not respond to stressful or arousing events (Mark et al., 1995). The same aversive stimulus also caused a decrease in dopamine release suggesting a dopaminergic modulation of the cholinergic interneurons of the nucleus accumbens (Mark et al., 1991). A small increase in ACh release in the rat hippocampus was observed during the aversive stimulation (electric shocks) in the acquisition session of a contextual fear conditioning but a sixfold increase occurred in the retention session (NailBoucherie et al., 2000). Investigating ACh release during the acquisition and recall of a step-down

Complex operant behaviors have been used to train rats to attentional efforts to obtain food reward and to investigate the relationship between attention and ACh release. Simple operant tasks have been used to investigate the release of ACh during memory formation. Orsetti et al. (1996) showed that in untrained rats placed in an operant chamber both cortical and hippocampal ACh release remained constant until the number of reinforced responses remained low but increased sharply, together with the rise in reinforced responses, when the rats were learning that by pressing the lever they could obtain food. Conversely, lever pressing for food reward in trained rats was not accompanied by an increase in ACh release. These ﬁndings demonstrate that activation of the forebrain cholinergic pathways only occurs during the acquisition of the operant response. Investigating extinction in a similar operant task, Izaki et al. (2001) observed that in rats trained to obtain food reward there was an increase in ACh release in the prefrontal cortex when lever pressing produced no reward, and there was an inverse relationship between extinction of the lever-press responses and ACh release. In the above experiments, the increase in ACh release was concomitant with the acquisition of an information which was positive in the ﬁrst case (reward after

392

lever-press, therefore keep pressing) and negative in the second (no reward after lever-press, therefore stop pressing). The relationship between working memory, reference memory, and ACh release was investigated by Hironaka et al. (2001) using two paradigms in an operant chamber: a delayed alternation task for working memory and a cued alternation for reference memory. During the performance of the delayed alternation task there was a twofold increase in ACh release in the prefrontal cortex with no signiﬁcant increase in the hippocampus, due to a large individual variability. Conversely, in the cued alternation there was a signiﬁcant ACh increase only in the hippocampus. No simple explanation can be offered for the different role of the cortical and hippocampal cholinergic systems in the two types of memory. However, this experiment show that the forebrain cholinergic neurons not always respond as a single group as during sensory stimulation (Inglis and Fibiger, 1995). An increase in hippocampal ACh release was also observed during testing in a rewarded spontaneous alternation task in a Y maze, with a slight decrease in the forth 5 min testing period. In contrast, striatum ACh levels increased gradually throughout the testing reaching the maximal values in the fourth period. According to Pych et al. (2005a), the progressive shift in ACh release from the hippocampus to the striatum coincides with a change in the rat’s strategy from alternation of place to a response perseveration strategy. The switching of cholinergic systems, from hippocampus to striatum, with changing strategy to obtain food reward had been observed also in rats trained in a cross maze (Chang and Gold, 2003). Rats initially showed learning on the basis of place, together with a rapid and steady increase in hippocampal ACh release. After extensive training they shifted to learning on the base of response, concomitantly with a gradual increase in striatal ACh release. The different role of the hippocampal versus the striatal cholinergic systems in learning a food-rewarded master was further investigated (Pych et al., 2005b) by testing the rats in cue rich and cue-poor conditions. Initially, ACh release increased similarly under both cue conditions, but declined during training on the cue-poor

condition, when spatial processing by the hippocampus was not suitable for solving the maze. It may be added that intrahippocampal glucose injections enhanced ACh increase in the hippocampus during training and improved the spontaneous alternation performance (Ragozzino et al., 1998), a further demonstration of the importance of the septo-hippocampal cholinergic system in spatial memory formation. The hippocampal cholinergic system is also involved in mastering a food reinforced radial-arm maze. Stancampiano et al. (1999) investigated ACh and 5-HT release from the hippocampus of rats trained to master a foodreinforced radial-arm maze and reported an increase in ACh release in the waiting period and a further increase during the performance of the task indicating the involvement of the cholinergic system in food expectation and during attention and memory. They also demonstrated that the increase in 5-HT release was implicated in feeding behavior but not in memory function. Fadda et al. (2000) observed that training rats for 12 days to master a radial maze for food resulted in a progressive increase in ACh release in the hippocampus from 139 to 245%, while the rats were performing the task, and a decrease in the number of errors, as if training was facilitating the activation of the cholinergic neurons. McIntyre et al. (2002) also observed that ACh release in the hippocampus increased while rats performed a spontaneous alternation task. ACh release in the hippocampus increased also when the rats acquired an amygdala-dependent food conditioned place-preference task. However, the magnitude of the increased release from the hippocampus was negatively correlated with good performance in the conditioned place-preference task for which the activation of hippocampal cholinergic system seem to be detrimental.

VI. Conclusions From the papers reviewed in this chapter, microdialysis appears to be a powerful tool for investigating the role of brain cholinergic system in speciﬁc areas and speciﬁc cognitive processes. Using this tool, it has been possible to conﬁrm in a

393

dynamic form the extensive involvement of the brain cholinergic pathways in the ongoing cognitive activity which had been previously demonstrated by selective lesions of the cholinergic pathways in animals and muscarinic and nicotinic receptor antagonists and agonists in animals and man (see ref. in Pepeu and Giovannini, 2006). The activation of the cholinergic neurons is a widespread event and may either be non-selective as in the cerebral cortex and hippocampus during sustained attention, or selective as in the hippocampus and in the striatum during spatial and reference memory formation, respectively. However, more investigations are needed in which ACh release is measured simultaneously from two or more circumscribed brain areas during the performance of tasks involving different types of memory to build a clear picture of the relative importance of the various cholinergic pathways. According to Hasselmo and McGaughy (2004), the increase in ACh release serves to enhance, mostly by volume transmission, the inﬂuence of afferent inputs on neuronal ﬁring while reducing the inﬂuence of feedback processing thus facilitating the maintenance of consciousness and memory formation.

References Acquas, E., Wilson, C. and Fibiger, H.C. (1996) Conditioned and unconditioned stimuli increase frontal cortical and hippocampal acetylcholine release: effects of novelty, habituation, and fear. J. Neurosci., 16: 3089–3096. Acquas, E., Wilson, C. and Fibiger, H.C. (1998) Pharmacology of sensory stimulation-evoked increases in frontal cortical acetylcholine release. Neuroscience, 85: 73–83. Aloisi, A.M., Casamenti, F., Scali, C., Pepeu, G. and Carli, G. (1997) Effects of novelty, pain and stress on hippocampal extracellular acetylcholine levels in male rats. Brain Res., 748: 219–226. Arnold, H.M., Burk, J.A., Hodgson, E.M., Sarter, M. and Bruno, J.P. (2002) Differential cortical acetylcholine release in rats performing a sustained attention task versus behavioral control tasks that do not explicitly tax attention. Neuroscience, 114: 451–460. Bailey, C.H., Kandel, E.R. and Si, K. (2004) The persistence of long-term memory: a molecular approach to self-sustaining changes in learning-induced synaptic growth. Neuron, 44: 49–57. Bellucci, A., Luccarini, I., Scali, C., Pepeu, G. and Casamenti, F. (2004). Alzheimer’s disease pathology and cholinergic dysfunction in TgCRND8 mice. Soc. Neurosci. Abstr., 337.13.

Benveniste, H. (1989) Brain microdialysis. J. Neurochem., 52: 1667–1678. Benveniste, H. and Diemer, N.H. (1987) Cellular reactions to implantation of a microdialysis tube in the rat hippocampus. Acta Neuropathol. (Berl.), 74: 234–238. Benveniste, H. and Huttemeier, P.C. (1990) Microdialysis-theory and application. Progr. Neurobiol., 35: 195–215. Bertorelli, R., Hallstrom, A., Hurd, Y.L., Karlsson, A., Consolo, S. and Ungerstedt, U. (1990) Anaesthesia effects on in vivo acetylcholine transmission; comparisons of radioenzymatic and HPLC assays. Eur. J. Pharmacol., 175: 79–83. Bianchi, L., Ballini, C., Colivicchi, M.A., Della, C.L., Giovannini, M.G. and Pepeu, G. (2003) Investigation on acetylcholine, aspartate, glutamate and GABA extracellular levels from ventral hippocampus during repeated exploratory activity in the rat. Neurochem. Res., 28: 565–573. Cangioli, I., Baldi, E., Mannaioni, P.F., Bucherelli, C., Blandina, P. and Passani, M.B. (2002) Activation of histaminergic H3 receptors in the rat basolateral amygdala improves expression of fear memory and enhances acetylcholine release. Eur. J. Neurosci., 16: 521–528. Casamenti, F., Deffenu, G., Abbamondi, A.L. and Pepeu, G. (1986) Changes in cortical acetylcholine output induced by modulation of the nucleus basalis. Brain Res. Bull., 16: 689–695. Celesia, G.G. and Jasper, H.H. (1966) Acetylcholine released from cerebral cortex in relation to state of activation. Neurology, 16: 1053–1063. Chang, Q. and Gold, P.E. (2003) Switching memory systems during learning: changes in patterns of brain acetylcholine release in the hippocampus and striatum in rats. J. Neurosci., 23: 3001–3005. Chudasama, Y., Dalley, J.W., Nathwani, F., Bouger, P. and Robbins, T.W. (2004) Cholinergic modulation of visual attention and working memory: dissociable effects of basal forebrain 192-IgG-saporin lesions and intraprefrontal infusions of scopolamine. Learn. Mem., 11: 78–86. Consolo, S., Wu, C.F., Fiorentini, F., Ladinsky, H. and Vezzani, A. (1987) Determination of endogenous acetylcholine release in freely moving rats by trans striatal dialysis coupled to a radioenzymatic assay: effect of drugs. J. Neurochem., 48: 1459–1465. Cuello, A.C. and Sofroniew, M.V. (1984) The anatomy of the CNS cholinergic neurons. Trends Neurosci., 7: 74–78. Dalley, J.W., McGaughy, J., O’Connell, M.T., Cardinal, R.N., Levita, L. and Robbins, T.W. (2001) Distinct changes in cortical acetylcholine and noradrenaline efﬂux during contingent and noncontingent performance of a visual attentional task. J. Neurosci., 21: 4908–4914. Damsma, G., Lammerts Van Bueren, D., Westerink, B.H.C. and Horn, A.S. (1987) Determination of acetylcholine in the femtomole range by means of HPLC, a post-column enzyme reactor, and electro-chemical detection. Chromatographia, 24: 827–831. Damsma, G., Westerink, B.H., de Boer, P., De Vries, J.B. and Horn, A.S. (1988) Basal acetylcholine release in freely moving rats detected by on-line trans-striatal dialysis: pharmacological aspects. Life Sci., 43: 1161–1168.

394 Day, J., Damsma, G. and Fibiger, H.C. (1991) Cholinergic activity in the rat hippocampus, cortex and striatum correlates with locomotor activity: an in vivo microdialysis study. Pharmacol. Biochem. Behav., 38: 723–729. Day, J.C., Kornecook, T.J. and Quirion, R. (2001) Application of in vivo microdialysis to the study of cholinergic systems. Methods, 23: 21–39. de Boer, P., Damsma, G., Fibiger, H.C., Timmerman, W., De Vries, J.B. and Westerink, B.H. (1990) Dopaminergic– cholinergic interactions in the striatum: the critical signiﬁcance of calcium concentrations in brain microdialysis. Naunyn Schmiedebergs Arch. Pharmacol., 342: 528–534. DeFeudis, F.D. (1974) Supporting studies on the central release of acetylcholine. In: DeFeudis, F.D. (Ed.), Central Cholinergic Systems and Behaviour. Academic Press, London. Dudai, Y. (1989) The Neurobiology of Memory. Oxford University Press, Oxford. Erb, C., Troost, J., Kopf, S., Schmitt, U., Loffelholz, K., Soreq, H. and Klein, J. (2001) Compensatory mechanisms enhance hippocampal acetylcholine release in transgenic mice expressing human acetylcholinesterase. J. Neurochem., 77: 638–646. Fadda, F., Cocco, S. and Stancampiano, R. (2000) Hippocampal acetylcholine release correlates with spatial learning performance in freely moving rats. Neuroreport, 11: 2265–2269. Fadel, J., Sarter, M. and Bruno, J.P. (2001) Basal forebrain glutamatergic modulation of cortical acetylcholine release. Synapse, 39: 201–212. Fisone, G., Wu, C.F., Consolo, S., Nordstrom, O., Brynne, N., Bartfai, T., Melander, T. and Hokfelt, T. (1987) Galanin inhibits acetylcholine release in the ventral hippocampus of the rat: histochemical, autoradiographic, in vivo, and in vitro studies. Proc. Natl. Acad. Sci. U.S.A., 84: 7339–7343. Fournier, G.N., Semba, K. and Rasmusson, D.D. (2004) Modality- and region-speciﬁc acetylcholine release in the rat neocortex. Neuroscience, 126: 257–262. Gaddum, J.H. (1961) Push-pull cannulae. J. Physiol. (Lond.), 155: 1P–2P. Giovannini, M.G., Bartolini, L., Kopf, S.R. and Pepeu, G. (1998) Acetylcholine release from the frontal cortex during exploratory activity. Brain Res., 784: 218–227. Giovannini, M.G., Casamenti, F., Nistri, A., Paoli, F. and Pepeu, G. (1991) Effect of thyrotropin releasing hormone (TRH) on acetylcholine release from different brain areas investigated by microdialysis. Br. J. Pharmacol., 102: 363–368. Giovannini, M.G., Pazzagli, M., Malmberg-Aiello, P., Della Corte, L., Rakovska, A.D., Cerbai, F., Casamenti, F. and Pepeu, G. (2005) Inhibition of acetylcholine-induced activation of extracellular regulated protein kinase prevents the encoding of an inhibitory avoidance response in the rat. Neuroscience, 136: 15–32. Giovannini, M.G., Rakovska, A., Benton, R.S., Pazzagli, M., Bianchi, L. and Pepeu, G. (2001) Effects of novelty and habituation on acetylcholine, GABA, and glutamate release from the frontal cortex and hippocampus of freely moving rats. Neuroscience, 106: 43–53.

Gotoh, M., Tajima, T., Suzuki, Y., Ikari, H., Iguchi, A., Kakumu, S. and Hirooka, Y. (1998) Swimming stress that causes hyperglycemia increases in vivo release of noradrenaline, but not acetylcholine, from the hypothalamus of conscious rats. Brain Res., 780: 74–79. Hajnal, A., Pothos, E.N., Lenard, L. and Hoebel, B.G. (1998) Effects of feeding and insulin on extracellular acetylcholine in the amygdala of freely moving rats. Brain Res., 785: 41–48. Hasselmo, M.E. and McGaughy, J. (2004) High acetylcholine levels set circuit dynamics for attention and encoding and low acetylcholine levels set dynamics for consolidation. Prog. Brain Res., 145: 207–231. Himmelheber, A.M., Fadel, J., Sarter, M. and Bruno, J.P. (1998) Effects of local cholinesterase inhibition on acetylcholine release assessed simultaneously in prefrontal and frontoparietal cortex. Neuroscience, 86: 949–957. Himmelheber, A.M., Sarter, M. and Bruno, J.P. (1997) Operant performance and cortical acetylcholine release: role of response rate, reward density, and non-contingent stimuli. Brain Res. Cogn. Brain Res., 6: 23–36. Himmelheber, A.M., Sarter, M. and Bruno, J.P. (2000) Increases in cortical acetylcholine release during sustained attention performance in rats. Brain Res. Cogn. Brain Res., 9: 313–325. Himmelheber, A.M., Sarter, M. and Bruno, J.P. (2001) The effects of manipulations of attentional demand on cortical acetylcholine release. Brain Res. Cogn. Brain Res., 12: 353–370. Hironaka, N., Tanaka, K., Izaki, Y., Hori, K. and Nomura, M. (2001) Memory-related acetylcholine efﬂux from rat prefrontal cortex and hippocampus: a microdialysis study. Brain Res., 901: 143–150. Hock, B.J. Jr. and Bunsey, M.D. (1998) Differential effects of dorsal and ventral hippocampal lesions. J. Neurosci., 18: 7027–7032. Imperato, A. and Di Chiara, G. (1984) Trans-striatal dialysis coupled to reverse phase high performance liquid chromatography with electrochemical detection: a new method for the study of the ‘‘in vivo’’ release of endogenous dopamine and metabolites. J. Neurosci., 4: 966–977. Inglis, F.M., Day, J.C. and Fibiger, H.C. (1994) Enhanced acetylcholine release in hippocampus and cortex during the anticipation and consumption of a palatable meal. Neuroscience, 62: 1049–1056. Inglis, F.M. and Fibiger, H.C. (1995) Increases in hippocampal and frontal cortical acetylcholine release associated with presentation of sensory stimuli. Neuroscience, 66: 81–86. Izaki, Y., Hori, K. and Nomura, M. (2001) Elevation of prefrontal acetylcholine is related to the extinction of learned behavior in rats. Neurosci. Lett., 306: 33–36. Jimenez-Capdeville, M.E. and Dykes, R.W. (1996) Changes in cortical acetylcholine release in the rat during day and night: differences between motor and sensory areas. Neuroscience, 71: 567–579. Jones, D.N. and Higgins, G.A. (1995) Effect of scopolamine on visual attention in rats. Psychopharmacology (Berl.), 120: 142–149.

395 Kametani, H. and Kawamura, H. (1990) Alterations in acetylcholine release in the rat hippocampus during sleep-wakefulness detected by intracerebral dialysis. Life Sci., 47: 421–426. Kanai, T. and Szerb, J.C. (1965) Mesencephalic reticular activating system and cortical acetylcholine output. Nature, 205: 80–82. Khan, S.H. and Shuaib, A. (2001) The technique of intracerebral microdialysis. Methods, 23: 3–9. Kozak, R., Bruno, J.P. and Sarter, M. (2006) Augmented prefrontal acetylcholine release during challenged attentional performance. Cereb. Cortex, 16: 9–17. Kurosawa, M., Sato, A. and Sato, Y. (1989) Stimulation of the nucleus basalis of Meynert increases acetylcholine release in the cerebral cortex in rats. Neurosci. Lett., 98: 45–50. Laplante, F., Morin, Y., Quirion, R. and Vaucher, E. (2005) Acetylcholine release is elicited in the visual cortex, but not in the prefrontal cortex, by patterned visual stimulation: a dual in vivo microdialysis study with functional correlates in the rat brain. Neuroscience, 132: 501–510. Lo Conte, G., Bartolini, L., Casamenti, F., Marconcini-Pepeu, I. and Pepeu, G. (1982) Lesions of cholinergic forebrain nuclei: changes in avoidance behavior and scopolamine actions. Pharmacol. Biochem. Behav., 17: 933–937. MacIntosh, F.C. and Oboring, P.E. (1955) Release of acetylcholine from the intact cerebral cortex. Abstr. XIX Int. Physiol. Congr., pp. 580–581. Mark, G.P., Blander, D.S. and Hoebel, B.G. (1991) A conditioned stimulus decreases extracellular dopamine in the nucleus accumbens after the development of a learned taste aversion. Brain Res., 551: 308–310. Mark, G.P., Rada, P., Pothos, E. and Hoebel, B.G. (1992) Effects of feeding and drinking on acetylcholine release in the nucleus accumbens, striatum, and hippocampus of freely behaving rats. J. Neurochem., 58: 2269–2274. Mark, G.P., Rada, P.V. and Shors, T.J. (1996) Inescapable stress enhances extracellular acetylcholine in the rat hippocampus and prefrontal cortex but not in the nucleus accumbens or amygdala. Neuroscience, 74: 767–774. Mark, G.P., Weinberg, J.B., Rada, P.V. and Hoebel, B.G. (1995) Extracellular acetylcholine is increased in the nucleus accumbens following the presentation of an aversively conditioned taste stimulus. Brain Res., 688: 184–188. McIntyre, C.K., Marriott, L.K. and Gold, P.E. (2003a) Cooperation between memory systems: acetylcholine release in the amygdala correlates positively with performance on a hippocampus-dependent task. Behav. Neurosci., 117: 320–326. McIntyre, C.K., Marriott, L.K. and Gold, P.E. (2003b) Patterns of brain acetylcholine release predict individual differences in preferred learning strategies in rats. Neurobiol. Learn. Mem., 79: 177–183. McIntyre, C.K., Pal, S.N., Marriott, L.K. and Gold, P.E. (2002) Competition between memory systems: acetylcholine release in the hippocampus correlates negatively with good performance on an amygdala-dependent task. J. Neurosci., 22: 1171–1176. Mesulam, M.M., Mufson, E.J., Levey, A.I. and Wainer, B.H. (1983) Cholinergic innervation of cortex by the basal

forebrain: cytochemistry and cortical connections of the septal area, diagonal band nuclei, nucleus basalis (substantia innominata), and hypothalamus in the rhesus monkey. J. Comp. Neurol., 214: 170–197. Meyer, J.J., Allen, D.D. and Yokel, R.A. (1996) Hippocampal acetylcholine increases during eyeblink conditioning in the rabbit. Physiol. Behav., 60: 1199–1203. Miranda, M.I. and Bermudez-Rattoni, F. (1999) Reversible inactivation of the nucleus basalis magnocellularis induces disruption of cortical acetylcholine release and acquisition, but not retrieval, of aversive memories. Proc. Natl. Acad. Sci. U.S.A., 96: 6478–6482. Miranda, M.I., Ramirez-Lugo, L. and Bermudez-Rattoni, F. (2000) Cortical cholinergic activity is related to the novelty of the stimulus. Brain Res., 882: 230–235. Mitsushima, D., Mizuno, T. and Kimura, F. (1996) Age-related changes in diurnal acetylcholine release in the prefrontal cortex of male rats as measured by microdialysis. Neuroscience, 72: 429–434. Mizuno, T., Endo, Y., Arita, J. and Kimura, F. (1991) Acetylcholine release in the rat hippocampus as measured by the microdialysis method correlates with motor activity and exhibits a diurnal variation. Neuroscience, 44: 607–612. Nail-Boucherie, K., Dourmap, N., Jaffard, R. and Costentin, J. (2000) Contextual fear conditioning is associated with an increase of acetylcholine release in the hippocampus of rat. Brain Res. Cogn. Brain Res., 9: 193–197. Nelson, C.L., Sarter, M. and Bruno, J.P. (2005) Prefrontal cortical modulation of acetylcholine release in posterior parietal cortex. Neuroscience, 132: 347–359. Nilsson, O.G., Kalen, P., Rosengren, E. and Bjorklund, A. (1990) Acetylcholine release in the rat hippocampus as studied by microdialysis is dependent on axonal impulse ﬂow and increases during behavioural activation. Neuroscience, 36: 325–338. Oldford, E. and Castro-Alamancos, M.A. (2003) Input-speciﬁc effects of acetylcholine on sensory and intracortical evoked responses in the ‘‘barrel cortex’’ in vivo. Neuroscience, 117: 769–778. Orsetti, M., Casamenti, F. and Pepeu, G. (1996) Enhanced acetylcholine release in the hippocampus and cortex during acquisition of an operant behavior. Brain Res., 724: 89–96. Passetti, F., Dalley, J.W., O’Connell, M.T., Everitt, B.J. and Robbins, T.W. (2000) Increased acetylcholine release in the rat medial prefrontal cortex during performance of a visual attentional task. Eur. J. Neurosci., 12: 3051–3058. Pepeu, G. (1973) The release of acetylcholine from the brain: an approach to the study of the central cholinergic mechanisms. Prog. Neurobiol., 2: 259–288. Pepeu, G. and Giovannini, M.G. (2004) Changes in acetylcholine extracellular levels during cognitive processes. Learn. Mem., 11: 21–27. Pepeu, G. and Giovannini, M. G. (2006) The role of cholinergic system in cognitive processes. In: Giacobini, E. and Pepeu, G. (Eds.), Brain Cholinergic Mechanisms. Tailor & Francis, Oxford.

396 Puig, D.P., Paez, X., Parada, M.A., Hernandez, L., Molina, G., Murzi, E. and Contreras, Q. (1997) Changes in dopamine and acetylcholine release in the rat lateral hypothalamus during deprivation-induced drinking. Neurosci. Lett., 227: 153–156. Pych, J.C., Chang, Q., Colon-Rivera, C. and Gold, P.E. (2005a) Acetylcholine release in hippocampus and striatum during testing on a rewarded spontaneous alternation task. Neurobiol. Learn. Mem., 84: 93–101. Pych, J.C., Chang, Q., Colon-Rivera, C., Haag, R. and Gold, P.E. (2005b) Acetylcholine release in the hippocampus and striatum during place and response training. Learn. Mem., 12: 564–572. Rada, P.V., Mark, G.P., Yeomans, J.J. and Hoebel, B.G. (2000) Acetylcholine release in ventral tegmental area by hypothalamic self-stimulation, eating, and drinking. Pharmacol. Biochem. Behav., 65: 375–379. Ragozzino, M.E., Pal, S.N., Unick, K., Stefani, M.R. and Gold, P.E. (1998) Modulation of hippocampal acetylcholine release and spontaneous alternation scores by intrahippocampal glucose injections. J. Neurosci., 18: 1595–1601. Robbins, T.W. and Everitt, B.J. (1995) Arousal system in attention. In: Gazzaniga, M.S. (Ed.), The Cognitive Neuroscience. The MIT Press, Cambridge, MA. Sarter, M. and Bruno, J.P. (1997) Trans-synaptic stimulation of cortical acetylcholine and enhancement of attentional functions: a rational approach for the development of cognition enhancers. Behav. Brain Res., 83: 7–14. Sarter, M. and Bruno, J.P. (2000) Cortical cholinergic inputs mediating arousal, attentional processing and dreaming: differential afferent regulation of the basal forebrain by telencephalic and brainstem afferents. Neuroscience, 95: 933–952. Sarter, M., Givens, B. and Bruno, J.P. (2001) The cognitive neuroscience of sustained attention: where top-down meets bottom-up. Brain Res. Brain Res. Rev., 35: 146–160. Stancampiano, R., Cocco, S., Cugusi, C., Sarais, L. and Fadda, F. (1999) Serotonin and acetylcholine release response in the rat hippocampus during a spatial memory task. Neuroscience, 89: 1135–1143. Szerb, J.C., Clow, K. and Rasmusson, D.D. (1994) Pharmacological but not physiological modulation of cortical acetylcholine release by cholinergic mechanisms in the nucleus basalis magnocellularis. Can. J. Physiol. Pharmacol., 72: 893–898. Tajima, T., Endo, H., Suzuki, Y., Ikari, H., Gotoh, M. and Iguchi, A. (1996) Immobilization stress-induced increase of hippocampal acetylcholine and of plasma epinephrine, norepinephrine and glucose in rats. Brain Res., 720: 155–158. Tang, Y. and Aigner, T.G. (1996) Release of cerebral acetylcholine increases during visually mediated behavior in monkeys. Neuroreport, 7: 2231–2235. Thiel, C.M., Huston, J.P. and Schwarting, R.K. (1998a) Cholinergic activation in frontal cortex and nucleus accumbens related to basic behavioral manipulations: handling, and the role of post-handling experience. Brain Res., 812: 121–132.

Thiel, C.M., Huston, J.P. and Schwarting, R.K. (1998b) Hippocampal acetylcholine and habituation learning. Neuroscience, 85: 1253–1262. Treisman, A. (2004) Attention – introduction. In: Gazzaniga, M.S. (Ed.), The Cognitive Neurosciences III. The MIT Press, Cambridge, MA. Tsukada, H., Sato, K., Kakiuchi, T. and Nishiyama, S. (2000) Age-related impairment of coupling mechanism between neuronal activation and functional cerebral blood ﬂow response was restored by cholinesterase inhibition: PET study with microdialysis in the awake monkey brain. Brain Res., 857: 158–164. Uutela, P., Reinila, R., Piepponen, P., Ketola, R.A. and Kostiainen, R. (2005) Analysis of acetylcholine and choline in microdialysis samples by liquid chromatography/tandem mass spectrometry. Rapid Commun. Mass Spectrom., 19: 2950–2956. Vazquez, J. and Baghdoyan, H.A. (2001) Basal forebrain acetylcholine release during REM sleep is signiﬁcantly greater than during waking. Am. J. Physiol. Regul. Integr. Comp. Physiol., 280: R598–R601. Vazquez, J. and Baghdoyan, H.A. (2004) GABAA receptors inhibit acetylcholine release in cat pontine reticular formation: implications for REM sleep regulation. J. Neurophysiol., 92: 2198–2206. Watanabe, H., Shimizu, H. and Matsumoto, K. (1990) Acetylcholine release detected by trans-striatal dialysis in freely moving rats correlates with spontaneous motor activity. Life Sci., 47: 829–832. Weinberger, N.M. and Lindsley, D.B. (1964) Behavioral and electroencephalographic arousal to contrasting novel stimulation. Science, 144: 1355–1357. Westerink, B.H. (1995) Brain microdialysis and its application for the study of animal behaviour. Behav. Brain Res., 70: 103–124. Westerink, B.H.C., Damsma, G., Rollema, H., De Vries, J.B. and Horn, A.S. (1987) Scope and limitations of in vivo brain dialysis: a comparison of its application to various neurotransmitter systems. Life Sci., 41: 1763–1776. Westerink, B.H., Hofsteede, H.M., Damsma, G. and De Vries, J.B. (1988) The signiﬁcance of extracellular calcium for the release of dopamine, acetylcholine and amino acids in conscious rats, evaluated by brain microdialysis. Naunyn Schmiedebergs Arch. Pharmacol., 337: 373–378. Whishaw, I.Q. and Vanderwolf, C.H. (1973) Hippocampal EEG and behavior: changes in amplitude and frequency of RSA (theta rhythm) associated with spontaneous and learned movement patterns in rats and cats. Behav. Biol., 8: 461–484. Williams, J.A., Comisarow, J., Day, J., Fibiger, H.C. and Reiner, P.B. (1994) State-dependent release of acetylcholine in rat thalamus measured by in vivo microdialysis. J. Neurosci., 14: 5236–5242. Wu, C.F., Bertorelli, R., Sacconi, M., Pepeu, G. and Consolo, S. (1988) Decrease of brain acetylcholine release in aging freely-moving rats detected by microdialysis. Neurobiol. Aging, 9: 357–361.

CHAPTER 5.1

Microdialysis in genetically altered animals Tatyana D. Sotnikova and Raul R. Gainetdinov Department of Cell Biology, Duke University Medical Center, Duke University, Durham, NC, USA

Abstract: Recent advances in genetic approaches signiﬁcantly extended horizon of opportunities to manipulate with expression of mammalian genes in vivo resulting in development of thousands of mouse strains carrying speciﬁc genetic alterations. Microdialysis shows a great utility to analyze consequences of such genetic manipulations on the processes governing synthesis, release, uptake, and neurotransmitter interactions in a physiologically or behaviorally relevant context. In this review, a history of microdialysis in mutant animals, practical considerations of microdialysis applications in mouse genetic models, and several examples of such applications are discussed. With the increasing power and ﬁdelity of genetic manipulations and the development of uncountable number of various mutant mice, an explosion of studies using microdialysis in genetic models could be expected. Speciﬁcally, the mouse genome can be manipulated to produce the targeted deletion, replacement, or overexpression of certain proteins. In an attempt to characterize functional role of novel genes in a physiologically relevant setting, development of at least 4,000 mutant mouse strains have been reported till date. As one of the most informative and reliable tool to measure extracellular neurotransmitter concentrations, microdialysis has shown great utility in characterization of such models. The ﬁrst reports describing application of microdialysis technique to genetically altered animals are dated as early as 1993 and 1994 (Nakahara et al., 1993; Nagatsu et al., 1994), when techniques to develop transgenic mice were in infancy. These investigations give ambiguous results, however, most likely due to limitations of these early genetic approaches used to overexpress proteins of interest. In these studies, attempts to demonstrate expected elevation in dopamine (DA) synthesizing capacity in tyrosine hydroxylase (TH) overexpressing mice revealed a little impact of this mutation on TH activity in vivo (Nakahara et al., 1993; Nagatsu et al., 1994). These studies, nonetheless, were instrumental to demonstrate feasibility of microdialysis studies

I. Introduction Recent advances in molecular biology and genetics have greatly enhanced the classical, physiological, and pharmacological tools that are used to characterize the molecular mechanisms responsible for normal or pathological functions of the biological systems. Sequencing of the human genome and developments of molecular approaches to manipulate with gene expression in vivo had a tremendous impact on virtually all ﬁelds of biology and medicine. In particular, these approaches could provide an unprecedented opportunity for the development of novel therapeutics. It has been suggested that, historically, all existing drugs have targeted a total of approximately 500 molecular targets. Some have estimated that there may be as many as 5,000 new drug targets within the genome (Zambrowicz and Sands, 2003), which current estimates consider as consisting of 25,000–30,000 genes (Cravchik et al., 2001). Mouse genetics has become a powerful approach for deﬁning gene function in the mammalian physiology in vivo. Corresponding author: E-mail: [email protected]. edu

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

399

DOI: 10.1016/S1569-7339(06)16021-X Copyright 2007 Elsevier B.V. All rights reserved

400

in genetic mouse models. However, until effectiveness of homologous recombination-induced targeted deletion of genes (commonly known as knockout technique) has not been convincingly demonstrated, these transgenic models in general, and application of microdialysis in particular, has been of limited use. The ﬁrst reports of dramatic, but predictable, alterations in the extracellular DA homeostasis in mice lacking the dopamine transporter (DAT) have essentially opened a new avenue of application of microdialysis technique (Gainetdinov et al., 1997; Jones et al., 1998a). Using these mice, a clear-cut advantage of using microdialysis in combination with other neurochemical techniques to evaluate impact of genetic manipulation on brain neurochemistry has been demonstrated. The status of the monoaminergic system can be evaluated by a combination of several neurochemical approaches. Global measurement of total tissue content of a neurotransmitter can provide only a general assessment of the status of monoamine in a given region as it is determined by such important processes as synthesis, storage, metabolism, and release. Although synthesis, storage, and metabolism can be probed using relatively simple manipulations (Bannon and Roth, 1983; Carlsson et al., 2001), a reliable measurement of the most physiologically important portion of monoamine in synaptic cleft remains challenging. Contribution of released neurotransmitter to total tissue content is usually minimal and tissue level of monoamine may or may not correlate with its extracellular concentration of neurotransmitter as best illustrated by DA dynamics in DAT knockout (DATKO) mice (Gainetdinov and Caron, 2003). Two general approaches are currently available to assess extracellular dynamics of a neurotransmitter – electrochemical detection in the tissue by voltammetric or amperometric detections (Michael and Wightman, 1999; Benoit-Marand et al., 2000) and in vivo microdialysis (Ungerstedt, 1991; Westerink, 1995; Di Chiara et al., 1996). Each of these techniques has certain advantages and drawbacks. In brief, electrochemical approaches simultaneously monitor evoked release and clearance of monoamines with millisecond time resolution and relatively minor tissue damage, however,

with relatively poor sensitivity and selectivity. Generally, voltammetric and amperometric measurements reveal micromolar concentrations of stimulated DA release and do not allow reliable assessment of basal levels, although recent improvements in electrode sensitivity demonstrate such possibility (Stuber et al., 2005). Microdialysis provides a reliable assessment of the extracellular levels of DA and other neurotransmitters in nanomolar range, however, with relatively limited temporal resolution. Generally, microdialysis probes are located relatively far from release sites and the concentrations measured most likely represent extrasynaptic neurotransmitter levels that are, nonetheless, well-reﬂective of neuronal activity and release (Ungerstedt, 1991; Westerink, 1995; Di Chiara et al., 1996). At the same time, excellent selectivity, sensitivity, reproducibility, and opportunity to perform experiments in behaviorally relevant settings in freely moving animals determine wide use of microdialysis in studies aimed to understand brain neurochemistry in vivo.

II. Practical aspects of microdialysis in mice Microdialysis procedures are based on the implantation of microdialysis probes into speciﬁc brain regions of animals and collection of perfusates for subsequent analysis. The preferred animal model for genetic manipulations today is a mouse. Thus, from the point of view of microdialysis technique, a major challenge for using genetic models is adaptation of procedures generally developed for rats to smaller size of mouse brain. Mouse brain atlas is currently available (Franklin and Paxinos, 1996), and several companies developed adopted (smaller) size microdialysis probes for mice (see e.g., CMA/7, CMA Microdialysis AB, Solna, Sweden) and adaptors for ﬁxation of mouse head in stereotaxic instruments. Thus, at present, no technical problems exist for accurate placement of microdialysis probe in the relatively large brain areas such as striatum or cerebellum. With some minor adaptations, a careful placement of a probe in other brain areas including frontal cortex, hippocampus, and nucleus accumbens (NAc) is also possible (Sillaber et al., 1998; Ihalainen et al.,

401

1999; Olive et al., 2000a). However, due to the small size of these brain areas in mice, a careful histological examination of the probe placement subsequent to the experiments is required. One advantage of performing surgery in mice, in comparison to rats, is their excellent recovery following operation. With appropriate experience, it is possible to achieve nearly 100% success rate of postoperational survival of mice (Ihalainen et al., 1999). One important practical aspect of performing surgery in genetically modiﬁed mice relates to choice of appropriate anesthesia. As manipulations with genes could potentially create a situation when mutants may become more or less sensitive to a particular anesthetic, it might be necessary to explore several anesthetics to ﬁnd the most effective and safe one for a particular genetic model. In many instances, chloral hydrate remains anesthetic of choice. As in case with rats, it is critical to perform in vivo microdialysis experiments in fully recovered freely moving animals. It is well known from microdialysis studies performed in anesthetized rats that basal levels of neurotransmitters and/or responses to drugs could be potently modulated by anesthetics. Furthermore, experiments performed on freely moving animals allow direct comparison with behavioral measurements (Westerink, 1995). Mice generally well recover after surgery and 24-h recovery period should be sufﬁcient to perform reliable microdialysis experiments in freely moving animals. As genetic manipulations may unpredictably change synthesis, storage, release, or metabolism of neurotransmitter, it is becoming critical to determine basal extracellular levels of a neurotransmitter. Conventional microdialysis is commonly used to assess relative changes in neurotransmitter levels in response to stimuli or drug challenge, but due to multiple technical problems related to probe recovery in rapidly changing tissue conditions in vivo, this approach does not allow reliable estimation of basal extracellular levels of neurotransmitters (Justice, 1993; Parsons and Justice, 1994; Bungay et al., 2003). Two major quantitative approaches were developed to measure ‘‘true’’ extracellular level of a neurotransmitter: ‘‘no net ﬂux’’ and ‘‘low perfusion rate’’ microdialysis (Justice, 1993; Parsons and Justice, 1994).

Although based on different principles, both techniques provide very similar estimates of steady state DA levels (Justice, 1993). ‘‘Low perfusion rate’’ microdialysis approach is based on the fact that by reducing ﬂow rate of the perfusion, it is possible to increase recovery in vivo to 100% and thus dialysate concentrations would equal extracellular ones. In the ‘‘no net ﬂux’’ (‘‘zero net ﬂux’’) method, the infusion of a few concentrations of the studied analyte allows the direct determination of both the extracellular concentration and the in vivo recovery. Prior to the assessment of responses to pharmacological or physiological stimuli in knockout mice, a reliable determination of basal extracellular levels of neurotransmitters by quantitative approaches is strongly recommended (Gainetdinov et al., 1998; Jones et al., 1999a; Shippenberg et al., 1999). After estimation of the basal extracellular levels, it is informative to test classical responses to K+-stimulation as well as tetrodotoxin- and Ca2+-sensitivity of measured neurotransmitter. It is critically important to validate such responses to determine neuronal origin and dependence of the extracellular neurotransmitter dynamics from neuronal activity (Westerink and De Vries, 1988). Several neurotransmitter systems have been analyzed in genetically altered mice. Although majority of studies are focused on monoamines such as DA, serotonin, and norepinephrine (NE), there are growing number of studies assessing acetylcholine (ACh), histamine, and other neurotransmitters in these models. Although several groups also attempted to analyze glutamate and GABA transmission in genetically altered mice, it should be pointed out that organization of synaptic structures (Timmerman and Westerink, 1997) and/or technical peculiarities of detection of these neurotransmitters (Rea et al., 2005) seriously question validity of dialysate glutamate and GABA measurements as reﬂection of their neuronal activitydependent release. Therefore, in this review, we prefer not to discuss microdialysis studies involving measurements of these neurotransmitters. Microdialysis has proven to be a valuable approach to address numerous questions important to understand functionally relevant aspects of neurotransmission. One of the most frequently used

402

applications of microdialysis is to examine responses to pharmacological treatments by using conventional approach. Numerous genetically altered mice have been tested for such responses to deﬁne primary or secondary targets of these drugs. Microdialysis also can be used to analyze status of presynaptic machinery governing extracellular neurotransmitter dynamics, such as plasma membrane transporter activity or functional status of autoreceptor regulation. Feasibility of measurements of products of oxidative stress in several mouse models of neurodegenerative disorders has been also demonstrated. Last, but not the least, microdialysis has proven to be an excellent approach to perform pharmacokinetic studies and availability of genetically altered mice with mutations in critical proteins strongly increase potential of such applications. This review summarizes these and other applications of microdialysis in genetic mouse models. Owing to the rapid progress in the ﬁeld, and the large number of genetically modiﬁed and transgenic animals created to date, it is almost impossible to detail the ﬁndings of each and every mouse model. Therefore, a particular emphasis will be given to the best characterized neurochemically mice lacking the DAT (Giros et al., 1996) and vesicular monoamine transporter 2 (VMAT2) (Wang et al., 1997). Several other studies in which application of microdialysis to genetically altered animals has provided important clues for understanding functional role of speciﬁc proteins in physiology and pathology are also discussed.

III. Microdialysis studies in mutant mice III.A. Microdialysis studies in mice lacking the DAT and VMAT2 The DAT effectively regulates the extracellular concentrations of DA by recapturing released monoamine into the presynaptic terminals. The DAT knockout mice (Giros et al., 1996) display distinct behavioral phenotype. The DAT-KO mice are hyperactive, dwarf, demonstrate cognitive deﬁcits, and disrupted sensorimotor gating (Gainetdinov and Caron, 2003). All these behavioral

alterations are determined by functional hyperdopaminergia as best evidenced by in vivo microdialysis approach. Pharmacological evidence would predict that elimination of the DAT would result in disrupted clearance of released DA and elevated extracellular DA levels. Voltammetric and amperometric studies convincingly demonstrated remarkably prolonged clearance (>100 times) of extracellular DA in the striatum of DATKO mice (Jones et al., 1998a; Benoit-Marand et al., 2000). However, to directly prove that this prolonged clearance results in alterations in extracellular DA concentrations, a quantitative ‘‘no net ﬂux’’ microdialysis technique in freely moving mice was performed (Gainetdinov et al., 1998; Jones et al., 1998a). These studies revealed that DAT-KO mice have a ﬁvefold elevation in basal extracellular DA, while heterozygous mice demonstrate a twofold elevation. Intriguingly, placement of mutants in a novel environment triggers hyperactivity, but this novelty-driven hyperactivity is not accompanied by further rise in already high-extracellular DA (Gainetdinov et al., 1999b; Pogorelov et al., 2005). From the initial characterization of these mutants, it has become clear that lack of DAT results not only in remarkable changes in extracellular DA dynamics but also in numerous alterations in pre- and postsynaptic components of DA signaling (Gainetdinov and Caron, 2003). Microdialysis is perfectly suited to analyze status of presynaptic regulation of DA transmission. In microdialysis experiments, the response to high-K+ stimulation was profoundly reduced (Gainetdinov et al., 1999a; Pogorelov et al., 2005), suggesting that the actual amount of releasable DA in the DATKO mice is decreased. Importantly, DA extracellular levels in DAT-KO mice were found to be impulse ﬂow-dependent and Ca2+-sensitive. Infusion of tetrodotoxin potently reduced extracellular DA to undetectable levels. Similarly, infusion of Ca2+-free artiﬁcial cerebrospinal ﬂuid (CSF) markedly decreased the levels of extracellular DA in mutants (Gainetdinov et al., 1999a). These data indicate that while extracellular DA levels are abnormally elevated in DAT-KO mice, they are still reﬂective of depolarization-dependent vesicular exocytosis.

403

Fig. 1. TH inhibition induces selective severe DA depletion in the striatum of DAT-KO mice. (A) Striatal tissue levels of DA in salinetreated wild type and DAT-KO mice. (B) Dynamics of TH inhibition by aMT (250 mg/kg, i.p.) on striatal tissue DA in wild type and DAT-KO mice. (C) Tissue levels of NE in the frontal cortex of saline-treated wild type and DAT-KO mice. (D) Dynamics of TH inhibition by aMT (250 mg/kg, i.p.) on tissue levels of NE in the frontal cortex of wild type and DAT-KO mice. (E) Effect of TH inhibition by aMT on extracellular DA levels in the striatum of wild-type mice measured using in vivo microdialysis in freely moving mice. Data are presented as a percentage of the average level of DA measured in at least three samples collected before the drug administration. (F) Effect of TH inhibition by aMT on extracellular levels of DA in the striatum of DAT-KO mice, measured by using in vivo microdialysis in freely moving mice. Note that the basal extracellular levels of DA in DAT-KO mice were about ﬁvefold higher than in wild-type mice (reproduced from Sotnikova et al., 2005).

Strikingly, total striatal tissue levels of DA are 20-fold lower in mutant mice (Fig. 1). Furthermore, these low levels of DA in the striatum are extremely sensitive to inhibition of TH by a-methyl-p-tyrosine (aMT), indicating that they predominantly represent newly synthesized pool of DA (Gainetdinov et al., 1998; Jones et al., 1998a; Gainetdinov and Caron, 2003; Sotnikova et al.,

2005). By comparison, this treatment induced similar partial depletion in NE tissue levels in the frontal cortex of wild type and mutant mice (Fig. 1). Extreme depletion of extracellular DA by TH inhibition has been also documented in microdialysis experiments (Fig. 1) (Sotnikova et al., 2005). These observations strongly suggest that the inward transport of DA through the DAT is a

404

major contributor to the large DA storage pool in the striatal terminals. Thus, a tight dependence of DA storage on recycled DA must exist (Gainetdinov et al., 1998; Jones et al., 1998a; Gainetdinov and Caron, 2003; Sotnikova et al., 2005). Interestingly, depletion of DA stores in DATKO mice occurs even though DA synthesis rate measured in vivo in NSD-1015 model is elevated twofolds (Jones et al., 1998a). The increased synthesis of DA may be explained by the disinhibition of TH, which under normal conditions is a subject of tonic feedback inhibition (Bannon and Roth, 1983). In the DAT-KO mice, intraneuronal DA is greatly reduced and this could result in a disinhibition of TH. Alternatively, activation of TH might indicate a loss of autoreceptor function (Bannon and Roth, 1983). Indeed, functional studies revealed marked desensitization in the major autoreceptor functions: regulation of neuronal ﬁring rate, DA release, and synthesis (Jones et al., 1999b). Microdialysis was instrumental to document that nerve terminal release-regulating autoreceptors are almost nonfunctional. The D2/D3 DA receptor agonist quinpirole elicited only a minor decrease in striatal DA release in mutant mice. To study the sensitivity of DA autoreceptors controlling DA synthesis, the effect of quinpirole on L-DOPA accumulation in the striatum of freely moving mice during infusion of the DOPA decarboxylase inhibitor NSD-1015 under cessation of DA neuron impulse ﬂow by gamma-butyrolactone (GBL) (Roth et al., 1980) was examined by microdialysis. Neither GBL nor quinpirole produced signiﬁcant alterations in the striatal DA synthesis rate in DAT-KO mice. These data clearly demonstrate loss of autoreceptor control in conditions of persistent hyperdopaminergia (Jones et al., 1999b). Many of the neurochemical alterations found in DAT mutants display a clear (but not linear) gene–dose effect (Jones et al., 1998a, 1999b; Zhuang et al., 2001). Generally, mice heterozygous for DAT deletion displayed an intermediate phenotype. For example, in these mice, striatal extracellular DA levels are twofold higher as measured by quantitative ‘‘no net ﬂux’’ microdialysis (Jones et al., 1998a, 1999b). Similar degree of elevation in basal extracellular DA (about twofold) was observed when ‘‘true’’ striatal extracellular levels

were measured by ‘‘low perfusion rate’’ microdialysis in mice expressing only 10% of the DAT (Zhuang et al., 2001). DAT is the major drug target for psychostimulants such as cocaine, methylphenidate, amphetamine, and methamphetamine. It is not surprising, therefore, that in the striatum of DAT-KO mice, these drugs are not able to affect extracellular DA levels in in vivo microdialysis studies (Fumagalli et al., 1998; Jones et al., 1998b; Gainetdinov et al., 1999b; Rocha et al., 2002). Furthermore, b-phenylethylamine, an endogenous amine that is found in trace amounts in the brain, shared the ability of amphetamine to increase striatal extracellular DA concentrations via reversal of the DAT-mediated DA transport in wild type but not in DAT-KO mice (Sotnikova et al., 2004). Quite unexpected observations were made when microdialysis studies were performed in the NAc of these mice. DAT-KO mice, despite the lack of the major target of cocaine, demonstrate robust self-administration (Rocha et al., 1998) and conditioned place preference of cocaine (Sora et al., 1998). While this psychostimulant does not affect extracellular DA dynamics in the striatum of mutant mice (Rocha et al., 1998), it still could be active in neuronal circuits responsible for rewarding properties of cocaine. In fact, it has been demonstrated that cocaine and amphetamine are able to elevate DA levels in the NAc – the primary dopaminergic brain area associated with rewardrelated behaviors (Carboni et al., 2001). Moreover, it has been reported that norepinephrine transporter (NET) inhibitor, reboxetine, increased DA in the NAc of DAT-KO but not wild-type mice, while DAT inhibitor GBR 12909 increased dialysate DA concentrations in wild type but not DAT-KO mice. These observations provided an explanation for the surprising persistence of cocaine reinforcement in DAT-KO mice and further emphasized a major role of NAc DA in drug reinforcement (Carboni et al., 2001). In addition, reboxetine-induced increase in DA release in the NAc in DAT-KO mice has been related to the hypothetical role of the NET in the clearance of DA in this brain region in the absence of DAT (Carboni et al., 2001). It should be noted, however, that voltammetry studies performed in NAc slices

405

have not documented any effect of locally applied cocaine or another NET inhibitor, desipramine, on DA dynamics in DAT-KO mice (Budygin et al., 2002). Furthermore, recent voltammetry and microdialysis experiments clearly demonstrated role of cocaine-induced modulation of the serotonin (5HT) system at the level of DA cell bodies in these unexpected actions of cocaine in DAT-KO mice (Mateo et al., 2004). Similar conclusion has been made when microdialysis was applied to investigate mechanism of preserved reinforcing properties of amphetamine in DAT-KO mice (Budygin et al., 2004). It has been found that while systemic amphetamine increases extracellular DA in the NAc in DAT-KO mice, a local infusion of amphetamine into the NAc does not produce this effect. Furthermore, voltammetry in NAc slices revealed lack of amphetamine-induced decrease in the rate of DA clearance. Therefore, it has been suggested that despite the absence of the DAT, amphetamine displays rewarding effects and causes an increase in extracellular DA in the NAc of DAT-KO mice, acting indirectly via 5-HT system (Budygin et al., 2004). DAT plays a critical role in the action of the dopaminergic neurotoxin 1-methyl-4-phenyl-1,2,3, 6-tetrahydropyridine (MPTP) (Javitch and Snyder, 1984). Accordingly, direct intrastriatal infusion of the active metabolite of MPTP, 1-methyl-4-phenylpyridinium, via a microdialysis probe produced a massive efﬂux of DA in wild-type mice while only a minimal effect was observed in DAT-KO mice. As it is known that this massive efﬂux is generally reﬂective of neurotoxic damage following penetration of this toxin into the DA terminal, these data indicate apparent lack of neurotoxicity in the mutants. These observations illustrated the requirement of DAT to exert MPTP toxicity in dopaminergic terminals (Gainetdinov et al., 1997). Effects of several other pharmacological agents were probed in DAT-KO mice in microdialysis studies. As mentioned earlier, high-DA levels in DAT-KO mice are extremely sensitive to pharmacological inhibition of TH and aMT effectively eliminated extracellular DA in the striatum of DAT-KO mice, while exerting only partial decrease in normal animals (Sotnikova et al., 2005).

This profound DA depletion was accompanied by a striking behavioral phenotype manifested as severe akinesia, rigidity, tremor, and ptosis. In these DA-depleted mice, microdialysis was used to demonstrate that apparent antiparkinsonian effects of large doses of amphetamine derivatives, such as MDMA in DA-depleted DAT-KO mice are independent of DA (Sotnikova et al., 2005). Microdialysis investigations have also been instrumental to show lack of correlation between extreme behavioral activation in DAT-KO mice induced by treatment with glutamate N-methyl-D-aspartate (NMDA) receptor antagonist, MK-801 and striatal extracellular DA dynamics (Gainetdinov et al., 2001). In another study, effects of lithium on the striatal extracellular DA levels were tested in DATKO mice. These experiments showed that potent inhibitory action of lithium on the hyperactivity of DAT-KO mice is not due to decrease in DA release but rather relates to interaction of lithium with postsynaptic DA receptor signaling mechanism. These investigations eventually revealed that DA can exert its behavioral effects by acting on D2-like DA receptors coupled to lithium-sensitive signaling cascade involving Akt/PKB and glycogen synthase kinase 3 (Beaulieu et al., 2004). The vesicular monoamine transporter 2 controls neuronal vesicular storage by pumping monoamine neurotransmitters such as DA, NE, 5-HT, and histamine from the cytoplasm into secretory vesicles. VMAT2 null mice cannot move and, therefore, feed poorly and die within a few days after birth. The brains of mutant mice showed a drastic reduction in monoamine storage and vesicular release (Wang et al., 1997). Mice lacking one allele of VMAT2 gene are viable into adulthood and display less dramatic neurochemical and behavioral changes. Microdialysis investigations were focused on VMAT2 heterozygous mice (Wang et al., 1997; Fumagalli et al., 1999). The brains of VMAT2 heterozygote mice contain reduced tissue monoamine levels, and K+ depolarization induces less DA release from heterozygous mice in microdialysis experiments (Wang et al., 1997). Furthermore, extracellular striatal DA level in these mutants as measured by ‘‘low perfusion rate’’ microdialysis is signiﬁcantly lower in comparison to wild-type mice (Wang et al., 1997).

406

It is known that amphetamines cause monoamine redistribution from synaptic vesicles into the cytoplasm via an interaction with VMAT2 and this process is critical for amphetamine-induced efﬂux of DA from terminals (Sulzer et al., 1995). Accordingly, in heterozygous mice, amphetamine induced less-pronounced elevation of striatal extracellular DA (Wang et al., 1997). The VMAT2 heterozygote mice were also used to clarify the role of vesicular storage in the neurotoxic action of methamphetamine (Fumagalli et al., 1999). It has been observed that methamphetamine-induced dopaminergic neurotoxicity was increased in striatum of VMAT2 heterozygous mice compared with wildtype mice. However, microdialysis experiments have shown that the ability of methamphetamine to increase striatal extracellular DA is reduced in the mutants. These results suggested that disruption of vesicular monoamine transport potentiates methamphetamine-induced neurotoxicity and this increased toxicity does not correlate with extracellular DA dynamics. These data were instrumental to suggest that intraneuronal DA redistribution from vesicular to cytoplasmic compartments rather than extraneuronal DA accumulation is primarily responsible for this effect (Fumagalli et al., 1999). III. B. Microdialysis studies on DA neurochemistry in other mutants Mice deﬁcient in each of the ﬁve DA receptors (D1–D5) are currently available. As D2 and D3 receptors are located both pre- and postsynaptically and can serve as autoreceptors, D2-KO and D3-KO mice were speciﬁcally investigated for regulation of DA release. Two studies employed microdialysis in D2-KO mice. It has been reported that basal and K+-evoked extracellular DA concentrations in the striatum were not affected in mutant mice (Dickinson et al., 1999; Rouge-Pont et al., 2002). However, increase in extracellular DA induced by cocaine and morphine was found to be higher in D2-KO than in wild-type littermates suggesting a role of D2, particularly D2short, DA receptors in the autoreceptor regulation of DA release (Rouge-Pont et al., 2002). In D3-KO mice, a conventional microdialysis studies of DA release in the ventral striatum

revealed higher basal levels of extracellular DA but inhibitory effects of a D3 DA receptor agonist PD 128907 was reported to be similar in mutant and wild-type mice (Koeltzow et al., 1998). In another study, a quantitative ‘‘no net ﬂux’’ microdialysis technique was applied to characterize basal DA concentration in the NAc of D3-KO mice. However, neither the extracellular DA concentration nor the in vivo extraction fraction, an indirect measure of basal DA uptake, was found to be altered in D3-KO mice. Similarly, no differences in K+- or cocaine-induced DA release were detected between the two genotypes. At the same time, when PD 128907 was applied to induce autoreceptor-mediated decrease in DA levels, no effect was found in D3-KO mice (Zapata et al., 2001). In another study, analysis of the basal DA concentrations by a quantitative in vivo ‘‘low perfusion rate’’ microdialysis revealed elevated striatal DA extracellular levels in D3-KO mice (Joseph et al., 2002). Taken together, these microdialysis data are consistent with the role of both D2 and D3 receptors in the presynaptic regulation of DA release with D3 receptor having a small but signiﬁcant role as a DA autoreceptor that partially regulates secretion, but not synthesis of DA (Joseph et al., 2002). DA receptor regulatory mechanisms have been also assessed using mutant mice. Desensitization of G-protein coupled DA receptors is mediated via phosphorylation by members of the family of Gprotein coupled receptor kinases (GRK1–GRK7). We observed that GRK6-KO mice are supersensitive to the locomotor-stimulating effect of psychostimulants, such as cocaine and amphetamine as well as direct DA agonists. In vivo microdialysis was used to determine if such supersensitivity is determined by pre- or postsynaptic mechanisms. Demonstration of unchanged basal striatal level of DA as evidenced by ‘‘low perfusion rate’’ quantitative microdialysis as well as unaltered responses to cocaine measured by a conventional microdialysis indicated that postsynaptic D2-like DA receptors are likely physiological targets for GRK6 (Gainetdinov et al., 2003). Functional role of major metabolizing enzymes in DA homeostasis, monoamine oxidase B (MAO-B), and catechol-O-methyltransferase (COMT) has

407

been also probed by using microdialysis in knockout mice. Surprisingly, in MAO-B knockout mice, no changes in extracellular levels of DA were observed in striatum. Similarly, the synthesis, storage, uptake, and release of DA were also unaltered (Chen et al., 1999; Fornai et al., 1999). Administration of a selective MAO-A inhibitor, clorgyline, increased extracellular DA levels and decreased levels of its metabolites similarly in mutant and wild-type mice, but a selective MAO-B inhibitor, deprenyl, had no such effect in mutants. Only high dose of L-DOPA produced a larger increase in DA in KO than in wild-type mice. These results were interpreted to suggest that in mice, DA may be metabolized by only MAO-A under basal conditions and by both MAO-A and B at high concentrations (Fornai et al., 1999). COMT catalyzes the O-methylation of catecholamines and its major function is elimination of biologically active amines and their metabolites. Under normal conditions, only catecholamine metabolites, and not DA or NE were affected in mutants in microdialysis experiments. However, when COMT-deﬁcient mice are challenged with L-DOPA, DAT inhibitor GBR 12909 or amphetamine, mutants demonstrate an extensive accumulation of DOPAC and some additional increase in striatal extracellular and/or tissue DA, revealing an important role for COMT in striatal DA metabolism under such situations (Huotari et al., 2002a, b, 2004). Several groups applied microdialysis technique to deﬁne role of opioid receptors in the regulation of mesolimbic DA release. It is believed that the reinforcing and psychomotor effects of morphine involve opiate stimulation of DA system via activation of m-opioid receptors (MOR). Conventional and quantitative ‘‘no net ﬂux’’ microdialysis were used to quantify basal and morphine-induced extracellular DA levels and the extraction fraction, which provides an estimate of the rate of DA uptake, in the NAc of mice lacking MOR or deltaopioid receptors (DOR) (Chefer et al., 2003). ‘‘No net ﬂux’’ studies revealed signiﬁcant decreases in the DA extraction fraction in both MOR-KO and DOR-KO mice, indicating decreased basal DA uptake in these mutants. Basal extracellular DA, however, was unchanged. MOR-KO mice failed to exhibit locomotor response to morphine, but

the ability of morphine to increase DA levels, however, was reduced but not blocked. No behavioral or neurochemical alteration in the effects of morphine was observed in DOR-KO mice. Furthermore, effects of cocaine were analyzed in these mutants (Chefer et al., 2004). Acute administration of cocaine increased DA levels in both MORKO and DOR-KO animals. Paradoxically, however, the magnitude of this effect was smaller in DOR-KO as compared with control or MOR-KO mice. In kappa-opioid receptor (KOR-1) deﬁcient mice, microdialysis revealed that basal DA release and DA extraction fraction are enhanced in KOR-1 mutant mice. Cocaine induced more potent increase of extracellular DA in mutants revealing the existence of an endogenous KOR-1 system that tonically inhibits mesolimbic DA neurotransmission (Chefer et al., 2005). In vivo microdialysis was also performed in mice lacking b-arrestin-2 (Bohn et al., 2003) that display enhanced sensitivity to morphine in tests of pain perception attributable to impaired desensitization of MOR (Bohn et al., 2000). Assessing the effects of morphine and cocaine on locomotor activity, conditioned place preference, and striatal DA release in b-arrestin-2 knockout mice revealed that cocaine effects were minimally changed in mutants. However, morphine induced more pronounced increases in striatal extracellular DA and reward in the conditioned place preference test, supporting functional role of this regulatory molecule in MOR desensitization (Bohn et al., 2003). Acetylcholine neurotransmission serves an important role in the CNS. While it is known that cholinergic activity can modulate central DA transmission, the nature of this interaction and the receptors involved remain undeﬁned. This complex regulation has been probed in mice lacking the M1 muscarinic ACh receptor. A quantitative ‘‘low perfusion rate’’ microdialysis studies revealed that M1 deﬁciency leads to elevated DA transmission in the striatum and signiﬁcantly increased locomotor activity. M1-deﬁcient mice also demonstrate increased dopaminergic and behavioral responses to amphetamine (Gerber et al., 2001). In another microdialysis study, an elevated DA basal level and enhanced DA

408

response to psychostimulants in the NAc of M4 receptor knockout mice was observed (Tzavara et al., 2004). Thus, role of M1 and M4 muscarinic receptors in the control of striatal dopaminergic activity has been convincingly demonstrated. The mesolimbic DA system mediates the reinforcing properties of many drugs of abuse, including nicotine. In vivo microdialysis was used to investigate the role of the a4 and b2 subunits of nicotinic ACh receptor in mediating the effects of nicotine on the DA transmission by using mice lacking these subunits. It was observed that mutant mice lacking a4 have twofold elevated basal striatal DA levels. While both wild type and a4 null mutant mice show a similar increase in DA release in response to high-K+ infusion, a nicotine-induced increase in DA levels was not observed in mutant mice. Similarly, nicotine was not able to increase DA release in the ventral striatum of b2-mutant mice. These results indicate that a4and b2-containing nAChRs exert a tonic control on basal DA release, which is likely mediated by a heterogeneous population of nicotinic ACh receptors (Picciotto et al., 1998; Marubio et al., 2003). In another microdialysis study, it has been shown that the rewarding properties of nicotine as well as enhancement in DA extracellular levels in the NAc by nicotine are disrupted in preproenkephalindeﬁcient mice suggesting that endogenous opioid peptides derived from preproenkephalin are involved in the rewarding properties of nicotine (Berrendero et al., 2005). Understanding the role of glutamate–DA interaction in the CNS functions has been a subject of intense research for many years. Recent development of mutant mice deﬁcient in glutamate receptors or other key molecules involved in glutamate receptor signaling provided several novel important observations. Particularly, intriguing data were gained in mice deﬁcient in glutamate NMDA receptors (Mohn et al., 1999). Mice expressing only 5% of normal levels of the essential NMDA receptor NR1 subunit, display striking behavioral abnormalities, including increased motor activity and deﬁcits in social interaction. These behavioral abnormalities recapitulate behaviors observed in pharmacologically induced animal models of schizophrenia involving NMDA antagonists and

can be ameliorated by treatment with antipsychotics. Importantly, a quantitative microdialysis documented that this hypoglutamatergic state is not accompanied with an increased extracellular DA in the striatum (Mohn et al., 1999), as has been postulated by some researchers. These ﬁndings support a model in which reduced NMDA receptor activity results in schizophrenia-related abnormal behaviors independently from changes in striatal extracellular DA (Carlsson et al., 2001). Another genetic mouse model of schizophrenia which has deﬁcient glutamatergic signaling, has been developed by forebrain-speciﬁc genetic deletion of calcineurin, a calcium- and calmodulindependent protein phosphatase (Miyakawa et al., 2003). Mutant mice showed increased locomotor activity, decreased social interaction, and impairments in prepulse inhibition and latent inhibition. In addition, these mutants displayed an increased response to the locomotor stimulating effects of MK-801. At the same time, striatal extracellular DA assessed by a quantitative ‘‘low perfusion microdialysis’’ was found to be intact strongly indicating that these abnormal behaviors are not related to changes in DA transmission (Miyakawa et al., 2003). Microdialysis has been successfully applied to several other mutants that have dysfunctions in molecules that are integral to the assembly and function of proteins regulating glutamate signaling and synaptic plasticity. Alterations in DA release in responses to various drugs and stimuli were observed in metabotropic glutamate receptor subtype 2 knockout mice (Morishima et al., 2005), Fyn tyrosine kinase-deﬁcient mice (Hironaka et al., 2002), and Homer2-KO mice (Szumlinski et al., 2005). Thorough characterization of the basal extracellular levels of DA as well as other monoamines by a quantitative microdialysis approaches were performed in 5-HT1B-KO mice (Shippenberg et al., 2000) and melanin-concentrating hormone knockout mice (Smith et al., 2005). Responsiveness of DA system to various treatments has been also assessed in several other mutants, including tumor necrosis factor-alpha (TNF-a) mutant mice (Nakajima et al., 2004), glial cell line-derived neurotrophic factor knockout mice (GDNF knockout mice)

409

(Airavaara et al., 2004, 2006), cannabinoid CB1 receptor knockout mice (Hungund et al., 2003), Ca(v)2.3 calcium channel knockout mice (Han et al., 2002), mice lacking opioid receptor-like receptor (ORL1) (Koizumi et al., 2004a, b), adenosine A2A receptor knockout mice (Dassesse et al., 2001), mint-1 knockout mice (Mori et al., 2002), stable tubule only polypeptide (STOP) null mice (Brun et al., 2005), glutathione-deﬁcient mice (Jacobsen et al., 2005), PKCepsilon-deﬁcient mice (Olive et al., 2000b), transgenic models of Huntington’s disease (Petersen et al., 2002; Vetter et al., 2003), a2A-adrenoceptor knockout mice (Ihalainen and Tanila, 2002, 2004), somatostatin receptor-2 knockout/lacZ knock-in mice (Allen et al., 2003), and parkin-deﬁcient mice (Goldberg et al., 2003).

III.C. Other microdialysis applications in mutant mice As in case with DA neurochemistry, the most remarkable changes in extracellular 5-HT dynamics were observed in mice lacking plasma membrane serotonin transporter (SERT). A quantitative ‘‘no net ﬂux’’ microdialysis in freely moving mice was utilized to investigate alterations in extracellular 5-HT in mutants (Mathews et al., 2004). Using this method, a 6- to 10-fold increase in extracellular levels of 5-HT has been found in the striatum and frontal cortex of mutant mice. Heterozygous mice showed intermediate values. Potassium stimulation resulted in greater depolarization-induced increases in striatal extracellular 5-HT in SERT-KO mice. Overall, these results corroborated observations made in DAT-KO mice by demonstrating a major role of SERT in the control of 5-HT levels in the extracellular space (Mathews et al., 2004). Several reports have been focused on the impact of 5-HT receptors deletion on the extracellular 5-HT levels. In the striatum of 5-HT1A-KO mice, no differences in the basal extracellular 5-HT levels were found, but ﬂuoxetine induced twofold greater increase in the 5-HT output in mutant mice. By comparison, no differences in K+-stimulated 5-HT release were seen (He et al., 2001). In another microdialysis study, these

mutants were used to examine the effects of stressful and pharmacological challenges on dorsal raphe 5-HT efﬂux. Again, baseline 5-HT concentrations were not changed in 5-HT1A-KO mice, but autoreceptor-mediated effect of 5HT1A, but not 5-HT1B agonist, was disrupted. Both a saline injection and handling of mice increased dialysate 5-HT levels in mutants. As in the striatum, ﬂuoxetine induced markedly more pronounced increase in dialysate 5-HT in the 5-HT1A-KO mice (Bortolozzi et al., 2004). Similar observations were also made in the frontal cortex and ventral hippocampus of 5-HT1A receptor mutant mice convincingly demonstrating presynaptic autoreceptor function of this receptor in regulating 5-HT release (Parsons et al., 2001). To test contribution of the 5-HT1B receptor subtype in the presynaptic regulation of 5-HT release and involvement in the effects of selective serotonin reuptake inhibitors (SSRIs), intracerebral in vivo microdialysis was applied by several groups to 5-HT1B receptor knockout mice (Trillat et al., 1997; Knobelman et al., 2001; Malagie et al., 2001, 2002; De Groote et al., 2002). Measurements of 5-HT dynamics in the hippocampus and frontal cortex in these mutants have consistently revealed a lack of inhibitory effect of 5-HT1B agonists on 5-HT release. Furthermore, SSRIs tested exerted more pronounced increase in the extracellular 5-HT levels in 5-HT1B-KO mice. These data clearly illustrated that 5-HT1B autoreceptors at nerve terminals limit the effects of SSRIs on extracellular 5-HT. An important example of thoughtful use of microdialysis in knockout mice as a tool to validate drug target is neurochemical characterization of 5-HT2C knockout mice (Cremers et al., 2004). To support pharmacological observations in rats demonstrating that both selective and nonselective 5-HT2C receptor antagonists are able to produce a robust augmentation of SSRI-induced elevations of hippocampal and cortical extracellular 5-HT levels mice lacking this receptor were investigated. Although 5-HT2C-KO mice did not have altered baseline extracellular 5-HT levels, they had signiﬁcantly enhanced ﬂuoxetine effects on 5-HT levels thereby strongly supporting an unanticipated pharmacological action of 5-HT2C receptors (Cremers et al., 2004).

410

As MAO-A is primarily responsible for the catabolism of 5-HT, it is not surprising that MAO-A deﬁcient mice demonstrate elevated basal extracellular 5-HT levels in ventral hippocampus, frontal cortex, and dorsal raphe nucleus. Furthermore, an acute SERT blockade by citalopram produced a much larger increase in extracellular 5-HT levels in these mutants (Evrard et al., 2002). Measurements of extracellular 5-HT dynamics in mice lacking dopamine beta-hydroxylase (DBH) (Cryan et al., 2004) resulted in quite unexpected outcome. Mice deﬁcient in this enzyme lack NE and fail to respond to the behavioral effects of various antidepressants, including the NE reuptake inhibitors, desipramine and reboxetine. However, DBH-KO mice were also insensitive to behavioral effects of SSRIs and this apparent insensitivity was accompanied by the disrupted ability of SSRIs to elevate hippocampal extracellular 5-HT. Analysis of central 5-HT transmission using microdialysis has been performed also in the corticotropin-releasing hormone receptor type 1 (CRH-R1)-deﬁcient mice (Penalva et al., 2002), in galanin-overexpressing mice (Yoshitake et al., 2004) and brain-derived neurotrophic factor (BDNF)-deﬁcient mice (Szapacs et al., 2004). Similar to DAT-KO and SERT-KO mice, mice lacking the plasma membrane NET showed prolonged clearance of NE in voltammetry experiments and twofold elevated extracellular levels of this catecholamine as measured by a quantitative ‘‘low perfusion rate’’ microdialysis in the cerebellum (Xu et al., 2000). Furthermore, effect of alpha2-adrenoceptor antagonist on the extracellular NE dynamics was enhanced in NET-KO mice (Vizi et al., 2004). Several reports were devoted to measurements of ACh release in various mutant mice. Among ﬁve known muscarinic receptors, M2 and M4 receptors are localized both post- and presynaptically and are believed to play a role as autoreceptors. In vivo microdialysis was applied to test the functional importance of these receptors in the regulation of ACh release by using M2-KO and M4-KO mice as well as M2/M4 receptor double KO mice (Tzavara et al., 2003). Basal ACh release in the hippocampus was signiﬁcantly increased in M4-KO and was elevated further in

M2/M4-KO double mutants. The increase in hippocampal ACh induced by local administration of scopolamine was markedly reduced in M2-KO and completely abolished in M2/M4-Kos, supporting autoreceptor role of both M2 and M4 receptors in regulating ACh release (Tzavara et al., 2003). Mice lacking endothelial and neuronal nitric oxide (NO) synthase (eNOS and nNOS) were used to characterize role of NO in hippocampal synaptic plasticity mediated by ACh. In wild-type mice, perfusion of NMDA via microdialysis probe induced a twofold stimulation of ACh release in the striatum. This phenomenon was little affected in mice lacking eNOS but was completely absent in mice lacking nNOS. These data suggest involvement of NO synthesized by nNOS in the glutamatergic stimulation of striatal cholinergic interneurons (Buchholzer and Klein, 2002, 2003). Signiﬁcant changes in the hippocampal extracellular ACH, as well as 5-HT and NE, were found in galanin-deﬁcient and galanin-overexpressing mice (Kehr et al., 2001; Laplante et al., 2004a). Status of central cholinergic transmission has been also probed by using microdialysis in dopamine D5 knockout mice (Laplante et al., 2004b) and human amyloid precursor protein knock-in/presenilin-1 transgenic mice (Hartmann et al., 2004). Microdialysis measurements were performed also to detect changes in the frontal cortex histamine transmission in orexin A knockout mice (Huang et al., 2001; Hong et al., 2005) and striatal adenosine levels in Huntington transgenic mice (Gianfriddo et al., 2004). One important application of in vivo microdialysis relates to the measurement of the conversion of 4-hydroxybenzoic acid to 3,4-dihydroxybenzoic acid as an index of ‘‘hydroxyl radical’’ production (Bogdanov et al., 1998a). This approach has been successfully applied to several genetic mouse models of neurodegenerative disorders, such as transgenic amyotrophic lateral sclerosis (ALS) mice with the G93A mutation in the enzyme copper/zinc superoxide dismutase-1 (Bogdanov et al., 1998b), transgenic mouse model of HD (Bogdanov et al., 1998a), and BCL-2 overexpressing mice (Bogdanov et al., 1999). Another interesting application of microdialysis technique has been

411

recently demonstrated by measurement of 8-hydroxy-20 -deoxyguanosine as an indication of increased oxidative DNA damage in mouse model of ALS (Aguirre et al., 2005). A novel application of microdialysis has been recently validated by successful measurements of prolonged half-life of soluble amyloid-b peptide levels in the interstitial ﬂuid of human amyloid precursor protein transgenic mice (Cirrito et al., 2003). This approach has been successfully applied to measure CSF and brain interstitial ﬂuid amyloid-b levels in apolipoprotein E- and clusterindeﬁcient mice (DeMattos et al., 2004). Finally, microdialysis was successfully used to investigate the role of mdr1a-encoded P-glycoprotein in the blood-brain barrier transport of morphine by using mdr1a-deﬁcient mice. To measure morphine unbound concentrations in brain extracellular ﬂuid, two methods of estimating in vivo recovery were applied: retrodialysis with nalorphine as a calibrator, and the ‘‘no net ﬂux’’ method (Xie et al., 1999).

IV. Conclusion Beyond obvious application of genetic models to deﬁne physiological role of speciﬁc protein in vivo, mutant mice becoming increasingly valuable to understand the contribution of individual neurotransmitter systems in the circuits critical for manifestation of normal or aberrant behaviors. An additional beneﬁt of genetically altered animals is that they may model human diseases thereby providing excellent test subjects for future drug development. Such models could provide opportunity to test potential therapeutic agents in animals with an understood brain pathology resulting from a planned genetic manipulation. Many of the reports discussed here illustrate that in genetically altered animals, an accepted transmitter/function/ behavior relationship may not be necessarily intact and careful examination of as many aspects of transmitter systems as possible is needed. In vivo microdialysis, as a tool to provide important link between neurochemistry and function/behavior, has proven to be indispensable in such studies employing genetically altered mice.

There are a number of techniques available for generation of genetically altered mice. Depending on the intended purpose, various designs of the targeting construct can be achieved. In addition to conventional knockout (Bronson and Smithies, 1994), which allows one to examine the consequences of inactivation of either one or both copies of the gene, a number of various other approaches providing temporal and spatial control over gene expression are currently available. Furthermore, development of the technique of knock-in now allows replacing the native gene with a mutated gene or a gene of interest at the normal genomic locus (Sullivan et al., 1997). A recent emergence of RNAi technology to downregulate expression of speciﬁc genes in the brain (Salahpour et al., 2006) with future developments to enhance its efﬁciency in vivo could dramatically increase application of microdialysis. This approach could be potentially used to perform genetic manipulations not only in mice but also in larger animals making it possible to analyze neurotransmitter dynamics in smaller brain regions and/or during more complex behavioral tasks. With the exponential increase in the number and variety of genetically altered animals available, the potential for the use of microdialysis in such models is virtually limitless.

References Aguirre, N., Beal, M.F., Matson, W.R. and Bogdanov, M.B. (2005) Increased oxidative damage to DNA in an animal model of amyotrophic lateral sclerosis. Free Radic. Res., 39: 383–388. Airavaara, M., Mijatovic, J., Vihavainen, T., Piepponen, T.P., Saarma, M. and Ahtee, L. (2006) In heterozygous GDNF knockout mice the response of striatal dopaminergic system to acute morphine is altered. Synapse, 59: 321–329. Airavaara, M., Planken, A., Gaddnas, H., Piepponen, T.P., Saarma, M. and Ahtee, L. (2004) Increased extracellular dopamine concentrations and FosB/DeltaFosB expression in striatal brain areas of heterozygous GDNF knockout mice. Eur. J. Neurosci., 20: 2336–2344. Allen, J.P., Hathway, G.J., Clarke, N.J., Jowett, M.I., Topps, S., Kendrick, K.M., Humphrey, P.P., Wilkinson, L.S. and Emson, P.C. (2003) Somatostatin receptor 2 knockout/lacZ knockin mice show impaired motor coordination and reveal sites of somatostatin action within the striatum. Eur. J. Neurosci., 17: 1881–1895. Bannon, M.J. and Roth, R.H. (1983) Pharmacology of mesocortical dopamine neurons. Pharmacol. Rev., 35: 53–68.

412 Beaulieu, J.M., Sotnikova, T.D., Yao, W.D., Kockeritz, L., Woodgett, J.R., Gainetdinov, R.R. and Caron, M.G. (2004) Lithium antagonizes dopamine-dependent behaviors mediated by an AKT/glycogen synthase kinase 3 signaling cascade. Proc. Natl. Acad. Sci. U.S.A., 101: 5099–5104. Benoit-Marand, M., Jaber, M. and Gonon, F. (2000) Release and elimination of dopamine in vivo in mice lacking the dopamine transporter: functional consequences. Eur. J. Neurosci., 12: 2985–2992. Berrendero, F., Mendizabal, V., Robledo, P., Galeote, L., Bilkei-Gorzo, A., Zimmer, A. and Maldonado, R. (2005) Nicotine-induced antinociception, rewarding effects, and physical dependence are decreased in mice lacking the preproenkephalin gene. J. Neurosci., 25: 1103–1112. Bogdanov, M.B., Ferrante, R.J., Kuemmerle, S., Klivenyi, P. and Beal, M.F. (1998a) Increased vulnerability to 3-nitropropionic acid in an animal model of Huntington’s disease. J. Neurochem., 71: 2642–2644. Bogdanov, M.B., Ferrante, R.J., Mueller, G., Ramos, L.E., Martinou, J.C. and Beal, M.F. (1999) Oxidative stress is attenuated in mice overexpressing BCL-2. Neurosci. Lett., 262: 33–36. Bogdanov, M.B., Ramos, L.E., Xu, Z. and Beal, M.F. (1998b) Elevated ‘‘hydroxyl radical’’ generation in vivo in an animal model of amyotrophic lateral sclerosis. J. Neurochem., 71: 1321–1324. Bohn, L.M., Gainetdinov, R.R., Lin, F.T., Lefkowitz, R.J. and Caron, M.G. (2000) Mu-opioid receptor desensitization by beta-arrestin-2 determines morphine tolerance but not dependence. Nature, 408: 720–723. Bohn, L.M., Gainetdinov, R.R., Sotnikova, T.D., Medvedev, I.O., Lefkowitz, R.J., Dykstra, L.A. and Caron, M.G. (2003) Enhanced rewarding properties of morphine, but not cocaine, in beta(arrestin)-2 knock-out mice. J. Neurosci., 23: 10265–10273. Bortolozzi, A., Amargos-Bosch, M., Toth, M., Artigas, F. and Adell, A. (2004) In vivo efﬂux of serotonin in the dorsal raphe nucleus of 5-HT1A receptor knockout mice. J. Neurochem., 88: 1373–1379. Bronson, S.K. and Smithies, O. (1994) Altering mice by homologous recombination using embryonic stem cells. J. Biol. Chem., 269: 27155–27158. Brun, P., Begou, M., Andrieux, A., Mouly-Badina, L., Clerget, M., Schweitzer, A., Scarna, H., Renaud, B., Job, D. and Suaud-Chagny, M.F. (2005) Dopaminergic transmission in STOP null mice. J. Neurochem., 94: 63–73. Buchholzer, M.L. and Klein, J. (2002) NMDA-induced acetylcholine release in mouse striatum: role of NO synthase isoforms. J. Neurochem., 82: 1558–1560. Buchholzer, M.L. and Klein, J. (2003) Nitrergic modulation of hippocampal acetylcholine release is independent of endothelial NO synthase activity. Neurochem. Res., 28: 587–591. Budygin, E.A., Brodie, M.S., Sotnikova, T.D., Mateo, Y., John, C.E., Cyr, M., Gainetdinov, R.R. and Jones, S.R. (2004) Dissociation of rewarding and dopamine transportermediated properties of amphetamine. Proc. Natl. Acad. Sci. U.S.A., 101: 7781–7786.

Budygin, E.A., John, C.E., Mateo, Y. and Jones, S.R. (2002) Lack of cocaine effect on dopamine clearance in the core and shell of the nucleus accumbens of dopamine transporter knock-out mice. J. Neurosci., 22: RC222. Bungay, P.M., Newton-Vinson, P., Isele, W., Garris, P.A. and Justice, J.B. (2003) Microdialysis of dopamine interpreted with quantitative model incorporating probe implantation trauma. J. Neurochem., 86: 932–946. Carboni, E., Spielewoy, C., Vacca, C., Nosten-Bertrand, M., Giros, B. and Di Chiara, G. (2001) Cocaine and amphetamine increase extracellular dopamine in the nucleus accumbens of mice lacking the dopamine transporter gene. J. Neurosci., 21(RC141): 141–144. Carlsson, A., Waters, N., Holm-Waters, S., Tedroff, J., Nilsson, M. and Carlsson, M.L. (2001) Interactions between monoamines, glutamate, and GABA in schizophrenia: new evidence. Annu. Rev. Pharmacol. Toxicol., 41: 237–260. Chefer, V.I., Czyzyk, T., Bolan, E.A., Moron, J., Pintar, J.E. and Shippenberg, T.S. (2005) Endogenous kappa-opioid receptor systems regulate mesoaccumbal dopamine dynamics and vulnerability to cocaine. J. Neurosci., 25: 5029–5037. Chefer, V.I., Kieffer, B.L. and Shippenberg, T.S. (2003) Basal and morphine-evoked dopaminergic neurotransmission in the nucleus accumbens of MOR- and DOR-knockout mice. Eur. J. Neurosci., 18: 1915–1922. Chefer, V.I., Kieffer, B.L. and Shippenberg, T.S. (2004) Contrasting effects of mu opioid receptor and delta opioid receptor deletion upon the behavioral and neurochemical effects of cocaine. Neuroscience, 127: 497–503. Chen, L., He, M., Sibille, E., Thompson, A., Sarnyai, Z., Baker, H., Shippenberg, T. and Toth, M. (1999) Adaptive changes in postsynaptic dopamine receptors despite unaltered dopamine dynamics in mice lacking monoamine oxidase. B. J. Neurochem., 73: 647–655. Cirrito, J.R., May, P.C., O’Dell, M.A., Taylor, J.W., Parsadanian, M., Cramer, J.W., Audia, J.E., Nissen, J.S., Bales, K.R., Paul, S.M., DeMattos, R.B. and Holtzman, D.M. (2003) In vivo assessment of brain interstitial ﬂuid with microdialysis reveals plaque-associated changes in amyloid-beta metabolism and half-life. J. Neurosci., 23: 8844–8853. Cravchik, A., Subramanian, G., Broder, S. and Venter, J.C. (2001) Sequence analysis of the human genome: implications for the understanding of nervous system function and disease. Arch. Neurol., 58: 1772–1778. Cremers, T.I., Giorgetti, M., Bosker, F.J., Hogg, S., Arnt, J., Mork, A., Honig, G., Bogeso, K.P., Westerink, B.H., den Boer, H., Wikstrom, H.V. and Tecott, L.H. (2004) Inactivation of 5-HT(2C) receptors potentiates consequences of serotonin reuptake blockade. Neuropsychopharmacology, 29: 1782–1789. Cryan, J.F., O’Leary, O.F., Jin, S.H., Friedland, J.C., Ouyang, M., Hirsch, B.R., Page, M.E., Dalvi, A., Thomas, S.A. and Lucki, I. (2004) Norepinephrine-deﬁcient mice lack responses to antidepressant drugs, including selective serotonin reuptake inhibitors. Proc. Natl. Acad. Sci. U.S.A., 101: 8186–8191.

413 Dassesse, D., Massie, A., Ferrari, R., Ledent, C., Parmentier, M., Arckens, L., Zoli, M. and Schiffmann, S.N. (2001) Functional striatal hypodopaminergic activity in mice lacking adenosine A(2A) receptors. J. Neurochem., 78: 183–198. De Groote, L., Olivier, B. and Westenberg, H.G. (2002) The effects of selective serotonin reuptake inhibitors on extracellular 5-HT levels in the hippocampus of 5-HT(1B) receptor knockout mice. Eur. J. Pharmacol., 439: 93–100. DeMattos, R.B., Cirrito, J.R., Parsadanian, M., May, P.C., O’Dell, M.A., Taylor, J.W., Harmony, J.A., Aronow, B.J., Bales, K.R., Paul, S.M. and Holtzman, D.M. (2004) ApoE and clusterin cooperatively suppress Abeta levels and deposition: evidence that ApoE regulates extracellular Abeta metabolism in vivo. Neuron, 41: 193–202. Di Chiara, G., Tanda, G. and Carboni, E. (1996) Estimation of in-vivo neurotransmitter release by brain microdialysis: the issue of validity. Behav. Pharmacol., 7: 640–657. Dickinson, S.D., Sabeti, J., Larson, G.A., Giardina, K., Rubinstein, M., Kelly, M.A., Grandy, D.K., Low, M.J., Gerhardt, G.A. and Zahniser, N.R. (1999) Dopamine D2 receptor-deﬁcient mice exhibit decreased dopamine transporter function but no changes in dopamine release in dorsal striatum. J. Neurochem., 72: 148–156. Evrard, A., Malagie, I., Laporte, A.M., Boni, C., Hanoun, N., Trillat, A.C., Seif, I., De Maeyer, E., Gardier, A., Hamon, M. and Adrien, J. (2002) Altered regulation of the 5-HT system in the brain of MAO-A knock-out mice. Eur. J. Neurosci., 15: 841–851. Fornai, F., Chen, K., Giorgi, F.S., Gesi, M., Alessandri, M.G. and Shih, J.C. (1999) Striatal dopamine metabolism in monoamine oxidase B-deﬁcient mice: a brain dialysis study. J. Neurochem., 73: 2434–2440. Franklin, K.B.J. and Paxinos, G. (1996) The Mouse Brain in Stereotaxic Coordinates. Academic Press, San Diego. Fumagalli, F., Gainetdinov, R.R., Valenzano, K.J. and Caron, M.G. (1998) Role of dopamine transporter in methamphetamine-induced neurotoxicity: evidence from mice lacking the transporter. J. Neurosci., 18: 4861–4869. Fumagalli, F., Gainetdinov, R.R., Wang, Y.M., Valenzano, K.J., Miller, G.W. and Caron, M.G. (1999) Increased methamphetamine neurotoxicity in heterozygous vesicular monoamine transporter 2 knock-out mice. J. Neurosci., 19: 2424–2431. Gainetdinov, R.R. and Caron, M.G. (2003) Monoamine transporters: from genes to behavior. Annu. Rev. Pharmacol. Toxicol., 43: 261–284. Gainetdinov, R.R., Bohn, L.M., Sotnikova, T.D., Cyr, M., Laakso, A., Macrae, A.D., Torres, G.E., Kim, K.M., Lefkowitz, R.J., Caron, M.G. and Premont, R.T. (2003) Dopaminergic supersensitivity in G protein-coupled receptor kinase 6-deﬁcient mice. Neuron, 38: 291–303. Gainetdinov, R.R., Fumagalli, F., Jones, S.R. and Caron, M.G. (1997) Dopamine transporter is required for in vivo MPTP neurotoxicity: evidence from mice lacking the transporter. J. Neurochem., 69: 1322–1325.

Gainetdinov, R.R., Jones, S.R. and Caron, M.G. (1999a) Functional hyperdopaminergia in dopamine transporter knock-out mice. Biol. Psychiatry, 46: 303–311. Gainetdinov, R.R., Jones, S.R., Fumagalli, F., Wightman, R.M. and Caron, M.G. (1998) Re-evaluation of the role of the dopamine transporter in dopamine system homeostasis. Brain Res. Brain Res. Rev., 26: 148–153. Gainetdinov, R.R., Mohn, A.R., Bohn, L.M. and Caron, M.G. (2001) Glutamatergic modulation of hyperactivity in mice lacking the dopamine transporter. Proc. Natl. Acad. Sci. U.S.A., 98: 11047–11054. Gainetdinov, R.R., Wetsel, W.C., Jones, S.R., Levin, E.D., Jaber, M. and Caron, M.G. (1999b) Role of serotonin in the paradoxical calming effect of psychostimulants on hyperactivity. Science, 283: 397–401. Gerber, D.J., Sotnikova, T.D., Gainetdinov, R.R., Huang, S.Y., Caron, M.G. and Tonegawa, S. (2001) Hyperactivity, elevated dopaminergic transmission, and response to amphetamine in M1 muscarinic acetylcholine receptor-deﬁcient mice. Proc. Natl. Acad. Sci. U.S.A., 98: 15312–15317. Gianfriddo, M., Melani, A., Turchi, D., Giovannini, M.G. and Pedata, F. (2004) Adenosine and glutamate extracellular concentrations and mitogen-activated protein kinases in the striatum of Huntington transgenic mice. Selective antagonism of adenosine A2A receptors reduces transmitter outﬂow. Neurobiol. Dis., 17: 77–88. Giros, B., Jaber, M., Jones, S.R., Wightman, R.M. and Caron, M.G. (1996) Hyperlocomotion and indifference to cocaine and amphetamine in mice lacking the dopamine transporter. Nature, 379: 606–612. Goldberg, M.S., Fleming, S.M., Palacino, J.J., Cepeda, C., Lam, H.A., Bhatnagar, A., Meloni, E.G., Wu, N., Ackerson, L.C., Klapstein, G.J., Gajendiran, M., Roth, B.L., Chesselet, M.F., Maidment, N.T., Levine, M.S. and Shen, J. (2003) Parkin-deﬁcient mice exhibit nigrostriatal deﬁcits but not loss of dopaminergic neurons. J. Biol. Chem., 278: 43628–43635. Han, W., Saegusa, H., Zong, S. and Tanabe, T. (2002) Altered cocaine effects in mice lacking Ca(v)2.3 (alpha(1E)) calcium channel. Biochem. Biophys. Res. Commun., 299: 299–304. Hartmann, J., Erb, C., Ebert, U., Baumann, K.H., Popp, A., Konig, G. and Klein, J. (2004) Central cholinergic functions in human amyloid precursor protein knock-in/presenilin-1 transgenic mice. Neuroscience, 125: 1009–1017. He, M., Sibille, E., Benjamin, D., Toth, M. and Shippenberg, T. (2001) Differential effects of 5-HT1A receptor deletion upon basal and ﬂuoxetine-evoked 5-HT concentrations as revealed by in vivo microdialysis. Brain Res., 902: 11–17. Hironaka, N., Yagi, T. and Niki, H. (2002) Light-potentiation of acoustic startle response (ASR) and monoamine efﬂux related to fearfulness in Fyn-deﬁcient mice. Brain Res. Mol. Brain Res., 98: 102–110. Hong, Z.Y., Huang, Z.L., Qu, W.M. and Eguchi, N. (2005) Orexin A promotes histamine, but not norepinephrine or serotonin, release in frontal cortex of mice. Acta Pharmacol. Sin., 26: 155–159.

414 Huang, Z.L., Qu, W.M., Li, W.D., Mochizuki, T., Eguchi, N., Watanabe, T., Urade, Y. and Hayaishi, O. (2001) Arousal effect of orexin A depends on activation of the histaminergic system. Proc. Natl. Acad. Sci. U.S.A., 98: 9965–9970. Hungund, B.L., Szakall, I., Adam, A., Basavarajappa, B.S. and Vadasz, C. (2003) Cannabinoid CB1 receptor knockout mice exhibit markedly reduced voluntary alcohol consumption and lack alcohol-induced dopamine release in the nucleus accumbens. J. Neurochem., 84: 698–704. Huotari, M., Garcia-Horsman, J.A., Karayiorgou, M., Gogos, J.A. and Mannisto, P.T. (2004) d-amphetamine responses in catechol-O-methyltransferase (COMT) disrupted mice. Psychopharmacology (Berl.), 172: 1–10. Huotari, M., Gogos, J.A., Karayiorgou, M., Koponen, O., Forsberg, M., Raasmaja, A., Hyttinen, J. and Mannisto, P.T. (2002a) Brain catecholamine metabolism in catechol-O-methyltransferase (COMT)-deﬁcient mice. Eur. J. Neurosci., 15: 246–256. Huotari, M., Santha, M., Lucas, L.R., Karayiorgou, M., Gogos, J.A. and Mannisto, P.T. (2002b) Effect of dopamine uptake inhibition on brain catecholamine levels and locomotion in catechol-O-methyltransferase-disrupted mice. J. Pharmacol. Exp. Ther., 303: 1309–1316. Ihalainen, J.A. and Tanila, H. (2002) In vivo regulation of dopamine and noradrenaline release by alpha2A-adrenoceptors in the mouse prefrontal cortex. Eur. J. Neurosci., 15: 1789–1794. Ihalainen, J.A. and Tanila, H. (2004) In vivo regulation of dopamine and noradrenaline release by alpha2A-adrenoceptors in the mouse nucleus accumbens. J. Neurochem., 91: 49–56. Ihalainen, J.A., Riekkinen, P. Jr. and Feenstra, M.G. (1999) Comparison of dopamine and noradrenaline release in mouse prefrontal cortex, striatum and hippocampus using microdialysis. Neurosci. Lett., 277: 71–74. Jacobsen, J.P., Rodriguiz, R.M., Mork, A. and Wetsel, W.C. (2005) Monoaminergic dysregulation in glutathione-deﬁcient mice: possible relevance to schizophrenia? Neuroscience, 132: 1055–1072. Javitch, J.A. and Snyder, S.H. (1984) Uptake of MPP(+) by dopamine neurons explains selectivity of parkinsonism-inducing neurotoxin, MPTP. Eur. J. Pharmacol., 106: 455–456. Jones, S.R., Gainetdinov, R.R. and Caron, M.G. (1999a) Application of microdialysis and voltammetry to assess dopamine functions in genetically altered mice: correlation with locomotor activity. Psychopharmacology (Berl.), 147: 30–32. Jones, S.R., Gainetdinov, R.R., Hu, X.T., Cooper, D.C., Wightman, R.M., White, F.J. and Caron, M.G. (1999b) Loss of autoreceptor functions in mice lacking the dopamine transporter. Nat. Neurosci., 2: 649–655. Jones, S.R., Gainetdinov, R.R., Jaber, M., Giros, B., Wightman, R.M. and Caron, M.G. (1998a) Profound neuronal plasticity in response to inactivation of the dopamine transporter. Proc. Natl. Acad. Sci. U.S.A., 95: 4029–4034. Jones, S.R., Gainetdinov, R.R., Wightman, R.M. and Caron, M.G. (1998b) Mechanisms of amphetamine action revealed in mice lacking the dopamine transporter. J. Neurosci., 18: 1979–1986.

Joseph, J.D., Wang, Y.M., Miles, P.R., Budygin, E.A., Picetti, R., Gainetdinov, R.R., Caron, M.G. and Wightman, R.M. (2002) Dopamine autoreceptor regulation of release and uptake in mouse brain slices in the absence of D(3) receptors. Neuroscience, 112: 39–49. Justice, J.B. Jr. (1993) Quantitative microdialysis of neurotransmitters. J. Neurosci. Methods, 48: 263–276. Kehr, J., Yoshitake, T., Wang, F.H., Wynick, D., Holmberg, K., Lendahl, U., Bartfai, T., Yamaguchi, M., Hokfelt, T. and Ogren, S.O. (2001) Microdialysis in freely moving mice: determination of acetylcholine, serotonin and noradrenaline release in galanin transgenic mice. J. Neurosci. Methods, 109: 71–80. Knobelman, D.A., Hen, R., Blendy, J.A. and Lucki, I. (2001) Regional patterns of compensation following genetic deletion of either 5-hydroxytryptamine(1A) or 5-hydroxytryptamine(1B) receptor in the mouse. J. Pharmacol. Exp. Ther., 298: 1092–1100. Koeltzow, T.E., Xu, M., Cooper, D.C., Hu, X.T., Tonegawa, S., Wolf, M.E. and White, F.J. (1998) Alterations in dopamine release but not dopamine autoreceptor function in dopamine D3 receptor mutant mice. J. Neurosci., 18: 2231–2238. Koizumi, M., Midorikawa, N., Takeshima, H. and Murphy, N.P. (2004a) Exogenous, but not endogenous nociceptin modulates mesolimbic dopamine release in mice. J. Neurochem., 89: 257–263. Koizumi, M., Sakoori, K., Midorikawa, N. and Murphy, N.P. (2004b) The NOP (ORL1) receptor antagonist Compound B stimulates mesolimbic dopamine release and is rewarding in mice by a non-NOP-receptor-mediated mechanism. Br. J. Pharmacol., 143: 53–62. Laplante, F., Crawley, J.N. and Quirion, R. (2004a) Selective reduction in ventral hippocampal acetylcholine release in awake galanin-treated rats and galanin-overexpressing transgenic mice. Regul. Pept., 122: 91–98. Laplante, F., Sibley, D.R. and Quirion, R. (2004b) Reduction in acetylcholine release in the hippocampus of dopamine D5 receptor-deﬁcient mice. Neuropsychopharmacology, 29: 1620–1627. Malagie, I., David, D.J., Jolliet, P., Hen, R., Bourin, M. and Gardier, A.M. (2002) Improved efﬁcacy of ﬂuoxetine in increasing hippocampal 5-hydroxytryptamine outﬂow in 5HT(1B) receptor knock-out mice. Eur. J. Pharmacol., 443: 99–104. Malagie, I., Trillat, A.C., Bourin, M., Jacquot, C., Hen, R. and Gardier, A.M. (2001) 5-HT1B autoreceptors limit the effects of selective serotonin re-uptake inhibitors in mouse hippocampus and frontal cortex. J. Neurochem., 76: 865–871. Marubio, L.M., Gardier, A.M., Durier, S., David, D., Klink, R., Arroyo-Jimenez, M.M., McIntosh, J.M., Rossi, F., Champtiaux, N., Zoli, M. and Changeux, J.P. (2003) Effects of nicotine in the dopaminergic system of mice lacking the alpha4 subunit of neuronal nicotinic acetylcholine receptors. Eur. J. Neurosci., 17: 1329–1337. Mateo, Y., Budygin, E.A., John, C.E. and Jones, S.R. (2004) Role of serotonin in cocaine effects in mice with reduced dopamine transporter function. Proc. Natl. Acad. Sci. U.S.A., 101: 372–377.

415 Mathews, T.A., Fedele, D.E., Coppelli, F.M., Avila, A.M., Murphy, D.L. and Andrews, A.M. (2004) Gene dose-dependent alterations in extraneuronal serotonin but not dopamine in mice with reduced serotonin transporter expression. J. Neurosci. Methods, 140: 169–181. Michael, D.J. and Wightman, R.M. (1999) Electrochemical monitoring of biogenic amine neurotransmission in real time. J. Pharm. Biomed. Anal., 19: 33–46. Miyakawa, T., Leiter, L.M., Gerber, D.J., Gainetdinov, R.R., Sotnikova, T.D., Zeng, H., Caron, M.G. and Tonegawa, S. (2003) Conditional calcineurin knockout mice exhibit multiple abnormal behaviors related to schizophrenia. Proc. Natl. Acad. Sci. U.S.A., 100: 8987–8992. Mohn, A.R., Gainetdinov, R.R., Caron, M.G. and Koller, B.H. (1999) Mice with reduced NMDA receptor expression display behaviors related to schizophrenia. Cell, 98: 427–436. Mori, A., Okuyama, K., Horie, M., Taniguchi, Y., Wadatsu, T., Nishino, N., Shimada, Y., Miyazawa, N., Takeda, S., Niimi, M., Kyushiki, H., Kondo, M. and Mitsumoto, Y. (2002) Alteration of methamphetamine-induced striatal dopamine release in mint-1 knockout mice. Neurosci. Res., 43: 251–257. Morishima, Y., Miyakawa, T., Furuyashiki, T., Tanaka, Y., Mizuma, H. and Nakanishi, S. (2005) Enhanced cocaine responsiveness and impaired motor coordination in metabotropic glutamate receptor subtype 2 knockout mice. Proc. Natl. Acad. Sci. U.S.A., 102: 4170–4175. Nagatsu, T., Nakahara, D., Kobayashi, K., Morita, S., Sawada, H., Mizuguchi, T. and Kiuchi, K. (1994) Peripherally administered (6R)-tetrahydrobiopterin increases in vivo tyrosine hydroxylase activity in the striatum measured by microdialysis both in normal mice and in transgenic mice carrying human tyrosine hydroxylase. Neurosci. Lett., 182: 44–46. Nakahara, D., Hashiguti, H., Kaneda, N., Sasaoka, T. and Nagatsu, T. (1993) Normalization of tyrosine hydroxylase activity in vivo in the striatum of transgenic mice carrying human tyrosine hydroxylase gene: a microdialysis study. Neurosci. Lett., 158: 44–46. Nakajima, A., Yamada, K., Nagai, T., Uchiyama, T., Miyamoto, Y., Mamiya, T., He, J., Nitta, A., Mizuno, M., Tran, M.H., Seto, A., Yoshimura, M., Kitaichi, K., Hasegawa, T., Saito, K., Yamada, Y., Seishima, M., Sekikawa, K., Kim, H.C. and Nabeshima, T. (2004) Role of tumor necrosis factor-alpha in methamphetamine-induced drug dependence and neurotoxicity. J. Neurosci., 24: 2212–2225. Olive, M.F., Mehmert, K.K. and Hodge, C.W. (2000a) Microdialysis in the mouse nucleus accumbens: a method for detection of monoamine and amino acid neurotransmitters with simultaneous assessment of locomotor activity. Brain Res. Brain Res. Protoc., 5: 16–24. Olive, M.F., Mehmert, K.K., Messing, R.O. and Hodge, C.W. (2000b) Reduced operant ethanol self-administration and in vivo mesolimbic dopamine responses to ethanol in PKCepsilon-deﬁcient mice. Eur. J. Neurosci., 12: 4131–4140. Parsons, L.H. and Justice, J.B. Jr. (1994) Quantitative approaches to in vivo brain microdialysis. Crit. Rev. Neurobiol., 8: 189–220.

Parsons, L.H., Kerr, T.M. and Tecott, L.H. (2001) 5-HT(1A) receptor mutant mice exhibit enhanced tonic, stress-induced and ﬂuoxetine-induced serotonergic neurotransmission. J. Neurochem., 77: 607–617. Penalva, R.G., Flachskamm, C., Zimmermann, S., Wurst, W., Holsboer, F., Reul, J.M. and Linthorst, A.C. (2002) Corticotropin-releasing hormone receptor type 1-deﬁciency enhances hippocampal serotonergic neurotransmission: an in vivo microdialysis study in mutant mice. Neuroscience, 109: 253–266. Petersen, A., Puschban, Z., Lotharius, J., NicNiocaill, B., Wiekop, P., O’Connor, W.T. and Brundin, P. (2002) Evidence for dysfunction of the nigrostriatal pathway in the R6/1 line of transgenic Huntington’s disease mice. Neurobiol. Dis., 11: 134–146. Picciotto, M.R., Zoli, M., Rimondini, R., Lena, C., Marubio, L.M., Pich, E.M., Fuxe, K. and Changeux, J.P. (1998) Acetylcholine receptors containing the beta2 subunit are involved in the reinforcing properties of nicotine. Nature, 391: 173–177. Pogorelov, V.M., Rodriguiz, R.M., Insco, M.L., Caron, M.G. and Wetsel, W.C. (2005) Novelty seeking and stereotypic activation of behavior in mice with disruption of the Dat1 gene. Neuropsychopharmacology, 30: 1818–1831. Rea, K., Cremers, T.I. and Westerink, B.H. (2005) HPLC conditions are critical for the detection of GABA by microdialysis. J. Neurochem., 94: 672–679. Rocha, B.A., Fumagalli, F., Gainetdinov, R.R., Jones, S.R., Ator, R., Giros, B., Miller, G.W. and Caron, M.G. (1998) Cocaine self-administration in dopamine-transporter knockout mice. Nat. Neurosci., 1: 132–137. Rocha, B.A., Goulding, E.H., O’Dell, L.E., Mead, A.N., Coufal, N.G., Parsons, L.H. and Tecott, L.H. (2002) Enhanced locomotor, reinforcing, and neurochemical effects of cocaine in serotonin 5-hydroxytryptamine 2C receptor mutant mice. J. Neurosci., 22: 10039–10045. Roth, R.H., Doherty, J.D. and Walters, J.R. (1980) Gammahydroxybutyrate: a role in the regulation of central dopaminergic neurons? Brain Res., 189: 556–560. Rouge-Pont, F., Usiello, A., Benoit-Marand, M., Gonon, F., Piazza, P.V. and Borrelli, E. (2002) Changes in extracellular dopamine induced by morphine and cocaine: crucial control by D2 receptors. J. Neurosci., 221: 3293–3301. Salahpour, A., Medvedev, I.O., Beaulieu, J.M., Gainetdinov, R.R. and Caron, M.G. (2006) Local knockdown of genes in the brain using small interfering RNA: a phenotypic comparison with knockout animals. Biol. Psychiatry, [Epub ahead of print]. Shippenberg, T.S., He, M. and Chefer, V. (1999) The use of microdialysis in the mouse: conventional versus quantitative techniques. Psychopharmacology (Berl.), 147: 33–34. Shippenberg, T.S., Hen, R. and He, M. (2000) Regionspeciﬁc enhancement of basal extracellular and cocaine-evoked dopamine levels following constitutive deletion of the Serotonin(1B) receptor. J. Neurochem., 75: 258–265.

416 Sillaber, I., Montkowski, A., Landgraf, R., Barden, N., Holsboer, F. and Spanagel, R. (1998) Enhanced morphineinduced behavioural effects and dopamine release in the nucleus accumbens in a transgenic mouse model of impaired glucocorticoid (type II) receptor function: inﬂuence of longterm treatment with the antidepressant moclobemide. Neuroscience, 85: 415–425. Smith, D.G., Tzavara, E.T., Shaw, J., Luecke, S., Wade, M., Davis, R., Salhoff, C., Nomikos, G.G. and Gehlert, D.R. (2005) Mesolimbic dopamine super-sensitivity in melaninconcentrating hormone-1 receptor-deﬁcient mice. J. Neurosci., 25: 914–922. Sora, I., Wichems, C., Takahashi, N., Li, X.F., Zeng, Z., Revay, R., Lesch, K.P., Murphy, D.L. and Uhl, G.R. (1998) Cocaine reward models: conditioned place preference can be established in dopamine- and in serotonin-transporter knockout mice. Proc. Natl. Acad. Sci. U.S.A., 95: 7699–7704. Sotnikova, T.D., Beaulieu, J.M., Barak, L.S., Wetsel, W.C., Caron, M.G. and Gainetdinov, R.R. (2005) Dopamineindependent locomotor actions of amphetamines in a novel acute mouse model of Parkinson disease. PLoS Biol., 3: e271. Sotnikova, T.D., Budygin, E.A., Jones, S.R., Dykstra, L.A., Caron, M.G. and Gainetdinov, R.R. (2004) Dopamine transporter-dependent and -independent actions of trace amine beta-phenylethylamine. J. Neurochem., 91: 362–373. Stuber, G.D., Wightman, R.M. and Carelli, R.M. (2005) Extinction of cocaine self-administration reveals functionally and temporally distinct dopaminergic signals in the nucleus accumbens. Neuron, 46: 661–669. Sullivan, P.M., Mezdour, H., Aratani, Y., Knouff, C., Najib, J., Reddick, R.L., Quarfordt, S.H. and Maeda, N. (1997) Targeted replacement of the mouse apolipoprotein E gene with the common human APOE3 allele enhances diet-induced hypercholesterolemia and atherosclerosis. J. Biol. Chem., 272: 17972–17980. Sulzer, D., Chen, T.K., Lau, Y.Y., Kristensen, H., Rayport, S. and Ewing, A. (1995) Amphetamine redistributes dopamine from synaptic vesicles to the cytosol and promotes reverse transport. J. Neurosci., 15: 4102–4108. Szapacs, M.E., Mathews, T.A., Tessarollo, L., Ernest Lyons, W., Mamounas, L.A. and Andrews, A.M. (2004) Exploring the relationship between serotonin and brain-derived neurotrophic factor: analysis of BDNF protein and extraneuronal 5-HT in mice with reduced serotonin transporter or BDNF expression. J. Neurosci. Methods, 140: 81–92. Szumlinski, K.K., Lominac, K.D., Oleson, E.B., Walker, J.K., Mason, A., Dehoff, M.H., Klugman, M., Cagle, S., Welt, K., During, M., Worley, P.F., Middaugh, L.D. and Kalivas, P.W. (2005) Homer2 is necessary for EtOH-induced neuroplasticity. J. Neurosci., 25: 7054–7061. Timmerman, W. and Westerink, B.H. (1997) Brain microdialysis of GABA and glutamate: what does it signify? Synapse, 27: 242–261.

Trillat, A.C., Malagie, I., Scearce, K., Pons, D., Anmella, M.C., Jacquot, C., Hen, R. and Gardier, A.M. (1997) Regulation of serotonin release in the frontal cortex and ventral hippocampus of homozygous mice lacking 5-HT1B receptors: in vivo microdialysis studies. J. Neurochem., 69: 2019–2025. Tzavara, E.T., Bymaster, F.P., Davis, R.J., Wade, M.R., Perry, K.W., Wess, J., McKinzie, D.L., Felder, C. and Nomikos, G.G. (2004) M4 muscarinic receptors regulate the dynamics of cholinergic and dopaminergic neurotransmission: relevance to the pathophysiology and treatment of related CNS pathologies. FASEB J., 18: 1410–1412. Tzavara, E.T., Bymaster, F.P., Felder, C.C., Wade, M., Gomeza, J., Wess, J., McKinzie, D.L. and Nomikos, G.G. (2003) Dysregulated hippocampal acetylcholine neurotransmission and impaired cognition in M2, M4 and M2/M4 muscarinic receptor knockout mice. Mol. Psychiatry, 8: 673–679. Ungerstedt, U. (1991) Microdialysis – principles and applications for studies in animals and man. J. Intern. Med., 230: 365–373. Vetter, J.M., Jehle, T., Heinemeyer, J., Franz, P., Behrens, P.F., Jackisch, R., Landwehrmeyer, G.B. and Feuerstein, T.J. (2003) Mice transgenic for exon 1 of Huntington’s disease: properties of cholinergic and dopaminergic presynaptic function in the striatum. J. Neurochem., 85: 1054–1063. Vizi, E.S., Zsilla, G., Caron, M.G. and Kiss, J.P. (2004) Uptake and release of norepinephrine by serotonergic terminals in norepinephrine transporter knock-out mice: implications for the action of selective serotonin reuptake inhibitors. J. Neurosci., 24: 7888–7894. Wang, Y.M., Gainetdinov, R.R., Fumagalli, F., Xu, F., Jones, S.R., Bock, C.B., Miller, G.W., Wightman, R.M. and Caron, M.G. (1997) Knockout of the vesicular monoamine transporter 2 gene results in neonatal death and supersensitivity to cocaine and amphetamine. Neuron, 19: 1285–1296. Westerink, B.H. (1995) Brain microdialysis and its application for the study of animal behaviour. Behav. Brain Res., 70: 103–124. Westerink, B.H. and De Vries, J.B. (1988) Characterization of in vivo dopamine release as determined by brain microdialysis after acute and subchronic implantations: methodological aspects. J. Neurochem., 51: 683–687. Xie, R., Hammarlund-Udenaes, M., de Boer, A.G. and de Lange, E.C. (1999) The role of P-glycoprotein in blood-brain barrier transport of morphine: transcortical microdialysis studies in mdr1a (/) and mdr1a (+/+) mice. Br. J. Pharmacol., 128: 563–568. Xu, F., Gainetdinov, R.R., Wetsel, W.C., Jones, S.R., Bohn, L.M., Miller, G.W., Wang, Y.M. and Caron, M.G. (2000) Mice lacking the norepinephrine transporter are supersensitive to psychostimulants. Nat. Neurosci., 3: 465–471. Yoshitake, T., Wang, F.H., Kuteeva, E., Holmberg, K., Yamaguchi, M., Crawley, J.N., Steiner, R., Bartfai, T., Ogren, S.O., Hokfelt, T. and Kehr, J. (2004) Enhanced hippocampal noradrenaline and serotonin release in galanin-overexpressing

417 mice after repeated forced swimming test. Proc. Natl. Acad. Sci. U.S.A., 101: 354–359. Zambrowicz, B.P. and Sands, A.T. (2003) Knockouts model the 100 best-selling drugs – will they model the next 100? Nat. Rev. Drug Discov., 2: 38–51. Zapata, A., Witkin, J.M. and Shippenberg, T.S. (2001) Selective D3 receptor agonist effects of (+)-PD 128907 on

dialysate dopamine at low doses. Neuropharmacology, 41: 351–359. Zhuang, X., Oosting, R.S., Jones, S.R., Gainetdinov, R.R., Miller, G.W., Caron, M.G. and Hen, R. (2001) Hyperactivity and impaired response habituation in hyperdopaminergic mice. Proc. Natl. Acad. Sci. U.S.A., 98: 1982–1987.

CHAPTER 5.2

The use of microdialysis in neuropsychiatric disease models William T. O’Connor1,2, 1

Applied Neurotherapeutics Research Group, School of Biomolecular and Biomedical Science, UCD Conway Institute, University College Dublin, Belfield, Dublin 4, Ireland 2 Analytical Drug and Data Ltd., UCDNova, University College Dublin, Belfield, Dublin 4, Ireland

Abstract: This review describes the recent use of microdialysis in several neuropsychiatric disease models including schizophrenia, stress, addiction and depression and demonstrates the ability of this tool to target neurotransmitter systems and brain regions implicated in psychiatric illness. Microdialysis studies in psychotropic drug induced, neurodevelopmental and the more recently evolving ‘connectivity’ models are mentioned with particular emphasis on animal models of schizophrenia. The use of microdialysis as a drug delivery tool in neuropsychiatric disease models is also addressed and is of particular value in designing disease-modifying therapies focused on modulating structural and behavioural endophenotypes by targeting speciﬁc neurotransmitter systems and genes. The validity of animal models including how animal models constitute theories about disorders is also addressed. methodology in behavioural research (Westerink, 1995). In addition, the use of microdialysis in animal models of neuropsychiatric disease allows for a multifactoral view with a focus on neurotransmitter interactions in complex neurocircuits as well as identifying faulty connectivity of a developmental origin. This review emphasises the importance of microdialysis in animal models of human psychiatric illness including schizophrenia, stress, addiction and depression but the same principles apply to other psychiatric illnesses and mentions ﬁndings from my own laboratory. Finally, perspectives and possible future developments of the use of microdialysis in neuropsychiatric disease models are discussed.

I. Introduction Complex neuropsychiatric illness cannot be exactly reproduced in non-human primates and rodents. Nevertheless, by providing new information on the role of chemical neurotransmission in behaviour animal models can be important in unravelling mechanisms involved in the aetiology, pathophysiology and symptomatology of such illnesses. Since there is now strong evidence that a primary disturbance in one neurotransmitter system will inevitably inﬂuence several other systems, a modest aim would be to describe aberrations in the patterns of psychological and biochemical events and to try to correlate such aberrations to behavioural events. Microdialysis can be particularly useful in this regard since at present it is the most versatile and practical method to study the neurotransmitter release during behaviour and is already a routine

II. Ethical and theoretical considerations When using microdialysis in neuropsychiatric disease models ethically responsible use of experimental animals should be of the utmost importance. The research scientist must explain how the results

Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

419

DOI: 10.1016/S1569-7339(06)16022-1 Copyright 2007 Elsevier B.V. All rights reserved

420

assay models such as those used in the development of new antipsychotic drugs. Behavioural neuropsychiatric disease models at minimum should have predictive validity and be reliable. B. Face validity refers to the phenomenological similarity between the behaviour exhibited by the animal and the speciﬁc conditions of the human condition. However, it is unrealistic to expect similar behaviours in rodents and humans and the aim is to search for relevant equivalents based on brain regions and neurotransmitter systems assumed to be involved. C. Construct validity is most relevant for the use of microdialysis in neuropsychiatric disease models as it refers to similarity in underlying mechanisms even though the precise expression of behaviours may be different between experimental animals and humans. D. Aetiological validity is an extension of construct validity and refers to the degree of similarity of aetiology between the changes seen in the experimental animals and those observed in the human. Neuropsychiatric disease is complex and animal models can be

may beneﬁt human clinical practice. As few animals as possible should be used and special attention should be paid to minimising pain or discomfort. It is clear that animals that endure pain or distress may provide erroneous data, particularly where subtle changes in neurotransmitter release are studied. II.A. What do animal models tell us? Although it may be impossible to model the entire complex symptomatic spectrum of a neuropsychiatric disease in an animal, selected symptoms may be mimicked and have some validity. Any animal model of neuropsychiatric disease should be discussed in terms of its predictive, face, construct and aetiological validity (Sarter and Bruno, 2002) (Fig. 1). These terms are now discussed. A. Predictive validity and reliability refer respectively to (a) the predictive value that observations made in animals will have for the human condition; (b) the accuracy with which both the experimental and clinical observations are made. Both predictive validity and reliability are important for

Construct Validity BASAL MEASURES TEST PARADIGMS

ENVIRONMENT, GENETICS

NEUROBIOLOGY

PHYSIOLOGY

Face Validity BEHAVIOUR

PHENOTYPES, TRAITS, STATES

PHARMACOLOGY /REVERSAL

Aetiological Validity

Predictive Validity and reliability

Fig. 1. A schematic diagram illustrating the current use of brain microdialysis in neuropsychiatric disease models (as indicated is the stippled area). Microdialysis can be employed to investigate similarities in underlying aetiology and mechanisms and thus contribute to both aetiological and construct validity even when precise expression of behaviours (face validity) may be different between experimental animals and humans. Microdialysis can also be used as a bioassay in drug development and can thus contribute to predictive validity and reliability. Both predictive validity and reliability are important for assay models such as those used in the development of new antipsychotic drugs.

421

used to test hypotheses about aetiology. As will become clear in this review the use of microdialysis can be particularly useful in this regard.

III. Microdialysis methodology in animal models Many of the microdialysis studies mentioned in this review address the role of monoamines and indolamines in neuropsychiatric disease. In this respect dopamine, noradrenaline and serotonin receive the most attention. However, amino acid neurotransmitters such as GABA, glutamate and aspartate are being increasingly investigated in these models. In this case, release is much more difﬁcult to measure and interpret since these amino acids also play important roles in general cell metabolism, in addition to serving as a neurotransmitter. Connectivity between brain regions employing dual probe dialysis is starting to be addressed and will help our understanding of nerve pathways and circuits in the aetiology of neuropsychiatric disease (O’Connor, 1998, 2001; Frantz et al., 2002). Methodology is an important factor in explaining discrepancies between studies and apparently conﬂicting results in the stress models may reﬂect the different stressors used. Glutamate release in the striatum appears to be particularly sensitive to the extent of the stress. Weak stressors such as novelty exposure have little effect while stronger stressors such as restraint or forced swimming are associated with increased glutamate release (Moghaddam, 1993; Ho et al., 2000). Also, the choice of strain or genetically modiﬁed animal is often a key factor in the success of the model. Morilak et al. (2005) reports that repeated exposure to mild intermittent cold stress differs dramatically between Wistar–Kyoto compared with Sprague–Dawley rats. Yadid et al. (2001) shows that in contrast to that observed in control Sprague–Dawley rats the stress-induced increase in mesomimbic dopamine release was weak or absent in a Flinders Sensitive Line rats, an animal model of depression. Taken together, these ﬁndings illustrate the kind of neurobiological variation that may also contribute to the development of stress-related neuropsychiatric disorders in susceptible individuals.

IV. Psychotropic drug induced models of schizophrenia The classical approach to induce a schizophrenia-like condition is by treating animals with psychomimetic agents, that is drugs that induce psychotic-like behaviours in humans. In this regard, two of the most widely used animal models of psychosis have been those that monitor locomotor hyperactivity induced by amphetamine or phencyclidine (PCP) in rats (Geyer and Markou, 1995). Thus, the two most important animal models of psychosis are those induced by hyperfunction of dopamine (hyperdopaminergia) and hypofunction of glutamate (hypogluamatergia) (Carlsson et al., 2001). While spontaneous locomotion is believed to model positive symptoms, prepulse inhibition (PPI) has been widely used to assess deﬁcits in the processing of sensory information (sensorimotor gating) that may underlie some of the cognitive deﬁcits associated with the illness. PPI refers to the habituation to a strong startle stimulus following exposure to a weak lead stimulus (Geyer et al., 1993). The usefulness of microdialysis in these animal models with particular emphasis on the neurotransmitter-based hypotheses of schizophrenia is now discussed.

IV.A. Hyperdopaminergia For more than 25 years the effect of psychotropic drugs such as amphetamine on locomotor activity in rodents has been used to model positive symptoms of schizophrenia, particularly psychosis. Given the strong evidence in support of hyperdopaminergic activity in patients, increased releasability of mesolimbic dopamine in animal models may also be considered as supporting face validity. In this regard Segal and Kuczenski (1997) showed that an escalating dose-run pattern of amphetamine administration is associated with a progressive reduction in extracellular dopamine and serotonin levels in the caudate-putamen, which may reﬂect a shift towards relative activation of mesolimbic dopamine and suggests that this model closely resembles clinical manifestations of amphetamine psychosis. Likewise the effects of NMDA receptor antagonists PCP and

422

ketamine have been attributed in part to a dysregualtion of mesolimbic dopamine transmission (Jentsch and Roth, 1999; Kegeles et al., 2000). It has been proposed that a disruption in the interaction between cortical and subcortical dopamine neurons is involved in the pathophysiology of schizophrenia. King et al. (1997) report that dopamine depletion in the medial prefrontal cortex potentiates the stress-induced increase in nucleus accumbens (shell) dopamine release suggesting that a disruption in the interaction between mesocortical and mesolimbic dopamine neurons is involved in those symptoms of schizophrenia that are inﬂuenced by stress. Gessa (2000) presents convincing evidence that extracellular dopamine in rat prefrontal cortex is largely derived from noradrenergic neurons suggesting that variations in dopaminergic tone in this region may reﬂect an expression of locus ceruleus activity and may then be regulated very differently from the dopamine neurons in ventral tegmental area (VTA). Other studies focused on the consequences of repeated psychostimulant administration in forebrain circuits. For example, repeated amphetamine administration sensitises cortical acetylcholine release (Nelson et al., 2000). Abnormally high increases in cortical acetylcholine release have been hypothesised to represent an integral component of the abnormal regulation of limbic–prefrontal circuits, and to mediate the impairments in information processing that contribute to the development of positive symptoms (Sarter and Bruno, 1999). With respect to the use of microdialysis in PPI studies, Young et al. (2005) show that amphetamine disruption of PPI is a function of a calcium-dependent increase in nucleus accumbens dopamine release at the time of conditioning and that other drugs exacerbating PPI may also increase dopamine release in this brain region.

IV.B. Hypoglutamatergia Elevation of mesolimbic dopamine release in schizophrenia may not necessarily be a primary phenomenon but could be secondary to, for example hypoglutamatergia. In determining the role of glutamate in the control of monoamine release Carlsson et al. (2000) proposed that cortical glutamate

regulates the activity on monaminergic brainstem neurons by means of a direct glutamatergic (i.e., the accelerator) and indirect glutamatergic/GABAergic pathway (i.e., the brake) and Miller and Abercrombie (1996) showed that superimposing an NMDA receptor antagonist upon amphetamine increases striatal dopamine release. However, other studies report that NMDA receptor antagonists alone give variable effects on dopamine release (Svensson, 2000). In addition, while PCP increases prefrontal dopamine and glutamate release and locomotion, the prefrontal glutamate release and the PCP-induced psychostimulation is reversed by an mGlu2 receptor agonist while the enhanced dopamine release remains unchanged (Moghaddam and Adams, 1998). Swanson and Schoepp (2003) also showed that an mGlu2/3 receptor agonist is associated with a prolonged blockade of a PCP-induced increase in locomotion but not a PCP-induced increase in nucleus accumbens dopamine release supporting a role for noradrenaline transmission in the nucleus accumbens in the actions of drugs such as PCP and suggesting that stress pathways associated with these drugs can be normalised by mGlu2/3 receptor activation.

IV.C. Hyperserotonergia NMDA receptor antagonists stimulate serotonin release more consistently than that observed for dopamine (Martin et al., 1998) which may be of interest in the context of postmortem observations suggesting a presynaptic hyperserotonergia in paranoid schizophrenic patients (Hansson et al., 1994). A possible role for serotonin in schizophrenia has also been suggested due to the ﬁnding that activation of a local serotonin 5HT2A receptor is suspected to be a potential mechanism for disruption of prefrontal function in cognition and is known to be involved in hallucinogenesis (Aghajanian and Marek, 1999, 2000). In fact, a major distinction in antipsychotic targets of action is the high afﬁnity of atypical antipsychotics at the 5HT2A receptor where they act as an antagonist. Also, microdialysis studies in both non-human primates and rodents suggest that an important mechanism for the ability of atypicals to ameliorate

423

positive symptoms may be an elevation of prefrontal dopamine release possibly via direct activation of the local 5HT1A receptor (Youngren et al., 1999; Ichikawa et al., 2001; Liegeois et al., 2002). Taken together, these ﬁndings suggest that a mechanism other than dopamine release (possibly serotonin) may account for the psychostimulant and psychogenic actions of NMDA receptor antagonists.

V. Cognition models of schizophrenia While positive and negative symptoms have long been considered the hallmark features of schizophrenia, recent clinical studies have highlighted cognitive dysfunction including deﬁcits in working memory, selective attention and mental ﬂexibility as a third major diagnostic category considered to be the core deﬁcit in the disorder. However, cognitive impairment together with negative symptoms appear to be very difﬁcult to model in rodents (Ellenbroek and Cools, 2000). In general, typical antipsychotic drugs such as clozapine have been associated with less extrapyramidal motor side-effects and assumed to attenuate more effectively the cognitive deﬁcit in schizophrenia than conventional dopamine D2 receptor antagonists such as haloperidol and Drew et al. (1990) reported regional differences in the effects of clozapine and haloperidol on GABA release in rat basal ganglia that may parallel the unique clinical proﬁles of these drugs. One microdialysis model of cognitive impairments in schizophrenia focused on the hyperattentional performance of amphetamine sensitised rats in attentional tasks. These impairments were hypothesised to be due to increases in the reactivity of cortical cholinergic inputs (Nelson et al., 2000) and associated increases in mesolimbic dopamine (Moore et al., 1999). Fujiwara and Egashira (2004) demonstrate that an inhibition of rat dorsal hippocampal glutamate release via activation of the CB1-cannabinoid receptor may be one mechanism involved in Delta9-THC-induced impairment of spatial memory and they suggest that a Delta9-THC analog might provide an effective treatment of psychosis and possibly cognition. While there is a concern that rodent tests of working memory are relevant to the same processes

that are compromised in schizophrenia, nonhuman primates provide an excellent model for the study of inﬂuences on prefrontal function and working memory due to the high degree of homology between human and non-human primates models in the relationship between prefrontal cortex and higher cognitive capacities. In this context, an important mechanism for the ability of atypicals to ameliorate positive and improve cognitive symptoms may be an elevation of prefrontal dopamine release possibly via direct activation of the local 5HT1A receptor and/or blockade of the local 5HT2A and D4 receptor (Youngren et al., 1999). VI. Neurodevelopmental models of schizophrenia In this approach one aims to model the disease by mimicking aetiological factors underlying the condition and there is already a large body of evidence suggesting that genetic and early (pre-, peri- and/ or postnatal) environmental factors interact to play an important aetiological role in schizophrenia. Indeed, one of the most inﬂuential theories on the aetiology of schizophrenia is the so-called neurodevelopmental hypothesis (Pilowsky et al., 1993; Weinberger, 1996). This hypothesis states that due to the interaction between genetic and early environmental factors the normal development of the CNS is disturbed which ultimately leads to the development of clinical symptoms. In fact, many of the novel neuropsychiatric disease models developed over the last 10–15 years have speciﬁcally focussed on this hypothesis. In the schizophrenia models such perinatal manipulations include excitotoxic lesions, diet and stress. The use of microdialysis in two of these neurodevelopmental models is now discussed. VI.A. Prenatal methylazoxymethanol exposure Prenatal methylazoxymethanol (MAM) treatment on gestational day 17 in rats disrupts neurogenesis and Flagstad et al. (2004) reported that (systemic) amphetamine increases dopamine release in the nucleus accumbens but not medial prefrontal cortex. Furthermore, intra-prefrontal cortical perfusion with amphetamine decreases nucleus

424

accumbens dopamine release in controls while increasing it in MAM-exposed rats suggesting that late gestational disruption of neurogenesis leads to a prefrontal dysregulation of subcortical dopamine neurotransmission. VI.B. Neonatal hippocampal and prefrontal lesions Excitotoxic damage to the ventral hippocampus in neonates demonstrates concordance with dopaminergic and prefrontal dysfunction in schizophrenia (Lipska and Weinberger, 2000). In this regard, Lillrank et al. (1999) show that neonatal ventral hippocampal lesions attenuate the responsiveness of mesocorticolimbic dopamine to environmental and pharmacological stressors. Brake et al. (2000) report that while neonatal hippocampal lesions attenuate nucleus accumbens dopamine release in response to stress, prefrontal lesions actually increase the effect of stress on nucleus accumbens dopamine release. Taken together, these ﬁndings suggest that early loss of prefrontal input to mesolimbic dopamine leads to increased dopamine release in response to stress and psychostimulants (Saunders and Kuczenski, 1998). However, the stress-induced increase in nucleus accumbens dopamine release may depend on the time of the lesion (Lillrank et al., 1999; Brake et al., 2000). VII. Environmental models of schizophrenia Derived from the hypothesis that brain maturation processes are more susceptible to permanent modiﬁcation when intervention occurs at an early age, several non-invasive animal models use manipulations of the early life social environment, mainly by isolating the animals from the social group for a certain period of time. Social isolation and maternal separation are the two different environmental manipulations whose effects on neurochemistry have been studied using microdialysis. VII.A. Social isolation Postweanling social isolation alters the maturation of cerebral structures such as those involved

in social play including cortico–striato–limbic circuitry and thus offers an approach for the study of neurodevelopment disorders such as schizophrenia and depression. The discovery of a PPI deﬁcit in isolates is of particular interest since schizophrenic patients also show this impairment (Geyer et al., 1993). Isolation-reared rats display a reduction in mesocortical dopamine release and serotoninergic alterations in the nucleus accumbens and amygdala (Fulford and Marsden, 1996; Lapiz et al., 2003; Leng et al., 2004).

VII.B. Maternal deprivation A single 24-h period of maternal deprivation early in life alters the development of the brain and the neuroendocrine system, which in turn leads to permanent behavioural and physiological modiﬁcations in the adult animal (Ellenbroek et al., 1998). These rats also possess an enhanced sensitivity to dopaminergic drugs and stress, as well as a reduction in latent inhibition suggesting that the maternally deprived rat might represent an interesting animal model for speciﬁc aspects of schizophrenia. Recent fundings from this laboratory report that both the isolated and maternally deprived rat possesses a prefrontal hypoglutamatergia while an additional prefrontal hyperGABAergia is observed in the maternally deprived rat (Mulvany et al., 2006). A strong PPI deﬁcit is also observed in the isolated rat which correlated with the prefrontal hypoglutamatergia while this behavioural deﬁcit is weak or absent in the maternally deprived rat. Brady et al. (2006) revealed that both isolated and maternally deprived rats show enhanced prefrontal glutamate release in response to locally perfused pergolide while an enhanced GABA release was also observed in maternally deprived rats (Fig. 2). A prefrontal dopamine hyper-responsiveness in isolated rats has also been reported by Heidbreder et al., 2001). Taken together, these ﬁndings suggest that elevated prefrontal GABA in maternal deprivation may be protective in PPI and illustrates the usefulness of microdialysis in unravelling the possible aetiology of sensorimotor gating deﬁcits observed in animal models and in schizophrenic patients.

425

A

Glutamate

Total release (60 min) nM

Total release (60 min) µM

50 8 7 6 5 4 3

**

2 1

***

0

GABA *

30

20

10

0 Social Control Isolated Rat

C % Change from basal

B 40

Maternally Deprived

Social Control Isolated Rat Maternally Deprived

Glutamate

D

400 350

Social Control Isolated Rat Maternally Deprived

***

350

***

300

300

250

250

200

200

***

150

150

100

100

50

50

0 -40

GABA

400

-20

0

20

40

60

80

100

0 -40

-20

0

20

40

60

80

100

Fraction (20 min) Fig. 2. Medial prefrontal cortical (mPFC) dialysate glutamate and GABA levels under (A+B) basal and (C+D) following local perfusion with pergolide (1 mM, 60 min) in control, isolated and maternally deprived rats. A+B (raw data): each histogram represents the mean7SEM representing the mean of three dialysate samples (3 20 min) prior to pergolide. Basal dialysate glutamate levels (micromolar) were 3.971.4 (n ¼ 6) in social controls and were reduced by 59% and 95% to 1.670.9 (n ¼ 6, p ¼ 0.003 vs. control) and to 0.270.04 (n ¼ 5, po0.0001) respectively in both isolated and maternally deprived rats. Basal dialysate GABA levels (nanomolar) were similar in the social control 16.8674.65 (n ¼ 6) and isolated rats (14.4174.79) but were increased by 92% to 32.4878.46 (n ¼ 6, p ¼ 0.011 vs. control) in maternally deprived rats. C+D (percent response): each data point represents the mean7SEM percent change from basal. The black bar indicates the period of pergolide perfusion (60 min). Intra-mPFC pergolide (1,000 nM, 60 min) had no effect on local glutamate release in the control but increased glutamate release by 41% in isolated (p ¼ 0.002) and by 141% in maternally deprived rats (po0.0001). Intra-mPFC pergolide had no effect on local GABA release in the control or isolated rats but increased GABA release by 153% in maternally deprived rats (po0.001). po0.05, ** po0.01, *** po0.001 (A+B vs. control) and (C+D vs. basal) (ANOVA).

VIII. Connectivity models Just as reductionism has emphasised the search for an individual culprit among the neurotransmitters, it also attempts to identify a particular brain region as the site of the ‘core’ lesion. During the past decade microdialysis has begun to address the interaction between neurotransmitters and their receptors in nerve pathways and circuits engaged in the control of complex behaviours and is proving useful in investigating neuronal connectivity between brain regions including the brainstem, basal ganglia, limbic cortex and neocortex and demonstrating their engagement in complex circuitries.

The use of microdialysis in some pathways implicated in schizophrenia is now discussed. VIII.A. Prefrontal cortical regulation of the mesolimbic circuit Glutamate-containing pyramidal neurons in the medial prefrontal cortex project to the VTA where they synapse on mesocorticolimbic dopaminecontaining cell bodies and GABA interneurons. Working with social control rats, Harte and O’Connor (2004) demonstrated a differential regulation of glutamate and GABA release in the prefrontal cortex and VTA by prefrontal dopamine

426

D1 and D2 receptors whereby activation of the prefrontal dopamine D1 receptor increases local glutamate release while activation of the prefrontal dopamine D2 receptor inhibits both local glutamate release and VTA glutamate release. A selective prefrontal GABAB receptor-mediated inhibition of VTA glutamate release was also demonstrated (Harte and O’Connor, 2005) suggesting that cortical GABAB receptor activation may be useful in the treatment of schizophrenia when coadministered with conventional neuroleptics. At clinical doses, atypical antipsychotics fully occupy cortical serotonin 5-HT2 receptors, which suggests a strong relationship with their therapeutic action. Pyramidal neurons in the medial prefrontal cortex which project to the VTA are excited via 5-HT2A receptor activation and Bortolozzi et al. (2005) indicates that the activity of VTA dopamine is under the excitatory control of prefrontal 5-HT2A receptors. This observation may help in the understanding of the therapeutic action of atypical antipsychotics. In the VTA Wang et al. (1994) demonstrated that an NMDA-induced increase in nucleus accumbens dopamine release can be increased or decreased depending upon the magnitude of NMDA receptor stimulation within the VTA indicating that activity of VTA dopamine is under excitatory cortical control. Further along the mesolimbic circuit, a tonic mesolimic dopamine mediated inhibition of ventral striopallidal GABA release was demonstrated by Ferre et al. (1994) whereby the dopamine D2 receptor antagonist raclopride increases pallidal GABA release and since dopamine D2 receptor blockade in the ventral striopallidal pathway is associated with the antipsychotic activity of neuroleptics but not with their extrapyramidal motor side-effects, it suggests that this pathway may be a useful for screening antipsychotic drugs (O’Connor, 2001). VIII.B. Ventral hippocampus regulation of mesocorticolimbic dopamine Sensorimotor, attentional, memory and emotional processes involve a functional interaction between the ventral hippocampus and mesolimbic dopamine. In this context, ventral hippocampal NMDA receptor stimulation is facilitatory on mesolimbic

dopamine release (Legault et al., 2000; Mitchell et al., 2000; Peleg-Raibstein et al., 2005) suggesting that aberrant hippocampal activity as found in schizophrenia and mood disorders may contribute to cognitive and emotional functions which are sensitive to imbalanced mesocorticolimbic dopamine transmission.

IX. Stress models Clinical studies provide evidence that stress plays a role in the aetiology of many neuropsychiatric diseases. However, the relationships between stress, neurotransmission and behaviour appear to be context related. For instance, exposure to an uncontrollable but not a controllable stressor results in a wide variety of behavioural outcomes known as learned helplessness. In this context, Bland et al. (2003) demonstrated profound increases in prefrontal monoamine release during inescapable but not escapable stress or no stress suggesting that prefrontal monoamine transmission is involved in stressor controllability and the sensitisation of monoamine neurons by inescapable stress. The usefulness of microdialysis in physical and psychological stress models in revealing the role of speciﬁc neurotransmitters and brain regions is now described. IX.A. Dopamine Regional changes in dopamine release following various types and intensities of stress have been suggested by several lines of evidence (Abercrombie et al., 1989). Very mild stress, such as psychological stress and mild electric foot shock, selectively activates mesocortical dopamine release (Deutch et al., 1985; Kaneyuki et al., 1991). In contrast, more severe stress, such as that associated with severe electric foot shock and cold restraint activates not only mesocortical dopamine but also mesolimbic and nigrostriatal dopamine systems (Deutch and Roth, 1990). Van der Elst et al. (2005) employed apomorphinesusceptible and apomorphine-nonsusceptible rats to demonstrate that a mild stressor (e.g., novelty) is associated with a more profound and prolonged

427

increase in nucleus accumbens dopamine release in apomorphine-susceptible compared with apomorphine-unsusceptible rats suggesting that mesolimbic dopamine is more sensitive to an environmental stressor in apomorphine-susceptible rats. Nanni et al. (2003) suggested that rat exposure to a chronic unavoidable stress is associated with a decrease in nucleus accumbens (shell) dopamine release by demonstrating the rats with previous acquisition of an appetitive behaviour exposed to chronic unavoidable stress showed a higher dopamine release than that observed in stressed controls.

IX.B. Noradrenaline Physical stressors such as immobilisation or electric foot shock more markedly increase noradrenaline release in the more extended brain regions as compared with psychological stressors such as conditioned fear which increases noradrenaline release preferentially in the hypothalamus, amygdala and locus coeruleus. Furthermore, rats unable to express aggression display a more profound and prolonged increase in noradrenaline release in the amygdala when compared with those allowed to express aggression while aged rats exhibited a more delayed return to pre-stress levels compared with younger controls. Randomly timed stress periods result in greater increases in noradrenaline release than those observed following continuous stress suggesting that the extent and context in which the stressor is applied has differential effects on noradrenaline release (Tanaka, 1999). Pacak et al. (1992) report that acute immobilisation stress enhances noradrenaline release in rat paraventricular nucleus while a decrease is observed following chronic exposure possibly via reduced release or enhanced reuptake. In addition, acute immobilisation stress also activates noradrenaline release in the central and medial amygdala, lateral bed nucleus of the stria terminalis, medial prefrontal cortex and lateral septum and these effects are strain-dependent (Morilak et al., 2005). These ﬁndings suggest that a dysregulation of noradrenergic transmission may be a factor in determining vulnerability to genetic and environmental sensitisation and to stress-related pathology.

IX.C. Serotonin Increases in serotonin release have been reported in many brain regions in response to several physiological, environmental and behavioural stimuli (Rueter et al., 1997) and probably reﬂects an increase in behavioural arousal/motor activity associated with the manipulation. Using C57BL/6N mice, Beekman et al. (2005) showed that exposure to a predator (rat) increases hippocampal, cortical, septal but not striatal serotonin release suggesting a differential activation of serotonergic transmission in higher brain structures probably involved in the coping response to this potentially lifethreatening situation. In the streptozotocin-elicited diabetes model Thorre et al. (1997) reported that restraint stress increases hippocampal serotonin release in control but not in streptozotocin-pretreated rats suggesting that the prevalence of diabetes among patients suffering affective disorders could be related to the lack of hippocampal serotonergic response to aversive stimuli and that in this respect the streptozotocin-elicited rat model of diabetes may be a useful animal model of affective disorder. IX.D. Amino acids Increased glutamate release of fast onset and short duration in the nucleus accumbens, hippocampus and prefrontal cortex is associated with novelty stress, tail pinch and restraint stress (Lowy et al., 1993; Saulskaya and Marsden, 1995; Wheeler et al., 1995). Furthermore, the stress-induced increase in glutamate and aspartate release is reported to be highest in the prefrontal cortex and diazepam pre-treatment blocks this increase in this brain region (Bagley and Moghaddam, 1997). X. Anxiety and panic models Kanno et al. (2003) employed a rat anxiety model to investigate the effects of repetitive transcranial magnetic stimulation (rTMS) on prefrontal dopamine and serotonin release during the plus-maze test and shows that chronic rTMS suppressed the increase in prefrontal serotonin but not dopamine

428

release associated with the plus-maze test suggesting that chronic rTMS of the frontal brain has anxiolytic effects in rats which are related to local serotonin release. The involvement of the neuropeptide cholecystokinin (CCK) in the mechanisms of anxiety was assessed by Becker et al. (2001) who employed microdialysis in a model of anticipation of social defeat in the rat and reported an increase in cortical release of CCK-like material in defeated intruders. Pretreatment with diazepam prevents both the anxiety-related behaviour and CCK release while CCKB receptor blockade reduces anxiety-like behaviour without affecting the increase in CCK release indicating that anticipation of social defeat induces a marked activation of cortical CCK release associated with anxiety-related behaviours in rats. Sodero et al. (2004) employed an animal model for screening antipanic agents whereby an increase in locus coeruleus activity is observed in perinatally protein-deprived rats and considering the ability of serotonin to inﬂuence the locus coeruleus also assessed the effect of ﬂuoxetine. Prefrontal noradrenaline release was higher in proteindeprived rats compared with controls and was reduced after ﬂuoxetine. The alpha2-adrenergic antagonist yohimbine increased noradrenaline release in control but not in protein-deprived rats suggesting a subsensitivity of the prefrontal alpha2-adrenergic autoreceptor in early protein malnourished animals.

XI. Obsessive–compulsive and attention deﬁcit hyperactivity disorder models Motor stereotypies are characterised by repetitive, topologically invariant, apparently purposeless behaviours and are a common in obsessive–compulsive disorder. In animals motor stereotypies can be induced or attenuated via pharmacological manipulation of striatal neurochemistry. In a mouse model of stereotypy using Peromyscus maniculatus (deer mice) which display stereotypies; Presti et al. (2004) showed that rearing behaviour was associated with increased striatal glutamate and aspartate release supporting a role for striatal

glutamate and aspartate in spontaneous stereotypic behaviour. Spontaneously hypertensive and naples high-excitability rats show behavioural traits featuring the main aspects of attention deﬁcit hyperactivity disorder (ADHD) in humans but differ in striatal and prefrontal dopamine release such that both the mesocortical and mesolimbic dopamine pathways are active in the spontaneously hypertensive rats whereas only the mesocortical pathway is active in the naples high excitability rats (Viggiano et al. 2004). Fujita et al. (2003) compared the responsiveness of nucleus accumbens and ventrolateral striatal dopamine release to dopamine receptor agonists in the spontaneously hypertensive and control progenitor Wistar–Kyoto rat and demonstrated a hyposensitive postsynaptic dopamine D1 and D2 and a hypersensitive presynaptic dopamine D2 receptor in the nucleus accumbens of the spontaneously hypertensive rat.

XII. Addiction models Drugs of abuse produce their behavioural reinforcement in part, by activating mesolimbic dopamine release even following repeated use (Pothos et al. 1991; Wise et al. 1995). A preferential increase in nucleus accumbens and ventromedial striatal dopamine release was also observed in the non-human primate after cocaine self-administration with microdialysis (Bradberry et al. 2000). Segal et al. (2005) studied the effects of a schedule of intravenous methamphetamine administration in rats and reported that dopamine release in the caudate was reduced 6 days after the ﬁnal methamphetamine injection and was only partially recovered to pre-treatment levels 30 days following cessation indicating long-term damage to those terminals maintaining functional dopamine transmission. Using a rat model of binge eating (enhanced rebound hyperphagia induced by space restriction) Inoue et al. (1998) report increased prefrontal and ventral striatal dopamine release while Di Chiara et al. (2004) and others propose that dopamine release in the nucleus accumbens (core) and prefrontal cortex acts as an interface

429

XIII. Depression models The monoamine hypothesis states that depression reﬂects a deﬁciency in one or other biogenic monoamines. In this regard, the ability of both typical and atypical antidepressant drugs to increase monoamine release in the prefrontal cortex is a useful predictive indicator of therapeutic effect (Jordan et al., 1994). Also in line with this view, both acute and chronic electroconvulsive shock (ECS) increases nucleus accumbens dopamine release. Nomikos et al. (1991) report that while chronic ECS did not inﬂuence the

50

40

% Change from basal

between motivation and action while dopamine release in nucleus accumbens (shell) is a sensitive site for dopamine-dependent reward. Exposure to stressful experiences is a powerful mediator of the response of an individual to drugs. In this context, Yamamoto et al. (2001) employed an animal model of drug dependence called a relapse/reinstatement model to show that the timing of drug-seeking responses can be predicted from ﬂuctuations in nucleus accumbens dopamine release. Also, an AMPA but not a dopamine receptor antagonist prevents cocaine-seeking behaviour suggesting that nucleus accumbens glutamate release may be important for the expression of craving or drug-seeking behaviour. Shaham et al. (1996) also used an animal model of relapse to show that while naloxone induces withdrawal symptoms and reduces nucleus accumbens dopamine release it did not reinstate drug-seeking when given to heroinmaintained animals. In fact, reinstatement of heroin-seeking behaviour during spontaneous withdrawal was initiated by a brief stressor (i.e., an exposure to foot shock) without affecting dopamine release indicating that an opiod mechanism(s) mediating dependence may not exclusively deploy mesolimbic dopamine. McBride (2002) and McBride et al. (2002) demonstrated that acute ethanol increases amygdala dopamine and serotonin release and long-lasting alterations in nucleus accumbens dopamine release following chronic alcohol consumption that followed a prolonged deprivation period and which may underlie alcohol relapse.

30

20

10

0

-10

-20

Control

GABA

Control

Glutamate

Fig. 3. Medial prefrontal cortical dialysate glutamate and GABA levels following local perfusion with amitriptyline (1 mM, 60 min). Animals perfused with ringer alone acted as controls. Each histogram represents mean7SEM percent change (maximal response) from basal pre-treatment level (n ¼ 5–7 animals per group). Basal dialysate glutamate and GABA levels were similar in control and drug-treated groups. *** po0.001 vs. control.

apomorphine-induced decrease in nucleus accumbens dopamine release it enhances and prolongs the amphetamine-induced increases in dopamine release indicating that chronic ECS increases mesolimbic dopamine function. In addition, both typical and atypical antidepressants decrease prefrontal glutamate release and increase GABA release (Fig. 3) suggesting that antidepressants induce adaptatory changes in prefrontal amino acid receptors seen after prolonged antidepressant treatment (Nowak et al., 1998; Golembiowska and Dziubina, 2001; Pilc and Nowak, 2005). More sophisticated depression models possessing a broader range of biomarkers from the immunological and endocrinological to neurochemical and behavioural have also emerged. The use of microdialysis in these holistic models has the potential to reveal an underlying aetiology and also afford more precise and predictive models of novel antidepressant therapies. The use of microdialysis in two of these animal models is now discussed.

430

XIII.A. Learned helplessness Learned helplessness is a behavioural depression caused by exposure to inescapable stress and is considered to be an animal model of human depressive disorder. Gambarana et al. (2001) report that exposure to chronic stress not only induces an escape deﬁcit but also decreases nucleus accumbens (shell) dopamine release and that both the behavioural and neurochemical deﬁcits are reverted by long-term treatment with antidepressants. Yadid et al. (2001) monitored mesolimbic dopaminergic adaptation to a stressful stimulus in a Flinders Sensitive Line, a rat model of depression after a forced swim test and reported an absence of a stress-induced elevation in nucleus accumbens dopamine release suggesting that this stressful stimuli appears to disrupt the functionality of mesolimbic dopamine transmission in this animal model. XIII.B. Olfactory bulbectomy Bilateral removal of the olfactory bulbs in rats results in a complex constellation of behavioural, neurochemical, neuroendocrine and neuroimmune alterations, many of which are also reported in patients with major depressive disorder. Masini et al. (2004) report higher basal ventral striatal dopamine but lower norepinephrine release following bulbectomy which are also associated with increased locomotor activity in response to novelty and foot shock stress. The ﬁnding of higher ventral striatal dopamine may explain ‘agitation-like’ behaviour a commonly observed phenomenon in this model and resemble symptoms of psychomotor agitation while the lower noradrenaline may reﬂect previous clinical evidence of noradrenergic dysregulation in affective disorders. More recently, Van der Stelt et al. (2005) report a profound and long-lasting decrease in dorsal hippocampal and basolateral amygdala serotonin release possibly contributing to the spatial memory deﬁcits in bulbectomised animals. XIV. Concluding remarks and outlook The translational value of neuropsychiatric disease models depends in large part on the degree to

which they reproduce patterns similar to those experienced in the human illness. In addition, while animal models may contribute to elucidating some aspect of neuropsychiatric diseases, they require consideration of the natural life of the animal species studied and of their social behaviour in an evolutionary perspective. In this regard, there is substantial evidence that early life events inﬂuence brain development and subsequent adult behaviour and play an important role in the causation of certain neuropsychiatric diseases. The underlying mechanism(s) of the effects of these early environmental factors is still not understood. However, the ﬁndings in this review indicate that microdialysis is a useful tool in understanding how these early life events affect the neurochemistry and neuronal connectivity that may underlie the abnormal brain development in the aetiology of neuropsychiatric disease. The use of microdialysis also allows for a better understanding of the neurobiological mechanisms underlying therapy including the prediction of novel and better treatments for neuropsychiatric disease. At present, there is a major unmet need for drugs to treat the cognitive deﬁcits in schizophrenia. In this regard, the use of microdialysis in the study of neurotransmitter interactions within and between brain regions can facilitate the development of novel compounds targeted to treat those cognitive deﬁcits not limited to a single pharmacological class. It is hoped that we are edging closer to a stage where we can predict drug efﬁcacy in mentally ill patients (against all classes of symptoms) from the efﬁcacy seen in neuropsychiatric disease models. This in turn should greatly facilitate the discovery of improved drug treatments for the disease. Acknowledgements Supported by an SFI Investigator Award, HEA, NDP, PRTLI and Wyeth. The researchers in my laboratory Ian, Olive, Mary, Aine, Aoife, Angela, Sean and Stuart for help with the environmental models. Thanks to Michael Harte for help with the antidepressant study and to Dave Reynolds for his kind gift of HPLC integrator equipment. Abstracts of some of the data from the author’s laboratory

431

have been presented previously at the Monitoring Molecules in Neuroscience Meeting in Sardinia 2006.

References Abercrombie, E.D., Keefe, K.A., DiFrischia, D.S. and Zigmond, M.J. (1989) Differential effect of stress on in vivo dopamine release in striatum, nucleus accumbens, and medial frontal cortex. J. Neurochem., 52: 1655–1658. Aghajanian, G.K. and Marek, G.J. (1999) Serotonin and hallucinogens. Neuropsychopharmacology, 21: 16S–23S. Aghajanian, G.K. and Marek, G.J. (2000) Serotonin model of schizophrenia: emerging role of glutamate mechanisms. Brain Res. Rev., 31: 302–312. Bagley, J. and Moghaddam, B. (1997) Temporal dynamics of glutamate efﬂux in the prefrontal cortex and in the hippocampus following repeated stress: effects of pretreatment with saline or diazepam. Neuroscience, 77: 65–73. Becker, C., Thiebot, M.H., Touitou, Y., Hamon, M., Cesselin, F. and Benoliel, J.J. (2001) Enhanced cortical extracellular levels of cholecystokinin-like material in a model of anticipation of social defeat in the rat. J. Neurosci., 21: 262–269. Beekman, M., Flachskamm, C. and Linthorst, A.C. (2005) Effects of exposure to a predator on behaviour and serotonergic neurotransmission in different brain regions of C57bl/6N mice. Eur. J. Neurosci., 21: 3453. Bland, S.T., Hargrave, D., Pepin, J.L., Amat, J., Watkins, L.R. and Maier, S.F. (2003) Stressor controllability modulates stress-induced dopamine and serotonin efﬂux and morphineinduced serotonin efﬂux in the medial prefrontal cortex. Neuropsychopharmacology, 28: 1589–1596. Bortolozzi, A., Dı´ az-Mataix, L., Cecilia Scorza, M., Celada, P. and Artigas, F. (2005) The activation of 5-HT2A receptors in prefrontal cortex enhances dopaminergic activity. J. Neurochem., 95: 1597–1607. Bradberry, C.W., Barrett-Larimore, R.L., Jatlow, P. and Rubino, S.R. (2000) Impact of self-administered cocaine and cocaine cues on extracellular dopamine in mesolimbic and sensorimotor striatum in rhesus monkeys. J. Neurosci., 20: 3874–3883. Brady, A.T., DeSouza, I.E.J., McCabe, O.M., Moran, M.P., Duffy, A.M., Mulvany, S.K. and O’Connor, W.T. (2006) Altered dopamine regulation of medial prefrontal glutamate and GABA release in the socially isolated and maternally deprived rat. An in vivo microdialysis study. Proceedings of the 11th International Congress on In Vivo Methods. May 17–21, Sardinia, Italy. Brake, W.G., Flores, G., Francis, D., Meaney, M.J., Srivanasta, L.K. and Gratton, A. (2000) Enhanced nucleus accumbens dopamine and plasma corticosterone stress responses in adult rats with neonatal excitotoxic lesions to the medial prefrontal cortex. Neuroscience, 96: 687–695. Carlsson, A., Waters, N., Holm-Waters, S., Tedroff, J., Nilsson, M. and Carlsson, M.L. (2001) Interactions between

monoamines, glutamate, and GABA in schizophrenia: new evidence. Annu. Rev. Pharmacol. Toxicol., 41: 237–260. Carlsson, A., Waters, N., Waters, S. and Carlsson, M. (2000) Network interactions in schizophrenia: therapeutic implications. Brain Res. Rev., 31: 342–349. Deutch, A.Y. and Roth, R.H. (1990) The determinants of stress-induced activation of the prefrontal cortical dopamine system. Prog. Brain Res., 85: 367–402. Deutch, A.Y., Tam, S.-Y. and Roth, R.H. (1985) Footshock and conditioned stress increase 3,4-dihydroxyphenylacetic acid (DOPAC) in the VTA but not substantia nigra. Brain Res., 333: 143–146. Di Chiara, G., Bassareo, V., Fenu, S., Antonietta De Luca, M., Spina, L., Cadoni, C., Acquas, E., Carboni, E., Valentini, V. and Lecca, D. (2004) Dopamine and drug addiction: the nucleus accumbens shell connection. Neuropharmacology, 47: 227–241. Drew, K.L., O’Connor, W.T., Kehr, J. and Ungersted, U. (1990) Regional speciﬁc effects of clozapine and haloperidol on GABA and dopamine release in rat basal ganglia. Eur. J. Pharmacol., 187: 385–397. Ellenbroek, B.A. and Cools, A.R. (2000) Animal models for the negative symptoms of schizophrenia. Behav. Pharmacol., 11: 223–233. Ellenbroek, B.A., van den Kroonenberg, P.T. and Cools, A.R. (1998) The effects of an early stressful life event on sensorimotor gating in adult rats. Schizophr. Res., 30: 251–260. Ferre´, S., O’Connor, W.T., Snaprud, P., Ungerstedt, U. and Fuxe, K. (1994) Antagonistic interaction between adenosine A2A receptors and dopamine D2 receptors in the ventral striopallidal system. Implications for the treatment of schizophrenia. Neuroscience, 63: 765–773. Flagstad, P., Mork, A., Glenthoj, B.Y., van Beek, J., MichaelTitus, A.T. and Didriksen, M. (2004) Disruption of neurogenesis on gestational day 17 in the rat causes behavioral changes relevant to positive and negative schizophrenia symptoms and alters amphetamine-induced dopamine release in nucleus accumbens. Neuropsychopharmacology, 29: 2052–2064. Frantz, K., Harte, M., Ungerstedt, U. and O’ Connor, W.T. (2002) A dual probe characterization of dialysate amino acid levels in the medial prefrontal cortex and ventral tegmental area of the awake freely moving rat. J. Neurosci. Methods, 119: 109–119. Fujita, S., Okutsu, H., Yamaguchi, H., Nakamura, S., Adachi, K., Saigusa, T. and Koshikawa, N. (2003) Altered pre- and postsynaptic dopamine receptor functions in spontaneously hypertensive rat: an animal model of attention-deﬁcit hyperactivity disorder. J. Oral. Sci., 45: 75–83. Fujiwara, M. and Egashira, N. (2004) New perspectives in the studies on endocannabinoid and cannabis: abnormal behaviors associate with CB1 cannabinoid receptor and development of therapeutic application. J. Pharmacol. Sci., 96: 362–366. Fulford, A.J. and Marsden, C.A. (1996) Effect of isolationrearing on conditioned dopamine release in vivo in the nucleus accumbens of the rat. J. Neurochem., 70: 384–390.

432 Gambarana, C., Tolu, P.L., Masi, F., Rinaldi, M., Giachetti, D., Morazzoni, P. and De Montis, M.G. (2001) A study of the antidepressant activity of Hypericum perforatum on animal models. Pharmacopsychiatry, 34(Suppl. 1): S42–S44. Gessa, G.L. (2000). Co-release of noradrenaline and dopamine from noradrenergic neurons in the prefrontal cortex. Proceedings Stockholm-Cagliari Joint Conference on Neurosciences, 1st, p. 22 (Abstract) Stockholm: Italian Institute of Culture. Geyer, M.A. and Markou, A. (1995) Animal models of psychiatric disorders. In: Bloom, F.E. and Kupfer, D.J. (Eds.), Psychopharmacology: The Fourth Generation of Progress. Raven, New York, pp. 787–798. Geyer, M.A., Wilkinson, L.S., Humby, T. and Robbins, T.W. (1993) Isolation rearing of rats produces a deﬁcit in prepulse inhibition of acoustic startle similar to that in schizophrenia. Biol. Psychiatry, 34: 361–372. Golembiowska, K. and Dziubina, A. (2001) Involvement of adenosine in the effect of antidepressants on glutamate and aspartate release in the rat prefrontal cortex. Naunyn Schmiedebergs Arch. Pharmacol., 363: 663–670. Hansson, L.O., Waters, N., Windblad, B., Gotfries, C.-G. and Carlsson, A. (1994) Evidence for biochemical heterogeneity in schizophrenia: a multivariate study of monoaminergic indices in human postmortem brain tissue. J. Neural. Transm., 98: 217–235. Harte, M. and O’Connor, W.T. (2004) Evidence for a differential medial prefrontal dopamine D1 and D2 receptor regulation of local and ventral tegmental glutamate and GABA release: a dual probe microdialysis study in the awake rat. Brain Res., 1017: 120–129. Harte, M. and O’Connor, W.T. (2005) Evidence for a selective prefrontal cortical GABAB receptor-mediated inhibition of glutamate release in the VTA: a dual probe microdialysis study in the awake rat. Neuroscience, 130: 215–222. Heidbreder, C.A., Foxton, R., Cilia, J., Hughes, Z.A., Shah, A.J., Atkins, A., Hunter, A.J., Hagan, J.J. and Jones, D.N. (2001) Increased responsiveness of dopamine to atypical, but not typical antipsychotics in the medial prefrontal cortex of rats reared in isolation. Psychopharmacology (Berl.), 156: 338–351. Ho, Y.-J., Chang, Y.-C., Liu, T.-M., Tai, M.-Y., Wong, C.-S. and Tsai, Y.-F. (2000) Striatal glutamate release during novelty exposure-induced hyperactivity in olfactory bulbectomized rats. Neurosci. Lett., 287: 117–120. Ichikawa, J., Ishii, H., Bonaccorosso, S., Fowler, W.L., O’Laughlin, I.A. and Metzler, H.Y. (2001) 5HT2A and D2 receptor receptor blockade increases cortical dopamine release via 5HT1A receptor activation: a possible mechanism of atypical antipsychotic action- induced cortical dopamine release. J. Neurochem., 76: 1521–1531. Inoue, K., Kiriike, N., Okuno, M., Fujisaki, Y., Kurioka, M., Iwasaki, S. and Yamagami, S. (1998) Prefrontal and striatal dopamine metabolism during enhanced rebound hyperphagia induced by space restriction – a rat model of binge eating. Biol. Psychiatry, 44: 1329–1336. Jentsch, J.D. and Roth, R.H. (1999) The neuropsychopharmacology of phencyclidene: from NMDA receptor hypofunction

to the dopamine hypothesis of schizophrenia. Neuropsychopharmacology, 200: 201–225. Jordan, S., Kramer, G.L., Zukas, P.K., Moeller, M. and Petty, F. (1994) In vivo biogenic amine efﬂux in medial prefrontal cortex with imipramine, ﬂuoxetine, and ﬂuvoxamine. Synapse, 4: 294–297. Kaneyuki, H., Yokoo, H., Tsuda, A., Yoshida, M., Mizuki, Y., Yamada, M. and Tanaka, M. (1991) Psychological stress increases dopamine turnover selectively in mesoprefrontal dopamine neurons of rats: reversal by diazepam. Brain Res., 557: 154–161. Kanno, M., Matsumoto, M., Togashi, H., Yoshioka, M. and Mano, Y. (2003) Effects of repetitive transcranial magnetic stimulation on behavioral and neurochemical changes in rats during an elevated plus-maze test. J. Neurol. Sci., 211: 5–14. Kegeles, L.S., Abi-Dargham, A., Zea-Ponce, Y., Rodenhiser-Hill, J., Mann, J.J., Van Heertum, R.L., Cooper, T.B., Carlsson, A. and Laruelle, M. (2000) Modulation of amphetamine-induced striatal dopamine release by ketamine in humans: implications for schizophrenia. Biol. Psychiatry, 48: 627–640. King, D., Zigmond, M.J. and Finlay, J.M. (1997) Effects of dopamine depletion in the medial prefrontal cortex on stressinduced increases in extracellular dopamine in the nucleus accumbens core and shell. Neuroscience, 77: 141–153. Lapiz, M.D., Fulford, A., Muchimapura, S., Mason, R., Parker, T. and Marsden, C.A. (2003) Inﬂuence of postweaning social isolation in the rat on brain development, conditioned behavior, and neurotransmission. Neurosci. Behav. Physiol., 33: 13–29. Legault, M., Rompre, P.P. and Wise, R.A. (2000) Chemical stimulation of the ventral hippocampus elevates nucleus accumbens dopamine by activating dopaminergic neurons of the VTA. J. Neurosci., 20: 1635–1642. Leng, A., Feldon, J. and Ferger, B. (2004) Long-term social isolation and medial prefrontal cortex: dopaminergic and cholinergic neurotransmission. Pharmacol. Biochem. Behav., 77(2): 371–379. Liegeois, J.F., Ichikawa, J. and Metzler, H.Y. (2002) 5HT2A receptor antagonism potentiates haloperidol-induced dopamine release in rat medial prefrontal cortex and inhibits that in the nucleus accumbens in a dose-dependent manner. Brain Res., 947: 154–165. Lillrank, S.M., Lipska, B.K., Kolachana, B.S. and Weinberger, D.R. (1999) Attenuated extracellular dopamine levels after stress and amphetamine in the nucleus accumbens of rats with neonatal hippocampal damage. J. Neural. Transm., 106: 183–196. Lipska, B.K. and Weinberger, D.R. (2000) To model a psychiatric disorder in animals. Schizophrenia as a reality test. Neuropsychopharmacology, 23: 223–239. Lowy, M.T., Gault, L. and Yamamoto, B.K. (1993) Adrenalectomy attenuates stress-induced elevations in extracellular glutamate concentrations in the hippocampus. J. Neurochem., 61: 1957–1960. Martin, P., Carlsson, M.L. and Hjort, S. (1998) Systemic PCP treatment elevates brain extracellular serotonin: a microdialysis study in awake rats. Neuroreport, 9: 1285–1288.

433 Masini, C.V., Holmes, P.V., Freeman, K.G., Maki, A.C. and Edwards, G.L. (2004) Dopamine overﬂow is increased in olfactory bulbectomized rats: an in vivo microdialysis study. Physiol. Behav., 81: 111–119. McBride, W.J. (2002) Central nucleus of the amygdala and the effects of alcohol and alcohol-drinking behavior in rodents. Pharmacol. Biochem. Behav., 71: 509–515. McBride, W.J., Le, A.D. and Noronha, A. (2002) Central nervous system mechanisms in alcohol relapse. Alcohol. Clin. Exp. Res., 26: 280–286. Miller, D.W. and Abercrombie, E.D. (1996) Effects of MK-801 on spontaneous and amphetamine-stimulated dopamine release in the striatum measures with in vivo microdialysis in awake rats. Brain Res. Bull., 40: 320–329. Mitchell, S.N., Yee, B.K., Feldon, J., Gray, J.A. and Rawlins, J.N.P. (2000) Activation of the retrohippocampal region in the rat causes dopamine release in the nucleus accumbens: disruption by fornix section. Eur. J. Pharmacol., 407: 131–138. Moghaddam, B. (1993) Stress preferentially increases extraneuronal levels of excitatory amino acids in the prefrontal cortex: comparison to hippocampus and basal ganglia. J. Neurochem., 60: 1650–1657. Moghaddam, B. and Adams, B.W. (1998) Reversal of phencyclidene effects by a group II metabotropic glutamate recepror agonist in rats. Science, 281: 1349–1352. Moore, H., Fadel, J., Sarter, M. and Bruno, J.P. (1999) Role of acumbes and cortical dopamine receptors in the regulation of cortical acetylcholine release. Neuroscience, 88: 811–822. Morilak, D.A., Barrera, G., Echevarria, D.J., Garcia, A.S., Hernandez, A., Ma, S. and Petre, C.O. (2005) Role of brain norepinephrine in the behavioral response to stress. Prog. Neuropsychopharmacol. Biol. Psychiatry, 29: 1214–1224. Mulvany, S.K., De Souza, I.E.J., Murphy, K.J., ter Horst, J., Cassidy, A., Seymour, C., Moran, M., McCabe, O., Brady, A., Marquis, K.L., Logue, S.F., Pangalos, M., Regan, C.M. and O’Connor, W.T. (2006) Developmental emergence of sensorimotor gating deﬁcits and adult hypoglutamatergia in two animal models of schizophrenia, (manuscript in preparation). Nanni, G., Scheggi, S., Leggio, B., Grappi, S., Masi, F., Rauggi, R. and De Montis, M.G. (2003) Acquisition of an appetitive behavior prevents development of stress-induced neurochemical modiﬁcations in rat nucleus accumbens. J. Neurosci. Res., 15: 573–580. Nelson, C.L., Sarter, M. and Bruno, L.P. (2000) Repeated pretreatment with amphetamine sensitizes increases in cortical acetylcholine release. Psychopharmacology (Berl.), 151: 406–415. Nomikos, G.G., Zis, A.P., Damsma, G. and Fibiger, H.C. (1991) Effects of chronic electroconvulsive shock on interstitial concentrations of dopamine in the nucleus accumbens. Psychopharmacology (Berl.), 105: 230–238. Nowak, G., Legutko, B., Skolnick, P. and Popik, P. (1998) Adaptation of cortical NMDA receptors by chronic treatment with speciﬁc serotonin reuptake inhibitors. Eur. J. Pharmacol., 342: 367–370.

O’Connor, W.T. (1998) Functional neuroanatomy of the basal ganglia as studied by dual-probe microdialysis. Nucl. Med. Biol., 25: 743–746. O’Connor, W.T. (2001) Functional neuroanatomy of the ventral striopallidal GABA pathway. New sites of intervention for neurotensin and CCK in the treatment of schizophrenia. J. Neurosci. Methods, 109: 31–39. Pacak, K., Armando, I., Fukuhara, K., Kvetnansky, R., Palkovits, M., Kopin, I.J. and Goldstein, D.S. (1992) Noradrenergic activation in the paraventricular nucleus during acute and chronic immobilization stress in rats: an in vivo microdialysis study. Brain Res., 589: 91–96. Peleg-Raibstein, D., Pezze, M.A., Ferger, B., Zhang, W.N., Murphy, C.A., Feldon, J. and Bast, T. (2005) Activation of dopaminergic neurotransmission in the medial prefrontal cortex by N-methyl-D-aspartate stimulation of the ventral hippocampus in rats. Neuroscience, 132: 219–232. Pilc, A. and Nowak, G. (2005) GABAergic hypotheses of anxiety and depression: focus on GABA-B receptors. Drugs Today (Barc.), 41: 755–766. Pilowsky, L.S., Kerwin, R.W. and Murray, R.M. (1993) Schizophrenia: a neurodevelopmental perspective. Neuropsychopharmacology, 1: 83–91. Pothos, E., Rada, P., Mark, G.P. and Hoebel, B.G. (1991) Dopamine microdialysis in the nucleus accumbens during acute and chronic morphine, naloxone-precipitated withdrawal and clonidine treatment. Brain Res., 566: 348–350. Presti, M.F., Watson, C.J., Kennedy, R.T., Yang, M. and Lewis, M.H. (2004) Behavior-related alterations of striatal neurochemistry in a mouse model of stereotyped movement disorder. Pharmacol. Biochem. Behav., 77: 501–507. Rueter, L.E., Fornal, C.A. and Jacobs, B.L. (1997) A critical review of serotonin brain microdialysis and behavior. Rev. Neurosci., 8: 117–137. Sarter, M. and Bruno, J.P. (1999) Abnormal regulation of corticopetal cholinergic neurons and impaired information processing in neuropsychiatric disorders. Trends Neurosci., 22: 67–74. Sarter, M. and Bruno, J.P. (2002) Animal models in biological psychiatry. In: D’haenen, H., Den Boer, J.A. and Wilner, P. (Eds.), Biological Psychiatry. John Wiley & Sons Ltd, London, UK. Saulskaya, N. and Marsden, C.A. (1995) Extracellular glutamate in the nucleus accumbens during a conditioned emotional response in the rat. Brain Res., 698(6): 114–120. Saunders, R.C. and Kuczenski, R. (1998) Neonatal lesions of the medial temporal lobe disrupt cortical regulation of striatal dopamine. Nature, 393: 169–171. Segal, D.S. and Kuczenski, R. (1997) Behavioural alterations in an escalating dose-binge model of amphetamine psychosis: behavioural and neurochemical characteristics. J. Neurosci., 17: 2551–2566. Segal, D.S., Kuczenski, R., O’Neil, M.L., Melega, W.P. and Cho, A.K. (2005) Prolonged exposure of rats to intravenous methamphetamine: behavioral and neurochemical characterization. Psychopharmacology, 180: 501–512.

434 Shaham, Y., Rajabi, H. and Stewart, J. (1996) Relapse to heroin-seeking in rats under opioid maintenance: the effects of stress, heroin priming, and withdrawal. J. Neurosci., 16: 1957–1963. Sodero, A.O., Valdomero, A., Cuadra, G.R., Ramirez, O.A. and Orsingher, O.A. (2004) Locus coeruleus activity in perinatally protein-deprived rats: effects of ﬂuoxetine administration. Eur. J. Pharmacol., 503: 35–42. Svensson, T.H. (2000) Dysfunctional brain dopamine systems induced by psychotomimetic NMDA-receptor antagonists and the effects of antipsychotic drugs. Brain Res. Rev., 31: 320–329. Swanson, C.J. and Schoepp, D.D. (2003) A role for noradrenergic transmission in the actions of phencyclidine and antipsychotics. Ann. N.Y. Acad. Sci., 1003: 309–317. Tanaka, M. (1999) Emotional stress and characteristics of brain noradrenaline release in the rat. Ind. Health, 37: 143–156. Thorre, K., Chaouloff, F., Sarre, S., Meeusen, R., Ebinger, G. and Michotte, Y. (1997) Differential effects of restraint stress on hippocampal serotonin metabolism and extracellular levels of serotonin in streptozotocin-diabetic rats. Brain Res., 772: 209–216. Van der Elst, M.C., Verheij, M.M., Roubos, E.W., Ellenbroek, B.A., Veening, J.G. and Cools, A.R. (2005) A single exposure to novelty differentially affects the accumbal dopaminergic system of apomorphine-susceptible and apomorphine-unsusceptible rats. Life Sci., 76: 1291–1406. Van der Stelt, H.M., Breuer, M.E., Olivier, B. and Westenberg, H.G.M. (2005) Permanent deﬁcits in serotonergic functioning of olfactory bulbectomized rats: an in vivo microdialysis study. Biol. Psychiatry, 7: 1061–1067. Viggiano, D., Vallone, D. and Sadile, A. (2004) Dysfunctions in dopamine systems and ADHD: evidence from animals and modeling. Neural Plast., 11: 97–114. Wang, T., O’Connor, W.T., Ungerstedt, U. and French, E.D. (1994) N-methyl-D-aspartic acid biphasically regulates

the biochemical and electrophysiological response of A10 dopamine neurons in the VTA: in vivo microdialysis and in vitro electrophysiological studies. Brain Res., 666: 255–262. Weinberger, D.R. (1996) On the plausibility of ‘‘the neurodevelopmental hypothesis’’ of schizophrenia. Neuropsychopharmacology, 14(Suppl 3): 1S–11S. Westerink, B.H.C. (1995) Brain microdialysis and its application for the study of animal behaviour. Behav. Brain Res., 70(2): 103–124. Wheeler, D., Boutelle, M.G. and Fillenz, M. (1995) The role of N-methyl-D-aspartate receptors in the regulation of physiologically released dopamine. Neuroscience, 65: 767–774. Wise, R.A., Newton, P., Leeb, K., Burnette, B., Pocock, D. and Justice, J.B. Jr. (1995) Fluctuations in nucleus accumbens dopamine concentration during intravenous cocaine self-administration in rats. Psychopharmacology (Berl.), 120: 10–20. Yadid, G., Overstreet, D.H. and Zangen, A. (2001) Limbic dopaminergic adaptation to a stressful stimulus in a rat model of depression. Brain Res., 896: 43–47. Yamamoto, T., Yabuuchi, K., Yamaguchi, T. and Nakamichi, M. (2001) Animal models of drug dependence using the drug self-administration method. Nippon Yakurigaku Zasshi, 117: 49–57. Young, A.M.J., Moran, P.M. and Joseph, M.H. (2005) The role of dopamine in conditioning and latent inhibition: what, when, where and how? Neurosci. Biobehav. Rev., 29: 963–976. Youngren, K.D., Inglis, F.M., Pivirotto, P.J., Jedema, H.P., Bradberry, C.W., Goldman-Rakic, P.S., Roth, R.H. and Moghaddam, B. (1999) Clozapine preferentially increases dopamine release in the rhesus monkey prefrontal cortex compared with the caudate nucleus. Neuropsychopharmacology, 20: 403–412.

CHAPTER 5.3

The use of microdialysis for the study of neurological disorders Ilse Smolders and Yvette Michotte Department of Pharmaceutical Chemistry, Vrije Universiteit Brussel, Brussels, Belgium

Abstract: Microdialysis occupies an eminent position within the in vivo techniques used in neurochemistry and neuropharmacology. It is most often applied complementary to other in vivo and in vitro neuromethods such as in vivo voltammetry, monitoring of behavioural alterations, physiological and electrophysiological parameters, immunohistochemistry and blotting techniques. To understand the pathophysiological mechanisms of neurological disease states, microdialysis is a world wide accepted tool to monitor neurobiological and metabolic alterations occurring within animal models of the disorder. Several dialysate compounds subsequently serve as biomarkers of disease. The technique is concomitantly applied in drug discovery to investigate the effects of many diverse therapeutic approaches on these biomarker substances. In the present chapter, an overview is presented of how researchers have introduced microdialysis into their scientiﬁc work in order to unravel mechanisms of disease states in a variety of animals and in order to elucidate innovative drug targets. We will give examples of studies demonstrating the usefulness of microdialysis in research focussing on animal models of epilepsy, focal and global ischaemia, Parkinson’s disease, Huntington’s disease, Alzheimer’s disease, traumatic brain and spinal cord injuries.

parameters. Microdialysis made it possible to collect high-quality reproducible quantitative data in a wide range of laboratory animal species, such as mice, rats, gerbils, hamsters, guinea pigs, rabbits, cats, dogs, sheep, pigs and monkeys. Of course, most of the laboratory work is still performed in rodents. The discovery of transgenic animals and the elegance of the transgenes to clarify functions of the knocked-in or knocked-out gene made experiments with mice recently more popular than ever before. The use of microdialysis in transgenic animals was beyond the scope of this present chapter but will be discussed in other chapters. Microdialysis in freely moving animals presents numerous advantages. The collected samples from the extracellular space from almost every tissue can be analysed for a wide range of substances such as neurotransmitters, neuropeptides, hormones, ions, nucleotides and metabolic markers. Drugs or other

I. Introduction The aim of several research laboratories all over the world is to understand human disease in order to ﬁnd a cure or to obtain an appropriate diseasemodifying effect. Therefore, researchers have designed a variety of animal models, which all mimic certain features of the human disorder. Although none of the animal models can fully imitate the clinical situation, they remain a major drug discovery approach to elucidate innovative drug targets. Microdialysis is one of the most widely used in vivo methods in these animal models because it causes minimal pain and distress, reduces the number of animals used in biomedical research, and even allows chronic assessment of physiological Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

435

DOI: 10.1016/S1569-7339(06)16023-3 Copyright 2007 Elsevier B.V. All rights reserved

436

exogenous substances can be simultaneously perfused through the microdialysis probe in order to study the effects of the exogenously administered compound on the endogenous molecules being monitored. Several microdialysis probes can be implanted in different tissues of one animal; dual probe intracerebral microdialysis is a standard procedure in many laboratories. Microdialysis can be used to answer both pharmacodynamic (PD) and pharmacokinetic (PK) questions and is consequently an important tool in PK/PD modelling. Its application in PK and PK/PD studies is also described in other chapters in this book. Since microdialysis is preferably performed in conscious animals, having ad libitum access to food and water, and having minimal restrictions in their movement, their behavioural alterations and reactions can be quantiﬁed and correlated with the simultaneously monitored neurobiochemical changes. Microdialysis is thus an elegant tool for the neuroscientist and the neuropharmacologist that can be used complementary to other techniques, such as behavioural quantiﬁcation, monitoring of electrophysiological activity, cerebral blood ﬂow or cerebral temperature, and immunocytochemistry and blotting methods, to unravel the scientiﬁc questions of interest. The aim of this chapter is not to give a complete overview of all the studies to date that have been using intracerebral microdialysis in a particular animal model of a neurological disorder, because there are too many interesting studies. In contrast, we would like to give the reader of this chapter an idea for which purposes a researcher could introduce microdialysis into his/her laboratory and we hope to convince him/her that many interesting processes and mechanisms in animal models of neurological disease can be unravelled with this technique.

II. Microdialysis, an indispensable tool for studying animal model of disease II.A. The epilepsies Epilepsy groups essentially excessive or abnormal sudden high-frequency discharges of the brain’s

neurons that disrupt periods of more or less normal electroencephalographic activity and behaviour. A key feature in all epilepsies is the recurrent episodes of hypersynchronous neuronal activities in one or more cortical or phylogenetically related areas of the brain. The existing knowledge about epilepsy is directly or indirectly derived from animal models. The large number of epilepsy models (Pitka¨nen et al., 2006) ﬁnds its origin in two reasons. First, there are many types of epileptic seizures and syndromes to be modelled. Second, important ﬁndings need conﬁrmation in several seizure models since none of them fully imitate clinical epilepsy. To ascertain the presence or absence of epileptic seizures and/or convulsions within the different animal models, monitoring of the electrophysiological activity of the brain’s cortex or one of the regions involved in seizure generation and/or semi-quantifying the characteristic behavioural seizure-related patterns, provide a decisive answer. At least one of these two approaches is routinely combined within the microdialysis laboratories. Intracerebral microdialysis has been widely used to characterise the neurobiological and metabolic changes during the seizures generated in the different models. The technique is concomitantly applied to investigate the effects of anti-epileptic drugs (AEDs) or potential anti-convulsant ligands on these seizure-induced alterations in transmitter or metabolic substances in the dialysates. Although there has often been debated whether high-extracellular glutamate levels are the cause or the consequence of seizure activity or excitotoxicity in neurological disorders (Obrenovitch, 1999), we – among others – have demonstrated seizurerelated increases in glutamate dialysate levels within different epilepsy models. Indeed, signiﬁcantly elevated extracellular glutamate levels were reported in rat hippocampus following maximal electroshock (Rowley et al., 1995), during the recurrent spontaneous seizure phase of the chronic kainate model (Wilson et al., 1996), during focally evoked pilocarpine-induced seizures (Smolders et al., 1997a), during kainate-evoked neonatal seizures precipitating after hypoxia (Yager et al., 2002) and during the seizures evoked by high doses of norﬂoxacin (Smolders et al., 2002a). Enhanced

437

glutamate levels sampled with microdialysis have also been shown in the striatum of rats displaying 4-aminopyridine-induced seizures (Kovacs et al., 2003), in the amygdala of seizure-prone rats during kindling (Shin et al., 2004), and also in the frontal cortex of rats exhibiting pentylenetetrazole-induced seizures (Feng et al., 2005). We have been using the increased hippocampal glutamate and GABA dialysate levels routinely as biomarkers of increased network activity, and demonstrated in several studies that these increases were abolished by both marketed AEDs and drugs in development (Fig. 1; Smolders et al., 1997b, 2002b, 2004; Khan et al., 1999, 2000; Lindekens et al., 2000). Sometimes improvements in analytical techniques, thereby ameliorating the temporal resolution of microdialysis sampling, could reveal previously undetermined increases in extracellular amino acid concentrations concomitant with the seizures (Parrot et al., 2004). Several of the above described microdialysis studies also monitored monoamine concentrations during seizure states (Smolders et al., 1997a; Khan et al., 1999; Kovacs et al., 2003; Shin et al., 2004). Moreover, Kokaia et al. (1989) showed that both focal and generalised hippocampal seizures evoked by electrical kindling stimulation led to a marked increase of transmitter release from noradrenergic but not from serotonergic neurons in the hippocampus. Both basal and pentylenetetrazole-stimulated dopamine release was enhanced in the prefrontal cortex, nucleus accumbens and striatum of pentylenetetrazole-kindled rats (Dazzi et al., 1997). Soman-evoked seizures, which could be blocked by systemic administration of the D1 receptor antagonist SCH23390, elevated striatal dialysate levels in the freely moving guinea pig (Bourne et al., 2001). Lower baseline concentrations of noradrenaline, dopamine and serotonin in amygdala and locus coeruleus of young cats correlated with subsequent increases in duration of focal and generalised after discharges as well as the number of behavioural seizures (Shouse et al., 2001). Jobe et al. (1999) already used microdialysis as one of their tools to show that noradrenergic and/or serotonergic deﬁcits may contribute to predisposition to epilepsies and depression and to propose a linkage between epilepsy and depression.

We recently unravelled more about the involvement of the monoaminergic systems in limbic seizures and demonstrated that, within a certain concentration range, dopamine and serotonin contribute independently to the prevention of hippocampal epileptogenesis via, respectively, D2 and 5-HT1A receptor activation (Clinckers et al., 2004a). In continuation of these data, the selective monoamine re-uptake blockers GBR-12909 and citalopram were shown to possess important anticonvulsant actions against pilocarpine-induced limbic seizures (Clinckers et al., 2004b). Besides the measurement of excitatory/inhibitory amino acids and monoamines in the dialysates obtained from animals exhibiting seizures, other transmitters and neuromodulators have also been determined. Microdialysis in the hippocampus of kindled rats demonstrated that kindling increases adenosine release and metabolism (Aden et al., 2004). Dialysate hippocampal purine levels also increased during bicuculline-, kainic acidand pentylenetetrazole-induced seizure activity (Berman et al., 2000). The NO end products, nitrite and nitrate, were measured by in vivo microdialysis combined with an automated NO end-product analyser and then used as indices of NO synthesis following intrahippocampal kainate injection (Yasuda et al., 2001). This study showed that the NO end-product levels immediately increased after kainate injection and that this elevation preceded the seizure discharges. Dialysis experiments combined with liquid chromatography linked to radioimmunoassay revealed an important activation of the opioid peptide systems by kainic acid-induced status epilepticus but showed reduced hippocampal extracellular opioid peptide levels 28 days after kainic acid administration (Rocha and Maidment, 2003). Also other peptidergic systems are markedly stimulated during seizures as revealed with microdialysis, for example somatostatin in the hippocampus of kindled rats (Marti et al., 2000a) and neuropeptide Y in rat hippocampus during kainic acid-induced seizures (Husum et al., 1998). Microdialysis is also an elegant tool to elucidate mechanisms and neuronal circuits involved in the generation and control of epileptic seizures. In the latter respect, Nail-Boucherie et al. (2005) recently

438 sham

10 mM pilocarpine

pilocarpine (i)

1 0 0.0

0.5

1.0

(ii)

1.5

2.0

2.5

(iii)

3.0

3.5

4.0

4.5

(iv) 5.0

5.5

6.0

6.5

7.0

6.0

6.5

7.0

glutamate GABA

200

**

EC amino acid level (%)

250

**

ECoG recording

**

Time (h)

*

150

500 µV

100 4s

50 0

(i) (ii) (iii) (iv) collection period (20 min)

(iii)

LY377770 pilocarpine (i)

1 0 0.0

0.5

1.0

(ii)

1.5

2.0

2.5

3.0

(iii) 3.5

4.0

4.5

(iv) 5.0

5.5

Time (h)

20

TSSS7

15 10

Pilocarpine LY37770+ Pilocarpine

5 * 0

EC amino acid level (%)

250

glutamate GABA

200 150

500 µV

100 4s

50 0

(i)

(ii)

(iii)

(iv)

(iii)

collection period (20 min)

Fig. 1. Top left panel shows a schematic representation of the experimental setting for focally evoked pilocarpine-induced seizures. A microdialysis probe is stereotaxically implanted into the hippocampus of the anaesthetised rat. The probe is continuously perfused with modiﬁed Ringer’s solution. During the experiment, dialysates are sampled at a ﬂow rate of 2 mL/min. Pilocarpine seizures are evoked by perfusing the hippocampus with 10 mM pilocarpine for 40 min. Simultaneously, an array of electrocorticographic (ECoG) electrodes can register the electrical activity of the cortex. Middle left panel: This photograph shows a rat in a plexiglass microdialysis cage, which has both an implanted microdialysis probe in the hippocampus and an array of six cortical ECoG electrodes. Since the seizures are always evoked in conscious rats, behavioural alterations are routinely scored on a standard seizure severity score scale. Bottom left panel shows a graph of the total seizure severity score (TSSS) for a control group of animals receiving pilocarpine (n ¼ 10; black bar) and a test group of animals receiving the GLUK5 kainate receptor antagonist LY377770 (n ¼ 8; grey bar). Note that the anti-convulsant activity of LY377770 is reﬂected in a signiﬁcant attenuation of the TSSS of the pilocarpine control group. Top right panel shows the mean glutamate and GABA dialysate levels (n ¼ 8) during baseline (i), after sham injection with physiological saline (ii), 20–40 min (iii) and 80–100 min (iv) after ending pilocarpine perfusion. A representative ECoG recording taken during collection 20–40 min following cessation of pilocarpine perfusion (iii) shows characteristic patterns of pilocarpine-induced seizure activity which accompanies signiﬁcant increases in extracellular glutamate and GABA levels in hippocampus. Bottom right panel clearly shows the absence of increases in glutamate and GABA dialysate levels as well as a lack of electrographic signs of seizure activity on the ECoG obtained from animals receiving an i.p. injection of 30 mg/kg LY377770 (n ¼ 6) 30 min before commencing pilocarpine perfusion. Adapted from Smolders et al. (2002b).

performed a microdialysis study in which they unravelled the role of the glutamatergic neurons projecting to the parafascicular nucleus of the thalamus in the control of generalised absence

seizures by the superior colliculus. Disinhibition of these neurons are thus probably involved in the nigral control of generalised epilepsy. For this purpose, they worked with the GAERS strain of

439

rats, which genetically exhibit absences, and measured extracellular GABA and glutamate levels within the parafascicular nucleus. Microdialysis was also applied to unravel hippocampal GABA transporter function in temporal lobe epilepsy (During et al., 1995). The search for innovative AEDs (Smolders et al., 2002b; Feng et al., 2005) remains relevant due to the large amount of refractory patients not responding to the current available medication and since several adverse effects are known for most of the marketed AEDs. Besides the search for new drug targets, microdialysis is often used to study yet unidentiﬁed mechanisms of action of the marketed AEDs. Moreover, the microdialysis technique is applied in studies investigating mechanisms of other anti-convulsant therapies, such as deep brain stimulation. For instance, by measuring the extracellular levels of the serotonin metabolite, 5-hydroxyindoleacetic acid, Ziai et al. (2005) showed that modulation of the serotonergic activity in the anterior thalamic nucleus underlies the efﬁcacy of the anti-convulsant thalamic deep brain stimulation. Anti-convulsant effects of tiagabine and two other non-marketed GABA uptake inhibitors were assessed within the maximal electroshock model in rats and mice, against audiogenic seizures in DBA/2 mice and against pentylentetrazole-induced tonic convulsions in NMRI mice (Dalby, 2000). In vivo microdialysis revealed higher ambient extracellular kynurenic acid levels and enhanced de novo formation of 7-chlorokynurenic acid in the entorhinal cortex and hippocampus in spontaneous epileptic rats after the peripheral administration of their transportable precursors kynurenine and 4-chlorokynurenine, respectively (Wu et al., 2005). Both kynurenic acid and its synthetic derivative, 7-chlorokynurenic acid, are antagonists of the glycine co-agonist site of the N-methyl-D-aspartate (NMDA)-receptor and these data therefore bode well for the use of 4-chlorokynurenine in the treatment of chronic seizure disorders. The causes and mechanisms underlying multidrug resistance in epilepsy are still elusive and may depend on inadequate drug concentrations in crucial brain areas. Therefore, several research groups are investigating the role of multidrug transporters

in seizure mechanisms and whether multidrug transporter blockers could enhance the transport of AEDs across the blood-brain barrier. Microdialysis is in these experiments mostly used to sample the AEDs from relevant brain areas. Rizzi et al. (2002) showed that limbic seizures induced Pglycoprotein expression in rodents and that these P-glycoprotein alterations signiﬁcantly affected the concentrations of phenytoin and carbamazepine in the hippocampus. The fact that certain AEDs, such as levetiracetam, are no substrate or P-glycoprotein or multidrug resistance proteins present at the blood-brain barrier may in part explain its favourable efﬁcacy in pharmacoresistant patients (Potschka et al., 2004). The microdialysis study conducted by Clinckers et al. (2005) investigated the inﬂuence of intrahippocampal perfusion of verapamil, a P-glycoprotein inhibitor, and probenecid, a multidrug resistance protein inhibitor, on the blood-brain barrier passage and anti-convulsant properties of oxcarbazepine in the focal pilocarpine model for limbic seizures. In this study, microdialysis was not only used as a tool to sample the AED but also to administer both the chemoconvulsant and the multidrug transporter blockers, and to simultaneously determine the hippocampal monoamine dialysate levels which are used as pharmacodynamic markers for the anticonvulsant activity. II.B. Ischaemia This type of cerebrovascular disorder is a temporary disruption of the blood supply to a brain area, which eliminates thus also this brain region’s supply of glucose and oxygen, resulting in neuronal cell death. Following the occlusion phase, reperfusion of the deprived brain tissue occurs. This re-circulation phase is believed not only to be beneﬁcial because of the reintroduction of the essential nutrients for neuronal survival but also to represent an important facet of ischaemia-induced cell death due to processes such as increased reactive oxygen species. To validate the different animal models of ischaemia and to quantify the size of the infarct, several approaches are used complementary to microdialysis. Besides laser Doppler ﬂowmetry to measure cerebral blood ﬂow before and

440

during the ischaemic event, and some behavioural tasks mostly performed after the insult, histological examinations to quantify the infarct volume and size are most commonly performed following a microdialysis experiment. Several compounds sampled with microdialysis during ischaemic events serve as biomarkers of the disease and many drug therapies aim at re-establishing these event-induced neurobiochemical alterations. The extracellular striatal excitatory amino acids, GABA, taurine and adenosine levels during middle cerebral artery occlusion (MCAO) by the intraluminal suture technique were determined in conscious freely moving rats and signiﬁcant correlations were found between the efﬂux of all transmitters, neurological score and striatal infarct volume (Melani et al., 1999). MCAO in the anaesthetised rat resulted in a biphasic response in the insular cortex, consisting of a transient increase in the extracellular concentration of glutamate, aspartate and GABA, followed by sustained elevations in glutamate and aspartate, but reduced GABA levels 4 h post-MCAO. This MCAO-induced excitation was completely blocked following the prior intravenous administration of estrogen (Saleh et al., 2004). Microdialysis was applied to map the efﬂux of glutamate and GABA from central core and peripheral zones of focal ischaemia in mouse brain following a standard MCAO protocol (Wang et al., 2001). Calculation of glutamate/ GABA ratios from these mice experiments demonstrated that imbalances in glutamate versus GABA efﬂux may be an initial trigger of excitotoxic brain damage in the core but not the peripheral zones of focal cerebral ischaemia. In a model of malignant infarction in cats by 3 h left MCAO, glutamate determinations in dialysates during ischaemia predicted fatal outcome, as did intracranial pressure and cerebral perfusion pressure measurements during the early reperfusion phase (Toyota et al., 2002). The AED zonisamide was shown to be effective in reducing neuronal damage by a mechanism involving decreased ischaemiainduced extracellular glutamate accumulation in the gerbil hippocampus and interruption of excitotoxic pathways (Owen et al., 1997). We investigated the striatal release of dopamine, glutamate and GABA in the endothelin-1 rat model for focal

transient cerebral ischaemia and showed that these three dialysate substances largely increased following ischaemia (Bogaert et al., 2000). Interestingly, the dopamine release preceded the elevations in striatal glutamate levels. Kainate receptors of the GluR5 (GLUK5) subtype play a central role in ischaemic brain damage following global and focal cerebral ischaemia, and microdialysis elucidated that the GluR5 antagonist LY377770 attenuated ischaemia-induced increases in extracellular levels of glutamate but not of dopamine (O’Neill et al., 2000). Inhibition of extracellular glutamate release in the cortex of rats immediately after MCAO probably contributes to the pronounced neuroprotective efﬁcacy of the 5-HT1A receptor agonist BAY x 3702 against ischaemic brain damage (Mauler et al., 2001). Besides in the popular occlusion models to induce global or focal ischaemia and the focal ischaemia induced by abluminal application of vasoconstrictive peptides such as endothelin-1, microdialysis has also revealed amino acid changes in trombo-embolic models, such a the photochemical method using the Rose Bengal dye as thrombogenic agent (Montalbetti et al., 1995). From the above paragraph, it is clear that determination of classical neurotransmitters in dialysates received most attention within the models of ischaemic brain damage. However, several other substances have been monitored and subsequently implied in the pathogenesis of ischaemic disease. Elevated generation of reactive oxygen species, at least in part due to dopamine autooxidation, has been demonstrated during ischaemia and reperfusion. Hydroxyl radical formation, estimated via microdialysis, is larger in striatal core than in penumbra in the rat MCAO model of ischaemic stroke (Liu et al., 2003). The sodium channel blocker AM-36 profoundly reduced reactive oxygen species formation and dopamine release in the striatum of conscious rats after endothelin-1-induced ischaemia, as was shown with microdialysis (Callaway et al., 2003). Inhibition of NO production is part of the neuroprotective effects exerted by several compounds tested in ischaemia models. Indeed, the calmodulin antagonist DY-9760e abolished the increased nitrite and nitrate dialysate levels in the hippocampal CA1 region of conscious gerbils after 10-min forebrain ischaemia induced

441

by bilateral common carotid artery occlusion (Hashiguchi et al., 2003). In vivo striatal NO production, estimated via microdialysis by quantiﬁcation of citrulline recovery after labelled arginine infusion, was attenuated by the kappa-opioid receptor agonist BRL 52537 following intraluminal MCAO in rats (Goyagi et al., 2003). Interest is growing in the role of adenosine triphosphate (ATP) on P2 receptors during hypoxic/ischaemic events. Melani et al. (2005) recently demonstrated that basal striatal ATP outﬂow is stimulated during ischaemia evoked by intraluminal MCAO and conﬁrmed that ATP may be an important mediator in brain ischaemia. The excitatory actions of corticotropin-releasing hormone (CRH) in the brain and the neuroprotective effects of CRH antagonists in models of ischaemia suggest a role for this peptide in the cascade of events leading to cellular damage. Microdialysis was used to assess time-dependent changes in CRH concentrations in 10 brain regions at three post-ischaemic intervals following global ischaemia as well as to study the impact of a pretreatment with a NMDA receptor antagonist on the CRH levels (Khan et al., 2004). An on-line microdialysis-graphite furnace atomic absorption spectrometry assay permitted determination of alterations in magnesium and zinc levels in the cortex of gerbils subjected to focal cerebral ischaemia by MCAO (Yang et al., 2004). Although many of the neuroprotective drug approaches seem promising within the animal models of ischaemia, none of the drugs has reached the clinic to date. Therefore, another therapeutic approach that gained great interest to reduce ischaemia-induced neuronal damage is hypothermia. Whole body hypothermia or local hypothermia, for instance with a helmet device (Hachimi-Idrissi et al., 2001), demonstrated signiﬁcantly improved survival rates and neurological outcomes within the clinical setting. Based on these facts, many experiments tried to unravel the involved mechanisms of action of brain cooling within animals. We showed for instance that post-ischaemic mild hypothermia signiﬁcantly attenuated the endothelin-1-induced glutamate release in the striatum but not in the penumbral cortical region; in the penumbra it however inhibited apoptosis (Fig. 2; Van Hemelrijck et al., 2003). In a subsequent study,

microdialysis was applied to measure several amino acids, the citrulline/arginine ratio and hydroxyl radical formation via 2,3 dihydroxybenzoic acid detection as part of the in vivo salicylate trapping method (Van Hemelrijck et al., 2005). These data demonstrated that, within the focal endothelin-1 model, post-ischaemic mild hypothermia inhibited apoptosis in the penumbral region by reducing neuronal NO synthase activity and thereby preventing endothelin-1-induced hydroxyl radical formation. Post-ischaemic mild hypothermia was also shown to reduce ischaemia-induced striatal glutamate and dopamine dialysate levels and astroglial cell proliferation during reperfusion after asphyxial cardiac arrest in rats (Hachimi-Idrissi et al., 2004). Brain cooling also proved its cerebroprotective effects in a model of three main arteries occlusion in the dog by an intravascular perfusion of cooled crystalloid solution using an extracorporeal cooling-ﬁltration system (Furuse et al., 2003). II.C. Parkinson’s disease Parkinson’s disease is a neurodegenerative disorder characterised by a progressive loss of the dopaminergic nigrostriatal neurons and the presence of cytoplasmatic occlusions of a-synuclein. Lesioning of the nigrostriatal tract in rodents with 6-hydroxydopamine (6-OHDA) and the administration of the neurotoxin 1-methyl-4-phenyl1,2,3,6-tetrahydropyridine (MPTP) to mice and primates remain the most popular animal models. Several behavioural methods to monitor motor functioning in the animals are routinely used in combination with microdialysis, such as amphetamine-induced rotation (Yuan et al., 2005) or performance on an accelerating rod (Bergquist et al., 2003). To quantify the size of the lesion, researchers commonly use microdialysis to sample the extracellular striatal levels of dopamine but also histological veriﬁcation and determination of whole tissue dopamine content. Microdialysis is a popular technique to monitor alterations in basal ganglia circuitry within animal models of Parkinson’s disease. Besides the expected decrease in striatal dopamine release following denervation of the nigrostriatal pathway by

442 Glutamate dialysate levels obtained from rat striatum 10.0

*

Glu (µM)

7.5

Hypothermia

*

5.0

$

* 2.5

$

$ $

0.0

-20 Bas

0

20

40

60

$

$

$

$

$

80 100 120 140 160 180 200

Time (min)

Et-1

Glutamate dialysate levels obtained from parietal cortex

7.5

2.00

Normothermia Hypothermia

1.75 1.50

5.0 $

2.5

Glu (µM)

fragmented nucleï/mm²

normothermia hypothermia sham

Hypothermia

normothermia hypothermia sham

1.25 1.00 0.75 0.50 0.25 0.00

0.0

striatum

-20 Bas 0

cortex

20

40

60

80 100 120 140 160 180 200

Time (min) Et-1

Fig. 2. Top left panel shows a microdialysis experiment performed on a sevoﬂurane (1.5%) anaesthetised rat having two microdialysis guide cannulas implanted, one adjacent to the middle cerebral artery (MCA) and one in the ipsilateral striatum. Focal ischaemia is evoked by infusing endothelin-1 (Et-1; 500 pmol/6 mL) through the guide cannula near the MCA. Dialysate levels of glutamate are sampled from a microdialysis probe inserted into the guide cannula of the striatum. During normothermic experiments, brain and rectal temperature are maintained at 37.070.51C throughout the experiments with a thermostatically controlled heating pad. Spraying alcohol onto the rat and cooling it to mild hypothermic target temperature (34.070.21C) with a fan achieved a total body cooling within 10 min.Top right panel shows a graph of the glutamate dialysate levels obtained from the rat striatum before and following Et-1 injection near the MCA under normothermic (n ¼ 6) and hypothermic (n ¼ 6) conditions. It is clear that post-ischaemic mild hypothermia attenuated striatal glutamate dialysate levels signiﬁcantly over time in comparison with the normothermic conditions. This means that mild hypothermia has a signiﬁcant effect on the glutamate release in the core of the infarct. Bottom right panel: Similar experiments but in which glutamate was sampled from the rat parietal cortex showed no differences in cortical glutamate levels over time between the normothermic (n ¼ 4) and the hypothermic (n ¼ 4) group. Bottom left panel: The neuroprotective effect of mild hypothermia in the cortex, which is the so-called penumbra of the infarct, is thus not associated with a reduction in cortical glutamate release but is caused by inhibition of apoptosis. Indeed, a count of fragmented nuclei in both striatum (n ¼ 4) and cortex (n ¼ 4) to determine the degree of apoptosis revealed that mild hypothermia only signiﬁcantly affects apoptotic neuronal cell death in the penumbra. Adapted from Van Hemelrijck et al. (2003).

6-OHDA, extracellular dopamine levels did not decrease in the substantia nigra (Jonkers et al., 2002). A combination of compensatory changes of the remaining neurons and dopamine originating from the ventral tegmental area probably maintain extracellular dopamine in the substantia nigra at near-normal levels following the lesion (Fig. 3;

Sarre et al., 2004). Within the animal models of Parkinson’s disease, a lot of attention has been paid to changes in amino acid dialysate concentrations. The loss of dopaminergic neurons of the substantia nigra is indeed associated with an imbalance in the activity of the so-called ‘direct’ and ‘indirect’ pathways of information ﬂow through

443 SN

VTA

SN

VTA

a

MTN b

c

Baseline dopamine dialysate levels in the substantia nigra

Baseline dopamine dialysate levels in the striatum

2.0 1.5 1.0

*

0.5 0.0

Intact

3 weeks PL

0.5

Dopamine dialysate concentrations (nM)

Dopamine dialysate concentrations (nM)

2.5

ND 5 weeks PL

0.4 0.3 0.2 0.1 0.0

Intact

3 weeks PL

5 weeks PL

Fig. 3. Top panel shows micrographs of tyrosine hydroxylase (TH) immunostainings of the ventral tegmental area (VTA) and the substantia nigra (SN) of experimental animals at the level of the medial terminal nucleus (MTN). Three experimental group of rats are shown: (a) intact rats, (b) rats killed 3 weeks after a unilateral 6-hydroxydopamine (6-OHDA) lesion of the medial forebrain bundle and (c) rats killed 5 weeks after a unilateral 6-OHDA lesion of the medial forebrain bundle. TH cell counts are a marker of the dopamine phenotype. From these micrographs, it is clear that, in the SN, there is a signiﬁcant decrease in the amount of TH-immunoreactive cells on the lesioned side as a function of time. Approximately 1% of the neurons remained 5 weeks after the 6-OHDA lesion. In the VTA, we noticed also a signiﬁcant decrease in TH-immunoreactive cells on the lesioned side as a function of time but the decrease was less pronounced that the one observed in the SN. Approximately 25% of the neurons remained 5 weeks after the 6-OHDA lesion of the medial forebrain bundle. Bottom left panel: As expected following a denervation of the nigrostriatal pathway, the baseline dopamine dialysate levels of the striatum of a 6OHDA lesioned rat were signiﬁcantly lower 3 weeks post-lesioning (PL) (n ¼ 23) than the levels measured in the intact rats (n ¼ 35). We did not determine (ND) levels of dopamine in the striatum of lesioned rats 5 weeks PL. Bottom right panel: Lesioning of the nigrostriatal pathway had no effects on the extracellular levels of dopamine in dialysates obtained from the SN. Indeed, there were no differences in baseline nigral dopamine dialysate concentrations in intact rats (n ¼ 15), 3 weeks PL (n ¼ 14) or 5 weeks PL (n ¼ 15). Compensatory alterations of the remaining nigral neurons and dopamine originating from the neurons of the VTA presumably maintain the extracellular dopamine levels in the SN at near-normal levels following a 6-OHDA lesion. Adapted from Jonkers et al. (2000) and Sarre et al. (2004).

the basal ganglia. Galefﬁ et al. (2003) studied whether the imbalance is reﬂected in changes in the release of GABA, aspartate and glutamate in the striatopallidal ‘indirect’ pathway and in the

striatonigral ‘direct’ pathway using dual probe microdialysis in freely moving rats. They found that 6-OHDA lesioning results in an elevation of the basal release of GABA in the striatopallidal

444

pathway and a reduction in the evoked release of GABA in the striatonigral pathway. Other microdialysis studies from the laboratory of L. Della Corte also contributed to the understanding of the role of the amino acids in basal ganglia circuitry (Bianchi et al., 1994, 1998). The group of M. Morari has recently published several microdialysis studies, performed in intact and 6-OHDA lesioned rats, pointing out the role of different receptor types involved in the striatal control of glutamate release in the substantia nigra pars reticulata and in the nigral control of glutamate release in the striatum (Marti et al., 2000b, 2002, 2005). Different changes in extracellular striatal glutamate levels have also been reported following acute or subchronic administration of the neurotoxin MPTP to mice (Robinson et al., 2003). Using nigral microdialysis and low-dose pilocarpineinduced tremulous jaw movements, the GABAergic tone in the substantia nigra was elucidated to play a role in tremor, one of the parkinsonian symptoms (Ishiwari et al., 2004). A few studies also focussed on the modulation of striatal acetylcholine release in experimental models of parkinsonism (Sato et al., 1994; Kurokawa et al., 1996). The therapeutic potential of many drugs have subsequently been tested in the various models as well as their effects on the affected neurotransmitter systems following lesioning. The L-DOPAinduced increase in glutamate in the striatum of 6-OHDA-lesioned rats was suppressed by MK801 (Jonkers et al., 2002). Recently, interest in serotonergic strategies to manage parkinsonian symptoms appeared. Several microdialysis studies unravelled the involvement of different 5-HT receptor subtypes in the control of dopamine release in the rat striatum (De Deurwaerdere et al., 1997; De Deurwaerdere and Spampinato, 1999; Lucas and Spampinato, 2000). A 6-OHDA-lesion resulted in elevated baseline striatal serotonin dialysate levels but abolished apomorphine- and amphetamine-induced serotonin increases (Balcioglu et al., 2003). Systemic administration of a 5-HT1A receptor agonist was shown to reduce glutamate neurotransmission in the dopaminedenervated striatum following 6-OHDA (Mignon and Wolf, 2005). In order to unravel some of the mechanisms of action implied in ameliorating

parkinsonian symptoms by adenosine A2A receptor antagonists, A2A receptor-mediated modulation of the release of GABA and glutamate in the substantia nigra pars reticulata was studied using in vivo microdialysis in 6-OHDA-lesioned rats (Ochi et al., 2004). Also the AED zonisamide appears to possess anti-parkinsonian properties, as conﬁrmed by a microdialysis study measuring dopamine and its major metabolites following combined L-DOPA-carbidopa-zonisamide administration to 6-OHDA-lesioned rats (Gluck et al., 2004). Another therapeutic approach, deep brain stimulation of the subthalamic nucleus, alleviates Parkinson’s disease symptoms in humans. Microdialysis in 6-OHDA-lesioned rats sustained that deep brain stimulation of the subthalamic nucleus disinhibits substantia nigra compact neurons via inhibition of GABA-ergic substantia nigra reticulata neurons (Meissner et al., 2002). Nowadays more and more investigations aim at unravelling the potential therapeutic role of neuropeptides and growth factors in Parkinson’s disease. GABA-opioid interactions have been studied in the globus pallidus of intact and MPTP-treated parkinsonian cats from which basal and potassium-evoked GABA dialysate levels were determined (Schroeder and Schneider, 2002). Within most of these studies, microdialysis is routinely applied not only to monitor changes in striatal dopamine levels (Tuncel et al., 2005) but also to follow the in vivo processing of the peptides themselves (Klintenberg and Andren, 2005). Induction and release of brain-derived neurotrophic factor (BDNF) by activated glial cells induced by group II metabotropic glutamate receptor agonist perfusion may account for its protective action against MPP+-induced dopaminergic terminal degeneration in the rat, as demonstrated by both immunohistochemical techniques and microdialysis (Matarredona et al., 2001). Among the various promising growth factors, glial cell-line derived neurotrophic factor (GDNF) currently receives a lot of attention. It seems that GDNF is signiﬁcantly more effective than BDNF for both correcting behavioural deﬁcits and protecting nigrostriatal dopaminergic neurons following intrastriatal injections of 6-OHDA (Sun et al., 2005). Microdialysis was applied to study the effect of

445

intrastriatal GDNF injection on the nigrostriatal dopaminergic function in the rat (Xu and Dluzen, 2000). Intraventricular injection of GDNF was proved to be effective in enhancing dopamine release in the basal ganglia structures of both MPTPlesioned and aged rhesus monkeys and showed concomitant amelioration of their motor behaviour (Gerhardt et al., 1999; Grondin et al., 2003). Microdialysis was also part of the methodological approaches used to study the effect of GDNF gene therapy within the 6-OHDA rat model (Gerin, 2002). Concerning gene therapy in Parkinson’s disease, the microdialysis technique has provided the scientiﬁc community with a quantitative tool to determine the optimal set of genes for effective therapy. It enabled to compare the effects of various genes involved in dopamine synthesis and processing, such as the tyrosine hydroxylase gene, the GTP cyclohydrolase gene and the aromatic L-amino acid decarboxylase gene, necessary for optimal dopamine replacement in rat models (for review see Kang et al., 2001). Within these studies microdialysis is commonly applied to sample L-DOPA, dopamine and its metabolites from the denervated striatum of the rats. II.D. Alzheimer’s disease Alzheimer’s disease is the most common cause of progressive cognitive decline in aged humans. Animal models of this disease have replicated several of the pathophysiological brain alterations but have of course mostly focussed on the loss of cholinergic innervation of hippocampus, cortex and other limbic regions, leaving other characteristic features, such as the b-amyloid plaques and Tau protein tangles, of this neurodegenerative disorder untouched. Nowadays, as in many ﬁelds, a lot is expected from the genetically engineered mouse models that mimic at least some of these key pathological protein changes in Alzheimer’s disease. These transgenes are expected to provide tools that will rapidly facilitate drug development. In contrast to quantifying seizure severity, infarct volume or size of a 6-OHDA lesion, quantifying cognitive decline and loss of memory function within animals is more difﬁcult and time consuming,

although a variety of elegant behavioural tasks and approaches have been developed. The clinical evidence that restoring cholinergic functioning enhances learning and memory in Alzheimer disease patients supports the important role of acetylcholine in cognitive functions. Microdialysis has thus been widely applied not only to measure acetylcholine in brain regions linked to cognitive processing and memory function, such as hippocampus, amygdala, several cerebral cortical and subcortical areas but also to striatum (Gold, 2003; Pepeu and Giovannini, 2004; Nelson et al., 2005). It has even been hypothesised that the magnitude of acetylcholine release in different neuronal systems regulates the relative contributions of these systems to learning (Gold, 2003). While studying the relationships between acetylcholine release and cognitive processing during different behavioural tasks, several methodological and conceptual complexities pose challenges to the interpretation of the experimental results, as was reviewed by Bruno et al. (1999) and Pepeu and Giovannini (2004). Indeed, several behaviours displayed by the animals in the experimental tasks, not required for learning and memory per se, may affect the cholinergic system. Investigations have indeed shown that concepts such as attention, novelty, motivation and motor activity also involve the cholinergic system (Bruno et al., 1999; Sarter et al., 2003; Pepeu and Giovannini, 2004). More information on this speciﬁc topic can be found in other chapters in this book. Computational modelling by the group of M.E. Hasselmo sustains that high acetylcholine levels are needed for attention and encoding of information, but that low acetylcholine concentrations allow consolidation and retrieval (Hasselmo and McGaughy, 2004). Recent microdialysis work by Elvander et al. (2004) also showed that it is too simplistic to think that an increase in hippocampal acetylcholine dialysate concentrations is always facilitatory for learning and memory. They showed that there exists a limited range of acetylcholine release for optimal hippocampal functioning and that also the septohippocampal GABAergic pathway plays an important modulatory role on both the neurotransmitter and the cognitive effects. Indeed, retrograde amnesia is a

446

well-known feature of the clinically used benzodiazepines acting at the GABAA receptors. The role of GABA in cognition was also highlighted by experimental studies with benzodiazepine inverse agonists, which were found to augment cortical acetylcholine efﬂux through interaction with cognition-associated activation of the forebrain cholinergic neurons (Sarter and Bruno, 1997). The role of glutamate in cognitive processing is clear but complex. Cognitive-impairing effects of NMDA receptor antagonists have been extensively described, although the partial NMDA antagonist memantine clinically offers signiﬁcant beneﬁts in advanced Alzheimer’s disease patients. It seems that the physiological activation of NMDA receptors may not be compromised but that glutamatergic overstimulation and excitotoxicity adds to cognitive impairment and the progression of the disease. The activation of the regulated kinase signal transduction pathway during acquisition of the step-down inhibitory avoidance response in the rat showed increases in cortical and hippocampal acetylcholine but not glutamate dialysate levels, indicating that cholinergic neurons projecting to the medial prefrontal cortex and ventral hippocampus were activated during acquisition of the task (Giovannini et al., 2005). Microdialysis in the rat showed that activation of the limbic dopamine reward pathway is not sufﬁcient for the phencyclidine psychotomimetic and cognitiveimpairing effects but glutamatergic hyperstimulation is also involved (Adams and Moghaddam, 1998). Cognitive processing may also depend on speciﬁc and distinct functions of the cortical noradrenergic and serotonergic systems (Dalley et al., 2001; Russell and Dias, 2002). Galanin, which is co-localised with acetylcholine in the septohippocampal cholinergic pathway, is also involved in cognition and has powerful modulatory effects on hippocampal neurotransmission (Elvander et al., 2004). Aged rats have often been used to model cognitive decline in the elderly. Baseline extracellular acetylcholine levels in the cerebral cortex and hippocampus were signiﬁcantly lower in old than in young rats (Scali et al., 1997). Aged rats exhibited a marked attenuation of the potassium (100 mM)stimulated acetylcholine efﬂux relative to young

adult rats (Herzog et al., 2003). Moreover, old rats could not discriminate between novel and familiar objects within the well-established object recognition test (Scali et al., 1997). Scopolamine-induced disruption of spatial learning and memory is a frequently used animal model for the study of cognitive decline. A combination of both microdialysis and behavioural experiments, such as performance in a Morris water maze or monitoring passive avoidance response, has become an elegant tool for characterising this model (Tottori et al., 2002; Elvander et al., 2004). Infusion of neurotoxic amyloid b fragments induces cholinergic and glutamatergic dysfunction and cognitive deﬁcits, such as impairment of spatial memory in the Morris water maze or the radial eight-arm maze task (Tran et al., 2002). To assess the impairment of cholinergic neurotransmission as one of the mechanisms of neurotoxicity within this model, microdialysis revealed that amyloid b attenuated nicotine-stimulated acetylcholine release (Tran et al., 2002). In vivo amyloid b infusion by way of microdialysis in the rat magnocellular nucleus basalis revealed peak extracellular concentrations of excitatory amino acid neurotransmitters within 20–30 min and triggered an excitotoxic cascade explaining at least part of amyloid b’s neurotoxicity (Harkany et al., 2000). A neurotoxic lesion of the basal forebrain, achieved by bilateral injections of ibotenic acid, is another frequently used animal model for Alzheimer’s disease (Abe et al., 1998). Bilateral infusion of the selective cholinotoxin 192 IgG-saporin into the frontoparietal cortex of rats is used as a partial deaffentiation model of cortical cholinergic inputs, resulting in a signiﬁcant reduction of basal cortical acetylcholine efﬂux, and is combined with validated tasks for measurement of sustained attention such as the darkness/cereal stimulus (Fadel et al., 1996). Several existing and potential anti-Alzheimer drugs have subsequently been tested within these different animal models. Very often these studies have also been set up to investigate whether the anti-Alzheimer drugs or compounds in development alter neurotransmitter overﬂow in the brain of conscious animals. A microdialysis study investigating the effects of chronic treatment of aging rats with the cholinesterase inhibitors metrifonate,

447

rivastigmine and donepezil demonstrated longlasting increases in cortical and hippocampal acetylcholine levels and revealed marked differences between the three drugs (Scali et al., 2002). The same group showed that in old rats, metrifonate brought about 85% inhibition of cholinesterase activity in the cortex and hippocampus, a fourfold increase in extracellular acetylcholine levels in the cortex only, and restored object recognition (Scali et al., 1997). Cholinomimetics were also shown to increase extracellular striatal glutamate levels via increased cortical pyramidal neuron function (Dijk et al., 1995). Blockade of presynaptic M2 muscarinic autoreceptors increased acetylcholine release from the striatum of conscious rats and was also active in the young rat passive avoidance response paradigm of cognition (Lachowicz et al., 2001). A dual inhibitor of acetylcholinesterase and the serotonin transporter simultaneously elevated the extracellular levels of acetylcholine and serotonin in the rat hippocampus as conﬁrmed by microdialysis and ameliorated the spatial memory deﬁcits in the two-platform task of a water maze in aged rats (Abe et al., 2003). The selective serotonin 5-HT1A receptor antagonist lecozotan signiﬁcantly potentiated the potassium chloride-stimulated release of glutamate and acetylcholine in the dentate gyrus of the rat hippocampus (Schechter et al., 2005). The same study showed that learning deﬁcits in marmosets induced by the glutamatergic antagonist MK-801 and by speciﬁc cholinergic lesions of the hippocampus were reversed by lecozotan. A combined sigma/5-HT1A receptor agonist improved scopolamine-induced learning impairments in the passive-avoidance task and memory impairment in the Morris water maze and increased acetylcholine release in the dorsal hippocampus of freely moving rats following both oral and local delivery (Tottori et al., 2004). Benzodiazepine inverse agonists alleviated the partial 192 IgG-saporin-induced impairment in sustained attention and enhanced activated cortical acetylcholine efﬂux sampled with microdialysis (Fadel et al., 1996). Benzodiazepine inverse agonists also signiﬁcantly ameliorated the basal forebrain-lesion-induced impairment of spatial memory in the water maze task and signiﬁcantly increased acetylcholine release in the frontoparietal cortex of

basal forebrain-lesioned rats (Abe et al., 1998). Finally, the in vivo microdialysis technique revealed that a positive AMPA receptor modulator prevented the disrupting effect of scopolamine on passive avoidance acquisition and induced a longlasting acetylcholine release in the hippocampus of aged rats (Rosi et al., 2004). II.E. Huntington’s disease Huntington’s disease is a genetic, progressive, fatal, neurodegenerative disorder characterised by both motor and cognitive deterioration. The rat models of intrastriatal injection of the excitotoxin quinolinic acid and chronic administration of the mitochondrial toxin 3-nitropropionic acid via osmotic minipumps, lead to striatal medium sized spiny GABAergic neuronal cell loss, and are accepted as ‘pathogenetic’ models of the disease. Although both quinolinic acid and 3-nitropropionic acid also destroy the GABAergic projection neurons after direct perfusion into the rat striatum, different alterations in striatal extracellular GABA concentrations within the ﬁrst 90 min reﬂected different toxic mechanisms for these two neurotoxins (Reynolds et al., 1997). Several studies point to beneﬁcial effects of adenosine A1 receptor agonists and adenosine A2A receptor antagonists in animal models of Huntington’s disease (Blum et al., 2003a). Changes in extracellular glutamate and adenosine levels have been studied within the quinolinic acid model as well as the modulatory actions on these transmitters by adenosine A2A antagonists (Gianfriddo et al., 2003). Microdialysis was one of the methods applied to investigate the differential effects of adenosine A2A receptors on the striatal neurodegeneration induced by 3nitropropionic acid (Blum et al., 2003b). II.F. Traumatic brain injury/spinal cord injury Dual-function microdialysis/electrophysiology probes were placed in rats following experimental ﬂuid percussion brain injuries, and even in a series of severely head-injured human patients, as a feasibility study to monitor concomitant changes in electrical activity, metabolic changes and extracellular glutamate (Alves et al., 2005). A combined

448

electroencephalographic-microdialysis study following controlled cortical impact injury showed that, despite unchanged isoﬂurane concentrations, spontaneous increases in neuronal activity coincided with elevated extracellular glutamate levels sampled in the pericontusional cortex and vice versa, implying that neuronal activity needs to be considered for the interpretation of glutamate dialysate levels (Sakowitz et al., 2002). Multiparametric neuromonitoring following controlled cortical impact has also been developed in pigs with microdialysis as one of the on-line monitoring techniques (Alessandri et al., 2003). Microdialysis work directly demonstrated in vivo that, following spinal cord injury, glutamate reaches toxic concentrations that kill the oligodendrocytes, and that AMPA/kainate receptors mediate this excitotoxic glutamate effect (Xu et al., 2004). Another study, measuring excitatory amino acids and GABA with microdialysis ﬁbres inserted 0.5 mm caudal from the edge of the impact region, showed that also metabotropic glutamate receptors play an important role in excitatory amino acid toxicity following spinal cord injury (Mills et al., 2001). II.G. Others In vivo microdialysis has also offered insights into the pathology of pyrithiamine-induced thiamine deﬁciency encephalopathy used as rat model of Wernicke’s encephalopathy (Todd and Butterworth, 2001). An increase in histamine dialysate concentrations and the number of granulocytes was observed in lateral and medial thalamus during pyrithiamine treatment and was associated with perivascular edema in these rat brain regions, pointing out the possible involvement of histamine in the vascular changes that precede the onset of the thalamic lesions within this animal model (McRee et al., 2000). Experimental allergic encephalomyelitis is an animal model of multiple sclerosis and induces elevated cerebrospinal glutamate dialysate levels, which were attenuated by delta-9-tetrahydrocannabinol (Fujiwara and Egashira, 2004). Clindamycin was shown to be neuroprotective and lowered extracellular hippocampal hydroxyl radicals and glutamate levels in a rabbit model of pneumococcal meningitis (Bottcher et al., 2004).

III. Conclusion The above-described studies in animal models of various neurological disorders demonstrate the usefulness of microdialysis to unravel neurochemical and neuropharmacological mechanisms and processes in the brains of several animal species. Microdialysis is deﬁnitely an elegant in vivo neuromethod that can be used complementary to other established techniques in neuroscience to provide decisive answers to your scientiﬁc questions of interest.

Acknowledgements Ilse Smolders is a postdoctoral fellow of the Fund for Scientiﬁc Research Flanders (FWO-Vlaanderen), Belgium. We thank the FWO-Vlaanderen and the Onderzoeksraad van de Vrije Universiteit Brussel for ﬁnancial support.

References Abe, Y., Aoyagi, A., Hara, T., Abe, K., Yamazaki, R., Kumagae, Y., Naruto, S., Koyama, K., Marumoto, S., Tago, K., Toda, N., Takami, K., Yamada, N., Ori, M., Kogen, H. and Kaneko, T. (2003) Pharmacological characterization of RS-1259, an orally active dual inhibitor of acetylcholinesterase and serotonin transporter, in rodents: possible treatment of Alzheimer’s disease. J. Pharmacol. Sci., 93: 95–105. Abe, K., Takeyama, C. and Yoshimura, K. (1998) Effects of S-8510, a novel benzodiazepine receptor partial inverse agonist, on basal forebrain lesioning-induced dysfunction in rats. Eur. J. Pharmacol., 347: 145–152. Adams, B. and Moghaddam, B. (1998) Corticolimbic dopamine neurotransmission is temporally dissociated from the cognitive and locomotor effects of phencyclidine. J. Neurosci., 18: 5545–5554. Aden, U., O’Connor, W.T. and Berman, R.F. (2004) Changes in purine levels and adenosine receptors in kindled seizures in the rat. Neuroreport, 15: 1585–1589. Alessandri, B., Heimann, A., Filippi, R., Kopacz, L. and Kempski, O. (2003) Moderate controlled cortical contusion in pigs: effects on multi-parametric neuromonitoring and clinical relevance. J. Neurotrauma., 20: 1293–1305. Alves, O.L., Bullock, R., Clausen, T., Reinert, M. and Reeves, T.M. (2005) Concurrent monitoring of cerebral electrophysiology and metabolism after traumatic brain injury: an experimental and clinical study. J. Neurotrama., 22: 733–749.

449 Balcioglu, A., Zhang, K. and Tarazi, F.I. (2003) Dopamine depletion abolishes apomorphine- and amphetamine-induced increases in extracellular serotonin levels in the striatum of conscious rats: a microdialysis study. Neuroscience, 119: 1054–1053. Bergquist, F., Shahabi, H.N. and Nissbrandt, H. (2003) Somatodendritic dopamine release in rat substantia nigra inﬂuences motor performance on the accelerating rod. Brain Res., 973: 81–91. Berman, R.F., Fredholm, B.B., Aden, U. and O’Connor, W.T. (2000) Evidence for increased dorsal hippocampal adenosine release and metabolism during pharmacologically induced seizures in rats. Brain Res., 872: 44–53. Bianchi, L., Colvicchi, M.A., Bolam, J.P. and Della Corte, L. (1998) The release of amino acids from rat neostriatum and substantia nigra in vivo: a dual microdialysis probe analysis. Neuroscience, 87: 171–180. Bianchi, L., Sharp, T., Bolam, J.P. and Della Corte, L. (1994) The effect of kainic acid on the release of GABA in rat neostriatum and substantia nigra. Neuroreport, 5: 1233–1236. Blum, D., Galas, M.C., Pintor, A., Brouillet, E., Ledent, C., Muller, C.E., Bantubungi, K., Galluzzo, M., Gall, D., Cuvelier, L., Rolland, A.S., Popoli, P. and Schiffmann, S.N. (2003b) A dual role of adenosine A2A receptors in 3-nitropropionic acid-induced striatal lesions: implications for the neuroprotective potential of A2A antagonists. J. Neurosci., 23: 5361–5369. Blum, D., Hourez, R., Galas, M.C., Popoli, P. and Schiffmann, S.N. (2003a) Adenosine receptors and Huntington’s disease: implications for pathogenesis and therapeutics. Lancet Neurol., 2: 366–374. Bogaert, L., Scheller, D., Moonen, J., Sarre, S., Smolders, I., Ebinger, G. and Michotte, Y. (2000) Neurochemical changes and laser Doppler ﬂowmetry in the endothelin-1 rat model for focal cerebral ischemia. Brain Res., 887: 266–275. Bottcher, T., Ren, H., Goiny, M., Gerber, J., Lykkesfeldt, J., Kuhnt, U., Lotz, M., Bunkowski, S., Werner, C., Schau, I., Spreer, A., Christen, S. and Nau, R. (2004) Clindamycin is neuroprotective in experimental Streptococcus pneumoniae meningitis compared with ceftriaxone. J. Neurochem., 91: 1450–1460. Bourne, J.A., Fosbraey, P. and Halliday, J. (2001) SCH 23390 affords protection against soman-evoked seizures in the freely moving guinea-pig: a concomitant neurochemical, electrophysiological and behavioural study. Neuropharmacology, 40: 279–288. Bruno, J.P., Sarter, M., Moore Arnold, H. and Himmelheber, A.M. (1999) In vivo neurochemical correlates of cognitive processes: methodological and conceptual challenges. Rev. Neurosci., 10: 25–48. Callaway, J.K., Lawrence, A.J. and Jarrott, B. (2003) AM-36, a novel neuroprotective agent, profoundly reduces reactive oxygen species formation and dopamine release in the striatum of conscious rats after endothelin-1-induced middle cerebral artery occlusion. Neuropharmacology, 44: 787–800. Clinckers, R., Smolders, I., Meurs, A., Ebinger, G. and Michotte, Y. (2004a) Anticonvulsant action of hippocampal dopamine and serotonin is independently mediated by D2 and 5-HT1A receptors. J. Neurochem., 89: 834–843.

Clinckers, R., Smolders, I., Meurs, A., Ebinger, G. and Michotte, Y. (2004b) Anticonvulsant action of GBR-12909 and citalopram against acute experimentally induced limbic seizures. Neuropharmacology, 47: 1053–1061. Clinckers, R., Smolders, I., Meurs, A., Ebinger, G. and Michotte, Y. (2005) Quantitative in vivo microdialysis study on the inﬂuence of multidrug transporters on the blood-brain barrier passage of oxcarbazepine: concomitant use of hippocampal monoamines as pharmacodynamic markers for the anticonvulsant activity. J. Pharmacol. Exp. Ther., 314: 725–731. Dalby, N.O. (2000) GABA-level increasing and anticonvulsant effects of three different GABA uptake inhibitors. Neuropharmacology, 39: 2399–2407. Dalley, J.W., McGaughy, J., O’Connell, M.T., Cardinal, R.N., Levita, L. and Robbins, T.W. (2001) Distinct changes in cortical acetylcholine and noradrenaline efﬂux during contingent and noncontingent performance of a visual attentional task. J. Neurosci., 21: 4908–4914. Dazzi, L., Serra, M., Porceddu, M.L., Sanna, A., Chessa, M.F. and Biggio, G. (1997) Enhancement of basal and pentylenetetrazol (PTZ)-stimulated dopamine release in the brain of freely moving rats by PTZ-induced kindling. Synapse, 26: 351–358. De Deurwaerdere, P. and Spampinato, U. (1999) Role of serotonin(2A) and serotonin(2B/2C) receptor subtypes in the control of accumbal and striatal dopamine release elicited in vivo by dorsal raphe nucleus electrical stimulation. J. Neurochem., 73: 1033–1042. De Deurwaerdere, P., L’hirondel, M., Bonhomme, N., Lucas, G., Cheramy, A. and Spampinato, U. (1997) Serotonin stimulation of 5-HT4 receptors indirectly enhances in vivo dopamine release in the rat striatum. J. Neurochem., 68: 195–203. Dijk, S.N., Francis, P.T., Stratmann, G.C. and Bowen, D.M. (1995) Cholinomimetics increase glutamate outﬂow via an action on the corticostriatal pathways: implications for Alzheimer’s disease. J. Neurochem., 65: 2165–2169. During, M.J., Ryder, K.M. and Spencer, D.D. (1995) Hippocampal GABA transporter function in temporal-lobe epilepsy. Nature, 376: 174–177. Elvander, E., Schott, P.A., Sandin, J., Bjelke, B., Kehr, J., Yoshitake, T. and Ogren, S.O. (2004) Intraseptal muscarinic ligands and galanin: inﬂuence on hippocampal acetylcholine and cognition. Neuroscience, 126: 541–557. Fadel, J., Moore, H., Sarter, M. and Bruno, J.P. (1996) Transsynaptic stimulation of cortical acetylcholine release after partial 192 IgG-saporin-induced loss of cortical cholinergic afferents. J. Neurosci., 16: 6592–6600. Feng, Y., LeBlanc, M.H. and Regunathan, S. (2005) Agmatine reduces extracellular glutamate during pentylenetetrazole-induced seizures in rat brain: a potential mechanism for the anticonvulsive effects. Neurosci. Lett., 390: 129–133. Fujiwara, M. and Egashira, N. (2004) New perspectives in the studies on endocannabinoid and cannabis: abnormal behaviors associate with CB1 cannabinoid receptor and development of therapeutic application. J. Pharmacol. Sci., 96: 362–366 Review.

450 Furuse, M., Ohta, T., Ikenaga, T., Liang, Y.M., Isono, N., Kuroiwa, T. and Preul, M.C. (2003) Effects of intravascular perfusion of cooled crystalloid solution on cold-induced brain injury using an extracorporeal cooling-ﬁltration system. Acta Neurochir. (Wien), 145: 983–992. Galefﬁ, F., Bianchi, L., Bolam, J.P. and Della Corte, L. (2003) The effect of 6-hydroxydopamine lesions on the release of amino acids in the direct and indirect pathways of the basal ganglia: a dual microdialysis probe analysis. Eur. J. Neurosci., 18: 856–868. Gerhardt, G.A., Cass, W.A., Huettl, P., Brock, S., Zhang, Z. and Gash, D.M. (1999) GDNF improves dopamine function in the substantia nigra but not the putamen of unilateral MPTP-lesioned rhesus monkeys. Brain Res., 817: 163–171. Gerin, C. (2002) Behavioral improvement and dopamine release in a Parkinsonian rat model. Neurosci. Lett., 330: 5–8. Gianfriddo, M., Corsi, C., Melani, A., Pezzola, A., Reggio, R., Popoli, P. and Pedata, F. (2003) Adenosine A(2A) antagonism increases striatal glutamate outﬂow in the quinolinic acid rat model of Huntington’s disease. Brain Res., 979: 225–229. Giovannini, M.G., Pazzagli, M., Malmberg-Aiello, P., Della Corte, L., Rakovska, A.D., Cerbai, F., Casamenti, F. and Pepeu, G. (2005) Inhibition of acetylcholine-induced activation of extracellular regulated protein kinase prevents the encoding of an inhibitory avoidance response in the rat. Neuroscience, 136: 15–32. Gluck, M.R., Santana, L.A., Granson, H. and Yahr, M.D. (2004) Novel dopamine releasing response of an anticonvulsant agent with possible anti-Parkinson’s activity. J. Neural Transm., 111: 713–724. Gold, P.E. (2003) Acetylcholine modulation of neural systems involved in learning and memory. Neurobiol. Learn. Mem., 80: 194–210. Goyagi, T., Toung, T.J., Kirsch, J.R., Traystman, R.J., Koehler, R.C., Hurn, P.D. and Bhardwaj, A. (2003) Neuroprotective kappa-opioid receptor agonist BRL 52537 attenuates ischemia-evoked nitric oxide production in vivo in rats. Stroke, 34: 1533–1538. Grondin, R., Cass, W.A., Zhang, Z., Stanford, J.A., Gash, D.M. and Gerhardt, G.A. (2003) Glial cell line-derived neurotrophic factor increases stimulus-evoked dopamine release and motor speed in aged rhesus monkeys. J. Neurosci., 23: 1974–1980. Hachimi-Idrissi, S., Corne, L., Ebinger, G., Michotte, Y. and Huyghens, L. (2001) Mild hypothermia induced by a helmet device: a clinical feasibility study. Resuscitation, 51: 275–281. Hachimi-Idrissi, S., Van Hemelrijck, A., Michotte, A., Smolders, I., Sarre, S., Ebinger, G., Huyghens, L. and Michotte, Y. (2004) Postischemic mild hypothermia reduces neurotransmitter release and astroglial cell proliferation during reperfusion after asphyxial cardiac arrest in rats. Brain Res., 1019: 217–225. Harkany, T., Abraham, I., Timmerman, W., Laskay, G., Toth, B., Sasvari, M., Konya, C., Sebens, J.B., Korf, J., Nyakas, C., Zarandi, M., Soos, K., Penke, B. and Luiten, P.G. (2000) Beta-amyloid neurotoxicity is mediated by a glutamate-triggered excitotoxic cascade in rat nucleus basalis. Eur. J. Neurosci., 12: 2735–2745.

Hashiguchi, A., Kawano, T., Yano, S., Morioka, M., Hamada, J., Sato, T., Shirasaki, Y., Ushio, Y. and Fukunaga, K. (2003) The neuroprotective effect of a novel calmodulin antagonist, 3-[2-[4-(3-chloro-2-methylphenyl)-1-piperazinyl]ethyl]5,6-dimethoxy-1-(4-imidazolylmethyl)-1H-indazole dihydrochloride 3.5 hydrate, in transient forebrain ischemia. Neuroscience, 121: 379–386. Hasselmo, M.E. and McGaughy, J. (2004) High acetylcholine levels set circuit dynamics for attention and encoding and low acetylcholine levels set dynamics for consolidation. Prog. Brain Res., 145: 207–231 Review. Herzog, C.D., Nowak, K.A., Sarter, M. and Bruno, J.P. (2003) Microdialysis without acetylcholinesterase inhibition reveals an age-related attenuation in stimulated cortical acetylcholine release. Neurobiol. Aging, 24: 861–863. Husum, H., Mikkelsen, J.D. and Mork, A. (1998) Extracellular levels of neuropeptide Y are markedly increased in the dorsal hippocampus of freely moving rats during kainic acidinduced seizures. Brain Res., 781: 351–354. Ishiwari, K., Mingote, S., Correa, M., Trevitt, J.T., Carlson, B.B. and Salamone, J.D. (2004) The GABA uptake inhibitor beta-alanine reduces pilocarpine-induced tremor and increases extracellular GABA in substantia nigra pars reticulata as measured by microdialysis. J. Neurosci. Meth., 140: 39–46. Jobe, P.C., Dailey, J.W. and Wernicke, J.F. (1999) A noradrenergic and serotonergic hypothesis of the linkage between epilepsy and affective disorders. Crit. Rev. Neurobiol., 13: 317–356 Review. Jonkers, N., Sarre, S., Ebinger, G. and Michotte, Y. (2000) MK801 inﬂuences L-DOPA-induced dopamine release in intact and hemi-Perkinson rats. Eur. J. Pharmacol., 407: 281–291. Jonkers, N., Sarre, S., Ebinger, G. and Michotte, Y. (2002) MK801 suppresses the L-DOPA-induced increase of glutamate in striatum of hemi-Parkinson rats. Brain Res., 926: 149–155. Kang, U.J., Lee, W.Y. and Chang, J.W. (2001) Gene therapy for Parkinson’s disease: determining the genes necessary for optimal dopamine replacement in rat models. Hum. Cell., 14: 39–48 Review. Khan, G.M., Smolders, I., Ebinger, G. and Michotte, Y. (2000) Anticonvulsant effect and neurotransmitter modulation of focal and systemic 2-chloroadenosine against the development of pilocarpine-induced seizures. Neuropharmacology, 39: 2418–2432. Khan, G.M., Smolders, I., Lindekens, H., Manil, J., Ebinger, G. and Michotte, Y. (1999) Effects of diazepam on extracellular brain neurotransmitters in pilocarpine-induced seizures in rats. Eur. J. Pharmacol., 373: 153–161. Khan, S., Milot, M., Lecompte-Collin, J. and Plamondon, H. (2004) Time-dependent changes in CRH concentrations and release in discrete brain regions following global ischemia: effects of MK-801 pretreatment. Brain Res., 1016: 48–57. Klintenberg, R. and Andren, P.E. (2005) Altered extracellular striatal in vivo biotransformation of the opioid neuropeptide dynorphin A(1–17) in the unilateral 6-OHDA rat model of Parkinson’s disease. J. Mass Spectrom., 40: 261–270.

451 Kokaia, M., Kalen, P., Bengzon, J. and Lindvall, O. (1989) Noradrenaline and 5-hydroxytryptamine release in the hippocampus during seizures induced by hippocampal kindling stimulation: an in vivo microdialysis study. Neuroscience, 32: 647–656. Kovacs, A., Mihaly, A., Komaromi, A., Gyengesi, E., Szente, M., Weiczner, R., Krisztin-Peva, B., Szabo, G. and Telegdy, G. (2003) Seizure, neurotransmitter release, and gene expression are closely related in the striatum of 4-aminopyridinetreated rats. Epilepsy Res., 55: 117–129. Kurokawa, M., Koga, K., Kase, H., Nakamaura, J. and Kuwana, Y. (1996) Adenosine A2a receptor-mediated modulation of striatal acetylcholine release in vivo. J. Neurochem., 66(5): 1882–1888. Lachowicz, J.E., Duffy, R.A., Ruperto, V., Kozlowski, J., Zhou, G., Clader, J., Billard, W., Binch, H. III., Crosby, G., Cohen-Willians, M., Strader, C.D. and Cofﬁn, V. (2001) Facilitation of acetylcholine release and improvement in cognition by a selective M2 muscarinic antagonist, SCH 72788. Life Sci., 68: 2585–2592. Lindekens, H., Smolders, I., Khan, G.M., Bialer, M., Ebinger, G. and Michotte, Y. (2000) In vivo study of the effect of valpromide and valnoctamide in the pilocarpine rat model of focal epilepsy. Pharm. Res., 17: 1408–1413. Liu, S., Liu, M., Peterson, S., Miyake, M., Vallyathan, V. and Liu, K.J. (2003) Hydroxyl radical formation is greater in striatal core than in penumbra in a rat model of ischemic stroke. J. Neurosci. Res., 71: 882–888. Lucas, G. and Spampinato, U. (2000) Role of striatal serotonin2A and serotonin2C receptor subtypes in the control of in vivo dopamine outﬂow in the rat striatum. J. Neurochem., 74: 693–701. Marti, M., Bregola, G., Morari, M., Gemignani, A. and Simonato, M. (2000a) Somatostatin release in the hippocampus in the kindling model of epilepsy: a microdialysis study. J. Neurochem., 74: 2497–2503. Marti, M., Manzalini, M., Fantin, M., Bianchi, C., Della Corte, L. and Morari, M. (2005) Striatal glutamate release evoked in vivo by NMDA is dependent upon ongoing neuronal activity in the substantia nigra, endogenous striatal substance P and dopamine. J. Neurochem., 93: 195–205. Marti, M., Mela, F., Bianchi, C., Beani, L. and Morari, M. (2002) Striatal dopamine-NMDA receptor interactions in the modulation of glutamate release in the substantia nigra pars reticulata in vivo: opposite role for D1 and D2 receptors. J. Neurochem., 83: 635–644. Marti, M., Sbrenna, S., Fuxe, K., Bianchi, C., Beani, L. and Morari, M. (2000b) Increased responsivity of glutamate release from the substantia nigra pars reticulata to striatal NMDA receptor blockade in a model of Parkinson’s disease. A dual probe microdialysis study in hemiparkinsonian rats. Eur. J. Neurosci., 12: 1848–1850. Matarredona, E.R., Santiago, M., Venero, J.L., Cano, J. and Machado, A. (2001) Group II metabotropic glutamate receptor activation protects striatal dopaminergic nerve terminals against MPP+-induced neurotoxicity along with brain-derived neurotrophic factor induction. J. Neurochem., 76: 351–360.

Mauler, F., Fahrig, T., Horvath, E. and Jork, R. (2001) Inhibition of evoked glutamate release by the neuroprotective 5-HT(1A) receptor agonist BAY x 3702 in vitro and in vivo. Brain Res., 888: 150–157. McRee, R.C., Terry-Ferguson, M., Langlais, P.J., Chen, Y., Nalwalk, J.W., Blumenstock, F.A. and Hough, L.B. (2000) Increased histamine release and granulocytes within the thalamus of a rat model of Wernicke’s encephalopathy. Brain Res., 858: 227–236. Meissner, W., Harnack, D., Paul, G., Reum, T., Sohr, R., Morgenstern, R. and Kupsch, A. (2002) Deep brain stimulation of subthalamic neurons increases striatal dopamine metabolism and induces contralateral circling in freely moving 6-hydroxydopamine-lesioned rats. Neurosci. Lett., 328: 105–108. Melani, A., Pantoni, L., Corsi, C., Bianchi, L., Monopoli, A., Bertorelli, R., Pepeu, G. and Pedata, F. (1999) Striatal outﬂow of adenosine, excitatory amino acids, gamma-aminobutyric acid, and taurine in awake freely moving rats after middle cerebral artery occlusion: correlations with neurological deﬁcit and histopathological damage. Stroke, 30: 2448–2454. Melani, A., Turchi, D., Vannucchi, M.G., Cipriani, S., Gianfriddo, M. and Pedata, F. (2005) ATP extracellular concentrations are increased in the rat striatum during in vivo ischemia. Neurochem. Int., 47: 442–448. Mignon, L.J. and Wolf, W.A. (2005) 8-Hydroxy-2-(di-n-propylamino)tetralin reduces striatal glutamate in an animal model of Parkinson’s disease. Neuroreport, 16: 699–703. Mills, C.D., Xu, G.Y., McAdoo, D.J. and Hulsebosch, C.E. (2001) Involvement of metabotropic glutamate receptors in excitatory amino acid and GABA release following spinal cord injury in rat. J. Neurochem., 79: 835–848. Montalbetti, L., Rozza, A., Rizzo, V., Favalli, L., Scavini, C., Lanza, E., Savoldi, F., Racagni, G. and Scelsi, R. (1995) Aminoacid recovery via microdialysis and photoinduced focal cerebral ischemia in brain cortex of rats. Neurosci. Lett., 192: 153–156. Nail-Boucherie, K., Le-Pham, B.T., Gobaille, S., Maitre, M., Aunis, D. and Depaulis, A. (2005) Evidence for a role of the parafascicular nucleus of the thalamus in the control of epileptic seizures by the superior colliculus. Epilepsia, 46: 141–145. Nelson, C.L., Sarter, M. and Bruno, J.P. (2005) Prefrontal cortical modulation of acetylcholine release in posterior parietal cortex. Neuroscience, 132: 347–359. Obrenovitch, T.P. (1999) High extracellular glutamate and neuronal death in neurological disorders. Cause, contribution or consequence? Ann. N.Y. Acad. Sci., 890: 273–286. Ochi, M., Shiozaki, S. and Kase, H. (2004) Adenosine A(2A) receptor-mediated modulation of GABA and glutamate release in the output regions of the basal ganglia in a rodent model of Parkinson’s disease. Neuroscience, 127: 223–231. O’Neill, M.J., Bogaert, L., Hicks, C.A., Bond, A., Ward, M.A., Ebinger, G., Ornstein, P.L., Michotte, Y. and Lodge, D. (2000) LY377770, a novel iGlu5 kainate receptor antagonist with neuroprotective effects in global and focal cerebral ischaemia. Neuropharmacology, 39(9): 1575–1588.

452 Owen, A.J., Ijaz, S., Miyashita, H., Wishart, T., Howlett, W. and Shuaib, A. (1997) Zonisamide as a neuroprotective agent in an adult gerbil model of global forebrain ischemia: a histological, in vivo microdialysis and behavioral study. Brain Res., 770: 115–122. Parrot, S., Sauvinet, V., Riban, V., Depaulis, A., Renaud, B. and Denoroy, L. (2004) High temporal resolution for in vivo monitoring of neurotransmitters in awake epileptic rats using brain microdialysis and capillary electrophoresis with laser-induced ﬂuorescence detection. J. Neurosci. Methods, 140: 29–38. Pepeu, G. and Giovannini, M.G. (2004) Changes in cetylcholine extracellular levels during cognitive processess. Learn. Mem., 11: 21–27. Pitka¨nen, A., Schwarzkroin, P. and Moshe´, S. (2006) Models of Seizures and Epilepsy. Elsevier-Academic Press, San Diego. Potschka, H., Baltes, S. and Lo¨scher, W. (2004) Inhibition of multidrug transporters by verapamil or probenecid does not alter blood-brain barrier penetration of levetiracetam in rats. Epilepsy Res., 58: 85–91. Reynolds, N.C. Jr., Lin, W., Meyer Cameron, C. and Roerig, D.L. (1997) Differential responses of extracellular GABA to intrastriatal perfusions of 3-nitropropionic acid and quinolinic acid in the rat. Brain Res., 778: 140–149. Rizzi, M., Caccia, S., Guiso, G., Richichi, C., Gorter, J.A., Aronica, E., Aliprandi, M., Bagnati, R., Fanelli, R., D’Incalci, M., Samanin, R. and Vezzani, A. (2002) Limbic seizures induce P-glycoprotein in rodent brain: functional implications for pharmacoresistance. J. Neurosci., 22: 5833–5839. Robinson, S., Freeman, P., Moore, C., Touchon, J.C., Krentz, L. and Meshul, C.K. (2003) Acute and subchronic MPTP administration differentially affects striatal glutamate synaptic function. Exp. Neurol., 180: 74–87. Rosi, S., Giovannini, M.G., Lestage, P.J., Munoz, C., Corte, L.D. and Pepeu, G. (2004) S 18986, a positive modulator of AMPA receptors with cognition-enhancing properties, increases ACh release in the hippocampus of young and aged rat. Neurosci. Lett., 361: 120–123. Rowley, H.L., Martin, K.F. and Marsden, C.A. (1995) Decreased GABA release following tonic-clonic seizures is associated with an increase in extracellular glutamate in rat hippocampus. Neuroscience, 68: 415–422. Russell, M.G. and Dias, R. (2002) Memories are made of this (perhaps): a review of serotonin 5-HT(6) receptor ligands and their biological functions. Curr. Top Med. Chem., 2: 643–654. Sakowitz, O.W., Unterberg, A.W. and Stover, J.F. (2002) Neuronal activity determined by quantitative EEG and cortical microdialysis is increased following controlled cortical impact injury in rats. Acta Neurochir. Suppl., 81: 221–223. Saleh, T.M., Connell, B.J., Legge, C. and Cribb, A.E. (2004) Estrogen attenuates neuronal excitability in the insular cortex following middle cerebral artery occlusion. Brain Res., 1018: 119–129. Sarre, S., Yuan, H., Jonkers, N., Van Hemelrijck, A., Ebinger, G. and Michotte, Y. (2004) In vivo characterization of somatodendritic dopamine release in the substantia nigra of 6hydroxydopamine-lesioned rats. J. Neurochem., 90: 29–39.

Sarter, M. and Bruno, J.P. (1997) Trans-synaptic stimulation of cortical acetylcholine and enhancement of attentional functions: a rational approach for the development of cognition enhancers. Behav. Brain Res., 83: 7–14. Sarter, M., Bruno, J.P. and Givens, B. (2003) Attentional functions of cortical cholinergic inputs: what does it mean for learning and memory? Neurobiol. Learn. Mem., 80: 245–256. Sato, K., Ueda, H., Okumura, F. and Misu, Y. (1994) Supersensitization of intrastriatal dopamine receptors involved in opposite regulation of acetylcholine release in Parkinson’s model rats. Neurosci. Lett., 173: 59–62. Scali, C., Casamenti, F., Bellucci, A., Costagli, C., Schmidt, B. and Pepeu, G. (2002) Effect of subchronic administration of metrifonate, rivastigmine and donepezil on brain acetylcholine in aged F344 rats. J. Neural Transm., 109: 1067–1080. Scali, C., Giovannini, M.G., Bartolini, L., Prosperi, C., Hinz, V., Schmidt, B. and Pepeu, G. (1997) Effect of metrifonate on extracellular brain acetylcholine and object recognition in aged rats. Eur. J. Pharmacol., 325: 173–180. Schechter, L.E., Smith, D.L., Rosenzweig-Lipson, S., Sukoff, S.J., Dawson, L.A., Marquis, K., Jones, D., Piesla, M., Andree, T., Nawoschik, S., Harder, J.A., Womack, M.D., Buccafusco, J., Terry, A.V., Hoebel, B., Rada, P., Kelly, M., Abou-Gharbia, M., Barrett, J.E. and Childers, W. (2005) Lecozotan (SRA-333): a selective serotonin 1A receptor antagonist that enhances the stimulated release of glutamate and acetylcholine in the hippocampus and possesses cognitive-enhancing properties. J. Pharmacol. Exp. Ther., 314: 1274–1289. Schroeder, J.A. and Schneider, J.S. (2002) GABA-opioid interactions in the globus pallidus: [d-Ala2]-Met-enkephalinamide attenuates potassium-evoked GABA release after nigrostriatal lesion. J. Neurochem., 82: 666–673. Shin, R.S., Anisman, H., Merali, Z. and McIntyre, D.C. (2004) Amygdala amino acid and monoamine levels in genetically fast and slow kindling rat strains during massed amygdala kindling: a microdialysis study. Eur. J. Neurosci., 20: 185–194. Shouse, M.N., Staba, R.J., Ko, P.Y., Saquib, S.F. and Farber, P.R. (2001) Monoamines and seizures: microdialysis ﬁndings in locus ceruleus and amygdala before and during amygdala kindling. Brain Res., 892: 176–192. Smolders, I., Bortolotto, Z.A., Clarke, V.R.J., Warre, R., Khan, G.M., O’Neill, M.J., Ornstein, P.L., Bleakman, D., Ogden, A.M., Weiss, B., Stables, J.P., Ho, K.H., Ebinger, G., Collingridge, G.L., Lodge, D. and Michotte, Y. (2002b) Antagonists of GLUK5-containing kainate receptors prevent pilocarpine-induced limbic seizures. Nat. Neurosci., 5: 796–804. Smolders, I., Gousseau, C., Marchand, S., Couet, W., Ebinger, G. and Michotte, Y. (2002a) Convulsant and subconvulsant doses of norﬂoxacin in the presence and absence of biphenylacetic acid alter extracellular hippocampal glutamate but not gamma-aminobutyric acid levels in conscious rats. Antimicrob. Agents Chemother., 46: 471–477.

453 Smolders, I., Khan, G.M., Lindekens, H., Prikken, S., Marvin, C.A., Manil, J., Ebinger, G. and Michotte, Y. (1997b) Effectiveness of vigabatrin against focally evoked pilocarpine-induced seizures and concomitant changes in extracellular hippocampal and cerebellar glutamate, GABA and dopamine levels, a microdialysis-electrocorticography study in freely moving rats. J. Pharmacol. Exp. Ther., 283: 1239–1248. Smolders, I., Khan, G.M., Manil, J., Ebinger, G. and Michotte, Y. (1997a) NMDA receptor-mediated pilocarpine-induced seizures: characterization in freely moving rats by microdialysis. Br. J. Pharmacol., 121: 1171–1179. Smolders, I., Lindekens, H., Clinckers, R., Meurs, A., O’Neill, M.J., Lodge, D., Ebinger, G. and Michotte, Y. (2004) In vivo modulation of extracellular glutamate and GABA levels and of limbic seizures by group I and II metabotropic glutamate receptor ligands. J. Neurochem., 88: 1068–1077. Sun, M., Kong, L., Wang, X., Lu, X.G., Gao, Q. and Geller, A.I. (2005) Comparison of the capability of GDNF, BDNF, or both, to protect nigrostriatal neurons in a rat model of Parkinson’s disease. Brain Res., 1052: 119–129. Todd, K.G. and Butterworth, R.F. (2001) In vivo microdialysis in an animal model of neurological disease: thiamine deﬁciency (Wernicke) encephalopathy. Methods, 23: 55–61 Review. Tottori, K., Nalai, M., Uwahodo, Y., Miwa, T., Yamada, S., Oshiro, Y., Kikuchi, T. and Altar, C.A. (2002) Attenuation of scopolamine-induced and age-associated memory impairments by the sigma and 5-hydroxytryptamine(1A) receptor agonist OPC-14523 (1-[3-[4-(3-chlorophenyl)-1-piperazinyl] propyl]-5-methoxy-3,4-dihydro-2[1H]-quinolinone monomethanesulfonate. J. Pharmacol. Exp. Ther., 301: 249–257. Toyota, S., Graf, R., Valentino, M., Yoshimine, T. and Heiss, W.D. (2002) Malignant infarction in cats after prolonged middle cerebral artery occlusion: glutamate elevation related to decrease of cerebral perfusion pressure. Stroke, 33: 1383–1391. Tran, M.H., Yamada, K. and Nabeshima, T. (2002) Amyloid bpeptide induces cholinergic dysfunction and cognitive deﬁcits: a minireview. Peptides, 23: 1271–1283. Tuncel, N., Sener, E., Cerit, C., Karasu, U., Gurer, F., Sahinturk, V., Baycu, C., Ak, D. and Filiz, Z. (2005) Brain mast cells and therapeutic potential of vasoactive intestinal peptide in a Parkinson’s disease model in rats: brain microdialysis, behavior, and microscopy. Peptides, 26: 827–836. Van Hemelrijck, A., Hachimi-Idrissi, S., Sarre, S., Ebinger, G. and Michotte, Y. (2005) Neuroprotective effect of N-acetylaspartyl-glutamate in combination with mild hypothermia in the endothelin-1 rat model of focal cerebral ischaemia. J. Neurochem., 95: 1287–1297.

Van Hemelrijck, A., Vermijlen, D., Hachimi-Idrissi, S., Sarre, S., Ebinger, G. and Michotte, Y. (2003) Effect of resuscitative mild hypothermia on glutamate and dopamine release, apoptosis and ischaemic brain damage in the endothelin-1 rat model for focal cerebral ischaemia. J. Neurochem., 87: 66–75. Wang, X., Shimizu-Sasamata, M., Moskowitz, M.A., Newcomb, R. and Lo, E.H. (2001) Proﬁles of glutamate and GABA efﬂux in core versus peripheral zones of focal cerebral ischemia in mice. Neurosci. Lett., 313: 121–124. Wilson, C.L., Maidment, N.T., Shomer, M.H., Behnke, E.J., Ackerson, L., Fried, I. and Engel, J. Jr. (1996) Comparison of seizure related amino acid release in human epileptic hippocampus versus a chronic, kainate rat model of hippocampal epilepsy. Epilepsy Res., 26: 245–254. Wu, H.Q., Rassoulpour, A., Goodman, J.H., Scharfman, H.E., Bertram, E.H. and Schwarcz, R. (2005) Kynurenate and 7-chlorokynurenate formation in chronically epileptic rats. Epilepsia, 46: 1010–1016. Xu, G.Y., Hughes, M.G., Ye, Z., Hulsebosch, C.E. and McAdoo, D.J. (2004) Concentrations of glutamate released following spinal cord injury kill oligodendrocytes in the spinal cord. Exp. Neurol., 187: 329–336. Xu, K. and Dluzen, D.E. (2000) The effect of GDNF on nigrostriatal dopaminergic function in response to a two-pulse K(+) stimulation. Exp. Neurol., 166: 450–457. Yager, J.Y., Armstrong, E.A., Miyashita, H. and Wirrell, E.C. (2002) Prolonged neonatal seizures exacerbate hypoxicischemic brain damage: correlation with cerebral energy metabolism and excitatory amino acid release. Dev. Neurosci., 24: 367–381. Yang, D.Y., Lee, J.B., Lin, M.C., Huang, Y.L., Liu, H.W., Liang, Y.J. and Cheng, F.C. (2004) The determination of brain magnesium and zinc levels by a dual-probe microdialysis and graphite furnace atomic absorption spectrometry. J. Am. Coll. Nutr., 23: 552S–555S. Yasuda, H., Fujii, M., Fujisawa, H., Ito, H. and Suzuki, M. (2001) Changes in nitric oxide synthesis and epileptic activity in the contralateral hippocampus of rats following intrahippocampal kainate injection. Epilepsia, 42: 13–20. Yuan, H., Sarre, S., Ebinger, G. and Michotte, Y. (2005) Histological, behavioural and neurochemical evaluation of medial forebrain bundle and striatal 6-OHDA lesions as rat models of Parkinson’s disease. J. Neurosci. Methods, 144: 35–45. Ziai, W.C., Sgerman, D.L., Bhardwaj, A., Zhang, N., Keyl, P.M. and Mirski, M.A. (2005) Target-speciﬁc catecholamine elevation induced by anti-convulsant thalamic deep brain stimulation. Epilepsia, 46: 878–888.

CHAPTER 5.4

Online glucose and lactate monitoring during physiological and pathological conditions Marianne Fillenz Department of Physiology, University of Oxford, Parks Road, Oxford, UK

Abstract: There exists a controversy concerning the role of glucose and lactate in brain metabolism. Microdialysis provides a technique for monitoring the temporal relationship of changes in the extracellular concentration of these metabolites in relation to the neuronal activity under a variety of conditions. In response to physiological neuronal activation, a decrease in glucose coincides with the increase in neuronal activity, as signaled by the increase in local cerebral blood ﬂow, whereas lactate shows a rise followed by a slow return to baseline. These ﬁndings suggest that glucose rather than lactate provides the additional energy required by the activated neurons. During hypoxia, there is a large increase in lactate, which decreases steeply on the readmission of air, which suggests that lactate makes a signiﬁcant contribution to the restoration of neuronal function under these conditions. Infusion of glutamate, to mimic the effects of brain damage resulting from ischemia, produces a large increase in lactate and localized brain lesions. Addition of L-lactate and not D-lactate to glutamate reduces the size of the lesion, which suggests that lactate, by acting as a metabolic substrate, is exerting a neuroprotective effect. Finally, the introduction of microdialysis combined with a new rapid sampling techniques into the clinic provides a much more accurate assessment of the metabolic state of brain metabolism in cases of human traumatic brain injury. increase in blood ﬂow that accompanies a local increase in neuronal activity was reported by Sherrington in 1980 (Roy and Sherrington, 1890), but it took a further 100 years before it became possible to measure changes in the rate of glucose and oxygen consumption during neuronal activation. Using positron emission tomography, brain glucose uptake, local cerebral blood ﬂow, and oxygen utilization were measured in the human visual cortex during visual stimulation. While glucose uptake and local cerebral blood ﬂow increased by 50%, there was only a 5% increase in oxygen utilization (Fox and Raichle, 1986; Fox et al., 1988). This implied that glucose was undergoing aerobic glycolysis. With the use of 1 H-NMR spectroscopy (Prichard et al., 1991), a rise in lactate in the human visual cortex during photic stimulation conﬁrmed the occurrence of glycolysis.

I. Introduction Glucose is an essential energy substrate for the brain. The function of the brain is to monitor changes in the external and internal environment, and plan appropriate responses to these changes; this requires the transmission of action potentials through complex neuronal networks. The setting up and transmission of nerve impulses involve the movement of ions along electrochemical gradients. The maintenance of these gradients entails the active transport of ions and accounts for the major part of energy consumed by brain. The energy for brain is derived from the oxidation of glucose; glucose and oxygen are delivered to the brain by the cerebral circulation. The local Corresponding author: E-mail: marianne.ﬁ[email protected]. ac.uk

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

455

DOI: 10.1016/S1569-7339(06)16024-5 Copyright 2007 Elsevier B.V. All rights reserved

456

II. Brain metabolism under physiological conditions II.A. ANLS hypothesis Neuronal activation is produced by glutamate, the main excitatory transmitter, which is released from presynaptic terminals and acts on postsynaptic receptors. The action of glutamate is very brief, as it is removed by glutamate transporters, localized on the astrocytes that surround the synapse. When glutamate is added to astrocytes in culture, there is an increase in glucose uptake and release of lactate (Pellerin and Magistretti, 1994, 1996). This led to the astrocyte-neurone-lactate shuttle (ANLS) hypothesis, which postulates that astrocytes provide the link between increased neuronal activity and the increase in glucose utilization (Pellerin et al., 1998). The central elements of the hypothesis are the uptake of synaptically released glutamate by astrocytes, an accompanying increase in the uptake of glucose, glycolysis of glucose, and the export of the resulting lactate; the lactate, taken up by neurons, provides the energy for energyrequiring processes of which the most important is ion pumping. The hypothesis has come under increasing challenge (Chih et al., 2001; Chih and Roberts, 2003; Mangia et al., 2003b; Dienel and Cruz, 2004); the controversy centers around the question whether the activated neurons use lactate or glucose for their energy production. Most of the evidence for the ANLS hypothesis is based on in vitro experiments using astrocytic cultures or the demonstration of the presence of the enzymes and transporters required by the hypothesis. The hypothesis postulates a sequence of events, which can only be established by using in vivo monitoring. Microdialysis provides a technique for establishing the temporal sequence of events.

II.B. Zero-net-flux method Microdialysis combined with an enzyme-packed bed system provides an online system for measuring changes in the extracellular ﬂuid (ECF) concentration of glucose and lactate (Boutelle et al., 1992). The concentration in the dialysate is not the

same as the ECF concentration but depends on the probe recovery. Recovery in vivo is a measure of the rate at which a solute is delivered to the perfusate. This in turn depends on diffusion barriers within the brain, on properties of the dialysis membrane, and on the rate of perfusion (Dykstra et al., 1992). The true ECF concentration of glucose and lactate, which reﬂect the balance between supply and utilization, can be established by the use of the zero-net-ﬂux method of Lo¨nnroth (Lo¨nnroth et al., 1987). Various concentrations of analyte are added to the perfusion ﬂuid and gain or loss by the brain is plotted against the infused concentration. The point of zero ﬂux across the probe, determined by regression analysis, is the point of equilibrium and is equal to the true ECF concentration (Fig. 1A). With the use of this technique, values for ECF glucose have been obtained that vary from 0.35 mM (Fray et al., 1997) to 0.47 mM (Fellows et al., 1992) for the rat striatum and from 1.0 to 1.2 mM for the rat hippocampus (McNay and Gold, 1999); the latter study demonstrated that ECF concentrations for glucose vary with the brain region and the strain of the rat. The use of a different zero-ﬂux method (Jacobson et al., 1985) gave a value of 1.66 mM for glucose in rat hippocampus (Abi-Saab et al., 2002). Values similar to those obtained with the zero-net-ﬂux method have been obtained either using an implanted glucose sensor (Lowry et al., 1998c) or differential normal pulse voltammetry (Netchiporouk et al., 2001). The ECF concentration of lactate using the zero-net-ﬂux technique in the rat striatum is 0.35 mM (Demestre et al., 1997); with the zeroﬂux method, the value for ECF lactate in rat hippocampus is 2.70 mM (Abi-Saab et al., 2002). A further complication is that the in vivo recovery depends on both passive and active processes, such as uptake (Boutelle and Fillenz, 1996); it will, therefore, be affected by changes in the rate of metabolism. The in vivo recovery is derived from the slope of the Lo¨nnroth curve. Administration of veratridine, which opens the voltagegated Na channels and so stimulates Na/K ATPase, leads to an increase in the energy metabolism. Addition of 50-mM veratridine to the perfusion ﬂuid produces a steep decrease in dialysate

457

Fig. 1. Lo¨nnroth zero-net-ﬂux technique for determining true extracellular concentration and in vivo turnover rate of glucose. (A) Lo¨nnroth curve showing zero-net-ﬂux point, which is the extracellular concentration. The slope of the curve is a measure of the in vivo recovery. (B) Dialysate concentration of glucose when 50 mM veratridine is added to the perfusion medium. (C) A time-resolved zeronet-ﬂux experiment showing the time course of changes in dialysate concentration, extracellular concentration, and in vivo recovery.

glucose concentration; this decrease has a delay of 7.5 min (Fellows et al., 1992; Fray et al., 1997; Osborne et al., 1997) (Fig. 1B). A zero-net-ﬂux experiment shows a dramatic decrease in the calculated extracellular concentration but an increase in in vivo recovery, as measured by the slope of the Lo¨nnroth curve. The dynamic effects of veratridine can be established using a timeresolved zero-net-ﬂux technique. This shows an initial steep increase in in vivo recovery, which decreases progressively as the extracellular concentration of glucose decreases (Fray et al., 1997). The conclusion from these experiments is that a substantial increase in glucose utilization leads to an increase in vivo recovery and a steep decline in extracellular glucose concentration; glucose turnover and, therefore, in vivo recovery decline as the

availability of glucose is exhausted (Fig. 1C). These results also suggest that there is no rapid replenishment from vascular glucose, in spite of its high glucose concentration and, therefore, argues against a direct supply of glucose to the ECF from the blood stream. Furthermore, the delayed decrease in dialysate glucose, while ECF glucose shows a steep decrease, is due to dynamic changes in in vivo recovery. As a result, microdialysis measures relative changes in extracellular glucose but does not provide an accurate measure of the time course of these changes. In contrast, an implanted glucose biosensor (Lowry et al., 1998b) has a 100% in vivo recovery and provides a realtime measure of the changes in extracellular glucose concentration (Lowry and Fillenz, 2001) and produces less tissue damage (Peters et al., 2004).

458

II.C. Physiological changes in lactate and glucose II.C.1. Lactate To test the ANLS hypothesis, changes in lactate and glucose in response to various forms of neuronal activation have been measured. One of the earliest reports of an increase in lactate in rat brain was the increase in hippocampal lactate in response to stress (Kuhr and Korf, 1988). In subsequent studies by this group, lactate changes were measured in different brain regions in response to different forms of activation. These included a 50-Hz current delivered through ear clips for a period of 1 s in chloral hydrate anesthetized rats and three forms of mild stress in unanesthetized rats. These stresses were handling, restraint, and cold exposure–all for periods of 5 min. In all these experiments, there was a rise in hippocampal dialysate lactate of 20–25% above baseline; this occurred during the period of activation and was followed by a slow return to baseline over periods of up to 15–20 min after the end of stimulation (Schasfoort et al., 1988). In another study, changes in lactate were measured in both hippocampus and striatum in response to various forms of stress and exercise. These manipulations were 5-min immobilization, placing the rat on a platform, which was lowered into a water bath (emotional stress), and 15-min forced swimming (exercise). All three forms of stress resulted in an increase in hippocampal lactate, but the only statistically signiﬁcant increase in striatal lactate was produced by exercise. As in the previous study, the rise in lactate occurred during the period of activation and was followed by a slow return to baseline (De Bruin et al., 1990). Changes in striatal lactate have also been measured in response to a 5-min tail pinch, produced by attaching a paper clip to the rat’s tail or induced grooming, produced by dripping water on the rat’s snout. The tail pinch stimulus produced a lactate increase of 70% above baseline during the application of the pinch and returned to baseline over a period of 45 min (Fellows et al., 1993; Fray et al., 1996) (Fig. 2A). A very similar pattern of lactate increase occurred in response to induced grooming (Demestre et al., 1997). The ANLS hypothesis, based on evidence from in vitro experiments (Pellerin and Magistretti,

1994), postulates that lactate release is derived from astrocytes and results from the uptake of glutamate. The rise in lactate in response to infusion of glutamate through the dialysis probe suggests that this phenomenon also occurs in vivo (Demestre et al., 1997). This leaves the question whether the release of lactate in response to physiological stimulation is a response of the glutamate uptake. The fact that abolition of the lactate increases in response to tail pinch (Fray et al., 1996) and induced grooming (Demestre et al., 1997), when a glutamate blocker is added to the perfusion medium, conﬁrms that the physiologically induced rise in lactate is derived from glutamate uptake. This appears to lend support to the ANLS hypothesis. However, a second essential element of the hypothesis is that the lactate thus released provides the energy substrate for the activated neurons (Pellerin and Magistretti, 2003). There is extensive evidence, most of it from in vitro experiments, that brain tissue can use lactate for metabolism. Some of the earlier evidences come from experiments with brain slices, carried out by McIlwain in 1950 (McIlwain, 1953, 1955). Larrabee, using embryonic sympathetic ganglia, compared the output of radiolabeled CO2 from radiolabeled glucose and lactate (Larrabee, 1983, 1995, 1996). He concluded that both substrates could be used and the rate of uptake depended on their relative concentrations. McKenna et al. (1998) demonstrated the uptake of lactate into synaptosomes. In a number of studies, the effect of the monocarboxylate transporter antagonist, alpha-cyano-4-hydroxycinnamate (4-CIN), has been taken as evidence for the role of lactate in metabolism. However, in primary cultures of neurons, 4-CIN depressed the oxidation of both glucose and lactate; this was attributed to the inhibition of mitochondrial oxidation of pyruvate derived from either lactate or glucose (McKenna et al., 2001). What the experiments using microdialysis provide is the temporal relationship of the rise in lactate to the period of neuronal activation. The most striking feature of the lactate change is that it is a rise, sometimes as much as 70% above baseline; this demonstrates supply rather than utilization of lactate. Furthermore, the rise occurs during the period of activation when the energy requirements

459

Fig. 2. Lactate and glucose changes during a 5-min tail pinch. *po0.05 compared with basal. (A) Time course of increased blood ﬂow measured as changes in tissue oxygen and changes in dialysate glucose in response to tail pinch. (B). Changes in glucose in response to tail pinch measured using microdialysis and a glucose sensor. (C) Changes in glucose in the same rat in response to tail pinch with aCSF or propranolol in the perfusion medium.

are at their highest (Fig. 2A). The failure to observe a decrease in lactate could be due to the low-time resolution of microdialysis. There are two reports of a stimulation-induced decrease in lactate. Using proton magnetic resonance spectroscopy, a 1-s visual stimulus in human subjects produced a decrease in lactate 5 s after the application of the stimulus (Mangia et al., 2003a). In another study, using an implanted enzyme-based lactate sensor, lactate changes in the dentate gyrus of the rat hippocampus were measured in response to a 5-s electrical stimulation of the perforant path. This led to a 7% decrease in lactate lasting 10–12 s from the onset of the stimulus; the decrease was followed by an increase in lactate to 149–200% of basal at 60 s after the stimulus and a return to basal levels over a period of 10–14 min (Hu and Wilson, 1997). These experiments, therefore, conﬁrm that the main change in lactate is a large production of lactate with very little evidence of a signiﬁcant utilization of lactate.

II.C.2. Glucose The next question is the role of glucose. Pharmacologically induced changes in neuronal activity produce changes in ECF glucose concentration. Thus, infusion of tetrodotoxin (TTX) (Fellows et al., 1992) or a systemic application of the anesthetic chloral hydrate (Fellows et al., 1992) or nembutal (Osborne et al., 1997), which all lead to a reduction in neuronal activity, leads to an increase in striatal dialysate glucose. In contrast, drugs added to the perfusion medium that increase neuronal activity, such as the glutamate receptor agonist, N-methylD-aspartate (NMDA) (Fellows et al., 1992) or veratridine, a drug that opens Na channels (Fellows et al., 1992; Fray et al., 1997; Osborne et al., 1997) lead to a profound decrease in dialysate glucose. One of the earliest studies of physiologically induced changes in glucose using microdialysis was that of van der Kuil and Korf (1991). They showed that a 50-Hz current applied through ear clips for a period of 1 s produced seizures and a rise

460

in hippocampal glucose. The increase in glucose occurred after a delay of 2 min and reached a level of 200–250% of basal. The increase declined slowly and basal levels were reached 30 min after the end of the stimulation. A milder from of stress, consisting of a 5 min immobilization, led to an initial 10% decrease of hippocampal glucose during the stress followed by a maximum increase to 125% of basal at 15 min after the stress and a return to basal at 40 min after the stress. A 5-min tail pinch also led to a small initial decrease in glucose followed by a delayed, prolonged increase in dialysate glucose (Fray et al., 1996). In some studies, tail pinch (Fellows and Boutelle, 1993) or a 7.5-min restraint (Osborne et al., 1997) produced only a delayed increase without the initial decrease in dialysate glucose. Finally, McNay et al. (2000) measured hippocampal glucose changes associated with the cognitive demands during a spatial task. The experiments involved spontaneous alternation tests of spatial working memory in one of two mazes. ECF glucose levels in the hippocampus decreased by 32% below baseline during the test period on the more complex maze, but by a maximum of only 11% on the less complex maze (McNay et al., 2000). II.D. Source and fate of ECF lactate and glucose The time course of the changes of both lactate and glucose bear no simple relationship to the time course of the increase in neuronal activity and raise questions concerning the source of both these metabolites. II.D.1. Lactate The ratio of plasma to brain ECF lactate in rats is 0.4 and the rise in lactate occurs after the major part of the increase in regional cerebral blood ﬂow (RCBF) (Fig. 2A); the lactate, therefore, must be generated in the brain. The blockade of the rise in lactate by inhibitors of the glutamate transporter (Fray et al., 1996; Demestre et al., 1997) conﬁrms the hypothesis that lactate is the result of glutamate uptake. However, the fact that the stimulated lactate increase occurs after the end of the period

of activation suggests that it is not the uptake of synaptically released glutamate. Another possible source of lactate is the reuptake of glutamate released from astrocytes (Fillenz, 2005). Astrocytes have on their surface membrane a number of G-protein coupled receptors, including metabotropic glutamate receptors, whose stimulation leads to a rise of [Ca2+]i, released from intracellular stores (Cornell-Bell et al., 1990). Experiments with astrocytes in culture have shown that the rise in astrocytic Ca2+ leads to the release of glutamate (Jeftinija et al., 1996; Innocenti et al., 2000). Astrocytic glutamate release may, therefore, be the source of the large rise in lactate in response to neuronal activation. The very slow decay of this lactate suggests that it does not undergo local oxidation and, therefore, does not contribute to the energy supply of activated neurons (Fillenz, 2005).

II.D.2. Glucose The ratio of plasma to ECF concentration for glucose is 4.7, which suggests that there is no free exchange between the plasma and the ECF. This is supported by the ﬁnding that an intravenous (i.v.) infusion of glucose, which raises plasma glucose to 11.5 mM, produces a very slow rise in ECF glucose that reaches a peak over 50–60 min (Abi-Saab et al., 2002). Neuronal activation leads to a biphasic change in ECF glucose – an early decrease followed by a late prolonged increase. The early decrease is underestimated by microdialysis, because of the lowtime resolution, and is much more pronounced when glucose is measured with an implanted glucose sensor, which has a higher in vivo recovery and provides a continuous record of the changes (Lowry and Fillenz, 1997) (Fig. 2B). The period of activation is accompanied by an increase in RCBF. This increase in blood ﬂow fails to compensate for the increased glucose utilization, which in the same way as the veratridine-induced depletion (see above), argues against a direct transfer of glucose from the vascular system to the ECF. The decrease in glucose coincides with the period of stimulation. In hippocampal slices, ﬂuorescence imaging of NADH reveals that electrical stimulation of Schaffer collaterals induces an early

461

oxidative phosphorylation conﬁned to neurons. This implies that in activated neurons, oxidation of glucose provides the additional energy required by activation. There is a delayed increase in NADH conﬁned to astrocytes, which signals the late glycolysis resulting in the formation of lactate (Kasischke et al., 2004). Although the early decrease in glucose signals its activity-dependent increase in utilization, there remains the question of the origin of the late rise in glucose. As it occurs after the increase in blood ﬂow, it cannot be derived from the blood stream. An alternative source is the astrocytic glycogen. In cultured astrocytes, stimulation of receptors coupled to adenylate cyclase leads to glycogenolysis (Sorg and Magistretti, 1991). When 50-mM 8-BrcAMP, a membrane-permeant analog of cAMP, is added to the perfusion medium, there is a rise in dialysate glucose. Isoprenaline, a b-adrenoceptor agonist, produces a similar increase, an effect that is blocked by the b-adrenoceptor antagonist, propranolol. When propranolol is added to the perfusion medium, the late rise in glucose in response to physiological stimulation disappears (Fray et al., 1996). Again, the effect is more striking when glucose changes are measured using an implanted glucose sensor (Fillenz and Lowry, 1998). With this technique, the presence of propranolol not only abolishes the late rise in glucose but also increases and prolongs the initial decrease in glucose (Fig. 2C). In the experiments where a spatial memory task produced a decrease in hippocampal glucose, systemic administration of glucose, although producing no change in baseline glucose, abolished the behaviorally induced decrease (McNay et al., 2000). This suggests that glucose enters a store, which can replace depleted ECF glucose. The use of microdialysis has demonstrated that neuronal activation leads to a relatively large increase in lactate and a much smaller biphasic change in glucose. The measurement of hippocampal glucose and lactate in response of electrical stimulation of the perforant path, using an implanted sensor, shows an early 21% decrease in glucose compared with a 7% decrease in lactate. Analysis of these changes supports the hypothesis that activated neurons use mainly, if not

exclusively, glucose as their metabolic substrate and the accompanying increase in lactate is a by-product of astrocytic glutamate release, which under physiological conditions makes no contribution to the metabolism of the activated neurons.

III. Glucose and lactate under pathological conditions III.A. Hypoxia Changes in glucose and lactate have also been investigated under nonphysiological conditions. One of these conditions is hypoxia. The high rate of brain energy metabolism has an absolute requirement for oxygen and glucose, the metabolic substrates for its energy production. When oxygen in the inspired air is replaced with nitrogen, spontaneous electrical activity in the brain disappears within 20 s and all evoked electrical activity by both physiological and electrical stimulation disappears within 2 min (Noell and Chinn, 1950). Chemical changes are very much slower, one of the earliest being an accumulation of lactate and a slow decrease in ATP (Albaum et al., 1953). A more moderate reduction of inspired oxygen causes an increase in lactate and a moderate increase in the NADH/NAD ratio in the absence of any change in ATP (Bachelard et al., 1974). Measurement of brain pO2 and extracellular glucose during graded reductions in inspired oxygen, measured using glucose selective microelectrodes, showed that with 10% oxygen, there was no change in glucose but a reduction to 5 or 3% for periods of 4 min caused a lowering of brain pO2 to almost zero and a reduction of glucose by 20–80%. When animals were returned to normal levels of oxygen, there was an overshoot of both tissue pO2 and ECF glucose (Silver and Erecinska, 1994). One of the key questions is whether lactate contributes to the negative effects of hypoxia or whether it contributes to recovery after the readmission of oxygen. There have been numerous in vitro studies using acute or cultured brain slices, which have yielded conﬂicting results. In these experiments, synaptic function is assessed by the size

462

of the electrically evoked population spike; brain slices are exposed to various periods of anoxia, and on reoxygenation, the incubation medium contains either glucose or lactate. On the basis of an extensive series of papers, Schurr claims that ‘brain lactate not glucose fuels the recovery of synaptic function from hypoxia upon reoxygenation’ (Schurr et al., 1988a, 1997a, b; Schurr and Rigor, 1998). Cater et al. (2003) also concluded that functional recovery during reoxygenation was greater with lactate than glucose. During reoxygenation following 10 min of hypoxia, 10 mM lactate maintained synaptic function (Takata et al., 2004). These results appeared to imply that neurons preferred lactate to glucose as their metabolic substrate. The conclusion has been challenged by other work. While in some experiments it was found that lactate could maintain neural activity in acute hippocampal slices (Schurr et al., 1988b; Fowler, 1993); in others, incubation in 10-mM lactate produced a gradual decrease in population spikes, which reached a minimum amplitude after 35 min followed by a spontaneous recovery to 80% by 60 min. During the decline in the population spike, there was a rise in intracellular Ca2+ and release of glutamate (Takata et al., 2001). A similar decline in the population spike was observed when glucose was replaced by other glycolytic metabolites (Kanatani et al., 1995; Wada et al., 1997). Both the recovery after transient block of population spikes following replacement of glucose with lactate and the maintenance of the population spike in lactate following hypoxia are dependent on the activation of NMDA receptors and voltage-dependent Ca2+ channels (Takata et al., 2001, 2004). The ﬁnding that in none of these experiments was there a decrease in ATP concentration led to the conclusion that although lactate oxidation leads to ATP production, this is not sufﬁcient to maintain synaptic function. The absence of a reduction in ATP, suggested that it is not the energy level that determines the maintenance of synaptic activity; it appears instead that it is glycolysis that is required. There is a similar preferential role for glycolysis in preventing anoxic depolarization of rat hippocampal pyramidal neurons (Allen et al., 2005).

Synaptic function depends on the maintenance of the ionic electrochemical gradients, which is the function of Na/K ATPase. There is evidence for a membrane-bound coupling of glycolytic enzymes to Na/K ATPase (Mercer and Dunham, 1981). Measurement of the activity of phosphofructokinase, one of the glycolytic enzymes, demonstrated a parallelism between the activity of this enzyme and the maintenance of synaptic activity (Li et al., 2000). It appears that release of glutamate and the resulting inﬂux of Ca2+ ions leads to a switch, which allows lactate to substitute for glucose. This appears to be the explanation for the fact that after hypoxia, lactate substitution for glucose allows the maintenance of synaptic function (Takata et al., 2004). Experiments with brain slices have revealed a number of important features of brain metabolism. However, in these in vitro experiments, there is no circulation and the concentrations of glucose used in the incubation media are well beyond the normal concentration of brain extracellular glucose. It is, therefore, important to test the conclusions of these experiments in vivo. Brain glucose and lactate, using microdialysis, were measured at 30-min intervals, in unanesthetized rats exposed to 7% oxygen for 90 min. As an indication of the severity of the hypoxia, a number of rats died during the exposure to hypoxia and those that survived showed few spontaneous movements and poor responses to stimulation. There was a small but signiﬁcant reduction in dialysate glucose, which returned to basal when air was readmitted. Dialysate lactate increased during hypoxia and returned to basal on readmission of air (Harada et al., 1992). In a later study, simultaneous changes in glucose and lactate in unanesthetized rats were measured at 3-min intervals, using microdialysis and a dual online enzymatic assay. The rats were exposed to a 15-min period of oxygen levels reduced to 8%. At this level of hypoxia, the rats remained quiet and showed no signs of discomfort. The reduction in oxygen led to an immediate rise in lactate; there was an initial steep slope, which gradually decreased. The rise continued throughout the 15-min period, reaching a maximum of 1200% of basal, at the end of the hypoxic period. On the readmission of air, there was a steep

463

Fig. 3. The effect of hypoxia on glucose and lactate in rat cortex. *po0.05 compared with basal. (A) Simultaneous measurement of changes in glucose and lactate when inspired oxygen is reduced to 8% for a period of 15 min. (B) A comparison of the time course of lactate changes in response to tail pinch and hypoxia. (C) The effect of 3-min puffs of nitrogen on tissue oxygen and glucose measured with an oxygen electrode and a glucose sensor.

decline to basal over a period of 20 min after the end of hypoxia. There was no decrease in dialysate glucose during hypoxia, but a delayed increase, which reached a peak of 310% of basal, 9 min after the end of hypoxia (Jones et al., 2000) (Fig. 3A). There are two features of note in these results. One is the rate of decline of the hypoxia-induced rise in lactate compared with the rise in activationinduced rise in lactate (Fig. 3B). While the evidence suggests that the rise in lactate following neuronal activation does not undergo local oxidative phosphorylation and, therefore, does not contribute to neuronal metabolism (see above), the steep decline of lactate on readmission of air after hypoxia suggests that it is rapidly oxidized and contributes to neuronal metabolism. This view is supported by the brain slice experiments. Two features of the glucose changes were surprising: one was the absence of a decrease in

glucose and the other was the delayed rise. Unlike the delayed rise in glucose, which was abolished by adding propranolol to the perfusion medium (Fray et al., 1996), propranolol had no such effect on the late rise in glucose following hypoxia (Jones et al., 2000). As demonstrated in the zero-net-ﬂux experiments, during rapid changes in in vivo recovery, dialysate concentrations are not an accurate reﬂection of changes in extracellular concentration (Fray et al., 1997). This is illustrated by the fact that a brief puff of nitrogen produces a rapid decrease in glucose when measured with an implanted glucose biosensor (Lowry et al., 1998a) (Fig. 3C), which has a 100% in vivo recovery (Lowry et al., 1998c). It seems possible, therefore, that the increased in vivo recovery due to increased glucose turnover during hypoxia masks the simultaneous decrease in extracellular glucose. The rise in glucose after the readmission of air is probably

464

due to a combination of preferential use of lactate and, therefore, reduced glucose utilization combined with a resulting decrease in in vivo recovery.

III.B. Ischemia In ischemia or stroke, which is the loss of blood supply to a brain region, there is a failure of glucose delivery in addition to the failure of oxygen delivery. The animal model for stroke is occlusion of the middle cerebral artery (MCAO). An early observation was a reduction in blood ﬂow, an increase in K+, and a reduction in the amplitude of the EEG (Strong et al., 1983). In later studies, in addition to the reduction in blood ﬂow, changes in pH, NADH glutamate, as well as glucose and lactate were recorded. A picture emerged of a ‘‘core’’ region of severe blood-ﬂow reduction, which progresses to irreversible injury, while the adjacent ‘‘penumbral’’ zone appears to represent an unstable region threatened with possible injury yet potentially amenable to therapeutic intervention (Ginsberg, 1990; Selman et al., 1990). This penumbral region is characterized by an increase in K+ concentration and a spreading depolarization. Propagated ﬂuorescence transients are used to detect and track this spreading depolarization (Strong et al., 1996). This spreading depolarization, termed the peri-infarct depolarization (PID), contrasts with cortical spreading depression (CSD) (Somjen, 2001). While in the latter, there is an increase in blood ﬂow; in PID, there is a reduction in blood ﬂow and a reduction in tissue oxygen. Recent evidence has shown that there is a linear relation between infarct size and the frequency of PIDs (Iijima et al., 1992; Busch et al., 1996). Furthermore, the frequency of ﬂuorescence transients increases with low plasma glucose levels (Strong et al., 2000). The increase in lactate and glutamate, and the decrease in glucose and pyruvate have been conﬁrmed using microdialysis (Yang et al., 2001). In macaque monkeys, the effect of changes in RCBF and cerebral metabolic rates of oxygen (CMRO2) on extracellular glucose and lactate was studied following a 2-h period of MCAO. There was a signiﬁcant correlation between glucose, RCBF, and CMRO2, whereas

lactate concentration was only correlated with CMRO2 (Frykholm et al., 2005). The independent development of two biosensor-based detection techniques with ﬂow-injection analysis has enabled the monitoring of changes in glucose and lactate in near real-time (Kaptein et al., 1998; Jones et al., 2000). Using the endothelin-1 rat model of transient focal cerebral ischemia (Bogaert et al., 2000; Hughes et al., 2003; Virley et al., 2004), the temporal changes in glucose and lactate were monitored at 1-min intervals in unanesthetized rats. There was a rapid increase in lactate while glucose decreased below detection levels. The changes were reversible and varied with the dose of endothelin-1. There was a correlation between lactate increase over a 2-h period and the striatal infarct size (Gramsbergen et al., 2004). In a similar study, the transient changes in cortical glucose and lactate levels associated with peri-infarct depolarization was studied in anesthetized cats following MCAO for a period of 3 h. MCAO caused an increase in lactate within the ﬁrst 20 min and remained at this level. In contrast, glucose showed a time-dependent decrease after MCAO. The incidence of ﬂuorescence transients was inversely related to plasma glucose concentration (Hopwood et al., 2005).

III.C. Glutamate excitotoxicity Ischemia leads to the release of glutamate (Benveniste et al., 1984; Globus et al., 1989; Globus et al., 1990) in concentrations, which are sufﬁcient to kill brieﬂy exposed neurons in culture (Choi, 1988). The release of glutamate is accompanied by large increases in lactate. There has been some debate concerning the contribution of this raised lactate to brain damage (MacMillan and Shankaran, 1984; Shimizu et al., 1993). Experiments with brain slices have suggested that lactate has a neuroprotective effect (Schurr et al., 1999). This question has been investigated in in vivo experiments using microdialysis. Infusion of glutamate, through a microdialysis probe, into the cortex of anesthetized rats produces a decrease in glucose and a rise in lactate. Examination of the resulting brain lesions demonstrated that the

465

Fig. 4. Simultaneous measurement of glucose and lactate in rat cortex in response to an infusion of 100 mM glutamate for 30 min.*po0.05 compared with basal. (A) Glutamate infusion produces a 50% reduction in dialysate glucose. (B) Glutamate infusion produces a rapid and sustained increase in lactate. (C) When a reduction of inspired oxygen to 8% is added to last 15 min of the glutamate infusion, there is a further large increase in lactate, which decreases steeply on the readmission of air.

100 mM glutamate was the minimum concentration that produced consistent lesions (Alessandri et al., 1996). These experiments were repeated in unanesthetized rats using the dual enzymatic assay that allowed simultaneous assays of glucose and lactate at 3-min intervals. Infusion of 100 mM glutamate for a period of 30 min produced no behavioral abnormalities. As soon as the glutamate reached the probe, there was a steep decline in glucose by around 50%; this level was maintained throughout the period of infusion and it did not return to control levels 30 min after the end of the

infusion (Fig. 4A). In contrast, lactate showed a rapid rise to 800% of control and was maintained at a plateau, throughout the period of glutamate infusion (Fig. 3B). When during the last 15 min of the glutamate infusion, inspired oxygen was reduced to 8%, there was an additional large rise in lactate, which reached 3,000% of basal (Fig. 3C). This suggested that during the plateau produced by glutamate infusion alone, lactate oxidation balanced the rate as glutamate release (Ros et al., 2002). Measurement of the lesions in the cortex showed that adding hypoxia to the glutamate

466

infusion produced a signiﬁcant decrease in the size of the lesion (Ros et al., 2002). This suggested the possibility that the additional lactate produced by hypoxia had a neuroprotective effect. This possibility was examined in a separate set of experiments in which 6 mM lactate was added to the perfusion medium. The combination of 100 mM glutamate with 6 mM lactate produced a smaller lesion than 100 mM glutamate alone. The insertion of the dialysis probe produces a small lesion. Perfusion with aCSF produced a larger lesion than perfusion with 6 mM lactate alone. To test the hypothesis that the neuroprotective role of lactate was due to its role as a metabolic substrate, the experiments were repeated with D-lactate, the nonmetabolisable isomer of L-lactate. The lesion produced by perfusion of D-lactate alone was no different from that produced by aCSF; D-lactate added to 100 mM glutamate produced a lesion that was larger than that produced by glutamate alone (Ros et al., 2001). The conclusion from the animal experiments using microdialysis is that under physiological conditions, there is minimal use of lactate as a metabolic substrate under both basal condition and during neuronal activation. Under pathological conditions, when there is a release of glutamate, a decrease in extracellular glucose, and activation of glutamate receptors, there is a switch and lactate becomes the preferred metabolic substrate.

IV. Microdialysis in the clinic Microdialysis developed by Ungerstedt was widely used for monitoring changes in the extracellular compartment of the brain in animals. In 1990, the technique was introduced into the clinic to monitor changes in the human brain (Hillered et al., 1990; Meyerson et al., 1990). The technique has been used to monitor adverse neurochemical changes in cases of subarachnoid hemorrhage, traumatic brain injury, thrombo-embolic stroke, and during neurosurgery. Based on the results from animal experiments, many of the measurements have concentrated on markers for disturbances in glucose metabolism: glucose, lactate, and

pyruvate. Typical patterns observed in early studies of brain injury were a marked increase in the lactate/pyruvate ratio and a reduction to near zero levels of glucose in the microdialysate (Langemann et al., 1995; Persson et al., 1996). The purpose of clinical microdialysis is to elucidate the mechanism of the secondary spread of injury and to develop strategies for preventing it. Clinical microdialysis encounters many problems not present in the application to animal experiments (Hillered et al., 2005). Prolonged monitoring and progressive changes in the brain, such as edema, tissue swelling, and therefore, shrinkage of the extracellular space, lead to changes in in vivo recovery (Boutelle and Fillenz, 1996). A partial compensation for such factors can be achieved by calculating the ratio of two parameters, which are both likely to be inﬂuenced by the underlying changes; examples are the metabolites, glucose and lactate. A change in opposite directions of these compounds suggests that they are not due to changes in in vivo recovery. Another problem for critical-care monitoring is the unknown value for basal levels of glucose and lactate in near-normal human brain. Basal values of 200–400 mM for dialysate lactate have been obtained during frontal lobe resection in tumor patients (Hillered et al., 1990), and values of 250 mM for dialysate lactate and 247 mM for dialysate glucose during operations for aneurysm (Langemann et al., 1995). There have also been attempts to determine the true extracellular concentrations of these compounds. Using the zero-net-ﬂux method of Lo¨nnroth (Lo¨nnroth et al., 1987), the value obtained for glucose was 1.66 mM and for lactate was 2.24 mM (Langemann et al., 2001), whereas the zero-ﬂux method of Jacobson (Jacobson et al., 1985) gave values of 1.57 mM for glucose and 5.10 mM for lactate (Abi-Saab et al., 2002). Another approach adopted to bring dialysate values closer to extracellular values has been to increase probe length to 10 mm and reduce the ﬂow through the probe to 0.3 mL/min. Attempts have been made to correlate the neurochemical changes with other changes of functional signiﬁcance, such as changes in pH and changes in tissue oxygen. In patients with severe traumatic brain injury, microdialysis samples were

467

collected at hourly intervals. During periods of tissue hypoxia, when brain tissue oxygen fell below 10 mmHg, dialysate glucose closely followed the oxygen level, while lactate levels were increased. Dialysate glutamate only increased when oxygen decreased to very low levels (Hlatky et al., 2004). In another study of severely head-injured patients, glucose and lactate were monitored by microdialysis with changes in intracranial pressure (ICP). Two probes were implanted at different distances from the main lesion. All the patients died due to uncontrollable rise in ICP. At the probe in the ‘‘better’’ position, biochemical deterioration occurred after the rise in ICP, whereas in the ‘‘worse’’ probe position, biochemical changes preceded the rise in ICP (Stahl et al., 2001). In spite of a history of 15 years of clinical microdialysis, the difﬁculty of interpreting the changes in brain function that underlie the neurochemical changes has meant that the technique has had only modest effects on clinical decisions. One reason is that the ex vivo assay of enormous numbers of samples is not only very labor intensive but also is open to errors (Hillered et al., 2005). The recent transfer to the clinic of an automated ﬂow-injection assay of microdialysate from cerebral cortex for glucose and lactate at 30-s intervals has transformed the situation. This ‘rapid-sampling microdialysis’ (rsMD) requires no expertise from the nursing staff. Furthermore, the high-time resolution means that brief transient events are recorded (Parkin et al., 2003). In view of the evidence from animal experiments that ischemia is associated with a reduction in EEG activity, attempts have been made to investigate whether such phenomena occur in brain-damaged patients. Electrocorticographic electrodes placed near foci of damaged tissue revealed two kinds of phenomena: one consisted of spontaneous, transient episodes of depressed electrocorticographic activity that propagated across the cortex at velocities reminiscent of cortical spreading depression; the other were transient suppressions of electrocorticographic activity that showed no evidence of spread (Strong et al., 2002). Based on these observations, microdialysis sampling is now combined with four channels of electrocorticogram from a subdural strip adjacent to the catheter. In

addition, intracranial pressure, systemic blood pressure, and arterial oxygen saturation are monitored. With this combination of techniques, a number of different patterns of rapid, brief spontaneous changes in these parameters have been identiﬁed (Parkin et al., 2005). Two important results have so far emerged from this assay system. One is the ability to ‘guide’ the neurosurgeon by monitoring dialysate glucose and restoring levels by i.v. glucose administration when values are unacceptably low (Parkin et al., 2003). The other is the establishment of the ‘metabolic signature’ of the PID (the fall in glucose with the associated rise in lactate) as a sign of PIDs. What is signiﬁcant is the fact that the increase in lactate is twice as large as the decrease in glucose. The authors attribute the increase in lactate to glycolysis (Hopwood et al., 2005). As PIDs are associated with an increase in glutamate, uptake of glutamate released from astrocytes could also make a signiﬁcant contribution (Fillenz, 2005). Fluorescence monitoring of PIDs has so far not been tried in clinical cases; but as their frequency is an index of deterioration in the penumbra (Hopwood et al., 2005), their occurrence could serve as a predictor of clinical outcome. V. Conclusion Microdialysis has proved to be an enormously useful technique, which has led to very important advances in both basic research and clinical investigations. Until recently, implanted enzyme-based sensors had the advantage of a very much higher time resolution. The advent of rsMD has largely removed that advantage. The main difference between the techniques is that microdialysis can simultaneously assay more than one analyte, whereas multiple implanted sensors can monitor simultaneous changes in more than one brain region. Acknowledgments I want to thank Martyn Boutelle for useful comments and Tyra Zetterstrøm who brought microdialysis to my lab in 1986.

468

References Abi-Saab, W.M., Maggs, D.G., Jones, T., Jacob, R., Srihari, V., Thompson, J., Kerr, D., Leone, P., Krystal, J.H., Spencer, D.D., During, M.J. and Sherwin, R.S. (2002) Striking differences in glucose and lactate levels between brain extracellular ﬂuid and plasma in conscious human subjects: effects of hyperglycemia and hypoglycemia. J. Cereb. Blood Flow Metab., 22: 271–279. Albaum, H.G., Noell, W.K. and Chinn, H.I. (1953) Chemical changes in rabbit brain during anoxia. Am. J. Physiol., 174: 408–412. Alessandri, B., Landolt, H., Langemann, H., Gregorin, J., Hall, J. and Gratzl, O. (1996) Application of glutamate in the cortex of rats: a microdialysis study. Acta Neurochir. Suppl., 67: 6–12. Allen, N.J., Karadottir, R. and Attwell, D. (2005) A preferential role for glycolysis in preventing the anoxic depolarization of rat hippocampal area CA1 pyramidal cells. J. Neurosci., 25: 848–859. Bachelard, H.S., Lewis, L.D., Ponten, U. and Siesjo, B.K. (1974) Mechanisms activating glycolysis in the brain in arterial hypoxia. J. Neurochem., 22: 395–401. Benveniste, H., Drejer, J., Schousboe, A. and Diemer, N.H. (1984) Elevation of the extracellular concentrations of glutamate and aspartate in rat hippocampus during transient cerebral ischemia monitored by intracerebral microdialysis. J. Neurochem., 43: 1369–1374. Bogaert, L., Scheller, D., Moonen, J., Sarre, S., Smolders, I., Ebinger, G. and Michotte, Y. (2000) Neurochemical changes and laser Doppler ﬂowmetry in the endothelin-1 rat model for focal cerebral ischemia. Brain Res., 887: 266–275. Boutelle, M.G., Fellows, L.K. and Cook, C. (1992) Enzyme packed bed system for the on-line measurement of glucose, glutamate, and lactate in brain microdialysate. Anal. Chem., 64: 1790–1794. Boutelle, M.G. and Fillenz, M. (1996) Clinical microdialysis: the role of on-line measurement and quantitative microdialysis. Acta Neurochir. Suppl., 67: 13–20. Busch, E., Gyngell, M.L., Eis, M., Hoehn-Berlage, M. and Hossmann, K.A. (1996) Potassium-induced cortical spreading depressions during focal cerebral ischemia in rats: contribution to lesion growth assessed by diffusion-weighted NMR and biochemical imaging. J. Cereb. Blood Flow Metab., 16: 1090–1099. Cater, H.L., Chandratheva, A., Benham, C.D., Morrison, B. and Sundstrom, L.E. (2003) Lactate and glucose as energy substrates during, and after, oxygen deprivation in rat hippocampal acute and cultured slices. J. Neurochem., 87: 1381–1390. Chih, C.P. and Roberts, E.L. Jr. (2003) Energy substrates for neurons during neural activity: a critical review of the astrocyte-neuron lactate shuttle hypothesis. J. Cereb. Blood Flow Metab., 23: 1263–1281. Chih, C.P., Lipton, P. and Roberts, E.L. Jr. (2001) Do active cerebral neurons really use lactate rather than glucose? Trends Neurosci., 24: 573–578.

Choi, D.W. (1988) Glutamate neurotoxicity and diseases of the nervous system. Neuron, 1: 623–634. Cornell-Bell, A.H., Finkbeiner, S.M., Cooper, M.S. and Smith, J.S. (1990) Glutamate induces calcium waves in cultured astrocytes: long-range glial signaling. Science, 247: 470–473. De Bruin, L.A., Schasfoort, E.M., Steffens, A.B. and Korf, J. (1990) Effects of stress and exercise on rat hippocampus and striatum extracellular lactate. Am. J. Physiol., 259: R773–R779. Demestre, M., Boutelle, M. and Fillenz, M. (1997) Stimulated release of lactate in freely moving rats is dependent on the uptake of glutamate. J. Physiol., 499: 825–832. Dienel, G.A. and Cruz, N.F. (2004) Nutrition during brain activation: does cell-to-cell lactate shuttling contribute signiﬁcantly to sweet and sour food for thought? Neurochem. Int., 45: 321–351. Dykstra, K.H., Hsiao, J.K., Morrison, P.F., Bungay, P.M., Mefford, I.N., Scully, M.M. and Dedrick, R.L. (1992) Quantitative examination of tissue concentration proﬁles associated with microdialysis. J. Neurochem., 58: 931–940. Fellows, L.K. and Boutelle, M.G. (1993) Rapid changes in extracellular glucose levels and blood ﬂow in the striatum of the freely moving rat. Brain Res., 604: 225–231. Fellows, L.K., Boutelle, M.G. and Fillenz, M. (1992) Extracellular brain glucose levels reﬂect local neuronal activity: a microdialysis study in awake, freely moving rats. J. Neurochem., 59: 2141–2147. Fellows, L.K., Boutelle, M.G. and Fillenz, M. (1993) Physiological stimulation increases nonoxidative glucose metabolism in the brain of the freely moving rat. J. Neurochem., 60: 1258–1263. Fillenz, M. (2005) The role of lactate in brain metabolism. Neurochem. Int., 47: 413–417. Fillenz, M. and Lowry, J.P. (1998) Studies of the source of glucose in the extracellular compartment of the rat brain. Dev. Neurosci., 20: 365–368. Fowler, J.C. (1993) Glucose deprivation results in a lactate preventable increase in adenosine and depression of synaptic transmission in rat hippocampal slices. J. Neurochem., 60: 572–576. Fox, P.T. and Raichle, M.E. (1986) Focal physiological uncoupling of cerebral blood ﬂow and oxidative metabolism during somatosensory stimulation in human subjects. Proc. Natl. Acad. Sci. U.S.A., 83: 1140–1144. Fox, P.T., Raichle, M.E., Mintun, M.A. and Dence, C. (1988) Nonoxidative glucose consumption during focal physiologic neural activity. Science, 241: 462–464. Fray, A.E., Boutelle, M. and Fillenz, M. (1997) Extracellular glucose turnover in the striatum of unanaesthetized rats measured by quantitative microdialysis. J. Physiol., 504: 721–726. Fray, A.E., Forsyth, R.J., Boutelle, M.G. and Fillenz, M. (1996) The mechanisms controlling physiologically stimulated changes in rat brain glucose and lactate: a microdialysis study. J. Physiol., 496: 49–57.

469 Frykholm, P., Hillered, L., Langstrom, B., Persson, L., Valtysson, J. and Enblad, P. (2005) Relationship between cerebral blood ﬂow and oxygen metabolism, and extracellular glucose and lactate concentrations during middle cerebral artery occlusion and reperfusion: a microdialysis and positron emission tomography study in nonhuman primates. J. Neurosurg., 102: 1076–1084. Ginsberg, M.D. (1990) Local metabolic responses to cerebral ischemia. Cerebrovasc. Brain Metab. Rev., 2: 58–93. Globus, M.Y., Busto, R., Dietrich, W.D., Martinez, E., Valdes, I. and Ginsberg, M.D. (1989) Direct evidence for acute and massive norepinephrine release in the hippocampus during transient ischemia. J. Cereb. Blood Flow Metab., 9: 892–896. Globus, M.Y., Busto, R., Martinez, E., Valdes, I. and Dietrich, W.D. (1990) Ischemia induces release of glutamate in regions spared from histopathologic damage in the rat. Stroke, 21: III43–III46. Gramsbergen, J.B., Skjoth-Rasmussen, J., Rasmussen, C. and Lambertsen, K.L. (2004) On-line monitoring of striatum glucose and lactate in the endothelin-1 rat model of transient focal cerebral ischemia using microdialysis and ﬂow-injection analysis with biosensors. J. Neurosci. Methods, 140: 93–101. Harada, M., Okuda, C., Sawa, T. and Murakami, T. (1992) Cerebral extracellular glucose and lactate concentrations during and after moderate hypoxia in glucose- and salineinfused rats. Anesthesiology, 77: 728–734. Hillered, L., Persson, L., Ponten, U. and Ungerstedt, U. (1990) Neurometabolic monitoring of the ischaemic human brain using microdialysis. Acta Neurochir. (Wien.), 102: 91–97. Hillered, L., Vespa, P.M. and Hovda, D.A. (2005) Translational neurochemical research in acute human brain injury: the current status and potential future for cerebral microdialysis. J. Neurotrauma, 22: 3–41. Hlatky, R., Valadka, A.B., Goodman, J.C., Contant, C.F. and Robertson, C.S. (2004) Patterns of energy substrates during ischemia measured in the brain by microdialysis. J. Neurotrauma, 21: 894–906. Hopwood, S.E., Parkin, M.C., Bezzina, E.L., Boutelle, M.G. and Strong, A.J. (2005) Transient changes in cortical glucose and lactate levels associated with peri-infarct depolarisations, studied with rapid-sampling microdialysis. J. Cereb. Blood Flow Metab., 25: 391–401. Hu, Y. and Wilson, G.S. (1997) Rapid changes in local extracellular rat brain glucose observed with an in vivo glucose sensor. J. Neurochem., 68: 1745–1752. Hughes, P.M., Anthony, D.C., Ruddin, M., Botham, M.S., Rankine, E.L., Sablone, M., Baumann, D., Mir, A.K. and Perry, V.H. (2003) Focal lesions in the rat central nervous system induced by endothelin-1. J. Neuropathol. Exp. Neurol., 62: 1276–1286. Iijima, T., Mies, G. and Hossmann, K.A. (1992) Repeated negative DC deﬂections in rat cortex following middle cerebral artery occlusion are abolished by MK-801: effect on volume of ischemic injury. J. Cereb. Blood Flow Metab., 12: 727–733. Innocenti, B., Parpura, V. and Haydon, P.G. (2000) Imaging extracellular waves of glutamate during calcium signaling in cultured astrocytes. J. Neurosci., 20: 1800–1808.

Jacobson, I., Sandberg, M. and Hamberger, A. (1985) Mass transfer in brain dialysis devices – a new method for the estimation of extracellular amino acids concentration. J. Neurosci. Methods, 15: 263–268. Jeftinija, S.D., Jeftinija, K.V., Stefanovic, G. and Liu, F. (1996) Neuroligand-evoked calcium-dependent release of excitatory amino acids from cultured astrocytes. J. Neurochem., 66: 676–684. Jones, D.A., Ros, J., Landolt, H., Fillenz, M. and Boutelle, M.G. (2000) Dynamic changes in glucose and lactate in the cortex of the freely moving rat monitored using microdialysis. J. Neurochem., 75: 1703–1708. Kanatani, T., Mizuno, K. and Okada, Y. (1995) Effects of glycolytic metabolites on preservation of high energy phosphate level and synaptic transmission in the granule cells of guinea pig hippocampal slices. Experientia, 51: 213–216. Kaptein, W.A., Zwaagstra, J.J., Venema, K. and Korf, J. (1998) Continuous ultraslow microdialysis and ultraﬁltration for subcutaneous sampling as demonstrated by glucose and lactate measurements in rats. Anal. Chem., 70: 4696–4700. Kasischke, K.A., Vishwasrao, H.D., Fisher, P.J., Zipfel, W.R. and Webb, W.W. (2004) Neural activity triggers neuronal oxidative metabolism followed by astrocytic glycolysis. Science, 305: 99–103. Kuhr, W.G. and Korf, J. (1988) Extracellular lactic acid as an indicator of brain metabolism: continuous on-line measurement in conscious, freely moving rats with intrastriatal dialysis. J. Cereb. Blood Flow Metab., 8: 130–137. Langemann, H., Alessandri, B., Mendelowitsch, A., Feuerstein, T., Landolt, H. and Gratzl, O. (2001) Extracellular levels of glucose and lactate measured by quantitative microdialysis in the human brain. Neurol. Res., 23: 531–536. Langemann, H., Mendelowitsch, A., Landolt, H., Alessandri, B. and Gratzl, O. (1995) Experimental and clinical monitoring of glucose by microdialysis. Clin. Neurol. Neurosurg., 97: 149–155. Larrabee, M.G. (1983) Lactate uptake and release in the presence of glucose by sympathetic ganglia of chicken embryos and by neuronal and nonneuronal cultures prepared from these ganglia. J. Neurochem., 40: 1237–1250. Larrabee, M.G. (1995) Lactate metabolism and its effects on glucose metabolism in an excised neural tissue. J. Neurochem., 64: 1734–1741. Larrabee, M.G. (1996) Partitioning of CO2 production between glucose and lactate in excised sympathetic ganglia, with implications for brain. J. Neurochem., 67: 1726–1734. Li, X., Yokono, K. and Okada, Y. (2000) Phosphofructokinase, a glycolytic regulatory enzyme has a crucial role for maintenance of synaptic activity in guinea pig hippocampal slices. Neurosci. Lett., 294: 81–84. Lo¨nnroth, P., Jansson, P.A. and Smith, U. (1987) A microdialysis method allowing characterization of interecellular water space in humans. Am. J. Physiol., 253: 652–654. Lowry, J.P. and Fillenz, M. (1997) Evidence for uncoupling of oxygen and glucose utilization during neuronal activation in rat striatum. J. Physiol., 498(Pt 2): 497–501.

470 Lowry, J.P. and Fillenz, M. (2001) Real-time monitoring of brain energy metabolism in vivo using microelectrochemical sensors: the effects of anesthesia. Bioelectrochemistry, 54: 39–47. Lowry, J.P., Demestre, M. and Fillenz, M. (1998a) Relation between cerebral blood ﬂow and extracellular glucose in rat striatum during mild hypoxia and hyperoxia. Dev. Neurosci., 20: 52–58. Lowry, J.P., Miele, M., O’Neill, R.D., Boutelle, M.G. and Fillenz, M. (1998b) An amperometric glucose-oxidase/ poly(o-phenylenediamine) biosensor for monitoring brain extracellular glucose: in vivo characterisation in the striatum of freely-moving rats. J. Neurosci. Methods, 79: 65–74. Lowry, J.P., O’Neill, R.D., Boutelle, M.G. and Fillenz, M. (1998c) Continuous monitoring of extracellular glucose concentrations in the striatum of freely moving rats with an implanted glucose biosensor. J. Neurochem., 70: 391–396. MacMillan, V. and Shankaran, R. (1984) Inﬂuence of lactate accumulation on Na+, K+-ATPase activity of ischemic and post-ischemic brain. Brain Res., 303: 125–132. Mangia, S., Garreffa, G., Bianciardi, M., Giove, F., Di Salle, F. and Maraviglia, B. (2003a) The aerobic brain: lactate decrease at the onset of neural activity. Neuroscience, 118: 7–10. Mangia, S., Giove, F., Bianciardi, M., Di Salle, F., Garreffa, G. and Maraviglia, B. (2003b) Issues concerning the construction of a metabolic model for neuronal activation. J. Neurosci. Res., 71: 463–467. McIlwain, H. (1953) Substances which support respiration and metabolic response to electrical impulses in human cerebral tissues. J. Neurol. Neurosurg. Psychiatry, 16: 257–266. McIlwain, H. (1955) A transitory, rapid, production of lactate in electrically excited cerebral tissues. Biochem. J., 60: xxxi. McKenna, M.C., Hopkins, I.B. and Carey, A. (2001) Alphacyano-4-hydroxycinnamate decreases both glucose and lactate metabolism in neurons and astrocytes: implications for lactate as an energy substrate for neurons. J. Neurosci. Res., 66: 747–754. McKenna, M.C., Tildon, J.T., Stevenson, J.H., Hopkins, I.B., Huang, X. and Couto, R. (1998) Lactate transport by cortical synaptosomes from adult rat brain: characterization of kinetics and inhibitor speciﬁcity. Dev. Neurosci., 20: 300–309. McNay, E.C. and Gold, P.E. (1999) Extracellular glucose concentrations in the rat hippocampus measured by zero-netﬂux: effects of microdialysis ﬂow rate, strain, and age. J. Neurochem., 72: 785–790. McNay, E.C., Fries, T.M. and Gold, P.E. (2000) Decreases in rat extracellular hippocampal glucose concentration associated with cognitive demand during a spatial task. Proc. Natl. Acad. Sci. U.S.A., 97: 2881–2885. Mercer, R.W. and Dunham, P.B. (1981) Membrane-bound ATP fuels the Na/K pump. Studies on membrane-bound glycolytic enzymes on inside-out vesicles from human red cell membranes. J. Gen. Physiol., 78: 547–568. Meyerson, B.A., Linderoth, B., Karlsson, H. and Ungerstedt, U. (1990) Microdialysis in the human brain: extracellular measurements in the thalamus of parkinsonian patients. Life Sci., 46: 301–308.

Netchiporouk, L., Shram, N., Salvert, D. and Cespuglio, R. (2001) Brain extracellular glucose assessed by voltammetry throughout the rat sleep-wake cycle. Eur. J. Neurosci., 13: 1429–1434. Noell, W. and Chinn, H.I. (1950) Failure of the visual pathway during anoxia. Am. J. Physiol., 161: 573–590. Osborne, P.G., Niwa, O., Kato, T. and Yamamoto, K. (1997) On-line, continuous measurement of extracellular striatal glucose using microdialysis sampling and electrochemical detection. J. Neurosci. Methods, 77: 143–150. Parkin, M., Hopwood, S., Jones, D.A., Hashemi, P., Landolt, H., Fabricius, M., Lauritzen, M., Boutelle, M.G. and Strong, A.J. (2005) Dynamic changes in brain glucose and lactate in pericontusional areas of the human cerebral cortex, monitored with rapid sampling on-line microdialysis: relationship with depolarisation-like events. J. Cereb. Blood Flow Metab., 25: 402–413. Parkin, M., Hopwood, S., Strong, A.J. and Boutelle, M. (2003) Resolving dynamic changes in brain metabolism using biosensors and on-line microdialysis. Trends Anal. Chem., 22: 487–497. Pellerin, L. and Magistretti, P.J. (1994) Glutamate uptake into astrocytes stimulates aerobic glycolysis: a mechanism coupling neuronal activity to glucose utilization. Proc. Natl. Acad. Sci. U.S.A., 91: 10625–10629. Pellerin, L. and Magistretti, P.J. (1996) Excitatory amino acids stimulate aerobic glycolysis in astrocytes via an activation of the Na+/K+ ATPase. Dev. Neurosci., 18: 336–342. Pellerin, L. and Magistretti, P.J. (2003) Food for thought: challenging the dogmas. J. Cereb. Blood Flow Metab., 23: 1282–1286. Pellerin, L., Pellegri, G., Bittar, P.G., Charnay, Y., Bouras, C., Martin, J.L., Stella, N. and Magistretti, P.J. (1998) Evidence supporting the existence of an activity-dependent astrocyteneuron lactate shuttle. Dev. Neurosci., 20: 291–299. Persson, L., Valtysson, J., Enblad, P., Warme, P.E., Cesarini, K., Lewen, A. and Hillered, L. (1996) Neurochemical monitoring using intracerebral microdialysis in patients with subarachnoid hemorrhage [see comments]. J. Neurosurg., 84: 606–616. Peters, J.L., Miner, L.H., Michael, A.C. and Sesack, S.R. (2004) Ultrastructure at carbon ﬁber microelectrode implantation sites after acute voltammetric measurements in the striatum of anesthetized rats. J. Neurosci. Methods, 137: 9–23. Prichard, J., Rothman, D., Novotny, E., Petroff, O., Kuwabara, T., Avison, M., Howseman, A., Hanstock, C. and Shulman, R. (1991) Lactate rise detected by 1H NMR in human visual cortex during physiologic stimulation. Proc. Natl. Acad. Sci. U.S.A., 88: 5829–5831. Ros, J., Jones, D., Pecinska, N., Alessandri, B., Boutelle, M., Landolt, H. and Fillenz, M. (2002) Glutamate infusion coupled with hypoxia has a neuroprotective effect in the rat. J. Neurosci. Methods, 119: 129–133. Ros, J., Pecinska, N., Alessandri, B., Landolt, H. and Fillenz, M. (2001) Lactate reduces glutamate-induced neurotoxicity in rat cortex. J. Neurosci. Res., 66: 790–794.

471 Roy, C.S. and Sherrington, C.S. (1890) On the regulation of the blood supply of the brain. J. Physiol. (Lond.), 11: 85–108. Schasfoort, E.M., De Bruin, L.A. and Korf, J. (1988) Mild stress stimulates rat hippocampal glucose utilization transiently via NMDA receptors, as assessed by lactography. Brain Res., 475: 58–63. Schurr, A. and Rigor, B.M. (1998) Brain anaerobic lactate production: a suicide note or a survival kit? Dev. Neurosci., 20: 348–357. Schurr, A., Dong, W.Q., Reid, K.H., West, C.A. and Rigor, B.M. (1988a) Lactic acidosis and recovery of neuronal function following cerebral hypoxia in vitro. Brain Res., 438: 311–314. Schurr, A., Miller, J.J., Payne, R.S. and Rigor, B.M. (1999) An increase in lactate output by brain tissue serves to meet the energy needs of glutamate-activated neurons. J. Neurosci., 19: 34–39. Schurr, A., Payne, R.S., Miller, J.J. and Rigor, B.M. (1997a) Brain lactate is an obligatory aerobic energy substrate for functional recovery after hypoxia: further in vitro validation. J. Neurochem., 69: 423–426. Schurr, A., Payne, R.S., Miller, J.J. and Rigor, B.M. (1997b) Brain lactate, not glucose, fuels the recovery of synaptic function from hypoxia upon reoxygenation: an in vitro study. Brain Res., 744: 105–111. Schurr, A., West, C.A. and Rigor, B.M. (1988b) Lactatesupported synaptic function in the rat hippocampal slice preparation. Science, 240: 1326–1328. Selman, W.R., Ricci, A.J., Crumrine, R.C., LaManna, J.C., Ratcheson, R.A. and Lust, W.D. (1990) The evolution of focal ischemic damage: a metabolic analysis. Metab. Brain Dis., 5: 33–44. Shimizu, H., Graham, S.H., Chang, L.H., Mintorovitch, J., James, T.L., Faden, A.I. and Weinstein, P.R. (1993) Relationship between extracellular neurotransmitter amino acids and energy metabolism during cerebral ischemia in rats monitored by microdialysis and in vivo magnetic resonance spectroscopy. Brain Res., 605: 33–42. Silver, I.A. and Erecinska, M. (1994) Extracellular glucose concentration in mammalian brain: continuous monitoring of changes during increased neuronal activity and upon limitation in oxygen supply in normo-, hypo-, and hyperglycemic animals. J. Neurosci., 14: 5068–5076. Somjen, G.G. (2001) Mechanisms of spreading depression and hypoxic spreading depression-like depolarization. Physiol. Rev., 81: 1065–1096. Sorg, O. and Magistretti, P.J. (1991) Characterization of the glycogenolysis elicited by vasoactive intestinal peptide, noradrenaline and adenosine in primary cultures of mouse cerebral cortical astrocytes. Brain Res., 563: 227–233. Stahl, N., Mellergard, P., Hallstrom, A., Ungerstedt, U. and Nordstrom, C.H. (2001) Intracerebral microdialysis and

bedside biochemical analysis in patients with fatal traumatic brain lesions. Acta Anaesthesiol. Scand., 45: 977–985. Strong, A.J., Fabricius, M., Boutelle, M.G., Hibbins, S.J., Hopwood, S.E., Jones, R., Parkin, M.C. and Lauritzen, M. (2002) Spreading and synchronous depressions of cortical activity in acutely injured human brain. Stroke, 33: 2738–2743. Strong, A.J., Harland, S.P., Meldrum, B.S. and Whittington, D.J. (1996) The use of in vivo ﬂuorescence image sequences to indicate the occurrence and propagation of transient focal depolarizations in cerebral ischemia. J. Cereb. Blood Flow Metab., 16: 367–377. Strong, A.J., Smith, S.E., Whittington, D.J., Meldrum, B.S., Parsons, A.A., Krupinski, J., Hunter, A.J., Patel, S. and Robertson, C. (2000) Factors inﬂuencing the frequency of ﬂuorescence transients as markers of peri-infarct depolarizations in focal cerebral ischemia. Stroke, 31: 214–222. Strong, A.J., Venables, G.S. and Gibson, G. (1983) The cortical ischaemic penumbra associated with occlusion of the middle cerebral artery in the cat: 1. Topography of changes in blood ﬂow, potassium ion activity, and EEG. J. Cereb. Blood Flow Metab., 3: 86–96. Takata, T., Sakurai, T., Yang, B., Yokono, K. and Okada, Y. (2001) Effect of lactate on the synaptic potential, energy metabolism, calcium homeostasis and extracellular glutamate concentration in the dentate gyrus of the hippocampus from guinea-pig. Neuroscience, 104: 371–378. Takata, T., Yang, B., Sakurai, T., Okada, Y. and Yokono, K. (2004) Glycolysis regulates the induction of lactate utilization for synaptic potentials after hypoxia in the granule cell of guinea pig hippocampus. Neurosci. Res., 50: 467–474. van der Kuil, J.H. and Korf, J. (1991) On-line monitoring of extracellular brain glucose using microdialysis and a NADPH-linked enzymatic assay. J. Neurochem., 57: 648–654. Virley, D., Hadingham, S.J., Roberts, J.C., Farnﬁeld, B., Elliott, H., Whelan, G., Golder, J., David, C., Parsons, A.A. and Hunter, AJ. (2004) A new primate model of focal stroke: endothelin-1-induced middle cerebral artery occlusion and reperfusion in the common marmoset. J. Cereb. Blood Flow Metab., 24: 24–41. Wada, H., Okada, Y., Nabetani, M. and Nakamura, H. (1997) The effects of lactate and beta-hydroxybutyrate on the energy metabolism and neural activity of hippocampal slices from adult and immature rat. Brain Res. Dev. Brain Res., 101: 1–7. Yang, D.Y., Tsai, T.H., Cheng, C.H., Lee, C.W., Chen, S.H. and Cheng, F.C. (2001) Simultaneous monitoring of extracellular glucose, pyruvate, lactate and glutamate in gerbil cortex during focal cerebral ischemia by dual probe microdialysis. J. Chromatogr. A., 913: 349–354.

CHAPTER 5.5

Microdialysis in pain research Carl-Olav Stiller1, Ernst Brodin2 and Bradley K. Taylor3, 1 Department of Medicine, Karolinska University Hospital, Stockholm, Sweden Department of Physiology and Pharmacology, Karolinska Institutet, Stockholm, Sweden 3 Department of Pharmacology, Tulane University Health Sciences Center, New Orleans, LA, USA 2

Abstract: In vivo microdialysis has aided preclinical pain research for more than a decade. This valuable tool has allowed correlations between behavioral signs of pain and CNS neurotransmitter release. Here, we review studies that describe the effects of noxious stimulation on neurotransmitter concentrations in microdialysate obtained from either spinal cord or pain modulatory centers in the brain. We conclude that such studies are currently evolving from models of acute pain to a more modern approach that assesses release in animal models of lasting painful conditions.

and in preceding chapters of this book, we here focus on microdialysis experiments in animal models of pain. We have organized our review by the speciﬁc region of dialysis: spinal cord (SC), thalamus, nucleus accumbens (nAcc), anterior cingulate cortex (ACC), periaqueductal gray (PAG), and rostral ventral medulla (RVM). We conclude with a tabular summary of the literature, organized with respect to pain model.

I. Introduction In vivo microdialysis has become the method of choice for estimating neurotransmitter release in a number of animal pain models. Microdialysis allows a correlation between behavioral signs of nociception and extracellular neurotransmitter levels in pain control centers. Such studies have yielded important insight into the mechanisms of presynaptic release that contribute to pain modulation. Most investigators have designed their studies to evaluate the effects of acute inﬂammatory stimulation on neurotransmitter release. To better model chronic pain, however, investigators are increasingly using microdialysis to evaluate release in models of peripheral nerve injury, diabetic neuropathy, arthritis, stroke, and cancer. The pain associated with these diseases is often resistant to the currently available analgesics, and questions regarding neurotransmitter release are critical for the identiﬁcation of novel pharmacological targets. As the technical aspects of microdialysis have been covered in our previous review (Stiller et al., 2003)

II. Microdialysis in the spinal cord following noxious stimulation Noxious somatosensory information from peripheral tissues is relayed by primary afferent sensory neurons to the dorsal horn of the SC. Spinal pain transmission neuron then relays these signals to higher brain centers such as the thalamus. These SC neurons are controlled by the intensity and coding of peripheral input, spinal interneurons, and descending signals from the brain (Millan, 2002; Almeida et al., 2004). The descending bulbospinal pathways were classically viewed as inhibitory, but recent evidence strongly implicates a

Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

473

DOI: 10.1016/S1569-7339(06)16025-7 Copyright 2007 Elsevier B.V. All rights reserved

474

contribution of supraspinal sites to descending facilitation (Gebhart, 2004; Vanegas and Schaible, 2004). Most microdialysis studies in pain research have focused on the modulation of neurotransmitter release in the SC, particularly at the lumbar level after unilateral stimulation of the hind paw. Three key approaches are typically used to collect microdialysate from the SC. The most widely used approach places a microdialysis ﬁber within the rat dorsal SC using a transverse approach (Sorkin et al., 1988). Although most commonly used in anesthetized rats, this method can be applied in awake animals, albeit with considerable difﬁculty (Gustafsson et al., 1999; Matos et al., 1992; Muth-Selbach et al., 1999; Zhang et al., 1999; Zahn et al., 2002). Although this method lacks spatial precision, postmortem histological examination can exclude animals with a poorly placed dialysis membrane, for example, in the ventral horn. A second approach to microdialysate collection introduces a conventional concentric microdialysis probe through laminae I–V, in the dorsal-ventral plane. An oblique, 451–601angle in the rostro-caudal orientation can be used to maximize the tissue volume of microdialysis (Stiller et al., 1996b). Although this approach is typically employed in decerebrated cats (Linderoth et al., 1992) or anaesthetized rats (Linderoth et al., 1994; Sundstro¨m et al., 1995; Dmitrieva et al., 2004), thus precluding correlations between behavior and release (Sundstro¨m et al., 1995; Dmitrieva et al., 2004), a similar technique can be used in awake rats (Gerin et al., 1994; Gerin and Privat, 1998). A third approach to obtain spinal microdialysate involves an intrathecal microdialysis loop catheter, introduced through the atlanto-occipital membrane and pushed to the lumbar level (Marsala et al., 1995). Because the catheter rests within the cerebrospinal ﬂuid on the dorsal surface of the SC, spinal tissue is not traumatized as in the other methods. This represents an important advantage. Furthermore, addition of an injection cannula to the loop assembly enables direct injection of pharmacological agents into the intrathecal space. Finally, compared to direct spinal superfusion (Yaksh and Tyce, 1980), intrathecal loop dialysis avoids sample contamination by

high-molecular-weight proteins that may interfere with the detection of transmitters by HPLC or RIA. On the downside, the loop method does not directly estimate parenchymal release, but rather estimates overﬂow into the cerebrospinal ﬂuid. As illustrated in Tables 1–3, spinal microdialysis can assess the extracellular concentration of numerous neurotransmitters following acute, inﬂammatory, and neuropathic conditions. Of these, most attention has been paid to the amino acids glutamate and GABA, the neuropeptides substance P and cholecystokinin (CCK), and the prostaglandin PGE2. A common ﬁnding is that noxious stimuli, either in the setting of acute or chronic pain, increases the concentration of these substances in spinal microdialysate. For example, substance P release in the SC can be induced by formalin (Calcutt et al., 2000) or by a high-intensity titanic stimulation of the sciatic nerve (Afrah et al., 2002). III. Microdialysis in brain regions of importance for pain transmission and modulation Microdialysis offers a powerful approach toward a better understanding of the neurochemical mechanisms of pain transmission and analgesia. The following section, summarized in Table 2, focuses on the most commonly studied brain areas. III.A. Brain areas contributing to ascending transmission of pain signals: thalamic nuclei Microdialysis has aided our understanding of the neurochemical pathways mediating the transmission of nociceptive messages from the SC to the thalamus. For example, intraplantar injection of formalin increased glutamate and aspartate in the ventral posterolateral thalamic nucleus (Silva et al., 2001), while analgesic doses of morphine decrease the levels of glutamate (Abarca et al., 2000). These data support the hypothesis that glutamatergic mechanisms in this area of the thalamus mediates the transmission of ongoing noxious inﬂammatory stimuli, a process that is subject to inhibition by opioid analgesics. Whether glutamatergic thalamic mechanisms also contribute to

475 Table 1. Acute pain models studied by microdialysis in the rat Model

CNS region

Transmitter

Reference

Plantar incision Intramuscular injection of acidic saline Hind paw pinch or saline injection Spinal capsaicin, spinal PGE2

Dorsal horn Dorsal horn

Asp, Glu, Ser, Asn, Gln, Gly Glu, Asp

Zahn et al. (2002) Skyba et al. (2005)

PAG Nucleus accumbens Intrathecal dialysis

Silva et al. (2000) Rouge´-Pont et al. (1998) Malmberg et al. (1995)

Hot water tail ﬂick Low- or high-frequency TNS Mustard oil

Intrathecal dialysis Dorsal horn (transverse probe) Spinal trigeminal nucleus

Glu, Arg, Asp Dopamine Glu, Asp, Tau, Gly, GABA, ethanolamine Asp, Glu Asp, Glu Glu, Asp, Ser, Gly

Electrical stimulation of A-beta or small A-delta myelinated ﬁbers Electrical high intensity stimulation of A-delta or C ﬁbers Tetanic stimulation of the sciatic nerve Hind paw pinch

Dorsal horn

5-HIAA, DOPAC, HVA

Wen et al. (2004) Sluka et al. (2005) Bereiter and Benetti (1996); Bereiter et al. (2002) Men and Matsui (1994a)

Dorsal horn

Noradrenaline

Men and Matsui (1994b)

Dorsal horn

SP

Afrah et al. (2002)

Dorsal horn Dorsal horn Intrathecal microdialysis

Glu, Asp, Arg, GABA 5-HT, noradrenaline PGE2

Dmitrieva et al. (2004) Lisi et al. (2003) Milne et al. (2001)

Dorsal horn – transverse probe

Glu, PGE2, NO

Formalin

Dorsal horn – transverse probe

PGE2

Formalin Formalin Formalin Formalin

Dorsal horn – transverse probe Dorsal horn – transverse probe Intrathecal dialysis Intrathecal dialysis

Glu, Asp Asp, Glu, Asn, Gly, Tau PGE2, Glu, Asp, Tau PGE2, Asp, Glu

Formalin Formalin Formalin

Intrathecal dialysis Intrathecal dialysis Intrathecal dialysis

Formalin Formalin 5%/10% Formalin Formalin Formalin Formalin Formalin Formalin Formalin Formalin Formalin Formalin Formalin Formalin

Intrathecal dialysis Intrathecal ldialysis Intrathecal dialysis Intrathecal dialysis Trigeminal nucleus PAG PAG PAG PAG RVM Accumbens Locus coeruleus Arcuate nucleus VPL

Glu, nitrite/nitrate Glu Glu, Asp, Tau, Gly, Cit, Ser, Asn, Gln, PGE2 Glu, Asp, Gly, Tau, Ser, PGE2 Glu Substance P PGE2 GABA Glu, Arg, Asp GABA, Glu Gln, Gly Anandamide 5-HT Met-enkephalin-LI, CCK-LI NA Beta-endorphin Glu, Asp, Arg

Scheuren et al. (1997); Vetter et al. (2001) Muth-Selbach et al. (1999); Geisslinger et al. (2000); Tegeder et al. (2001) Skilling et al. (1988) Skilling et al. (1990) Hua et al. (1997) Malmberg and Yaksh (1995a); Dirig et al. (1997) Watanabe et al. (2003) Buerkle et al. (1998) Malmberg and Yaksh (1995b); Marsala et al. (1995) Malmberg et al. (1997) Okuda et al. (2001) Calcutt et al. (2000) Freshwater et al. (2002) Viggiano et al. (2004) Silva et al. (2000) Maione et al. (1999) Maione et al. (2000) Walker et al. (1999) Taylor and Basbaum (1995) Lapeyre et al. (2001) Sajedianfard et al. (2005) Zangen et al. (1998) Silva et al. (2001)

Intrathecal strychnine+hair deﬂection Formalin

Note: Glu, glutamate; Asp, aspartate; Gln, glutamine; Asn, asparagine; Ser, serine; Cit, citrulline; Arg, arginine; Thr, threonine; PGE2, prostaglandin E2; PAG, periaqueductal gray; VPL, ventral posterolateral thalamus; RVM, rostral ventromedial medulla; TNS, transcutaneous nerve stimulation.

476 Table 2. Acute and sustained peripheral inﬂammation and microdialysis Inﬂammation model treatment

Region

Neurotransmitter

Reference

Carragenan hind paw Carragenan hind paw Carragenan hind paw, naloxone (i.c.v.) Carragenan hind paw, pinching of hind paw Carragenan hind paw Carrageenan Kaolin/carragenan, knee joint Kaolin/carragenan, knee joint Kaolin/carragenan, knee joint Kaolin/carragenan, knee joint Kaolin/carragenan, knee joint CFA

Dorsal horn Dorsal horn Dorsal horn

CCK-LI PGE2 5-HT, 5-HIAA

de Araujo Lucas et al. (1998) Nakayama et al. (2002) Zhang et al. (2000)

Dorsal horn

Glu, Asp, Arg, GABA

Dmitieva et al. (2004)

PAG ACC Dorsal horn

5-HT, 5-HIAA CCK-LI Glu, Asp, Gln

Zhang et al. (2000) Erel et al. (2004) Sluka and Westlund (1993)

Dorsal horn

Glu, Asp, Gln, Ser, Gly

Sluka et al. (1994)

Dorsal horn, monkey

Glu, Asp, Gly, Ser, Gln, Tau

Sorkin et al. (1992)

Intrathecal dialysis

Glu, Asp, Cit, PGE2

Yang et al. (1996)

Intrathecal dialysis

PGE2, 6-ketoPGF1a

Dirig and Yaksh (1999)

PAG

GABA, NT met-enkephalin

Turpentine induced oral inﬂammation

Trigeminal ganglion

Substance P

Williams et al. (1995); Renno and Beitz (1999) Neubert et al. (2000)

Note: NT, neurotensin; ACC, anterior cingulate cortex.

Table 3. Nerve injury and microdialysis (rat) Nerve injury

Region

Neurotransmitter

Reference

Axotomy of the sciatic nerve Axotomy of the sciatic nerve Axotomy of the sciatic nerve Axotomy of the sciatic nerve Spinal nerve ligation L5, L6 brushing Partial sciatic nerve ligation, SCS Partial sciatic nerve ligation Partial sciatic nerve ligation Loose ligation of inferior alveolar nerve

Dorsal horn Dorsal horn Dorsal horn Cingulate cortex Intrathecal dialysis Dorsal horn Dorsal horn Intrathecal dialysis Trigeminal nucleus caudalis

CCK-LI SP-LI CCK-LI CCK-LI PGE2, Glu GABA, glutamate SP-LI Glu, Asp Glu, Asp

Gustafsson et al. (1998) Wallin and Scho¨tt (2002) Afrah et al. (2001) Gustafsson et al. (2000) Hefferan et al. (2003) Stiller et al. (1996a); Cui et al. (1997) Wallin and Scho¨tt (2002) Skilling et al. (1992) Fujita et al. (2004)

Note: CFA, complete Freunds adjuvans; SCS, spinal cord stimulation.

the supraspinal transmission of pain signals induced by peripheral nerve injury is not clear. However, in a rat model of neuropathic pain (tight ligation of the L5 and L6 spinal nerves) a decreased 5-HT release was detected in the contralateral ventrobasal (VB) thalamus (Goettl et al., 2002). The authors suggest that a decreased 5-HT release may result in a decreased descending inhibition.

III.B. Brain areas contributing to the affective component of pain: nucleus accumbens and anterior cingulate cortex Microdialysis also provides important clues as to the neurochemical mechanisms underlying the modulation of pain signals by brain areas controlling affect. In the nAcc, dopamine is an important

477

pain modulator, as suggested by pharmacological studies demonstrating that dopaminergic agonist injection into the nAcc reduce formalin-induced nociception (Taylor et al., 2003), and by the ﬁnding that injection of a dopamine antagonist into the nAcc blocks pain-induced analgesia (Altier and Stewart, 1999). On the basis of these latter studies, Schmidt et al. (2002) evaluated the effects of noxious intraplantar capsaicin administration on accumbal dopamine concentrations and pain-induced antinociception. They found that capsaicin increased dialysate dopamine, and suggested the provocative hypothesis that dopamine release in the nAcc yields antinociception. Furthermore, the behavioral signs of pain triggered by formalin injection are accompanied by an immediate and sustained increase of met-enkephalin and CCK within the nAcc (Lapeyre et al., 2001). CCK injection into nAcc antagonizes the antinociceptive effect of morphine. Furthermore, CCK release in this region was prevented by administration of the opioid receptor antagonist naloxone. These data suggest that noxious formalin increases accumbal enkephalins, which most likely results in an antinociceptive effect. Activation of opioid receptors in this region elicits the local release of CCK, which counteracts analgesia. Human imaging studies convincingly demonstrate that painful stimulation activates the ACC (Casey et al., 1994; Hsieh et al., 1995). Neuronal activity within the ACC appears to be closely related to the subjective experience of pain unpleasantness (Rainville et al., 1997). Nociceptive neurons have also been found in the rat ACC (Yamamura et al., 1996), and we found that both peripheral nerve damage (Gustafsson et al., 2000) and monoarthrithis (Erel et al., 2004) increased the release of CCK in the rat ACC. Such data suggest that increased CCK release in is the ACC contributes to the affective component of pain.

III.C. Brain areas contributing to the descending modulation of pain: periaqueductal gray and rostral ventral medulla The PAG in the midbrain and RVM in the brainstem coordinate the descending modulation of

nociceptive signaling While electrophysiological, pharmacological, and anatomical studies have yielded a tremendous amount of information describing the pathway from the PAG to the RVM to the dorsal horn (Basbaum and Fields, 1984; Urban and Gebhart, 1999; Ossipov et al., 2000), questions regarding the neurotransmitters released in this pathway require techniques such as microdialysis. We organize our review of the key studies in the literature into those that describe changes in release associated with opioid analgesia or noxious stimulation.

III.C.1. Opioid analgesia Systemic morphine decreased the veratridine-induced release of GABA in the lateral, but not medial, PAG (Renno et al., 1992). Local administration of morphine into the PAG also decreased the local release of GABA (Stiller et al., 1996a). In control rats, systemic administration of morphine, but not saline, released histamine in the PAG (Barke and Hough, 1992). Taylor and Basbaum (2003) found that intravenous morphine produced thermal antinociception and increased RVM dialysate 5-HT in a naloxone-reversible manner. These studies suggest that 5-HT release in the RVM, and GABA and histamine release in the PAG, contributes to opioid analgesia.

III.C.2. Noxious stimulation Peripheral inﬂammation induced by injection of complete Freund’s adjuvant (CFA) into the hind paw signiﬁcantly decreased GABA release (Renno and Beitz, 1999) and increased neurotensin and met-enkephalin release in the PAG (Williams et al., 1995). Formalin (but not handling or saline injection controls) increased glutamate, arginine, and aspartate concentration in PAG dialysates (Silva et al., 2000). Taken together, these studies suggest that GABA, neurotensin, met-enkephalin, and excitatory amino acids contribute to nociceptive modulation by regulating the outﬂow of PAG neurons. Likewise, intraplantar formalin increased dialysate 5-HT in RVM, suggesting that this neurotransmitter contributes to nociceptive modulation by regulating the outﬂow of

478

the rostral ventromedial medulla neurons (Taylor and Basbaum, 1995). IV. Conclusion In vivo microdialysis is now an established method for monitoring the extracellular level of neurotransmitters in pain-related regions of the CNS. Several regions of pain transmission and pain modulation have been studied with regard to the release of amino acids, monoamines, neuropeptides, and prostanoids, largely in the context of acute nociceptive pain models. More recently, however, microdialysis has begun to be used to elucidate the neurochemical mechanisms underlying chronic pain conditions. A dialog between the pain clinician and the basic pain scientist may result in the design of microdialysis studies that will lead to a new pharmacotherapy for clinical pain conditions. References Abarca, C., Silva, E., Sepu´lveda, M.J., Oliva, P. and Contreras, E. (2000) Neurochemical changes after morphine, dizocilpine or riluzole in the ventral posterolateral thalamic nuclei of rats with hyperalgesia. Eur. J. Pharmacol., 403: 67–74. Afrah, A.W., Fiska˚, A., Gjerstad, J., Gustafsson, H., Tjølsen, A., Olgart, L., Stiller, C.O., Hole, K. and Brodin, E. (2002) Spinal substance P release in vivo during the induction of long-term potentiation in dorsal horn neurons. Pain, 96: 49–55. Afrah, A.W., Gustafsson, H., Olgart, L., Brodin, E. and Stiller, C.O. (2001) Changes in spinal cholecystokinin release after peripheral axotomy. Neuroreport, 12: 49–52. Almeida, T.F., Roizenblatt, S. and Tuﬁk, S. (2004) Afferent pain pathways: a neuroanatomical review. Brain Res., 1000: 40–56. Altier, N. and Stewart, J. (1999) The role of dopamine in the nucleus accumbens in analgesia. Life Sci., 65: 2269–2287. Barke, K.E. and Hough, L.B. (1992) Morphine-induced increases of extracellular histamine levels in the periaqueductal grey in vivo: a microdialysis study. Brain Res., 572: 146–153. Basbaum, A.I. and Fields, H.L. (1984) Endogenous pain control systems: brainstem spinal pathways and endorphin circuitry. Annu. Rev. Neurosci., 7: 309–338. Bereiter, D.A. and Benetti, A.P. (1996) Excitatory amino release within spinal trigeminal nucleus after mustard oil injection into the temporomandibular joint region of the rat. Pain, 67: 451–459. Bereiter, D.A., Shen, S. and Benetti, A.P. (2002) Sex differences in amino acid release from rostral trigeminal subnucleus

caudalis after acute injury to the TMJ region. Pain, 98: 89–99. Buerkle, H., Marsala, M. and Yaksh, T.L. (1998) Effect of continuous spinal remifentanil infusion on behaviour and spinal glutamate release evoked by subcutaneous formalin in the rat. Br. J. Anaesth., 80: 348–353. Calcutt, N.A., Stiller, C.O., Gustafsson, H. and Malmberg, A.B. (2000) Elevated substance-P-like immunoreactivity levels in spinal dialysates during the formalin test in normal and diabetic rats. Brain Res., 856: 20–27. Casey, K.L., Minoshima, S., Berger, K.L., Koeppe, R.A., Morrow, T.J. and Frey, K.A. (1994) Positron emission tomographic analysis of cerebral structures activated speciﬁcally by repetitive noxious heat stimuli. J. Neurophysiol., 71: 802–807. Cui, J.G., O’Connor, W.T., Ungerstedt, U., Linderoth, B. and Meyerson, B.A. (1997) Spinal cord stimulation attenuates augmented dorsal horn release of excitatory amino acids in mononeuropathy via a GABAergic mechanism. Pain, 73: 87–95. de Araujo Lucas, G., Alster, P., Brodin, E. and WiesenfeldHallin, Z. (1998) Differential release of cholecystokinin by morphine in rat spinal cord. Neurosci. Lett., 245: 13–16. Dirig, D.M., Konin, G.P., Isakson, P.C. and Yaksh, T.L. (1997) Effect of spinal cyclooxygenase inhibitors in rat using the formalin test and in vitro prostaglandin E2 release. Eur. J. Pharmacol., 331: 155–160. Dirig, D.M. and Yaksh, T.L. (1999) Spinal synthesis and release of prostanoids after peripheral injury and inﬂammation. Adv. Exp. Med. Biol., 469: 401–408. Dmitrieva, N., Rodrı´ guez-Malaver, A.J., Pe´rez, J. and Herna´ndez, L. (2004) Differential release of neurotransmitters from superﬁcial and deep layers of the dorsal horn in response to acute noxious stimulation and inﬂammation of the rat paw. Eur. J. Pain, 8: 245–252. Erel, U., Arborelius, L. and Brodin, E. (2004) Increased cholecystokinin release in the rat anterior cingulate cortex during carrageenan-induced arthritis. Brain Res., 1022: 39–46. Freshwater, J.D., Svensson, C.I., Malmberg, A.B. and Calcutt, N.A. (2002) Elevated spinal cyclooxygenase and prostaglandin release during hyperalgesia in diabetic rats. Diabetes, 51: 2249–2255. Fujita, T., Kamisaki, Y. and Yonehara, N. (2004) Nitric oxideinduced increase of excitatory amino acid levels in the trigeminal nucleus caudalis of the rat with tactile hypersensitivity evoked by the loose-ligation of the inferior alveolar nerves. J. Neurochem., 91: 558–567. Gebhart, G.F. (2004) Descending modulation of pain. Neurosci. Biobehav. Rev., 27: 729–737. Geisslinger, G., Muth-Selbach, U., Coste, O., Vetter, G., Schro¨dter, A., Schaible, H.G., Brune, K. and Tegeder, I. (2000) Inhibition of noxious stimulus-induced spinal prostaglandin E2 release by ﬂurbiprofen enantiomers: a microdialysis study. J. Neurochem., 74: 2094–2100. Gerin, C., Legrand, A. and Privat, A. (1994) Study of 5-HT release with a chronically implanted microdialysis probe in

479 the ventral horn of the spinal cord of unrestrained rats during exercise on a treadmill. J. Neurosci. Methods, 52: 129–141. Gerin, C. and Privat, A. (1998) Direct evidence for the link between monoaminergic descending pathways and motor activity: II. A study with microdialysis probes implanted in the ventral horn of the spinal cord. Brain Res., 794: 169–173. Goettl, V.M., Huang, Y., Hackshaw, K.V. and Stephens, R.L. (2002) Reduced basal release of serotonin from the ventrobasal thalamus of the rat in a model of neuropathic pain. Pain, 99: 359–366. Gustafsson, H., de Arau´jo Lucas, G., Scho¨tt, E., Stiller, C.O., Alster, P., Wiesenfeld-Hallin, Z. and Brodin, E. (1998) Peripheral axotomy inﬂuences the in vivo release of cholecystokinin in the spinal cord dorsal horn-possible involvement of cholecystokinin-B receptors. Brain Res., 790: 141–150. Gustafsson, H., Lucas, G., Schott, E., Stiller, C.O., Alster, P., Wiesenfeld-Hallin, Z. and Brodin, E. (1999) Measurement of cholecystokinin release in vivo in the rat spinal dorsal horn. Brain Research Protocols, 4: 192–200. Gustafsson, H., Stiller, C.O. and Brodin, E. (2000) Peripheral axotomy increases cholecystokinin release in the rat anterior cingulate cortex. Neuroreport, 11: 3345–3348. Hefferan, M.P., Carter, P., Haley, M. and Loomis, C.W. (2003) Spinal nerve injury activates prostaglandin synthesis in the spinal cord that contributes to early maintenance of tactile allodynia. Pain, 101: 139–147. Hsieh, J.C., Belfrage, M., Stone-Elander, S., Hansson, P. and Ingvar, M. (1995) Central representation of chronic ongoing neuropathic pain studied by positron emission tomography. Pain, 63: 225–236. Hua, X.Y., Calcutt, N.A. and Malmberg, A.B. (1997) Neonatal capsaicin treatment abolishes formalin-induced spinal PGE2 release. Neuroreport, 8: 2325–2329. Lapeyre, S., Mauborgne, A., Becker, C., Benoliel, J.J., Cesselin, F., Hamon, M. and Bourgoin, S. (2001) Subcutaneous formalin enhances outﬂow of met-enkephalin- and cholecystokinin-like materials in the rat nucleus accumbens. Naunyn Schmiedebergs Arch. Pharmacol., 363: 399–406. Linderoth, B., Gazelius, B., Franck, J. and Brodin, E. (1992) Dorsal column stimulation induces release of serotonin and substance P in the cat dorsal horn. Neurosurgery, 31: 289–296. Linderoth, B., Stiller, C.O., Gunasekera, L., O’Connor, W.T., Ungerstedt, U. and Brodin, E. (1994) Gamma-aminobutyric acid is released in the dorsal horn by electrical spinal cord stimulation: an in vivo microdialysis study in the rat. Neurosurgery, 34: 484–488. Lisi, T.L., Westlund, K.N. and Sluka, K.A. (2003) Comparison of microdialysis and push-pull perfusion for retrieval of serotonin and norepinephrine in the spinal cord dorsal horn. J. Neurosci. Methods, 126: 187–194. Maione, S., Marabese, I., Oliva, P., de Novellis, V., Stella, L., Rossi, F., Filippelli, A. and Rossi, F. (1999) Periaqueductal gray matter glutamate and GABA decrease following subcutaneous formalin injection in rat. Neuroreport, 10(7): 1403–1407. Maione, S., Marabese, I., Rossi, F., Berrino, L., Palazzo, E. and Trabace, L. (2000) Effects of persistent nociception on

periaqueductal gray glycine release. Neuroscience, 97: 311–316. Malmberg, A.B. and Yaksh, T.L. (1995a) The effect of morphine on formalin-evoked behaviour and spinal release of excitatory amino acids and prostaglandin E2 using microdialysis in conscious rats. Br. J. Pharmacol., 114: 1069–1075. Malmberg, A.B. and Yaksh, T.L. (1995b) Cyclooxygenase inhibition and the spinal release of prostaglandin E2 and amino acids evoked by paw formalin injection: a microdialysis study in unanesthetized rats. J. Neurosci., 15: 2768–2776. Malmberg, A.B., Hamberger, A. and Hedner, T. (1995) Effects of prostaglandin E2 and capsaicin on behavior and cerebrospinal ﬂuid amino acid concentrations of unanesthetized rats: a microdialysis study. J. Neurochem., 65: 2185–2193. Malmberg, A.B., Hedner, T., Fallgren, B. and Calcutt, N.A. (1997) The effect of alpha-trinositol (D-myo-inositol 1,2,6trisphosphate) on formalin-evoked spinal amino acid and prostaglandin E2 levels. Brain Res., 747: 160–164. Marsala, M., Malmberg, A.B. and Yaksh, T.L. (1995) The spinal loop dialysis catheter: characterization of use in the unanesthetized rat. J. Neurosci. Methods, 62: 43–53. Matos, F.F., Rollema, H., Brown, J.L. and Basbaum, A.I. (1992) Do opioids evoke the release of serotonin in the spinal cord? An in vivo microdialysis study of the regulation of extracellular serotonin in the rat. Pain, 48: 439–447. Men, D.S. and Matsui, Y. (1994a) Peripheral nerve stimulation increases serotonin and dopamine metabolites in rat spinal cord. Brain Res. Bull., 33: 625–632. Men, D.S. and Matsui, Y. (1994b) Activation of descending noradrenergic system by peripheral nerve stimulation. Brain Res. Bull., 34: 177–182. Millan, M.J. (2002) Descending control of pain. Prog. Neurobiol., 66: 355–474. Milne, B., Hall, S.R., Sullivan, M.E. and Loomis, C. (2001) The release of spinal prostaglandin E2 and the effect of nitric oxide synthetase inhibition during strychnine-induced allodynia. Anesth. Analg., 93: 728–733. Muth-Selbach, U.S., Tegeder, I., Brune, K. and Geisslinger, G. (1999) Acetaminophen inhibits spinal prostaglandin E2 release after peripheral noxious stimulation. Anesthesiology, 91: 231–239. Nakayama, Y., Omote, K. and Namiki, A. (2002) Role of prostaglandin receptor EP1 in the spinal dorsal horn in carrageenan-induced inﬂammatory pain. Anesthesiology, 97: 1254–1262. Neubert, J.K., Maidment, N.T., Matsuka, Y., Adelson, D.W., Kruger, L. and Spigelman, I. (2000) Inﬂammation-induced changes in primary afferent-evoked release of substance P within trigeminal ganglia in vivo. Brain Res., 871: 181–191. Okuda, K., Sakurada, C., Takahashi, M., Yamada, T. and Sakurada, T. (2001) Characterization of nociceptive responses and spinal releases of nitric oxide metabolites and glutamate evoked by different concentrations of formalin in rats. Pain, 92: 107–115. Ossipov, M.H., Lai, J., Malan, T.P. and Porreca, F. (2000) Spinal and supraspinal mechanisms of neuropathic pain. Ann. N.Y. Acad. Sci., 909: 12–24.

480 Rainville, P., Duncan, G.H., Price, D.D., Carrier, B. and Bushnell, M.C. (1997) Pain affect encoded in human anterior cingulate but not somatosensory cortex. Science, 277: 968–971. Renno, W.M. and Beitz, A.J. (1999) Peripheral inﬂammation is associated with decreased veratridine-induced release of GABA in the rat ventrocaudal periaqueductal gray: microdialysis study. J. Neurol. Sci., 163: 105–110. Renno, W.M., Mullett, M.A. and Beitz, A.J. (1992) Systemic morphine reduces GABA release in the lateral but not the medial portion of the midbrain periaqueductal gray of the rat. Brain Res., 594: 221–232. Rouge´-Pont, F., Deroche, V., Le Moal, M. and Piazza, P.V. (1998) Individual differences in stress-induced dopamine release in the nucleus accumbens are inﬂuenced by corticosterone. Eur. J. Neurosci., 10: 3903–3907. Sajedianfard, J., Khatami, S., Semnanian, S., Naghdi, N. and Jorjani, M. (2005) In vivo measurement of noradrenaline in the locus coeruleus of rats during the formalin test: a microdialysis study. Eur. J. Pharmacol., 512: 153–156. Scheuren, N., Neupert, W., Ionac, M., Neuhuber, W., Brune, K. and Geisslinger, G. (1997) Peripheral noxious stimulation releases spinal PGE2 during the ﬁrst phase in the formalin assay of the rat. Life Sci., 60: 295–300. Schmidt, B.L., Tambeli, C.H., Barletta, J., Luo, L., Green, P., Levine, J.D. and Gear, R.W. (2002) Altered nucleus accumbens circuitry mediates pain-induced antinociception in morphine-tolerant rats. J. Neurosci., 22: 6773–6780. Silva, E., Hernandez, L., Contreras, Q., Guerrero, F. and Alba, G. (2000) Noxious stimulation increases glutamate and arginine in the periaqueductal gray matter in rats: a microdialysis study. Pain, 87: 131–135. Silva, E., Quin˜ones, B., Freund, N., Gonzalez, L.E. and Hernandez, L. (2001) Extracellular glutamate, aspartate and arginine increase in the ventral posterolateral thalamic nucleus during nociceptive stimulation. Brain Res., 923: 45–49. Skilling, S.R., Harkness, D.H. and Larson, A.A. (1992) Experimental peripheral neuropathy decreases the dose of substance P required to increase excitatory amino acid release in the CSF of the rat spinal cord. Neurosci. Lett., 139: 92–96. Skilling, S.R., Smullin, D.H. and Larson, A.A. (1990) Differential effects of C- and N-terminal substance P metabolites on the release of amino acid neurotransmitters from the spinal cord: potential role in nociception. J. Neurosci., 10: 1309–1318. Skilling, S.R., Smullin, D.H., Beitz, A.J. and Larson, A.A. (1988) Extracellular amino acid concentrations in the dorsal spinal cord of freely moving rats following veratridine and nociceptive stimulation. J. Neurochem., 51: 127–132. Skyba, D.A., Lisi, T.L. and Sluka, K.A. (2005) Excitatory amino acid concentrations increase in the spinal cord dorsal horn after repeated intramuscular injection of acidic saline. Pain, 119: 142–149. Sluka, K.A. and Westlund, K.N. (1993) Spinal cord amino acid release and content in an arthritis model: the effects of pretreatment with non-NMDA, NMDA, and NK1 receptor antagonists. Brain Res., 627: 89–103.

Sluka, K.A., Vance, C.G. and Lisi, T.L. (2005) High-frequency, but not low-frequency, transcutaneous electrical nerve stimulation reduces aspartate and glutamate release in the spinal cord dorsal horn. J. Neurochem., 95: 1794–1801. Sluka, K.A., Willis, W.D. and Westlund, K.N. (1994) Inﬂammation-induced release of excitatory amino acids is prevented by spinal administration of a GABAA but not by a GABAB receptor antagonist in rats. J. Pharmacol. Exp. Ther., 271: 76–82. Sorkin, L.S., Steinman, J.L., Hughes, M.G., Willis, W.D. and McAdoo, D.J. (1988) Microdialysis recovery of serotonin released in spinal cord dorsal horn. J. Neurosci. Methods, 23: 131–138. Sorkin, L.S., Westlund, K.N., Sluka, K.A., Dougherty, P.M. and Willis, W.D. (1992) Neural changes in acute arthritis in monkeys. IV. Time-course of amino acid release into the lumbar dorsal horn. Brain Res. Brain Res. Rev., 17: 39–50. Stiller, C.O., Bergquist, J., Beck, O., Ekman, R. and Brodin, E. (1996a) Local administration of morphine decreases the extracellular level of GABA in the periaqueductal gray matter of freely moving rats. Neurosci. Lett., 209: 165–168. Stiller, C.O., Cui, J.G., O’Connor, W.T., Brodin, E., Meyerson, B.A. and Linderoth, B. (1996b) Release of gamma-aminobutyric acid in the dorsal horn and suppression of tactile allodynia by spinal cord stimulation in mononeuropathic rats. Neurosurgery, 39: 367–374. Stiller, C.O., Taylor, B.K., Linderoth, B., Gustafsson, H., Warsame Afrah, A. and Brodin, E. (2003) Microdialysis in pain research. Adv. Drug Deliv. Rev., 55: 1065–1079. Sundstro¨m, E., Mo, L.L. and Seiger, A. (1995) In vivo studies on NMDA-evoked release of amino acids in the rat spinal cord. Neurochem. Int., 27: 185–193. Taylor, B.K. and Basbaum, A.I. (1995) Neurochemical characterization of extracellular serotonin in the rostral ventromedial medulla and its modulation by noxious stimuli. J. Neurochem., 65: 578–589. Taylor, B.K. and Basbaum, A.I. (2003) Systemic morphineinduced release of serotonin in the rostroventral medulla is not mimicked by morphine microinjection into the periaqueductal gray. J. Neurochem., 86: 1129–1141. Taylor, B.K., Joshi, C. and Uppal, H. (2003) Stimulation of dopamine D2 receptors in the nucleus accumbens inhibits inﬂammatory pain. Brain Res., 987: 135–143. Tegeder, I., Niederberger, E., Vetter, G., Bra¨utigam, L. and Geisslinger, G. (2001) Effects of selective COX-1 and -2 inhibition on formalin-evoked nociceptive behaviour and prostaglandin E(2) release in the spinal cord. J. Neurochem., 79: 777–786. Urban, M.O. and Gebhart, G.F. (1999) Supraspinal contributions to hyperalgesia. Proc. Natl. Acad. Sci. U.S.A., 96: 7687–7692. Vanegas, H. and Schaible, H.G. (2004) Descending control of persistent pain: inhibitory or facilitatory? Brain Res. Brain Res. Rev., 46: 295–309. Vetter, G., Geisslinger, G. and Tegeder, I. (2001) Release of glutamate, nitric oxide and prostaglandin E2 and metabolic

481 activity in the spinal cord of rats following peripheral nociceptive stimulation. Pain, 92: 213–218. Viggiano, A., Monda, M., Viggiano, A., Chiefari, M., Aurilio, C. and De Luca, B. (2004) Evidence that GABAergic neurons in the spinal trigeminal nucleus are involved in the transmission of inﬂammatory pain in the rat: a microdialysis and pharmacological study. Eur. J. Pharmacol., 496: 87–92. Walker, J.M., Huang, S.M., Strangman, N.M., Tsou, K. and San˜udo-Pen˜a, M.C. (1999) Pain modulation by release of the endogenous cannabinoid anandamide. Proc. Natl. Acad. Sci. U.S.A., 96: 12198–12203. Wallin, J. and Scho¨tt, E. (2002) Substance P release in the spinal dorsal horn following peripheral nerve injury. Neuropeptides, 36: 252–256. Watanabe, C., Okuda, K., Sakurada, C., Ando, R., Sakurada, T. and Sakurada, S. (2003) Evidence that nitric oxide-glutamate cascade modulates spinal antinociceptive effect of morphine: a behavioural and microdialysis study in rats. Brain Res., 990: 77–86. Wen, Z.H., Chang, Y.C., Cherng, C.H., Wang, J.J., Tao, P.L. and Wong, C.S. (2004) Increasing of intrathecal CSF excitatory amino acids concentration following morphine challenge in morphine-tolerant rats. Brain Res., 995: 253–259. Williams, F.G., Mullet, M.A. and Beitz, A.J. (1995) Basal release of Met-enkephalin and neurotensin in the ventrolateral periaqueductal gray matter of the rat: a microdialysis study of antinociceptive circuits. Brain Res., 690: 207–216.

Yaksh, T.L. and Tyce, G.M. (1980) Resting and K+-evoked release of serotonin and norephinephrine in vivo from the rat and cat spinal cord. Brain Res., 192: 133–146. Yamamura, H., Iwata, K., Tsuboi, Y., Toda, K., Kitajima, K., Shimizu, N., Nomura, H., Hibiya, J., Fujita, S. and Sumino, R. (1996) Morphological and electrophysiological properties of ACCx nociceptive neurons in rats. Brain Res., 735: 83–92. Yang, L.C., Marsala, M. and Yaksh, T.L. (1996) Characterization of time course of spinal amino acids, citrulline and PGE2 release after carrageenan/kaolin-induced knee joint inﬂammation: a chronic microdialysis study. Pain, 67: 345–354. Zahn, P.K., Sluka, K.A. and Brennan, T.J. (2002) Excitatory amino acid release in the spinal cord caused by plantar incision in the rat. Pain, 100: 65–76. Zangen, A., Herzberg, U., Vogel, Z. and Yadid, G. (1998) Nociceptive stimulus induces release of endogenous beta-endorphin in the rat brain. Neuroscience, 85: 659–662. Zhang, C., Davies, M.F., Guo, T.Z. and Maze, M. (1999) The analgesic action of nitrous oxide is dependent on the release of norepinephrine in the dorsal horn of the spinal cord. Anesthesiology, 91: 1401–1407. Zhang, Y.Q., Gao, X., Zhang, L.M. and Wu, G.C. (2000) The release of serotonin in rat spinal dorsal horn and periaqueductal gray following carrageenan inﬂammation. Neuroreport, 11: 3539–3543.

CHAPTER 6.1

The role of microdialysis in drug discovery: focus on antipsychotic agents M.J. Millan, F. Panayi, J.M. Rivet, B. Di Cara, L. Cistarelli, R. Billiras, S. Girardon and A. Gobert Department of Psychopharmacology, Institut de Recherches Servier, Croissy sur Seine, Paris, France

Abstract: Since its inception some two decades ago, microdialysis has rapidly assumed a crucial role as an interface between cellular models of drug actions in vitro and studies of their behavioural effects in vivo. Microdialysis provides invaluable information regarding the mechanisms of action of psychotropic agents and their inﬂuence upon endogenous modulators implicated in the aetiology and treatment of CNS disorders. In addition, measures of extracellular levels of speciﬁc neurotransmitters are complementary to behavioural parameters in the characterisation of experimental models of psychiatric and neurological diseases. Further, microdialysis can be used for quantiﬁcation of levels of psychotropic agents themselves in speciﬁc brain regions. Though microdialysis techniques are increasingly being applied to a broad variety of cellular mediators, most studies have to date focussed on monoamines, acetylcholine and amino acids like glycine, glutamate and GABA; that is, neurotransmitters strongly implicated in the pathogenesis and control of depression, anxiety, schizophrenia and other psychiatric states. Recent years have seen substantial improvements in the sensitivity of systems for their detection, which now permit, for example, the simultaneous quantiﬁcation of dopamine (DA), serotonin (5-HT) and noradrenaline (NA) levels in single dialysis samples; quantiﬁcation of acetylcholine (ACh) levels in the absence of acetylcholinesterase inhibitors; and determination of glutamate and GABA levels concurrently with glycine and a miscellany of related amino acids. The present chapter provides an overview of how microdialysis can be applied to the discovery and evaluation of centrally active drugs. Furthermore, it specifically focuses on the application of novel microdialysis techniques to the characterisation of antipsychotic agents for the improved treatment of schizophrenia.

Kloft, 2005). This period has also witnessed the extension of microdialysis studies from rodents to primates, a diversiﬁcation of techniques employed for analysis, and dramatic improvements in sensitivity, precision and kinetics of measurements (Chaurasia, 1999; Bradberry, 2000; Kennedy et al., 2002; Parrot et al., 2003; Powell and Ewing, 2005; Uutela et al., 2005; Zhang and Beyer, 2006). Hence, it has become feasible to apply microdialysis measures to a plethora of endogenous mediators including the intracellular messenger, cGMP (Pepicelli et al., 2004), cytokines and growth

I. Introduction: aims of review One of the most striking and productive developments in neuroscience over the past two decades has been the increasingly broad use of microdialysis in the characterisation of psychotropic agents (Di Chiara, 1990; Ungerstedt, 1991; Westerink, 1995; Di Chiara et al., 1996; Khan and Shuaib, 2001; Bourne, 2003; Fillenz, 2005; Plock and Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

485

DOI: 10.1016/S1569-7339(06)16026-9 Copyright 2007 Elsevier B.V. All rights reserved

486

factors (Clough, 2005), neurosteroids (Serra et al., 2002), corticosterone (Linthorst et al., 1995; De Groote et al., 2005), neuropeptides like neurotensin and galanin (Frankel et al., 2005; Hilke et al., 2005) and histamine (Westerink et al., 2002). However, a clear majority of studies till date have been devoted to the significance of monoamines, acetylcholine (ACh) and amino acids like GABA, glutamate and glycine in the aetiology and treatment of psychiatric and neurological disorders (Millan et al., 2000b; Westerink, 2002; Parrot et al., 2003; Fillenz, 2005; Millan, 2005a; see citations below). The present manuscript highlights the importance of microdialysis in the characterisation of novel psychotropic drugs. It focuses, in particular, on the application of microdialysis to psychotic states and their control by antipsychotic agents. A related point of emphasis is the frontal cortex (FCX) in that a deregulation of this structure is strongly implicated in the mood, cognitive and motor deficits of schizophrenia and other CNS disorders (Meltzer et al., 2003; Millan, 2003, 2005a, b, 2006; Laruelle et al., 2005; Sarter et al., 2005). Advances in the application of microdialysis to drug discovery are also exempliﬁed by three complementary and innovative detection systems for the determination of monoamines, ACh, glycine and other amino acids in freely moving rats. Finally, before embarking on a detailed discussion of antipsychotic agents, several general principles concerning the use of microdialysis in the discovery and characterisation of centrally active drugs are summarised.

II. The role of microdialysis in the characterisation of psychotropic agents II.A. Measures of CNS levels of drugs For all classes of psychotropic agent, irrespective of their mechanism of action and therapeutic indication, it is crucial to establish that they actually enter and act in the CNS. Furthermore, for drugs possessing active metabolites with significant half-lives, such as the antidepressants, ﬂuoxetine (norfluoxetine) and imipramine (desimipramine),

and the antipsychotics, clozapine (N-desmethylclozapine (NDMC)) and risperidone (9-OH-risperidone), it is also useful to document the concentrations of metabolites in the CNS (Aravagiri et al., 1998; Ejsing et al., 2005). Despite the utility of in vitro cellular models of blood-brain-barrier (BBB) penetration (Van der Sandt et al., 2001; Lundquist et al., 2002; El Ela et al., 2004; Garberg et al., 2005; Kusch-Poddar et al., 2005; Maine et al., 2005; Josserand et al., 2006) and in silico models (Vayer et al., 2004; Goodwin and Clark, 2005; Allen and Geldenhuys, 2006), direct knowledge of drug presence and activity in the brain is indispensable. In parallel with behavioural parameters and, if possible, imaging studies (Shah and Marsden, 2004; Pomper and Lee, 2005; Sossi and Ruth, 2005; Steward et al., 2005) microdialysis yields direct, quantitative and kinetic information on the cerebral levels of drugs in discrete CNS regions as compared with the circulation (Boschi and Scherrmann, 2000; Hammarlund-Udenaes, 2000; Sawchuk and Elmquist, 2000; De Lange et al., 2005). Though drug levels in dialysates do not correspond to those actually present in extracellular compartments of the brain, several calculations have been proposed to reﬁne their estimation (Deguchi, 2002). Microdialysis-derived information on cerebral levels of drugs in speciﬁc brain regions is informative as numerous factors modify their central availability. These include (1), lipophilicity, amphiphilicity, molecular weight, polarity and plasma protein binding; (2), passive and active inﬂux and efﬂux across the BBB; (3), diffusion through the cerebrospinal ﬂuid, uptake in neurones and other cell types; (4), local degradation and (5), potency and kinetics of association to/dissociation from speciﬁc proteins, notably receptors and other targets (Deguchi, 2002; De Lange and Danhof, 2002; Partridge, 2003). As concerns multiple classes of transporters, which actively encourage or deny entry into the brain (Kusuhara and Sugiyama, 2001a, b; Partridge, 2003; Chan et al., 2004), certain classes of transporters may facilitate the access of drugs, for example, amino acid transporters assist entry of the transmitter and potential antipsychotic adjunct, glycine (Tamai and Tsuji, 2000) (vide infra). However, from the point of view of

487

drug discovery, of inﬁnitely greater importance are the several ATP-driven pumps, which actively exclude and extrude drugs from the brain. The most notorious and prominent mediator of drug efflux is the ‘‘Multidrug-Resistance-Gene’’ encoded P-glycoprotein (P-gp), a 170-kDa glycoprotein of the ‘‘ATP-binding cassette’’ (ABC) superfamily of membrane transporters (CordonCardo et al., 1989; Kusuhara and Sugiyama, 2001a, b; Stouch and Gudmundsson, 2002). Its ‘‘MDR1’’ isoform is localised in endothelial cells, astrocytes and microglia and is enriched in the BBB as well as in the gut (Rao et al., 1999; Watanabe et al., 2002, 2004; Mercier et al., 2004; Ronaldson, 2004; Takano et al., 2006). P-gp exerts a profound (and sometimes unforeseen) inﬂuence upon brain levels of many agents necessitating their determination by microdialysis wherever possible. Though rejection of some peripherally active drugs may delimit unwanted central side effects, curtailing of BBB access is a serious problem for psychotropic drugs designed to treat central disorders, in particular, in view of functionally distinct polymorphisms of P-gp in man (Eichelbaum et al., 2004). Major substrates of P-gp include the antidepressants, ﬂuvoxamine, paroxetine and citalopram (5-HT reuptake inhibitors), nortriptyline (a NA reuptake inhibitor) (Uhr and Grauer, 2003; El Ela et al., 2004; Ejsing et al., 2005), GR205,171 (a selective neurokinin1 receptor antagonist) (Smith et al., 2001; Rupniak et al., 2003) and ﬂesinoxan (a 5-HT1A receptor agonist) (Van der Sandt et al., 2001). As regards antipsychotics, whereas haloperidol and clozapine are not particularly susceptible to elimination by P-gp (Boulton et al., 2002; Maine et al., 2005), good substrates include the clozapine metabolite, NDMC (El Ela et al., 2004), the benzamides, amisulpride and sulpiride (Watanabe et al., 2002; Ha¨rtter et al., 2003; El Ela et al., 2004; Schmitt et al., 2006), and the atypical agents, risperidone (Boulton et al., 2002; Yasui-Furukori et al., 2004), quetiapine (Boulton et al., 2002) and olanzapine (Boulton et al., 2002; El Ela et al., 2004; Wang et al., 2004). Accordingly, systematic comparisons of peripheral vs. central concentrations of drugs in rats are of considerable interest (Wang et al., 2004; Ejsing et al., 2005; Schmitt et al., 2006).

In addition, in order to clarify the precise inﬂuence of P-gp upon cerebral concentrations of drugs, it is possible to examine the effects of speciﬁc P-gp inhibitors in rats, and to characterise differences between wild-type mice and conspecifics with genetically deleted P-gp (Boschi and Scherrmann, 2000; Deguchi, 2002; Potschka et al., 2002; Uhr and Grauer, 2003; De Lange et al., 2005). In addition to the fundamental question of whether a drug enters the brain per se, more subtle issues can be addressed with microdialysis, for example, information on the absolute concentrations of a drug in the brain relative to its afﬁnities for various classes of receptor help reﬁne assessments of its likely actions – both anticipated and unsuspected – at speciﬁc classes of receptor and other proteins. Further, central concentrations of drugs at speciﬁc doses can be related to effects in pharmacological models predictive of therapeutic properties as compared with undesirable side effects. Microdialysis determination of central drug levels can, thus, be instructive in deﬁning the molecular mechanisms of action of centrally active agents (see further below). In addition, dual-probe microdialysis allows for concomitant measurement of drug levels at two cerebral sites. This can be important where preferential actions in speciﬁc regions are desirable as compared with others, for example, for an antidepressant agent, an inhibition of 5-HT and NA reuptake in the FCX and hippocampus favourably inﬂuences mood whereas actions in other regions, such as the hypothalamus or brainstem may be less desirable (Mir and Taylor, 1997; Pacher and Kecskemeti, 2004). As concerns D2 receptor blockade by antipsychotics, preferential actions at mesolimbic vs. striatal populations should optimise the window for therapeutic (‘‘anti-positive’’) vs. undesirable (extrapyramidal) side effects. For example, the antipsychotic, amisulpride, has been claimed to possess a predominantly limbic proﬁle of dopaminergic actions, and direct dialysis measures of its concentrations in the nucleus accumbens vs. the striatum would provide insights into the validity of this hypothesis and its mechanistic bases (Perrault et al., 1997; Schoemaker et al., 1997; Mo¨ller, 2003). Finally, a related complication is that certain (psychotropic and other) drugs, whether substrates

488

or not, inhibit P-gp; examples include the antidepressants, nefazodone, sertraline and paroxetine (5-HT reuptake inhibitors) (Stormer et al., 2001; Weiss et al., 2003) and the antipsychotics, quetiapine, haloperidol and clozapine (El Ela et al., 2004; Maine et al., 2005; Weber et al., 2005). Microdialysis determination of the inﬂuence of such agents upon P-gp substrates with which they are likely to be co-administered are, then, advisable. To summarise, in parallel with cellular models of BBB passage and P-gp activity, microdialysis provides an excellent approach to determine the access, levels and distribution of novel drugs and their metabolites in the brain, as well as their inﬂuence upon other agents with which they may be co-administered.

II.B. Characterisation of mechanisms of drug actions Results acquired in cellular models of drug activity cannot be automatically extrapolated to the brain as many variables, such as coupling patterns to G-proteins and other intracellular signals, can differ markedly to those seen at pure populations of cloned receptors. Microdialysis offers an attractive and robust approach for characterisation of the actions of drugs in the brain at deﬁned populations of receptors and other targets. At the cellular level, the utility of microdialysis is exempliﬁed by measures of cGMP, arginine and citrulline, providing an indication of nitric oxide (NO) synthase activity in vivo (Pepicelli et al., 2004; Kodama and Koyama, 2006). Among innumerable examples of neurotransmitter studies, perhaps the most compelling are those of the effects of drugs at autoreceptors localised on monoaminergic and other classes of cerebral neurones (Gobert el al., 1998; Millan et al., 2000b; Blier, 2001; Celada et al., 2004). First, highly sensitive 5-HT1A autoreceptors on serotonergic perikarya exert an inhibitory inﬂuence upon the activity of ascending serotonergic pathways that is easily revealed in dialysis studies of 5-HT release in the FCX and other structures (Gardier et al., 1996; Gobert el al., 1998; Millan et al., 2000b; Celada et al., 2004). Thus, measures

of 5-HT release are ideal for characterising the actions of 5-HT1A antagonists and partial agonists, potential pro-cognitive and anxiolytic agents (Gobert et al., 1995; Schechter et al., 2002; Millan, 2003; Millan et al., 2004b; Luttgen et al., 2005). Down-regulation of 5-HT1A autoreceptors may be necessary for expression of full efﬁcacy of antidepressant agents and this can also be revealed upon their long-term administration (Gardier et al., 1996; Blier, 2001; Celada et al., 2004; Millan, 2006). Second, as likewise shown by dialysis studies, presynaptic a2-adrenoceptors (AR)s on the cell bodies and terminals of adrenergic pathways exert a marked tonic, inhibitory inﬂuence upon adrenergic transmission, while a2-heteroceptors exert a phasic inhibitory control of serotonergic neurones: accordingly, the progressive down-regulation of a2-auto- and a2-hetero-receptors by long-term exposure to antidepressants correlates with increased monoamine release and the gradual onset of antidepressant properties (Gobert et al., 1998; Linner et al., 1999; Millan et al., 2000b; Invernizzi and Garattini, 2004; Millan, 2006). Dialysis studies are also useful in demonstrating that antidepressants possessing antagonist properties at a2-ARs, such as mirtazapine, rapidly enhance the cerebral release of NA (Millan et al., 2000a; Invernizzi and Garattini, 2004). In addition, dialysis work suggests that combining a2-AR blockade with NA/5-HT reuptake suppression may yield highly effective and rapidly acting antidepressant agents (Millan et al., 2000b, 2001; Invernizzi and Garattini, 2004; Andres et al., 2005; Millan, 2006). Third, as pointed out below, dopaminergic pathways bear tonically active, inhibitory D2/D3 autoreceptors on their terminals and cell bodies; their activation provides a useful measure of agonist and antagonist properties at D2/D3 receptors in the brain. Activation of presynaptic D2/D3 sites by antiparkinson agents likely underlies their neuroprotective properties, and provides a ‘‘surrogate’’ measure of their activation of postsynaptic D2/D3 receptors, recruitment of which underlies relief of motor dysfunction (Millan et al., 2004a; Presgraves et al., 2004; Goetz et al., 2005; Van Kampen and Robertson, 2005). Evaluation of the actions of antipsychotic agents at nigrostriatal D2/D3 autoreceptors (enhanced striatal DA release) also

489

provides an indirect measure of their propensity to elicit extrapyramidal motor symptoms, though such effects are generally offset by other receptor-binding properties (Millan, 2000; Kapur and Mamo, 2003; Abi-Dargham and Laruelle, 2005). Finally, by analogy, dialysis approaches are useful for exploring the inﬂuence of drugs acting at histamine H3 and glutamatergic (metabotropic) receptors inhibitory to the release of histamine and glutamate, respectively (Schoepp, 2001; Westerink et al., 2002; Lamberty et al., 2003). These observations sufﬁce to illustrate the use of dialysis in establishing that mechanisms of drugs actions seen in vitro are indeed expressed in the brain. II.C. Prediction of the clinical actions of drugs The above comments lead on to a further important and broad use of microdialysis; that is, in exploring the potential therapeutic and undesirable actions of drugs. For example, as implied above, microdialysis has been ubiquitously applied to show that all classes of currently employed antidepressant reinforce corticolimbic serotonergic, adrenergic and/or dopaminergic transmission – with their precise patterns of effect depending upon their mechanism of action (Millan et al., 2000b; Blier, 2001; Celada et al., 2004; Millan, 2006). Even for novel antidepressants that do not directly interact with monoaminergic receptors, reuptake sites or catabolic enzymes, analysis of their inﬂuence upon monoaminergic transmission would appear obligatory (Millan, 2004). In addition, microdialysis offers the possibility of examining their effects upon other modulators implicated in the pathogenesis and control of depressed states, such as NO, cytokines and neuropeptides (see Section I). As concerns anxiolytic agents, the discovery and development of 5-HT1A partial agonists, like buspirone, was partially founded on their agonist actions at 5-HT1A autoreceptors, which translate, as pointed out above, into a reduction in dialysis levels of 5-HT. In a clinical perspective, this effect is thought to be predictive of a moderation of the excessive, inappropriate (and in some cases sustained) elevation in 5-HT release in the

hippocampus, periaqueductal gray and amygdala, which is triggered by exposure to otherwise anodyne stimuli or moderate stressors in pathologically sensitive individuals (Griebel, 1995; Sramek et al., 2002; Millan, 2003). Similarly, an inhibitory inﬂuence of benzodiazepines, corticotropin-releasing factor (CRF)1 antagonists and other classes of anxiolytic agent upon acute, stress-induced pulses of NA release in the FCX and limbic system provides a paradigm for evaluation of potential therapeutic utility (Millan, 2003; Morilak and Frazer, 2004; Lorrain et al., 2005; Page et al., 2005). Interestingly, it has also been shown that chronic administration with ﬂuoxetine and mirtazapine moderates the acute, transient, stressinduced release of NA and DA in FCX, an empirical model of the utility of long-term treatment with antidepressants in the clinical management of anxiety disorders (Dazzi et al., 2001, 2002; Millan, 2003, 2006; Morilak and Frazer, 2004). Dialysis studies may also be applicable to the demonstration that potential anxiolytic and/ or antidepressant drugs reduce the exaggerated, stress-related release of cytokines, glutamate, CRF and other modulators incriminated in the induction of anxio-depressive states (Holmes et al., 2003; Millan, 2003, 2006). As pointed out below, a rather different example of the use of dialysis is offered by measures of the frontocortical and hippocampal release of ACh, an instructive correlate of cognitive function. Dialysis measures of neurotransmitters and other mediators with well-deﬁned functions have, then, become indispensable in the characterisation of the therapeutic potential of psychotropic agents. Their use is all the more important in that numerous authorities have lamented the limitations of behavioural parameters ostensibly predictive of antidepressant, anxiolytic, antipsychotic and other therapeutic properties (Geyer and Ellenbroek, 2003; O’Neil and Moore, 2003; see Millan, 2006). It might be contended that neurochemical procedures are no better validated than their behavioural counterparts as concerns their ability to predict the therapeutic utility of novel drugs. Indeed, one universal problem – and something of a ‘‘Catch 22’’ – is that experimental procedures are generally established with clinically

490

active drugs, yet they are designed to reveal novel mechanisms of drug action. However, drugs inactive in preclinical models will never make it to the clinic, while for drugs possessing innovative mechanisms of action, there is still a dearth of therapeutic information. For example, all clinically validated antidepressants act directly via monoaminergic mechanisms, and deﬁnitive clinical data is still awaited for other novel approaches, such as CRF1, neurokinin (NK)1 or melanin concentrating hormone (MCH)1 antagonists (Borowsky et al., 2002; Holmes et al., 2003; Millan, 2006). It makes no sense to bemoan the inadequacies of experimental models when unambiguous feedback from the clinic – the only genuine ‘‘validation’’ – remains outstanding. In any case, adopting neurochemical in addition to behavioural criteria should improve our ability to predict the therapeutic use of novel classes of psychotropic agents. Dialysis studies can also be useful for predicting potentially undesirable effects. For example, a sudden and intense elevation in DA release in the nucleus accumbens warns of a potential problem of abuse and addiction (Bradberry, 2002; Volkow et al., 2005). As a further example, protracted glutamate release in the hippocampus or striatum is associated with deleterious neurotoxic effects (Moghaddam and Jackson, 2004; Koch et al., 2005; Millan, 2005a). To summarise, microdialysis is becoming as important as behavioural models in exploring the potential beneﬁcial and detrimental actions of novel drugs. Neither neurochemical nor behavioural variables can be considered sufﬁcient alone, but their association affords a powerful strategy for characterisation of the potential therapeutic effects of novel agents.

II.D. Characterisation of experimental models of CNS disorders Predictions of therapeutic activity need not necessarily be based upon drug actions in models of CNS disorders: dialysis studies are extensively and instructively performed on ‘‘normal’’ subjects. Nevertheless, one would ideally wish to show that antidepressant or antipsychotic agents, for

example, normalise neurochemical and behavioural perturbations in experimental models of depression and schizophrenia, respectively. Accordingly, a related and persistent challenge in the ﬁeld of neuropsychiatry is the need for wellfounded models of CNS disorders. Such paradigms, whether of the disorder in general, or of speciﬁc symptoms, should present features similar and relevant to those displayed by patients and, in general, they should be responsive to therapeutically validated drugs and other modes of therapy, such as electroconvulsive shock therapy (Geyer and Ellenbroek, 2003; O’Neil and Moore, 2003; Dekeyne, 2005). The major emphasis of many models – be it an exposure to stressors or other environmental manipulations, developmental lesions, or genetic modiﬁcations of mice – is on behavioural changes and, in certain cases, biochemical alterations in cellular markers, such as transcription factors (op. cit.). However, in parallel with the recently introduced techniques of small animal imaging like fMRI (providing a ‘‘bridge’’ to the clinic) (Shah and Marsden, 2004; Steward et al., 2005), a complementary approach is provided by microdialysis studies of changes in the levels of neurotransmitters and other mediators. As a simple example, acute, anxiety-provoking stimuli elicit a rapid and transient induction of monoamine, glutamate and CRF release in structures, such as the FCX and amygdala, while chronic models of depression may be accompanied by reductions in extracellular levels of 5-HT and DA in FCX and hippocampus (Dazzi et al., 2001, 2002; Millan, 2003; Page et al., 2005; Lorrain et al., 2005) (see also above citations). Further examples outlined below comprise: (1) the inﬂuence of psychostimulants upon corticolimbic levels of monoamines (Millan et al., 1999; Hori et al., 2000; Navailles et al., 2004) and (2) alterations in the response of corticolimbic pools of DA to antipsychotics and stress following neonatal isolation or lesions of the hippocampus, developmental models of schizophrenia (Lillrank et al., 1999; Hori et al., 2000; Heidbreder et al., 2001; Chrapusta et al., 2003; Kosten et al., 2005). There remains a need for the broader application of microdialysis in the validation and exploitation of experimental models of CNS disorders.

491

III. Focus on antipsychotic agents III.A. Simultaneous quantification of DA, NA and 5-HT levels III.A.1. Significance of monoaminergic transmission Monoaminergic transmission is pivotal to the control of mood, motor function and cognition, all of which are profoundly disrupted in schizophrenia. Dopaminergic pathways are the most ﬁrmly incriminated in the induction of psychotic states. Broadly speaking, an overactivity or hypersensitivity of mesolimbic dopaminergic pathways may contribute to positive symptoms, a hypoactivity of mesocortical projections is implicated in cognitive and negative symptoms, whereas their nigrostriatal counterparts are little affected (Kapur and Mamo, 2003; Mattay et al., 2003; Abi-Dargham and Laruelle, 2005; Tanaka, 2006). The challenge is, then, to normalise (moderate) activity of mesolimbic dopaminergic pathways, while simultaneously enhancing the activity of their mesocortical counterparts and preserving the integrity of nigrostriatal projections. As regards adrenergic mechanisms, things are rather less clear: elevations in NA release in the nucleus accumbens may aggravate psychosis, but increased NA release in the FCX (and hippocampus) might be expected to enhance cognitive performance (Svensson, 2003; Arnsten, 2004; Rossetti and Carboni, 2005; Carboni et al., 2006). As regards 5-HT, an acceleration of corticolimbic release might, upon long-term drug administration, be expected to alleviate negative symptoms and co-morbid (though not synonymous) depressive traits (Meltzer et al., 2003; Millan, 2006). However, no currently available antipsychotic is known to enhance 5-HT release; on the contrary, several reduce 5-HT release, primarily due to agonist actions at inhibitory 5HT1A autoreceptors (Millan, 2000; Meltzer et al., 2003). In theory, co-morbid anxious symptoms might be improved but there is no compelling clinical evidence for this. In contrast, a more convincing case can be made for an association between reduced striatal 5-HT release and a diminished risk of extrapyramidal impact (Millan et al., 1999; Millan, 2000; Meltzer et al., 2003;

Bardin et al., 2006). These comments are of necessity rather schematic but they serve as a simple framework for understanding the significance of microdialysis studies of the inﬂuence of antipsychotics upon monoaminergic transmission.

III.A.2. Technical aspects Historically, microdialysis was often devoted to measures of monoamine metabolites and/or DA in the striatum (Zetterstrom et al., 1983; Imperato and Di Chiara, 1984; Sharp et al., 1986; Westerink and Tuinte, 1986; Wood and Altar, 1988; Parent et al., 2001; see also Westerink et al., 1987). This technique was extended a few years latter to the measurement of 5-HT and its metabolites (Kalen et al., 1988; Carboni and Di Chiara, 1989) and to NA and its metabolites (L’Heureux et al., 1986; Routledge and Marsden, 1987; Abercrombie et al., 1988). Methods of detection employed included radioenzymatic assays and HPLC coupled to ﬂuorimetric or electrochemical detection. However, with the improvement of electrochemical detection techniques (amperometric and coulometric), the routine quantiﬁcation of individual monoamines in dialysis samples of the FCX and other discrete structures is now possible (see Section I for citations). One common drawback is a general inability to measure all three monoamines concurrently in single samples with realistic run times and acceptable sensitivity relative to basal levels in the FCX and other regions. However, Fig. 1 shows a currently available solution to the rapid, reliable and simultaneous quantiﬁcation of DA, NA and 5-HT levels in single dialysis samples based on electrochemical detection. This procedure is, logically enough, three times more ‘‘rapid’’ than those requiring measures of monoamines separately. Furthermore, in studies of combined drug administration (whether antagonistic or potentiating), a role of pharmacokinetic interactions can more easily be discounted. For example, where one drug enhances the inﬂuence of a second agent upon DA levels – but not NA and 5-HT levels – in the same samples, this cannot reﬂect an increase in drug exposure (Gobert et al., 1998, 2000; Millan et al., 2000b; Millan, 2006). Technical difﬁculties and the limited availability of requisite detection cells have not allowed for the generalisation of this microdialysis system but, in

492

Standard Dialysate, Cortex 5-HT

DA

NA

CURRENT (nA)

1.0

0.5

0 0

5

10

15 20 TIME (MIN)

25

30

Fig. 1. Dialysis and detection system employed for simultaneous quantiﬁcation of NA, DA and 5-HT levels in single dialysate samples of the frontal cortex (FCX) and other cerebral structures of freely moving rats. The dotted chromatogram represent a standard mixture (3 pg) of noradrenaline (NA), dopamine (DA) and serotonin (5-HT) injected onto the column using chromatographic conditions as described in Gobert et al. (2004). The continuous chromatogram represents 33 mL of a 20 mL basal microdialysate cortex (plus 20 mL mobile phase). The levels of NA, DA and 5-HT in cortex were 2.1, 0.9 and 1.0 pg, respectively.

the not too distant future, improvements in the sensitivity of mass spectrometry and other approaches will probably also allow for concurrent quantiﬁcation of all three monoamines in single dialysate samples (Hows et al., 2004). The problem of long separation times can be resolved by coupling capillary electrophoresis to laser-induced ﬂuorescence detection. However, this technique does not permit the simultaneous measurement of three monoamines and suffers from limited sensitivity due to low dialysate volumes (Robert et al., 1995; Bowser and Kennedy, 2001; Benturquia et al., 2005).

III.A.3. Actions of antipsychotic agents In line with the implication of mesolimbic dopaminergic pathways in psychosis, microdialysis studies have shown that cocaine and amphetamine (which interfere with monoamine reuptake and activate intracellular release mechanisms) provoke behavioural changes in parallel with increases in DA and NA release in the nucleus accumbens

(Millan et al., 1999; Navailles et al., 2004; AbiDargham and Laruelle, 2005). The causal relationship between DA release and behavioural effects is underpinned by their blockade by antipsychotics acting as antagonists at postsynaptic D2 and, possibly, D3 receptors in mesolimbic structures (Kapur and Mamo, 2003; Joyce and Millan, 2005). Unfortunately, the neuroleptic, haloperidol, also blocks inhibitory D2/D3 receptors on dopaminergic neurones, thereby increasing DA release in the nucleus accumbens itself. This action opposes its postsynaptic D2/D3 receptor blockade and may underlie certain cases of resistance (Navailles et al., 2004; Joyce and Millan, 2005). Dialysis studies suggest an innovative approach to antipsychotic actions – at least against positive symptoms, that is, activation of 5-HT2C receptors on GABAergic interneurones apposed to DA pathways leads to a reduction in mesolimbic DA release (Gobert et al., 2000; Di Matteo et al., 2001; Navailles et al., 2004; Millan, 2005b). Such drugs have recently entered development (Dunlop et al., 2005). A drawback of 5-HT2C agonists would be reduced DA release both in the striatum, which may aggravate extrapyramidal symptoms, and likewise in the FCX, which would exacerbate negative and cognitive symptoms (Gobert et al., 2000; Millan et al., 2000b; Di Matteo et al., 2001; De Deurwaerdere et al., 2004; Millan, 2005b). In contrast, many microdialysis studies have focussed on the ability of antipsychotics to preferentially buttress frontocortical dopaminergic transmission in the FCX as compared with the nucleus accumbens and striatum (Westerink, 2002; Meltzer et al., 2003). Thus, the atypical agent, clozapine, more markedly elevates DA release in FCX than in subcortical structures (Chung et al., 2004). Enhanced frontocortical DA release is also seen with its metabolite, NDMC (Li et al., 2005b), and with many other newer antipsychotics, including aripiprazole (a D2/D3 partial agonist) and ziprasidone (a D2/D3 antagonist) both of which possess 5-HT1A agonist properties (Fig. 2) (Meltzer et al., 2003; Li et al., 2004; Zocchi et al., 2005). Nevertheless, from a mechanistic point of view, it is unlikely that a single common mechanism underlies the preferential induction of frontocortical vs. subcortical DA release by many

493

MONOAMINE LEVELS (% of baseline)

250

250

NA

250

DA

ZIPRASIDONE (10.0)

200

200

200

150

150

150

100

100

100

50

600

0

60

120

50

600

NA

0

60

120

50

600

DA

500

500

500

400

400

400

300

300

300

200

200

200

100

100 0

60

120

VEHICLE ARIPIPRAZOLE (10.0)

5-HT

0

60

120

VEHICLE

5-HT

CLOZAPINE (10.0) N-DESMETHYLCLOZAPINE (10.0)

100 0

60

120

0

60

120

TIME (MIN) Fig. 2. Inﬂuence of the ‘‘atypical’’ antipsychotic agents, aripiprazole, ziprasidone, clozapine and N-desmethylclozapine (NDMC) upon extracellular level of monoamines in the FCX of freely moving rats. Levels are expressed as a percentage of basal values (deﬁned as 100%). Asterisks indicate significance of drug vs. vehicle values (po0.05). Doses in mg/kg, s.c. Data for clozapine are from Millan et al. (1998). Other data are unpublished.

mechanistically distinct antipsychotics. Apart from D2/D3 receptor antagonism and agonist actions at 5-HT1A receptors, blockade of 5-HT2C (and, controversially, 5-HT2A) receptors as well as antagonism of a2-ARs have been implicated (Kuroki et al., 1999; Millan et al., 2000b; Bonaccorso et al., 2002; Ichikawa et al., 2002a, c; Li et al., 2003, 2005a; Meltzer et al., 2003; Chung et al., 2004). It is interesting to note that all of these mechanisms could account for dialysis ﬁndings of enhanced frontocortical release of NA by antipsychotics, which are also illustrated in Fig. 2. For at least one, zotepine, its marked increase in NA levels may reﬂect blockade of NA transporters, an action suggested to account for its putative antidepressant properties (Rowley et al., 1998; Tatsumi et al., 1999). It should be noted that microdialysis levels of NA and DA in FCX often move in parallel owing to the surfeit of NA vs. DA transporters in this region, that is, NA reuptake sites actually play the dominant role in clearing DA (Millan et al., 2000b; Carboni and

Silvagni, 2004; Devoto et al., 2005; Carboni et al., 2006). Interestingly, ziprasidone has a modest afﬁnity for 5-HT transporters (Tatsumi et al., 1999), but any increase in 5-HT levels mediated via suppression of 5-HT reuptake appears to be overwhelmed by its above-mentioned agonist actions at 5-HT1A receptors by analogy to aripiprazole, quetiapine and, at high doses, clozapine (Millan et al., 1998; Zocchi et al., 2005). As mentioned above, 5-HT1A autoreceptor-mediated reductions in 5-HT release are also seen in the striatum and have been correlated with a reduced propensity for an extrapyramidal syndrome even in the face of pronounced D2 receptor blockade, though the mechanistic basis for this remains uncertain (Millan, 2000; Meltzer et al., 2003). These comments illustrate the utility of microdialysis for determining the inﬂuence of antipsychotic agents upon extracellular levels of NA, DA and 5-HT in cortex and subcortical structures, information of considerable help in interpreting their patterns of pharmacological and clinical effects.

494

III.B. Quantification of frontocortical levels of acetylcholine in the absence of acetylcholinesterase inhibitors III.B.1. Significance of frontocortical cholinergic transmission Currently available antipsychotic agents show reasonable efﬁcacy against positive symptoms in the majority of patients. However, based on experimental and clinical observations, most experts concur that there remains a need for drugs more effective than haloperidol, clozapine, olanzapine, risperidone, and so forth in the control of cognitive symptoms (Harvey et al., 2003; Hagan and Jones, 2005; Stip et al., 2005; Levin and Christopher, 2006). Mnemonic impairment in schizophrenia embraces deficits in social memory, verbal memory, executive function, working memory and attention/vigilance (Green et al., 2004; Honey and Fletcher, 2006). The latter three elements appear to involve a major contribution of the FCX wherein, over a deﬁned range of activity, cholinergic pathways exert a favourable inﬂuence upon cognitive performance, acting in interaction with adrenergic, glutamatergic, histaminergic, dopaminergic and other mechanisms (Bruno et al., 1999; Gold, 2003; Ito, 2004; Pepeu and Giovannini, 2004; Sarter et al., 2005). One reason accounting for the mitigated inﬂuence of clozapine, olanzapine, risperidone and other drugs upon cognitive performance is their antagonist properties at histaminergic (H1) receptors, activation of which exerts a favourable inﬂuence upon cognition attentional function (Kay, 2000; Tashiro et al., 2002). In addition, though the clozapine metabolite, NDMC, is an agonist at M1 sites (Weiner et al., 2004; Li et al., 2005b), clozapine and other antipsychotics like olanzapine behave as antagonists at multiple subtypes of muscarinic receptor: antagonism of muscarinic M2/M4 autoreceptors may enhance ACh release but blockade of postsynaptic muscarinic (M1 and others) receptors prejudices cognitive function (Bymaster et al., 1996; Shirazi-Southall et al., 2002; Millan et al., 2004b; Johnson et al., 2005; Li et al., 2005b). In light of the above comments, there is enormous current interest in improved mechanisms for ameliorating cognition in schizophrenia. One approach is to

reinforce cholinergic transmission in the FCX, and an increase in ACh release in the hippocampus may also be advantageous inasmuch as cholinergic processes in this structure fulﬁl an important role in long-term memory formation (Gold, 2003; Parent and Baxter, 2004). III.B.2. Technical aspects Acetylcholinesterase is one of the most efﬁcient enzymes known, and extracellular levels of choline are far higher than those of ACh, complicating quantiﬁcation. Accordingly, most dialysis studies employing electrochemical detection have resorted to the addition of acetylcholinesterase inhibitors to the perfusate to boost resting levels of ACh relative to those of choline. However, as discussed elsewhere, a major disadvantage is the modiﬁcation of autoreceptor sensitivity, physiological status and the actions of exogenously applied drugs (Fujii et al., 1997; Bruno et al., 1999; Ichikawa et al., 2000, 2002b; Herzog et al., 2003; Millan et al., 2004b). Efforts to ﬁnd alternative solutions to the determination of ACh levels include mass spectrometry (Hows et al., 2002) as well as an innovative technique (Fig. 3) pioneered by Ichikawa et al. (2000, 2002a–c) and subsequently adopted by our group (Gobert et al., 2003; Millan et al., 2004b). Thus, introduction of choline-oxidase immobilised on a ‘‘pre-column’’ for the elimination of choline and of peroxidase coated on a carbon electrode for H2O2 detection achieves a marked (10-fold) increase in sensitivity and – even under basal conditions – permits reliable and reproducible measurement of ACh in dialysate samples of the FCX and other cerebral structures (op. cit.). III.B.3. Actions of ligands differentiating D3 and D2 receptor subtypes Several studies have reported significant elevations in extracellular levels of ACh with the atypical antipsychotic, clozapine, both in the FCX and the hippocampus, whereas levels are not affected in the nucleus accumbens (Ichikawa et al., 2002a, b; Shirazi-Southall et al., 2002; Li et al., 2003, 2005b; Chung et al., 2004; Johnson et al., 2005). This regional speciﬁcity for strengthening froncortical vs. nucleus accumbens cholinergic transmission

495 Injection (Dialysate)

Pre-IMER column

Separative Column

+

Detection (Amperometric) HRP coated cell

Ch oxidase

AChE

ACh

ACh

Post-IMER column

H2O2

Ch

Reduction (+ 100mV)

Ch oxidase

H2O + ½O2 + e-

Choline (Ch)

H2O2 Mobile phase (pH 8.2 ; 140 µl/min) New technique: sensitivity × 10

Dialysate Cortex

Standard ACh 20.0 pg 10.0 pg 5.0 pg 1.0 pg 0.5 pg 0.2 pg 0.1 pg

1.0 0.8

Ch

0.6

1.2 1.0 0.8 0.6

0.4

0.4

0.2

0.2

0

0

ACh

CURRENT (nA)

1.2

-0.2

0

1

2

3

4

5

6

7

8

9

10

0

1

2

3

4

5

6

7

8

9

10

Fig. 3. Dialysis and detection system employed for quantiﬁcation of ACh in dialysates of freely moving rats without addition of acetylcholinesterase inhibitors to the perfusate. The upper schema represents the separation and detection of ACh as described by Ichikawa et al. (2000). ACh is separated on a cation exchange column (separative column) at a ﬂow rate of 140 mL/min. A redox polymer ﬁlm containing horseradish peroxidase (HRP) is coated onto the surface of a glassy carbon electrode. It operates at a potential of 100 mV vs. Ag/AgCl and reduces hydrogen peroxide generated by the post-IMER column. This system improves sensitivity as compared with a platinum electrode. The introduction of a pre-IMER column eliminating choline (Ch) allows for quantiﬁcation of ACh in dialysates without the use of acetylcholine inhibitors. The lower left chromatogram represents 20 mL of standards (0.1, 0.2, 0.5, 1.0, 5.0, 10.0 and 20.0 pg) of acetylcholine (ACh) injected onto the column using chromatographic conditions as described in Millan et al. (2004b). The lower right chromatogram represents 20 mL of a 20-mL basal microdialysate cortex (plus 10 mL acetic acid). The level of ACh in cortex was 2.0 pg.

has been reproduced by a surprising diversity of other ‘‘multi-receptorial’’ antipsychotics including NDMC, risperidone, ziprasidone, olanzapine, quetiapine, melperidone, iloperidone and loxapine. NDMC, olanzapine, risperidone and ziprasidone were also found to be active in the hippocampus (Ichikawa et al., 2002b, c; Shirazi-Southall et al., 2002; Li et al., 2003, 2004, 2005b; Johnson et al., 2005). Inasmuch as the ‘‘typical’’ antipsychotics, thioridazine and chlorpromazine, only enhanced hippocampal ACh release at high doses, and antipsychotic potency correlated with afﬁnity for M2

sites, Johnson et al. (2005) suggested that blockade of M2 autoreceptors was responsible. In contrast, direct agonist actions at M1 receptors were proposed to underlie the increases in FCX levels of ACh provoked by NDMC – though it is not clear which population is involved and how this is brought about (Li et al., 2005b). Though 5-HT1A receptors modulate corticolimbic ACh release (Millan et al., 2004b), the modest partial agonist actions of clozapine at 5-HT1A receptors (Millan, 2000) do not appear to be involved in its induction of ACh release in the hippocampus or FCX

496

(Ichikawa et al., 2002a; Chung et al., 2004). 5-HT1A receptors are also not implicated in the effects of melperone or iloperidone, though their engagement appears to mediate the inﬂuence of quetiapine upon frontocortical release of ACh (Ichikawa et al., 2002c). Curiously, their potential role in the effects of ziprasidone does not appear to have been documented (Li et al., 2003). Clearly, there is a need for further studies of receptorial mechanisms eliciting corticohippocampal ACh release by antipsychotics. In contrast, there is a general agreement that haloperidol does not significantly induce ACh release in FCX or hippocampus (see aforegoing citations) (Fig. 4). Like all other clinically available antipsychotics, haloperidol is a mixed antagonist at D2 and D3 receptors – with a modest preference for the former. Intriguingly, in contrast to the selective D2 receptor antagonist, L741,626, selective D3 antagonists, such as S33084 and SB277,011 increase FCX (though not hippocampal) levels of ACh

(Lacroix et al., 2003; Di Cara et al., 2005; Joyce and Millan, 2005). As haloperidol and sulpiride (which also blocks D2 and D3 receptors) are ineffective (op. cit.), D2 receptor blockade may somehow interfere with the facilitatory inﬂuence of D3 antagonism upon frontocortical ACh release. This further remains to be clariﬁed. In any case, the induction of FCX release of ACh by D3 antagonists is speciﬁc in that frontocortical levels of monoamines and amino acids are not affected (unpublished observations). Further, enhanced ACh release is likely related to recent pharmacological studies in rats, and reports from genetically modiﬁed mice, that selective interference with D3 vs. D2 receptors exerts a favourable inﬂuence upon cognitive function (Bernaerts and Tirelli, 2003; Glickstein et al., 2005; Joyce and Millan, 2005; Laszy et al., 2005). Though the lack of inﬂuence of D3 antagonists upon hippocampal release of ACh may be indicative of a less broad pattern of procognitive properties than D1 receptor agonists

ACETYLCHOLINE LEVELS (% of baseline)

VEHICLE HALOPERIDOL (0.63)

VEHICLE SB277,011 (40.0)

VEHICLE S33084 (0.63) 300

300

300

250

250

250

200

200

200

150

150

150

100

100

100

50

50 0

60

120

50 0

60

VEHICLE

VEHICLE

L741,626 (10.0)

SKF82958 (0.63)

120

0

300

300

250

250

250

200

200

200

150

150

150

100

100

100

50

50 60

120

120

VEHICLE GALANTAMINE (0.16)

300

0

60

50 0

60

120

0

60

120

TIME (MIN) Fig. 4. Inﬂuence of dopaminergic ligands as compared with the acetylcholinesterase inhibitor, galantamine, upon extracellular levels of ACh in the FCX of freely moving rats. S33084 and SB277,011 are selective antagonists at dopamine D3 receptors, haloperidol and L741,626 are preferential antagonists at D2 receptors, and SKF82958 is a selective agonist at D1 receptors. Levels are expressed as a percentage of basal values (deﬁned as 100%). Asterisks indicate significance of drug vs. vehicle values (po0.05). Doses in mg/kg, s.c. (i.p., SB277,011). All data are unpublished.

497

(Fig. 4) (Goldman-Rakic et al., 2004; Hersi et al., 2005), their lack of desensitisation and excellent tolerance would be advantageous in the treatment of the cognitive deficits of schizophrenia and other CNS (Joyce and Millan, 2005). Evidently, ACh is not the only neurotransmitter involved in the complex pattern of cognitive deficits seen in schizophrenia, and there are many other possible strategies for their amelioration. Nevertheless, the exploitation of D3 receptor antagonism in the control of mnemonic impairment in schizophrenia would offer a good example of a dialysis-driven hypothesis for the improved management of this disorder.

III.C. Quantification of glycine and other amino acids III.C.1. Significance of the NMDA receptor co-agonist, glycine Inhibitory GABAergic neurones are ubiquitous in the CNS and, via actions at ionotropic GABAA and metabotropic GABAB receptors, GABA fulﬁls an important role in the control of mood and cognition (Maubach, 2003; Millan, 2003, 2006; Bullock, 2005). There is a large body of evidence suggesting that perturbed GABAergic transmission is involved, though not necessarily causally, in psychotic states (Carlsson et al., 2001; Suzuki et al., 2004). Nevertheless, antipsychotic agents interacting directly with GABAergic mechanisms do not appear to be in development, and questions persist concerning the pertinence of dialysis studies of GABA to neuronal events at the synaptic cleft (Timmerman and Westerink, 1997; Del Arco et al., 2003; Drew et al., 2004; Rea et al., 2005). Similar queries have been raised concerning the signiﬁcance of dialysis measures of extracellular levels of glutamate, which is derived both from neuronal and from glial sources (Timmerman and Westerink, 1997; Danbolt, 2001; Del Arco et al., 2003; Drew et al., 2004; Millan, 2005a). Nevertheless, the massive volume of data relating schizophrenia to alterations in glutamatergic transmission, and to changes in the functional status of ionotropic and metabotropic receptors (Millan, 2002, 2005a; Heresco-Levy, 2003; Javitt, 2004; Marcus et al.,

2005), accentuates the importance of reﬁning in vivo measures of glutamatergic transmission. For example, it has been suggested that increases in glutamate release (measured by dialysis) provoked by phencyclidine in the cortex may be related to its pro-psychotic properties (Moghaddam and Adams, 1998; Adams and Moghaddam, 2001; Moghaddam, 2003; Millan, 2005a). Though this remains contentious, the pro-psychotic properties of phencyclidine and other open channel blockers at N-methyl-D-aspartate receptors (NMDA receptors) have been assimilated into the notion that a hypoactivity at NMDA receptors is involved in the aetiology of psychotic states (Millan, 2002, 2005a; Heresco-Levy, 2003; Moghaddam, 2003; Javitt, 2004). Ipso facto, there is a considerable interest in drugs, which enhance their operation, a concept underpinned by reports that glycine and other agonists of the co-agonist GlycineB site on NMDA receptors may, upon adjunctive administration, improve the inﬂuence of antipsychotics upon negative and, possibly, cognitive and positive symptoms (Millan, 2002, 2005a; Heresco-Levy, 2003; Heresco-Levy and Javitt, 2004; Javitt, 2004). These observations also emphasise the interest of dialysis techniques for the quantiﬁcation of extracellular levels of the endogenous GlycineB agonists, glycine and D-Serine, which are released from glial cells enclosing synapses bearing NMDA receptors (Millan, 2005a; Mothet et al., 2005).

III.C.2. Technical aspects The above reﬂections prompted us to set up a simple, robust and sensitive ﬂuorescent technique for quantiﬁcation of extracellular levels of glycine, simultaneously with those of glutamate, GABA and other amino acids. For amino acid measurements in microdialysates, the ortho-phthalaldehyde (OPA)/mercaptoethanol derivatisation (Lindroth and Mopper, 1979) procedure is not favoured because of the rather unpleasant odour of the mercapto compounds. One choice is OPA/ sulﬁte derivatisation with electrochemical detection (Jacobs, 1987; Rowley et al., 1995). This method is sensitive enough to measure low basal levels of GABA. However, with sensitive electrochemical detection, separation of the derivatised

498

Alanine

Taurine GABA

uta

ma te AA m Ad i n ipi o c Gl uta mi ne Se rin e Gl yc ine

FLUORESCENCE

400

Standard

Gl

As pa rta

te

500

Dialysate

300 200 100 Standard Dialysate, Cortex

0 10

15

20

25

30

35

TIME(MIN) Fig. 5. Dialysis and detection system employed for the simultaneous quantiﬁcation of amino acids in single dialysates of the FCX and other cerebral structures of freely moving rats. The dotted chromatogram represents a standard mixture (1.5 ng) of glutamate, glycine and GABA injected onto the column. The continuous chromatogram represents 5 mL of a 20-mL basal microdialysate cortex (plus 30 mL of derivatisation mix). In addition, aspartate, glutamine, serine, alanine and taurine can be quantiﬁed in a 40-min run. Amino adipic acid was used as an internal standard.

amino acids must be performed under isocratic conditions allowing optimum separation of only a few amino acids per dialysate sample. Further, the lack of stability of OPA derivatives necessitates automatisation of the derivatisation process just before injection. For these reasons, we have developed an HPLC system with binary gradient separation using a C18 reverse phase column and ﬂuorescence detection. Under these chromatographic conditions, 20 amino acids can be routinely separated (Fig. 5). The naphthalene2,3-dicarboxaldehyde (NDA) ﬂuorophore in the presence of cyanide is used to derivatise amino acids (Carlson et al., 1986; De Montigny et al., 1987; Shah et al., 1999). These NDA derivatives are very stable with a high ﬂuorescence efﬁciency and permit the measurement of amino acids in the nanomolar range comparable to capillary electrophoresis (Bert et al., 1996; Robert et al., 1998; Bowser and Kennedy, 2001; Shou et al., 2004). One weakness of this system is its inability to distinguish the co-agonist at NMDA receptors,

D-serine,

from its isomer, L-serine. To resolve this issue, with a slight modiﬁcation of the previous chromatographic conditions, an aliquot of the dialysate sample was derivatised using OPA and N-acetyl-L-cysteine (Kutla`n et al., 2002). Under those conditions, the two (‘‘D’’ and ‘‘L’’) derivatised serine diastereoisomers were rapidly (15 min) and fully separated. This method allows determination of basal D- and L-serine in rat brain dialysates (unpublished observations).

III.C.3. Influence of antipsychotic agents upon dialysate levels of glycine As reviewed recently (Millan, 2005a), clozapine reinforces activity at cortical populations of NMDA receptors by a multitude of complex and poorly deﬁned mechanisms, an action implicated in its ‘‘atypical’’ proﬁle and relief of cognitivedeficit symptoms. It has been suggested that one component of its activity may be suppression of glial uptake of glycine (Schwieler et al., 2004;

499

Williams et al., 2004; Javitt et al., 2005; Millan, 2005a), though we have observed no inﬂuence of clozapine upon extracellular levels of glycine in FCX dialysates of rats (not shown). In view of (1) the facilitation of NMDA receptor activity by clozapine, (2) adjunctive use of glycine in schizophrenia and (3) reports that glycine exerts antipsychotic effects in certain experimental models (Javitt et al., 2005; Millan, 2005a), it appears important to evaluate its neurochemical actions in rodents by microdialysis. As shown in Fig. 6, doses of glycine corresponding to those used experimentally and clinically elevate extracellular levels of glycine in FCX, and this action is selective in that levels of glutamate and GABA are not affected. However, to our surprise, we have found that glycine also provokes dose-dependent, pronounced and concomitant increases in monoamine and ACh levels in FCX, which are resistant to GlycineB antagonists (unpublished observations).

This observation raises questions concerning the speciﬁcity and mechanisms of action of glycine justiﬁes further study. A widely adopted approach to increasing cerebral levels of glycine comprises selective blockers of the glial glycine T1 transporter, which controls glycine levels at NMDA receptor-bearing synapses, and which is putatively inhibited by clozapine (Gadea and Lopez-Colome´, 2001; Schwieler et al., 2004; Williams et al., 2004; Millan, 2005a). Indeed, by analogy to glycine itself (though less markedly), selective glycine T1 inhibitors elevate extracellular levels of glycine; furthermore, this action is highly selective inasmuch as levels of GABA, glutamate, other amino acids, monoamines and ACh are not affected (Fig. 6 and data not shown). The clinical destiny of such agents remains to be established, and it is unlikely that they will be useful antipsychotic drugs alone, but other neurochemical and behavioural studies

VEHICLE

AMINO ACID LEVELS (% of baseline)

500

500

GLYCINE

500

GLUTAMATE

400

400

400

300

300

300

200

200

200

100

100 0

60

100 0

120

GLYCINE (800) ALX5407 (10.0)

GABA

60

120

0

60

120

VEHICLE L-trans-PDC (1 mM) 500

500

GLYCINE

400

400

300

300

200

200

100

100

1100

GLUTAMATE

NIPECOTIC ACID (0.5mM)

GABA

900 700 500

0

60

120

300 100 0

60

120

0

60

120

TIME (MIN) Fig. 6. Inﬂuence of glycine and of drugs modulating the reuptake of amino acids upon extracellular levels of glycine, glutamate and GABA in the FCX of freely moving rats. ALX 5407 is a selective blocker of the Glycine T1 transporter; L-trans-PDC is an inhibitor of glial glutamate reuptake and nipecotic acid is an inhibitor of GABA reuptake. Levels are expressed as a percentage of basal values (deﬁned as 100%). Asterisks indicate significance of drug vs. vehicle values. Doses of glycine and ALX5407 are in mg/kg, s.c. L-transPDC and nipecotic acid were administered via the dialysis probe over 60 min.The basal levels of glycine, glutamate and GABA were 8.3, 1.8 and 0.06 mM, respectively. All data are unpublished.

500

in rodents also support their adjunctive administration with antipsychotic agents (Harsing et al., 2003; Kinney et al., 2003; Depoortere et al., 2005; Millan, 2005a). Irrespective of the fate of Glycine T1 reuptake inhibitors, microdialysis appears ideally suited to the further investigation of their therapeutic potential and mechanisms of action.

IV. Some perspectives for future research In concluding this overview, the following general comments should be made. First, virtually all aﬁcionados will testify that, apart from its invasive nature and inherent complexity, microdialysis can be infuriatingly capricious and time-consuming; indeed, output can often be erratic and modest relative to other approaches for evaluating central actions of drugs. In this respect, technical improvements in detection techniques, sensitivity and throughput should continue to enhance efﬁciency and reliability. For example, mass spectrometry will inevitably assume a dominant role in detection systems. This strategy can generate substantial volumes of data rapidly and is applicable not only to monoamines (Hows et al., 2004), ACh (Hows et al., 2002; Uutela et al., 2005) and amino acids (Takada et al., 1995), but also to many other mediators, such as neuropeptides (Andre´n and Caprioli, 1999; Nyitrai et al., 2003; Baseski et al., 2005). Second, the present review focuses on insights afforded by microdialysis into the role of monoamines, ACh and amino acids in the aetiology and treatment of schizophrenia and other CNS disorders. These modulators remain primordial targets for the development of improved psychotropic agents. However, there is increasing interest in other strategies; in particular, drugs interacting with neuropeptidergic transmission, such as antagonists at NK1/3, MCH1, neuropeptide Y (NPY)1 and CRF1 receptors (Rupniak and Kramer, 1999; Borowsky et al., 2002; Millan, 2003, 2006; Gobbi and Blier, 2005). Our knowledge of possible changes in the release and turnover of their endogenous ligands (Substance P, MCH, NPY and CRF) in CNS disorders remains rudimentary. Microdialysis, notably in tandem with

mass spectrometry, should be brought to bear on such questions. Likewise, there is a need for more systematic and extensive microdialysis studies of alterations in other important mediators controlling mood, such as histamine (Westerink et al., 2002) and corticosterone (Oshim et al., 2003). Third, as regards clinical applications, it has been questioned whether CSF levels of drugs are genuinely pertinent to their actions at speciﬁc neuronal sites in the human brain (De Lange and Danhof, 2002) and imaging techniques (PET, fMRI etc) are better adapted than microdialysis to the evaluation of loci and mechanisms of action of drugs in man, and to determination of their active dose ranges; such studies can ideally be undertaken following preparative imaging studies in animals (Steward et al., 2005). Nevertheless, microdialysis is useful for the determination of drug levels in speciﬁc extra-CNS tissues (Langer and Muller, 2004; Joukhadar and Muller, 2005). Fourth, together with other ‘‘multi-modal’’ methods of brain monitoring, microdialysis will likely assume an important role in neurointensive care for the neurochemical monitoring of changes provoked by (1), subarachnoid haemorrhage; (2) thromboembolic stroke and (3) traumatic, focal CNS lesions – both within the lesion itself and in the surrounding penumbra (Peerdeman et al., 2003; Bellander et al., 2004; De Georgia and Deogaonkar, 2005; Engstrom et al., 2005). Mediators quantiﬁed include glutamate, NO and reactive oxygen species (indexes of neurotoxicity); phosphoethanol amine, glycerol (components of cell membranes) and taurine, elevations of which reﬂect cellular degeneration; glucose, lactate and pyruvate (markers of energy metabolism and the redox state of cells) (Ungerstedt and Rostami, 2004; Engstrom et al., 2005; Hillered et al., 2005). Similarly, determinations of the neurochemical status of the brain can be undertaken in the course of neurosurgery, though there remains a need for additional studies of accompanying changes. Further, applicability may not be restricted to neurological trauma. One can imagine the fruitful extension of such work to dialysis analyses of neurotransmitter release at loci of deep brain stimulation (Kopell et al., 2004; McIntyre et al., 2004) for (1) a major depression (subgenual cingulate

501

region, Brodmann area 25) (Aouizerate et al., 2005; Mayber et al., 2005); (2) obsessive-compulsive disorders (anterior limbs of the internal capsules, and ventral pallidum) (Abelson et al., 2005; Dell’Osso et al., 2005) or (3), refractory Parkinson’s disease (subthalamic nucleus, and the internal segment of the globus pallidus (Volkmann, 2004; Sestini et al., 2005). This could provide novel insights into the mechanisms of action to be subsequently exploited by novel drug classes. V. Summary and conclusions As illustrated herein, primarily in reference to antipsychotic agents, microdialysis has assumed a broad and important role in the discovery and characterisation of novel psychotropic agents. Its application ranges from quantiﬁcation of drug levels in speciﬁc structures; validation of molecular mechanisms of drug action in the brain; evaluation of beneﬁcial and undesirable properties of drugs; to characterisation of experimental models of CNS disorders. Indeed, though behavioural models remain invaluable, neurochemical procedures are no less important in the characterisation, discovery and development of agents for the management of schizophrenia and other psychiatric and neurological disorders. Together with complementary imaging strategies, microdialysis will continue to play a leading role, as in the interface between cellular and behavioural models, in generating data critical to the creative, efﬁcacious and successful development of centrally active agents. References Abelson, J.L., Curtis, G.C., Sagher, O., Albucher, R.C., Harrigan, M., Taylor, S.F., Martis, B. and Giordani, B. (2005) Deep brain stimulation for refractory obsessive-compulsive disorder. Biol. Psychiatry, 57: 510–516. Abercrombie, E.D., Keller, R.W. Jr. and Zigmond, M.J. (1988) Characterization of hippocampal norepinephrine release as measured by microdialysis perfusion: pharmacological and behavioral studies. Neuroscience, 27: 897–904. Abi-Dargham, A. and Laruelle, M. (2005) Mechanism of action of second generation antipsychotic drugs in schizophrenia: insights from brain imaging studies. Eur. Psychiatry, 20: 15–27. Adams, B.W. and Moghaddam, B. (2001) Effect of clozapine, haloperidol, or M100907 on phencyclidine-activated

glutamate efﬂux in the prefrontal cortex. Biol. Psychiatry, 50: 750–757. Allen, D.D. and Geldenhuys, W.J. (2006) Molecular modeling of blood-brain barrier nutrient transporters: in silico basis for evaluation of potential drug delivery to the central nervous system. Life Sci., 78: 1029–1033. Andre´n, P.E. and Caprioli, R.M. (1999) Determination of extracellular release of neurotensin in discrete rat brain regions utilizing in vivo microdialysis/electrospray mass spectrometry. Brain Res., 845: 123–129. Andres, J.I., Alcazar, J., Alonso, J.M., Alvarez, R.M., Bakker, M.H., Biesmans, I., Cid, J.M., De Lucas, A.I., Fernandez, J., Font, L.M., Hens, K.A., Iturrino, L., Lenaerts, I., Martinez, S., Megens, A.A., Pastor, J., Vermote, P.C. and Steckler, T. (2005) Discovery of a new series of centrally active tricyclic isoxazoles combining serotonin (5-HT) reuptake inhibition with alpha2adrenoceptor blocking activity. J. Med. Chem., 48: 2054–2071. Aouizerate, B., Martin-Guehl, C., Cuny, E., Guehl, D., Amieva, H., Benazzouz, A., Fabrigoule, C., Bioulac, B., Tignol, J. and Burbaud, P. (2005) Deep brain stimulation for OCD and major depression. Am. J. Psychiatry, 162: 2192. Aravagiri, M., Yuwiler, A. and Marder, S.R. (1998) Distribution after repeated oral administration of different dose levels of risperidone and 9-hydroxy-risperidone in the brain and other tissues of rat. Psychopharmacology, 139: 356–363. Arnsten, A.F. (2004) Adrenergic targets for the treatment of cognitive deficits in schizophrenia. Psychopharmacology, 174: 25–31. Bardin, L., Kleven, M.S., Barret-Grevoz, C., Depoortere, R. and Newman-Tancredi, A. (2006). Antipsychotic-like vs. cataleptogenic actions in mice of novel antipsychotics having D2 antagonist and 5-HT1A agonist properties. Neuropsychopharmacology, in press. Baseski, H.M., Watson, C.J., Cellar, N.A., Shackman, J.G. and Kennedy, R.T. (2005) Capillary liquid chromatography with MS3 for the determination of enkephalins in microdialysis samples from the striatum of anesthetized and freely-moving rats. J. Mass Spectrom., 40: 146–153. Bellander, B.M., Cantais, E., Enblad, P., Hutchinson, P., Nordstrom, C.H., Robertson, C., Sahuquillo, J., Smith, M., Stocchetti, N., Ungerstedt, U., Unterberg, A. and Olsen, N.V. (2004) Consensus meeting on microdialysis in neurointensive care. Intensive Care Med., 30: 2166–2169. Benturquia, N., Couderc, F., Sauvinet, V., Orset, C., Parrot, S., Bayle, C., Renaud, B. and Denoroy, L. (2005) Analysis of serotonin in brain microdialysates using capillary electrophoresis and native laser-induced ﬂuorescence detection. Electrophoresis, 26: 1071–1079. Bernaerts, P. and Tirelli, E. (2003) Facilitatory effect of the dopamine D4 receptor agonist PD168,077 on memory consolidation of an inhibitory avoidance learned response in C57BL/6J mice. Behav. Brain Res., 142: 41–52. Bert, L., Robert, F., Denoroy, L., Stoppini, L. and Renaud, B. (1996) Enhance temporal resolution for the microdialysis monitoring of catecholamines and excitatory amino acids using capillary electrophoresis with laser-induced ﬂuorescence detection analytical developments and in vitro validations. J. Chromatogr. A, 755: 99–111.

502 Blier, P. (2001) Crosstalk between the norepinephrine and serotonin systems and its role in the antidepressant response. J. Psychiatry Neurosci., 26: S3–S10. Bonaccorso, S., Meltzer, H.Y., Li, Z., Dai, J., Alboszta, A.R. and Ichikawa, J. (2002) SR46349-B, a 5-HT2A/2C receptor antagonist, potentiates haloperidol-induced dopamine release in rat medial prefrontal cortex and nucleus accumbens. Neuropsychopharmacology, 27: 430–441. Borowsky, B., Durkin, M.M., Ogozale, K., Marzabadi, M.R., DeLeon, J., Lagu, B., Heurich, R., Lichtblau, H., Shaposhnik, Z., Daniewska, I., Blackburn, T.P., Branchek, T.A., Gerald, C., Vaysse, P.J. and Forray, C. (2002) Antidepressant, anxiolytic and anorectic effects of a melanin-concentrating hormone-1 receptor antagonist. Nat. Med., 8: 825–830. Boschi, G. and Scherrmann, J. (2000) Microdialysis in mice for drug delivery research. Adv. Drug Deliv. Rev., 45: 271–281. Boulton, D.W., DeVane, C.L., Liston, H.L. and Markowitz, J.S. (2002) In vitro P-glycoprotein afﬁnity for atypical and conventional antipsychotics. Life Sci., 71: 163–169. Bourne, J.A. (2003) Intracerebral microdialysis: 30 years as a tool for the neuroscientist. Clin. Exp. Pharmacol. Physiol., 30: 16–24. Bowser, M.T. and Kennedy, R.T. (2001) In vivo monitoring of amine neurotransmitters using microdialysis with on-line capillary electrophoresis. Electrophoresis, 22: 3668–3676. Bradberry, C.W. (2000) Applications of microdialysis methodology in nonhuman primates: practice and rationale. Crit. Rev. Neurobiol., 14: 143–163. Bradberry, C.W. (2002) Dynamics of extracellular dopamine in the acute and chronic actions of cocaine. Neuroscientist, 8: 315–322. Bruno, J.P., Sarter, M., Moore Arnold, H. and Himmelheber, A.M. (1999) In vivo neurochemical correlates of cognitive processes: methodological and conceptual challenges. Rev. Neurosci., 10: 25–48. Bullock, R. (2005) SGS-742. Curr. Opin. Investig. Drugs, 6: 108–113. Bymaster, F.P., Hemrick-Luecke, S.K., Perry, K.W. and Fuller, R.W. (1996) Neurochemical evidence for antagonism by olanzapine of dopamine, serotonin, alpha1-adrenergic and muscarinic receptors in vivo in rats. Psychopharmacology, 124: 87–94. Carboni, E. and Di Chiara, G. (1989) Serotonin release estimated by transcortical dialysis in freely-moving rats. Neuroscience, 32: 637–645. Carboni, E. and Silvagni, A. (2004) Dopamine reuptake by norepinephrine neurons: exception or rule? Crit. Rev. Neurobiol., 16: 121–128. Carboni, E., Silvagni, A., Vacca, C. and Di Chiara, G. (2006) Cumulative effect of norepinephrine and dopamine carrier blockade on extracellular dopamine increase in the nucleus accumbens shell, bed nucleus of stria terminalis and prefrontal cortex. J. Neurochem., 96: 473–481. Carlson, R.G., Srinivasachar, R.S., Givens, R.S. and Matuszewski, B.K. (1986) New derivatizing agents for amino acids and peptides. 1. Facile synthesis of N-substituted

1-cyanobenz[f]isoindoles and their spectroscopic properties. J. Org. Chem., 51: 3978–3983. Carlsson, A., Waters, N., Holm-Waters, S., Tedroff, J., Nilsson, M. and Carlsson, M.L. (2001) Interactions between monoamines, glutamate, and GABA in schizophrenia: new evidence. Annu. Rev. Pharmacol. Toxicol., 41: 237–260. Celada, P., Puig, M., Amargos-Bosch, M., Adell, A. and Artigas, F. (2004) The therapeutic role of 5-HT1A and 5HT2A receptors in depression. J. Psychiatry Neurosci., 29: 252–265. Chan, L.M., Lowes, S. and Hirst, B.H. (2004) The ABCs of drug transport in intestine and liver: efﬂux proteins limiting drug absorption and bioavailability. Pharm. Sci., 21: 25–51. Chaurasia, C.S. (1999) In vivo microdialysis sampling: theory and applications. Biomed. Chromatogr., 13: 317–332. Chrapusta, S.J., Egan, M.F., Wyatt, R.J., Weinberger, D.R. and Lipska, B.K. (2003) Neonatal ventral hippocampal damage modiﬁes serum corticosterone and dopamine release responses to acute footshock in adult Sprague-Dawley rats. Synapse, 47: 270–277. Chung, Y.C., Li, Z., Dai, J., Meltzer, H.Y. and Ichikawa, J. (2004) Clozapine increases both acetylcholine and dopamine release in rat ventral hippocampus: role of 5-HT1A receptor agonism. Brain Res., 1023: 54–63. Clough, G.F. (2005) Microdialysis of large molecules. AAPS J., 7: E686–E692. Cordon-Cardo, C., O’Brien, J.P., Casals, D., Rittman-Grauer, L., Biedler, J.L., Melamed, M.R. and Bertino, J.R. (1989) Multidrug-resistance gene (P-glycoprotein) is expressed by endothelial cells at blood-brain barrier sites. Proc. Natl. Acad. Sci. U.S.A., 86: 695–698. Danbolt, N.C. (2001) Glutamate uptake. Prog. Neurobiol., 65: 1–105. Dazzi, L., Ladu, S., Spiga, F., Vacca, G., Rivano, A., Pira, L. and Biggio, G. (2002) Chronic treatment with imipramine or mirtazapine antagonizes stress- and FG7142-induced increase in cortical norepinephrine output in freely moving rats. Synapse, 43: 70–77. Dazzi, L., Serra, M., Spiga, F., Pisu, M.G., Jentsch, J.D. and Biggio, G. (2001) Prevention of the stress-induced increase in frontal cortical dopamine efﬂux of freely moving rats by long-term treatment with antidepressant drugs. Eur. Neuropsychopharmacol., 11: 343–349. De Deurwaerdere, P., Navailles, S., Berg, K.A., Clarke, W.P. and Spampinato, U. (2004) Constitutive activity of the serotonin2C receptor inhibits in vivo dopamine release in the rat striatum and nucleus accumbens. J. Neurosci., 24: 3235–3241. De Georgia, M.A. and Deogaonkar, A. (2005) Multimodal monitoring in the neurological intensive care unit. Neurologist, 11: 45–54. De Groote, L., Penalva, R.G., Flachskamm, C., Reul, J.M. and Linthorst, A.C. (2005) Differential monoaminergic, neuroendocrine and behavioural responses after central administration of corticotropin-releasing factor receptor type 1 and type 2 agonists. J. Neurochem., 94: 862–864.

503 Deguchi, Y. (2002) Application of in vivo brain microdialysis to the study of blood-brain barrier transport of drugs. Drug Metab. Pharmacokinet., 17: 395–407. Dekeyne, A. (2005) Behavioural models for the characterisation of established and innovative antidepressant agents. The´rapie, 60: 477–484. De Lange, E.C. and Danhof, M. (2002) Consideration in the use of cerebrospinal ﬂuid pharmacokinetics to predict brain target concentrations in the clinical setting: implications of the barriers between blood and brain. Clin. Pharmacokinet., 41: 691–703. De Lange, E.C., Ravenstijn, P.G., Groenendaal, D. and van Steeg, T.J. (2005) Towards the prediction of CNS drug-effect proﬁles in physiological and pathological conditions using microdialysis and mechanism-based pharmacokinetic-pharmacodynamic modelling. AAPS J., 7: E532–E543. Del Arco, A., Segovia, G., Fuxe, K. and Mora, F. (2003) Changes in dialysate concentrations of glutamate and GABA in the brain: an index of volume transmission mediated actions? J. Neurochem., 85: 23–33. Dell’Osso, B., Altamura, A.C., Allen, A. and Hollander, E. (2005) Brain stimulation techniques in the treatment of obsessive-compulsive disorder: current and future directions. CNS Spectr., 10: 966–979. De Montigny, P., Stobaugh, J.F., Givens, R.S., Carlson, R.G., Srinivasachar, K., Sternson, L.A. and Higuchi, T. (1987) Naphtalene-2,3-dicarboxaldehyde/cyanide ion: a rationally designed ﬂuorogenic reagent for primary amines. Anal. Chem., 59: 1096–1101. Depoortere, R., Dargazanli, G., Estenne-Bouhtou, G., Coste, A., Lanneau, C., Desvignes, C., Poncelet, M., Heaulme, M., Santucci, V., Decobert, M., Cudennec, A., Voltz, C., Boulay, D., Terranova, J.P., Stemmelin, J., Roger, P., Marabout, B., Sevrin, M., Vige, X., Biton, B., Steinberg, R., Francon, D., Alonso, R., Avenet, P., Oury-Donat, F., Perrault, G., Griebel, G., George, P., Soubrie, P. and Scatton, B. (2005) Neurochemical, electrophysiological and pharmacological proﬁles of the selective inhibitor of the glycine transporter1 SSR504734, a potential new type of antipsychotic. Neuropsychopharmacology, 30: 1963–1985. Devoto, P., Flore, G., Saba, P., Fa, M. and Gessa, G.L. (2005) Co-release of noradrenaline and dopamine in the cerebral cortex elicited by single train and repeated train stimulation of the locus coeruleus. BMC Neurosci., 6: 31. Di Cara, B., Panayi, F., Gobert, A. and Millan, M.J. (2005) Dopamine D3 receptors tonically inhibit, and D1 receptors tonically excite, frontocortical acetylcholine release: a dialysis study in rats. Am. Soc. Neurosci. Abstr., 31: 913.6. Di Chiara, G. (1990) In vivo brain dialysis of neurotransmitters. Trends Pharmacol. Sci., 11: 116–121. Di Chiara, G., Tanda, G. and Carboni, E. (1996) Estimation of in vivo neurotransmitter release by brain microdialysis: the issue of validity. Behav. Pharmacol., 7: 640–657. Di Matteo, V., De Blasi, A., Di Giulio, C. and Esposito, E. (2001) Role of 5-HT2C receptors in the control of central dopamine function. Trends Pharmacol. Sci., 22: 229–232.

Drew, K.L., Pehek, E.A., Rasley, B.T., Ma, Y.L. and Green, T.K. (2004) Sampling glutamate and GABA with microdialysis: suggestions on how to get the dialysis membrane closer to the synapse. J. Neurosci. Methods, 140: 127–131. Dunlop, J., Sabb, A.L., Mazandarani, H., Zhang, J., Kalgaonker, S., Shukhina, E., Sukoff, S., Vogel, R.L., Stack, G., Schechter, L., Harrison, B.L. and Rosenzweig-Lipson, S. (2005) WAY-163909 [(7bR, 10aR)-1,2,3,4,8,9,10,10a-octahydro-7bH-cyclopenta-[b][1,4]diazepino[6,7,1hi] indo1e], a novel 5-hydroxytryptamine2C receptor-selective agonist with anorectic activity. J. Pharmacol. Exp. Ther., 313: 862–869. Eichelbaum, M., Fromm, M.F. and Schwab, M. (2004) Clinical aspects of the MDR1 (ABCB1) gene polymorphism. Ther. Drug Monit., 26: 180–185. Ejsing, T.B., Pedersen, A.D. and Linnet, K. (2005) P-glycoprotein interaction with risperidone and 9-OH-risperidone studied in vitro, in knock-out mice and in drug–drug interaction experiments. Hum. Psychopharmacol., 20: 493–500. El Ela, A.A., Hartter, S., Schmitt, U., Hiemke, C., SpahnLangguth, H. and Langguth, P. (2004) Identiﬁcation of Pglycoprotein substrates and inhibitors among psychoactive compounds – implications for pharmacokinetics of selected substrates. J. Pharm. Pharmacol., 56: 967–975. Engstrom, M., Polito, A., Reinstrup, P., Romner, B., Ryding, E., Ungerstedt, U. and Nordstrom, C.H. (2005) Intracerebral microdialysis in severe brain trauma: the importance of catheter location. J. Neurosurg., 102: 460–469. Fillenz, M. (2005) In vivo neurochemical monitoring and the study of behaviour. Neurosci. Biobehav. Rev., 29: 949–962. Frankel, P.S., Hoonakker, A.J., Hanson, G.R., Bush, L., Keefe, K.A. and Alburges, M.E. (2005) Differential neurotensin responses to low and high doses of methamphetamine in the terminal regions of striatal efferents. Eur. J. Pharmacol., 522: 47–54. Fujii, T., Yoshizawa, M., Nakai, K., Fujimoto, K., Suzuki, T. and Kawashima, K. (1997) Demonstration of the facilitatory role of 8-OH-DPAT on cholinergic transmission in the rat hippocampus using in vivo microdialysis. Brain Res., 761: 244–249. Gadea, A. and Lopez-Colome´, A.M. (2001) Glial transporters for glutamate, glycine, and GABA III. Glycine transporters. J. Neurosci. Res., 64: 218–222. Garberg, P., Ball, M., Borg, N., Cecchelli, R., Fenart, L., Hurst, R.D., Lindmark, T., Mabondzo, A., Nilsson, J.E., Raub, T.J., Stanimirovic, D., Terasaki, T., Oberg, J. and Osterberg, T. (2005) In vitro models for the blood-brain barrier. Toxicol. In Vitro, 19: 299–334. Gardier, A.M., Malagie, I., Trillat, A.C., Jacquot, C. and Artigas, F. (1996) Role of 5-HT1A autoreceptors in the mechanism of action of serotoninergic antidepressant drugs: recent ﬁndings from in vivo microdialysis studies. Fundam. Clin. Pharmacol., 10: 16–27. Geyer, M.A. and Ellenbroek, B. (2003) Animal behavior models of the mechanisms underlying antipsychotic atypicality. Prog. Neuropsychopharmacol. Biol. Psychiatry, 27: 1071–1079.

504 Glickstein, S.B., Desteno, D.A., Hof, P.R. and Schmauss, C. (2005) Mice lacking dopamine D2 and D3 receptors exhibit differential activation of prefrontal cortical neurons during tasks requiring attention. Cereb. Cortex, 15: 1016–1024. Gobbi, G. and Blier, P. (2005) Effect of neurokinin-1 receptor antagonists on serotoninergic, noradrenergic and hippocampal neurons: comparison with antidepressant drugs. Peptides, 26: 1383–1393. Gobert, A., Billiras, R., Cistarelli, L. and Millan, M.J. (2004) Quantiﬁcation and pharmacological characterization of dialysate levels of noradrenaline in the striatum of freely-moving rats: release from adrenergic terminals and modulation by a2-autoreceptors. J. Neurosci. Methods, 140: 141–152. Gobert, A., Di Cara, B., Cistarelli, L. and Millan, M.J. (2003) Piribedil enhances frontocortical and hippocampal release of acetylcholine in freely moving rats by blockade of a2Aadrenoceptors: a dialysis comparison to talipexole and quinelorane in the absence of acetylcholinesterase inhibitors. J. Pharmacol. Exp. Ther., 305: 338–346. Gobert, A., Lejeune, F., Rivet, J.M., Audinot, V., NewmanTancredi, A. and Millan, M.J. (1995) Modulation of the activity of central serotoninergic neurones by novel serotonin1A receptor agonists and antagonists: a comparison to adrenergic and dopaminergic neurones in individual rats. J. Pharmacol. Exp. Ther., 273: 1032–1046. Gobert, A., Rivet, J.M., Audinot, V., Newman-Tancredi, A., Cistarelli, L. and Millan, M.J. (1998) Simultaneous quantiﬁcation of serotonin, dopamine and noradrenaline levels in single frontal cortex dialysates of freely-moving rats reveals a complex pattern of reciprocal auto- and hetero-receptor mediated control of release. Neuroscience, 84: 413–429. Gobert, A., Rivet, J.M., Lejeune, F., Newman-Tancredi, A., Adhumeau-Auclair, A., Nicolas, J.P., Cistarelli, L., Melon, C. and Millan, M.J. (2000) Serotonin2C receptors tonically suppress the activity of mesocortical dopaminergic and adrenergic, but not serotonergic, pathways: a combined dialysis and electrophysiological analysis in the rat. Synapse, 36: 205–211. Goetz, C.G., Poewe, W., Rascol, O. and Sampaio, C. (2005) Evidence-based medical review update: pharmacological and surgical treatments of Parkinson’s disease: 2001 to 2004. Mov. Disord., 20: 523–539. Gold, P.E. (2003) Acetylcholine modulation of neural systems involved in learning and memory. Neurobiol. Learn. Mem., 80: 194–210. Goldman-Rakic, P.S., Castner, S.A., Svensson, T.H., Siever, L.J. and Williams, G.V. (2004) Targeting the dopamine D1 receptor in schizophrenia: insights for cognitive dysfunction. Psychopharmacology, 174: 3–16. Goodwin, J.T. and Clark, D.E. (2005) In silico predictions of blood-brain barrier penetration: considerations to ‘‘keep in mind’’. J. Pharmacol. Exp. Ther., 315: 477–483. Green, M.F., Nuechterlein, K.H., Gold, J.M., Barch, D.M., Cohen, J., Essock, S., Fenton, W.S., Frese, F., Goldberg, T.E., Heaton, R.K., Keefe, R.S., Kern, R.S., Kraemer, H., Stover, E., Weinberger, D.R., Zalcman, S. and Marder, S.R. (2004) Approaching a consensus cognitive battery for clinical

trials in schizophrenia: the NIMH-MATRICS conference to select cognitive domains and test criteria. Biol. Psychiatry, 56: 301–307. Griebel, G. (1995) 5-Hydroxytryptamine-interacting drugs in animal models of anxiety disorders: more than 30 years of research. Pharmacol. Ther., 65: 319–395. Hagan, J.J. and Jones, D.N. (2005) Predicting drug efﬁcacy for cognitive deficits in schizophrenia. Schizophr. Bull., 31: 830–853. Hammarlund-Udenaes, M. (2000) The use of microdialysis in CNS drug delivery studies. Pharmacokinetic perspectives and results with analgesics and antiepileptics. Adv. Drug Deliv. Rev., 45: 283–294. Harsing, L.G. Jr., Gacsalyi, I., Szabo, G., Schmidt, E., Sziray, N., Sebban, C., Tesolin-Decros, B., Matyus, P., Egyed, A., Spedding, M. and Levay, G. (2003) The glycine transporters1 inhibitors NFPS and Org 24461: a pharmacological study. Pharmacol. Biochem. Behav., 74: 811–825. Ha¨rtter, S., Huwel, S., Lohmann, T., Abou El Ela, A., Langguth, P., Hiemke, C. and Galla, H.J. (2003) How does the benzamide antipsychotic amisulpride get into the brain? – an in vitro approach comparing amisulpride with clozapine. Neuropsychopharmacology, 28: 1916–1922. Harvey, P.D., Green, M.F., McGurk, S.R. and Meltzer, H.Y. (2003) Changes in cognitive functioning with risperidone and olanzapine treatment: a large-scale, double-blind, randomized study. Psychopharmacology, 169: 404–411. Heidbreder, C.A., Foxton, R., Cilia, J., Hughes, Z.A., Shah, A.J., Atkins, A., Hunter, A.J., Hagan, J.J. and Jones, D.N. (2001) Increased responsiveness of dopamine to atypical, but not typical antipsychotics in the medial prefrontal cortex of rats reared in isolation. Psychopharmacology, 156: 338–351. Heresco-Levy, U. (2003) Glutamatergic neurotransmission modulation and the mechanism of antipsychotic atypicality. Prog. Neuropsychopharmacol. Biol. Psychiatry, 27: 1113–1123. Heresco-Levy, U. and Javitt, D.C. (2004) Comparative effects of glycine and D-cycloserine on persistent negative symptoms in schizophrenia: a retrospective analysis. Schizophr. Res., 66: 89–96. Hersi, A.I., Richard, J.W., Gaudreau, P. and Quirion, R. (2005) Local modulation of hippocampal acetylcholine release by dopamine D1 receptors: a combined receptor autoradiography and in vivo dialysis study. J. Neurosci., 15: 7150–7157. Herzog, C.D., Nowak, K.A., Sarter, M. and Bruno, J.P. (2003) Microdialysis without acetylcholinesterase inhibition reveals an age-related attenuation in stimulated cortical acetylcholine release. Neurobiol. Aging, 24: 861–863. Hilke, S., Theodorsson, A., Fetissov, S., Aman, K., Holm, L., Hokfelt, T. and Theodorsson, E. (2005) Estrogen induces a rapid increase in galanin levels in female rat hippocampal formation – possibly a nongenomic/indirect effect. Eur. J. Neurosci., 21: 2089–2099. Hillered, L., Vespa, P.M. and Hovda, D.A. (2005) Translational neurochemical research in acute human brain injury: the current status and potential future for cerebral microdialysis. J. Neurotrauma., 22: 3–41.

505 Holmes, A., Heilig, M., Rupniak, N.M., Steckler, T. and Griebel, G. (2003) Neuropeptide systems as novel therapeutic targets for depression and anxiety disorders. Trends Pharmacol. Sci., 24: 580–588. Honey, G.D. and Fletcher, P.C. (2006) Investigating principles of human brain function underlying working memory: what insights from schizophrenia? Neuroscience, 139: 59–71. Hori, T., Subramaniam, S., Srivastava, L.K. and Quirion, R. (2000) Behavioral and neurochemical alterations following repeated phencyclidine administration in rats with neonatal ventral hippocampal lesions. Neuropharmacology, 39: 2478–2491. Hows, M.E., Lacroix, L., Heidbreder, C., Organ, A.J. and Shah, A.J. (2004) High-performance liquid chromatography/ tandem mass spectrometric assay for the simultaneous measurement of dopamine, norepinephrine, 5-hydroxytryptamine and cocaine in biological samples. J. Neurosci. Methods, 138: 123–132. Hows, M.E., Organ, A.J., Murray, S., Dawson, L.A., Foxton, R., Heidbreder, C., Hughes, Z.A., Lacroix, L. and Shah, A.J. (2002) High-performance liquid chromatography/tandem mass spectrometric assay for the rapid high sensitivity measurement of basal acetylcholine from microdialysates. J. Neurosci. Methods, 121: 33–39. Ichikawa, J., Dai, J. and Meltzer, H.Y. (2000) Acetylcholinesterase inhibitors are neither necessary nor desirable for microdialysis studies of brain acetylcholine. Curr. Sep., 19: 37–44. Ichikawa, J., Dai, J. and Meltzer, H.Y. (2002a) 5-HT1A and 5-HT2A receptors minimally contribute to clozapine-induced acetylcholine release in rat medial prefrontal cortex. Brain Res., 939: 34–42. Ichikawa, J., Dai, J., O’Laughlin, I.A., Fowler, W.L. and Meltzer, H.Y. (2002b) Atypical, but not typical, antipsychotic drugs increase cortical acetylcholine release without an effect in the nucleus accumbens or striatum. Neuropsychopharmacology, 26: 325–339. Ichikawa, J., Li, Z., Dai, J. and Meltzer, H.Y. (2002c) Atypical antipsychotic drugs, quetiapine, iloperidone, and melperone, preferentially increase dopamine and acetylcholine release in rat medial prefrontal cortex: role of 5-HT1A receptor agonism. Brain Res., 956: 349–357. Imperato, A. and Di Chiara, G. (1984) trans-Striatal dialysis coupled to reverse phase high performance liquid chromatography with electrochemical detection: a new method for the study of the in vivo release of endogenous dopamine and metabolites. J. Neurosci., 4: 966–977. Invernizzi, R.W. and Garattini, S. (2004) Role of presynaptic a2-adrenoceptors in antidepressant action: recent ﬁndings from microdialysis studies. Prog. Neuropsychopharmacol. Biol. Psychiatry, 28: 819–827. Ito, C. (2004) The role of the central histaminergic system in schizophrenia. Drug News Perspect., 17: 383–387. Jacobs, W.A. (1987) O-phthalaldehyde-sulﬁte derivatization of primary amines for liquid chromatography-electrochemistry. J. Chromatogr., 392: 435–441.

Javitt, D.C. (2004) Glutamate as a therapeutic target in psychiatric disorders. Mol. Psychiatry, 9: 984–997. Javitt, D.C., Duncan, L., Balla, A. and Sershen, H. (2005) Inhibition of system A-mediated glycine transport in cortical synaptosomes by therapeutic concentrations of clozapine: implications for mechanisms of action. Mol. Psychiatry, 10: 275–287. Johnson, D.E., Nedza, F.M., Spracklin, D.K., Ward, K.M., Schmidt, A.W., Iredale, P.A., Godek, D.M. and Rollema, H. (2005) The role of muscarinic receptor antagonism in antipsychotic-induced hippocampal acetylcholine release. Eur. J. Pharmacol., 506: 209–219. Josserand, V., Pe´lerin, H., de Bruin, B., Jego, B., Kuhnast, B., Hinnen, F., Duconge´, F., Boisgard, R., Beuvon, F., Chassoux, F., Daumas-Duport, C., Dolle´, E.E.F., Mabondzo, A. and Tavitian, B. (2006) Evaluation of drug penetration into the brain: a double study by in vivo imaging with positron emission tomography and using an in vitro model of the human blood-brain barrier. J. Pharmacol. Exp. Ther., 316: 79–86. Joukhadar, C. and Muller, M. (2005) Microdialysis: current applications in clinical pharmacokinetic studies and its potential role in the future. Clin. Pharmacokinet., 44: 895–913. Joyce, J.N. and Millan, M.J. (2005) Dopamine D3 receptor antagonists as therapeutic agents. Drug Discov. Today, 10: 917–925. Kalen, P., Strecker, R.E., Rosengren, E. and Bjorklund, A. (1988) Endogenous release of neuronal serotonin and 5-hydroxyindoleacetic acid in the caudate-putamen of the rat as revealed by intracerebral dialysis coupled to high-performance liquid chromatography with ﬂuorimetric detection. J. Neurochem., 51: 1422–1435. Kapur, S. and Mamo, D. (2003) Half a century of antipsychotics and still a central role for dopamine D2 receptors. Prog. Neuropsychopharmacol. Biol. Psychiatry, 27: 1081–1090. Kay, G.G. (2000) The effects of antihistamines on cognition and performance. J. Allergy Clin. Immunol., 105: S622–S627. Kennedy, R.T., Watson, C.J., Haskins, W.E., Powell, D.H. and Strecker, R.E. (2002) In vivo neurochemical monitoring by microdialysis and capillary separations. Curr. Opin. Chem. Biol., 6: 659–665. Khan, S.H. and Shuaib, A. (2001) The technique of intracerebral microdialysis. Methods, 23: 3–9. Kinney, G.G., Sur, C., Burno, M., Mallorga, P.J., Williams, J.B., Figueroa, D.J., Wittmann, M., Lemaire, W. and Conn, P.J. (2003) The glycine transporter type 1 inhibitor N-[3-(40 ﬂuorophenyl)-3-(40 -phenylphenoxyl]sarcosine potentiates NMDA receptor-mediated responses in vivo and produces an antipsychotic proﬁle in rodent behavior. J. Neurosci., 23: 7586–7591. Koch, H.J., Uyanik, G. and Fischer-Barnicol, D. (2005) Memantine: a therapeutic approach in treating Alzheimer’s and vascular dementia. Curr. Drug Targets CNS Neurol. Disord., 4: 499–506. Kodama, T. and Koyama, Y. (2006) Nitric oxide from the laterodorsal tegmental neurons: its possible retrograde modulation on norepinephrine release from the axon terminal of the locus coeruleus neurons. Neuroscience, 138: 245–256.

506 Kopell, B.H., Greenberg, B. and Rezai, A.R. (2004) Deep brain stimulation for psychiatric disorders. J. Clin. Neurophysiol., 21: 51–67. Kosten, T.A., Zhang, X.Y. and Kehoe, P. (2005) Neurochemical and behavioral responses to cocaine in adult male rats with neonatal isolation experience. J. Pharmacol. Exp. Ther., 314: 661–667. Kuroki, T., Meltzer, H.Y. and Ichikawa, J. (1999) Effects of antipsychotic drugs on extracellular dopamine levels in rat medial prefrontal cortex and nucleus accumbens. J. Pharmacol. Exp. Ther., 288: 774–781. Kusch-Poddar, M., Drewe, J., Fux, I. and Gutmann, H. (2005) Evaluation of the immortalized human brain capillary endothelial cell line BB19 as a human cell culture model for the blood-brain barrier. Brain Res., 1064: 21–31. Kusuhara, H. and Sugiyama, Y. (2001a) Efﬂux transport systems for drugs at the blood-brain barrier and blood-cerebrospinal ﬂuid barrier (part 1). Drug Discov. Today, 6: 150–156. Kusuhara, H. and Sugiyama, Y. (2001b) Efﬂux transport systems for drugs at the blood-brain barrier and blood-cerebrospinal ﬂuid barrier (part 2). Drug Discov. Today, 6: 206–212. Kutla`n, D., Presits, P. and Molnar-Perl, I. (2002) Behavior and characteristics of amine derivatives obtained with ophthaldehyde/3-mercaptopropionic acid and with o-phthaldehyde/N-acetyl-L-cysteine reagents. J. Chromatogr. A, 949: 235–248. Lacroix, L.P., Hows, M.E., Shah, A.J., Hagan, J.J. and Heidbreder, C.A. (2003) Selective antagonism at dopamine D3 receptors enhances monoaminergic and cholinergic neurotransmission in the rat anterior cingulate cortex. Neuropsychopharmacology, 28: 839–849. Lamberty, Y., Margineanu, D.G., Dassesse, D. and Klitgaard, H. (2003) H3 agonist immepip markedly reduces cortical histamine release, but only weakly promotes sleep in the rat. Pharmacol. Res., 48: 193–198. Langer, O. and Muller, M. (2004) Methods to assess tissuespeciﬁc distribution and metabolism of drugs. Curr. Drug Metab., 5: 463–481. Laruelle, M., Frankle, W.G., Narendran, R., Kegeles, L.S. and Abi-Dargham, A. (2005) Mechanism of action of antipsychotic drugs: from dopamine D2 receptor antagonism to glutamate NMDA facilitation. Clin. Ther., 27: S16–S24. Laszy, J., Laszlovszky, I. and Gyertyan, I. (2005) Dopamine D3 receptor antagonists improve the learning performance in memory-impaired rats. Psychopharmacology (Berl.), 179: 567–575. Levin, E.D. and Christopher, N.C. (2006) Effects of clozapine on memory function in the rat neonatal hippocampal lesion model of schizophrenia. Prog. Neuropsychopharmacol. Biol. Psychiatry, 30: 223–229. L’Heureux, R., Dennis, T., Curet, O. and Scatton, B. (1986) Measurement of endogenous noradrenaline release in the rat cerebral cortex in vivo by transcortical dialysis: effects of drugs affecting noradrenergic transmission. J. Neurochem., 46: 1794–1801. Li, Z., Huang, M., Ichikawa, J., Dai, J. and Meltzer, H.Y. (2005b) N-Desmethylclozapine, a major metabolite of

clozapine, increases cortical acetylcholine and dopamine release in vivo via stimulation of M1 muscarinic receptors. Neuropsychopharmacology, 30: 1986–1995. Li, Z., Ichikawa, J. and Meltzer, H.Y. (2003) A comparison of the effects of loxapine with ziprasidone and thioridazine on the release of dopamine and acetylcholine in the prefrontal cortex and nucleus accumbens. Psychopharmacology, 167: 315–323. Li, Z., Ichikawa, J., Dai, J. and Meltzer, H.Y. (2004) Aripiprazole, a novel antipsychotic drug, preferentially increases dopamine release in the prefrontal cortex and hippocampus in rat brain. Eur. J. Pharmacol., 493: 75–83. Li, Z., Ichikawa, J., Huang, M., Prus, A.J., Dai, J. and Meltzer, H.Y. (2005a) ACP-103, a 5-HT2A/2C inverse agonist, potentiates haloperidol-induced dopamine release in rat medial prefrontal cortex and nucleus accumbens. Psychopharmacology, 183: 144–153. Lillrank, S.M., Lipska, B.K., Kolachana, B.S. and Weinberger, D.R. (1999) Attenuated extracellular dopamine levels after stress and amphetamine in the nucleus accumbens of rats with neonatal ventral hippocampal damage. J. Neural Transm., 106: 183–196. Lindroth, P. and Mopper, K. (1979) High performance liquid chromatographic determination of subpicomole amounts of amino acids by precolumn ﬂuorescence derivatization with o-phthaldialdehyde. Anal. Chem., 51: 1667–1674. Linthorst, A.C., Flachskamm, C., Muller-Preus, P., Holsboer, F. and Reul, J.M. (1995) Effect of bacterial endotoxin and interleukin-1 beta on hippocampal serotonergic neurotransmission, behavioural activity, and free corticosterone levels: an in vivo microdialysis study. J. Neurosci., 15: 2920–2934. Linner, L., Arborelius, L., Nomikos, G.G., Bertilsson, L. and Svensson, T.H. (1999) Locus coeruleus neuronal activity and noradrenaline availability in the frontal cortex of rats chronically treated with imipramine: effect of a2-adrenoceptor blockade. Biol. Psychiatry, 46: 766–774. Lorrain, D.S., Baccei, C.S., Correa, L.D. and Bristow, L.J. (2005) Comparison of the effects of diazepam, the CRF1 antagonist CP-154,526 and the group II mGlu receptor agonist LY379268 on stress-evoked extracellular norepinephrine levels. Neuropharmacology, 48: 927–935. Lundquist, S., Renftel, M., Brillault, J., Fenart, L., Cecchelli, R. and Dehouck, M.P. (2002) Prediction of drug transport through the blood-brain barrier in vivo: a comparison between two in vitro cell models. Pharmaceutical Res., 19: 976–981. Luttgen, M., Elvander, E., Madjid, N. and Ogren, S.O. (2005) Analysis of the role of 5-HT1A receptors in spatial and aversive learning in the rat. Neuropharmacology, 48: 830–852. Maine, L.W., Antonetti, D.A., Wolpert, E.B. and Smith, C.D. (2005) Evaluation of the role of P-glycoprotein in the uptake of paroxetine, clozapine, phenytoin and carbamazapine by bovine retinal endothelial cells. Neuropharmacology, 49: 610–617. Marcus, M.M., Jardemark, K.E., Wadenberg, M.L., Langlois, X., Hertel, P. and Svensson, T.H. (2005) Combined alpha2

507 and D2/3 receptor blockade enhances cortical glutamatergic transmission and reverses cognitive impairment in the rat. Int. J. Neuropsychopharmacol., 8: 315–327. Mattay, V.S., Goldberg, T.E., Fera, F., Hariri, A.R., Tessitore, A., Egan, M.F., Kolachana, B., Callicott, J.H. and Weinberger, D.R. (2003) Catechol O-methyltransferase val158met genotype and individual variation in the brain response to amphetamine. Proc. Natl. Acad. Sci. U.S.A., 100: 6186–6191. Maubach, K. (2003) GABAA receptor subtype selective cognition enhancers. Curr. Drug Targets CNS Neurol. Disord., 2: 233–239. Mayber, H.S., Lozano, A.M., Voon, V., McNeely, H.E., Seminowicz, D., Hamani, C., Schwalb, J.M. and Kennedy, S.H. (2005) Deep brain stimulation for treatment resistant depression. Neuron, 45: 651–660. McIntyre, C.C., Savasta, M., Walter, B.L. and Vitek, J.L. (2004) How does deep brain stimulation work? Present understanding and future questions. J. Clin. Neurophysiol., 21: 40–50. Meltzer, H.Y., Li, Z., Kaneda, Y. and Ichikawa, J. (2003) Serotonin receptors: their key role in drugs to treat schizophrenia. Prog. Neuropsychopharmacol. Biol. Psychiatry, 27: 1159–1172. Mercier, C., Masseguin, C., Roux, F., Gabrion, J. and Scherrmann, J.M. (2004) Expression of P-glycoprotein (ABCB1) and Mrp1 (ABCC1) in adult rat brain: focus on astrocytes. Brain Res., 1021: 32–40. Millan, M.J. (2000) Improving the treatment of schizophrenia: focus on serotonin 5-HT1A receptors. J. Pharmacol. Exp. Ther., 295: 853–861. Millan, M.J. (2002) N-Methyl-D-aspartate receptor-coupled glycine B receptors in the pathogenesis and treatment of schizophrenia: a critical review. Curr. Drug Targets CNS Neurol. Disord., 1: 191–213. Millan, M.J. (2003) The neurobiology and control of anxious states. Prog. Neurobiol., 70: 83–244. Millan, M.J. (2004) The role of monoamines in the actions of established and ‘‘novel’’ antidepressant agents: a critical review. Eur. J. Pharmacol., 500: 371–384. Millan, M.J. (2005a) N-Methyl-D-aspartate receptors as a target for improved antipsychotic agents: novel insights and clinical perspectives. Psychopharmacology, 179: 30–53. Millan, M.J. (2005b) Serotonin 5-HT2C receptors as a target for the treatment of depressive and anxious states: focus on novel therapeutic strategies. The´rapie, 60: 441–460. Millan, M.J. (2006) Multi-target strategies for the improved treatment of depressive states: conceptual foundations and neuronal substrates, drug discovery and therapeutic application. Pharmacol. Ther., 110: 1–236. Millan, M.J., Brocco, M., Gobert, A., Joly, F., Bervoets, K., Rivet, J.M., Newman-Tancred, A., Audinot, V. and Maurel, S. (1999) Contrasting mechanisms of action and sensitivity to antipsychotics of phencyclidine versus amphetamine: importance of nucleus accumbens-localized 5-HT2A sites for induction of locomotion by PCP in the rat. Eur. J. Neurosci., 11: 4419–4432. Millan, M.J., Di Cara, B., Hill, M., Jackson, M., Joyce, J.N., Brotchie, J., McGuire, S., Crossman, A., Smith, L., Jenner,

P., Gobert, A., Peglion, J.L. and Brocco, M. (2004a) S32504, a novel naphtoxazine agonist at dopamine D3/D2 receptors: II. Actions in rodent, primate and cellular models of antiparkinson activity in comparison to ropinirole. J. Pharmacol. Exp. Ther., 309: 921–935. Millan, M.J., Gobert, A., Lejeune, F., Newman-Tancredi, A., Cussac, D., Audinot, V., Nicolas, J.P., Boutin, J.A., Lacoste, J.M. and Cordi, A. (2001) The novel, potential antidepressants, S35966 and S35967, act as serotonin and norepinephrine reuptake inhibitors and antagonists at a2-adrenoceptors. Am. Soc. Neurosci. Abstr., 27: 975.18. Millan, M.J., Gobert, A., Rivet, J.M., Adhumeau-Auclair, A., Cussa, Newman-Tancredi, A., Dekeyne, A., Nicolas, J.P. and Lejeune, F. (2000a) Mirtazapine enhances frontocortical dopaminergic and corticolimbic adrenergic, but not serotonergic, transmission by blockade of e2-adrenergic and serotonin2C receptors: a comparison to citalopram. Eur. J. Neurosci., 12: 1079–1095. Millan, M.J., Gobert, A., Roux, S., Porsolt, R., Meneses, A., Carli, M., Di Cara, B., Jaffard, R., Rivet, J.M., Lestage, P., Mocaer, E., Peglion, J.L. and Dekeyne, A. (2004b) The serotonin1A receptor partial agonist S15535 [4-(benzodioxan-5yl) 1-(indan-2-yl) piperazine] enhances cholinergic transmission and cognitive function in rodents: a combined neurochemical and behavioural analysis. J. Pharmacol. Exp. Ther., 311: 190–203. Millan, M.J., Lejeune, F. and Gobert, A. (2000b) Reciprocal autoreceptor and heteroceptor control of serotonergic, dopaminergic and noradrenergic transmission in the frontal cortex: relevance to the actions of antidepressant agents. J. Psychopharmacol., 14: 114–138. Millan, M.J., Schreiber, R., Dekeyne, A., Rivet, J.M., Bervoets, K., Mavridis, M., Sebban, C., Maurel-Remy, S., NewmanTancredi, A., Spedding, M., Muller, O., Lavielle, G. and Brocco, M. (1998) S16924 ((+)-2-{1-[2-(2,3-dihydro-benzo[1,4] dioxin-5-yloxy)-ethyl]-pyrrolidin-3yl}-1-(4-ﬂuoro-phenyl)-ethanone), a novel, potential antipsychotic with marked serotonin (5-HT)1A agonist properties I. Receptorial and neurochemical proﬁle in comparison with clozapine and haloperidol. J. Pharmacol. Exp. Ther., 286: 1341–1355. Mir, S. and Taylor, D. (1997) The adverse effects of antidepressants. Curr. Opin. Psychiatry, 10: 88–94. Moghaddam, B. (2003) Bringing order to the glutamate chaos in schizophrenia. Neuron, 40: 881–884. Moghaddam, B. and Adams, B.W. (1998) Reversal of phencyclidine effects by a group II metabotropic glutamate receptor agonist in rats. Science, 281: 1349–1352. Moghaddam, B. and Jackson, M. (2004) Effect of stress on prefrontal cortex function. Neurotox. Res., 6: 73–78. Mo¨ller, H.-J. (2003) Amisulpride: limbic speciﬁcity and the mechanism of antipsychotic atypicality. Prog. Neuropsychopharmacol. Biol. Psychiatry, 27: 1101–1111. Morilak, D.A. and Frazer, A. (2004) Antidepressants and brain monoaminergic systems: a dimensional approach to understanding their behavioural effects in depression and anxiety disorders. Int. J. Neuropsychopharmacol., 7: 193–218.

508 Mothet, J.P., Pollegioni, L., Ouanounou, G., Martineau, M., Fossier, P. and Baux, G. (2005) Glutamate receptor activation triggers a calcium-dependent and SNARE proteindependent release of the gliotransmitter, D-serine. Proc. Natl. Acad. Sci. U.S.A., 102: 5606–5611. Navailles, S., De Deurwaerdere, P., Porras, G. and Spampinato, U. (2004) In vivo evidence that 5-HT2C receptor antagonist but not agonist modulates cocaine-induced dopamine outﬂow in the rat nucleus accumbens and striatum. Neuropsychopharmacology, 29: 319–326. Nyitrai, G., Kekesi, K.A., Emri, Z., Szarics, E., Juhasz, G. and Kardos, J. (2003) GABAB receptor antagonist CGP-36742 enhances somatostatin release in the rat hippocampus in vivo and in vitro. Eur. J. Pharmacol., 478: 111–119. O’Neil, M.F. and Moore, N.A. (2003) Animal models of depression: are there any? Hum. Psychopharmacol., 18: 239–254. Oshim, A., Flachskamm, C., Reul, J.M., Holsboer, F. and Linthorst, A.C. (2003) Altered serotonergic neurotransmission but normal hypothalamic-pituitary-adrenocortical axis activity in mice chronically treated with the corticotropinreleasing hormone receptor type 1 antagonist NBI 30775. Neuropsychopharmacology, 28: 2148–2159. Pacher, P. and Kecskemeti, V. (2004) Trends in the development of new antidepressants. Is there a light at the end of the tunnel? Curr. Med. Chem., 11: 925–943. Page, M.E., Szeliga, P., Gasparini, F. and Cryan, J.F. (2005) Blockade of the mGlu5 receptor decreases basal and stressinduced cortical norepinephrine in rodents. Psychopharmacology, 179: 240–246. Parent, M.B. and Baxter, M.G. (2004) Septohippocampal acetylcholine: involved in but not necessary for learning and memory? Learn. Mem., 11: 9–20. Parent, M., Bush, D., Rauw, G., Master, S., Vaccarino, F. and Baker, G. (2001) Analysis of amino acids and catecholamines, 5-hydroxytryptamine and their metabolites in brain areas in the rat using in vivo microdialysis. Methods, 23: 11–20. Parrot, S., Bert, L., Mouly-Badina, L., Sauvinet, V., ColussiMas, J., Lambas-Senas, L., Robert, F., Bouilloux, J.P., Suaud-Chagny, M.F., Denoroy, L. and Renaud, B. (2003) Microdialysis monitoring of catecholamines and excitatory amino acids in the rat and mouse brain: recent development based on capillary electrophoresis with laser-induced ﬂuorescence detection – a mini-review. Cell. Mol. Neurobiol., 23: 793–804. Partridge, W.M. (2003) Blood-brain barrier drug targeting: the future of brain drug development. Mol. Interv., 3: 90–105. Peerdeman, S.M., van Tulder, M.W. and Vandertop, W.P. (2003) Cerebral microdialysis as a monitoring method in subarachnoid haemorrhage patients, and correlation with clinical events – a systematic review. J. Neurol., 250: 797–805. Pepeu, G. and Giovannini, G. (2004) Changes in acetylcholine extracellular levels during cognitive processes. Learn. Mem., 23: 21–27. Pepicelli, O., Raiteri, M. and Fedele, E. (2004) The NOS/sGC pathway in the rat central nervous system: a microdialysis overview. Neurochem. Int., 45: 787–797.

Perrault, G., Depoortere, R., Morel, E., Sanger, D.J. and Scatton, B. (1997) Psychopharmacological proﬁle of amisulpride: an antipsychotic drug with presynaptic D2/D3 dopamine antagonist activity and limbic selectivity. J. Pharmacol. Exp. Ther., 280: 73–82. Plock, N. and Kloft, C. (2005) Microdialysis – theoretical background and recent implementation in applied lifesciences. Eur. J. Pharm. Sci., 25: 1–24. Pomper, M.G. and Lee, J.S. (2005) Small animal imaging in drug development. Curr. Pharm. Des., 11: 3247–3272. Potschka, H., Fedrowitz, M. and Loscher, W. (2002) P-glycoprotein-mediated efﬂux of phenobarbital, lamotrigine, and felbamate at the blood-brain barrier: evidence from microdialysis experiments in rats. Neurosci. Lett., 327: 173–176. Powell, P.R. and Ewing, A.G. (2005) Recent advances in the application of capillary electrophoresis to neuroscience. Anal. Bioanal. Chem., 382: 581–591. Presgraves, S.P., Borwege, S., Millan, M.J. and Joyce, J.N. (2004) Involvement of dopamine D2/D3 receptors and BDNF in the neuroprotective effects of S32504 and pramipexole against 1-methyl-4-phenylpyridinium in terminally differentiated SH-SY5Y cells. Exp. Neurol., 190: 157–170. Rao, V.V., Dahlheimer, J.L., Bardgett, M.E., Snyder, A.Z., Finch, R.A., Sartorelli, A.C. and Piwnica-Worms, D. (1999) Choroid plexus epithelial expression of MDRI P glycoprotein and multidrug resistance-associated protein contribute to the blood-cerebrospinal-ﬂuid drug-permeability barrier. Proc. Natl. Acad. Sci. U.S.A., 96: 3900–3905. Rea, K., Cremers, T.I. and Westerink, B.H. (2005) HPLC conditions are critical for the detection of GABA by microdialysis. J. Neurochem., 94: 672–679. Robert, F., Bert, L., Denoroy, L. and Renaud, B. (1995) Capillary zone electrophoresis with laser-induced ﬂuorescence detection for the determination of nanomolar concentrations of noradrenaline and dopamine: application to brain microdialysis analysis. Anal. Chem., 67: 1838–1844. Robert, F., Bert, L., Parrot, S., Denoroy, L., Stoppini, L. and Renaud, B. (1998) Coupling on-line brain microdialysis, precolumn derivatization and capillary electrophoresis for routine minute sampling of O-phosphoethanolamine and excitatory amino acids. J. Chromatogr. A, 817: 195–203. Ronaldson, P.T. (2004) Cellular localization and functional expression of P-glycoprotein in rat astrocyte cultures. J. Neurochem., 89: 788–800. Rossetti, Z.L. and Carboni, S. (2005) Noradrenaline and dopamine elevations in the rat prefrontal cortex in spatial working memory. J. Neurosci., 25: 2322–2329. Routledge, C. and Marsden, C.A. (1987) Comparison of the effects of selected drugs on the release of hypothalamic adrenaline and noradrenaline measured in vivo. Brain Res., 426: 103–111. Rowley, H.L., Kilpatrick, I.C., Needham, P.L. and Heal, D.J. (1998) Elevation of extracellular cortical noradrenaline may contribute to the antidepressant activity of zotepine: an in vivo microdialysis study in freely moving rats. Neuropharmacology, 37: 937–944.

509 Rowley, H.L., Martin, K.F. and Marsden, C.A. (1995) Determination of in vivo amino acid neurotransmitters by high-performance liquid chromatography with o-phthalaldehyde-sulﬁte derivatisation. J. Neurosci. Methods, 57: 96–99. Rupniak, N.M., Fisher, A., Boyce, S., Clarke, D., Pike, A., O’Connor, D. and Watt, A. (2003) P-glycoprotein efﬂux reduces the brain concentration of the substance P (NK1 receptor) antagonists SR140333 and GR205171: a comparative study using mdr1a–/– and mdr1a+/+ mice. Behav. Pharmacol., 14: 457–463. Rupniak, N.M.J. and Kramer, M.S. (1999) Discovery of the antidepressant and anti-emetic efﬁcacy of substance P receptor (NK1) antagonists. Trends Pharmacol. Sci., 20: 485–490. Sarter, M., Nelson, C.L. and Bruno, J.P. (2005) Cortical cholinergic transmission and cortical information processing in schizophrenia. Schizophr. Bull., 31: 117–138. Sawchuk, R.J. and Elmquist, W.F. (2000) Microdialysis in the study of drug transporters in the CNS. Adv. Drug Deliv. Rev., 45: 295–307. Schechter, L.E., Dawson, L.A. and Harder, J.A. (2002) The potential utility of 5-HT1A receptor antagonists in the treatment of cognitive dysfunction associated with Alzheimer’s disease. Curr. Pharm. Des., 8: 139–145. Schmitt, U., Abou, E.-E., Guo, L.J., Galvinas, H., Krajcsi, P., Baron, J.M., Tillmann, C., Hiemke, C., Langguth, P. and Hartter, S. (2006) Cyclosporine A (CsA) affects the pharmacodynamics and pharmacokinetics of the atypical antipsychotic amisulpride probably via inhibition of P-glycoprotein (P-gp). J. Neural. Transm., 113: 787–801. Schoemaker, H., Claustre, Y., Fage, D., Rouquier, L., Chergui, K., Curet, O., Oblin, A., Gonon, F., Carter, C., Benavides, J. and Scatton, B. (1997) Neurochemical characteristics of amisulpride, and atypical dopamine D2/D3 receptor antagonist with both presynaptic and limbic selectivity. J. Pharmacol. Exp. Ther., 280: 83–97. Schoepp, D.D. (2001) Unveiling the functions of presynaptic metabotropic glutamate receptors in the central nervous system. J. Pharmacol. Exp. Ther., 299: 12–20. Schwieler, L., Engberg, G. and Erhardt, S. (2004) Clozapine modulates midbrain dopamine neuron ﬁring via interaction with the NMDA receptor complex. Synapse, 52: 114–122. Serra, M., Pisul, M.G., Dazzi, L., Purdy, R.H. and Biggio, G. (2002) Prevention of the stress-induced increase in the concentration of neuroactive steroids in rat brain by long-term administration of mirtazapine but not of ﬂuoxetine. J. Psychopharmacol., 16: 133–138. Sestini, S., Ramat, S., Formiconi, A.R., Ammannati, F., Sorbi, S. and Pupi, A. (2005) Brain networks underlying the clinical effects of long-term subthalamic stimulation for Parkinson’s disease: a 4-year follow-up study with rCBF SPECT. J. Nucl. Med., 46: 1444–1454. Shah, A.J., de Biasi, V., Taylor, S.G., Roberts, C., Hemmati, P., Munton, R., West, A., Routledge, C. and Camilleri, P. (1999) Development of a protocol for the automated analysis of amino acids in brain tissue samples and microdialysates. J. Chromatogr. B, 735: 133–140.

Shah, Y.B. and Marsden, C.A. (2004) The application of functional magnetic resonance imaging to neuropharmacology. Curr. Opin. Pharmacol., 4: 517–521. Sharp, T., Zetterstrom, T. and Ungerstedt, U. (1986) An in vivo study of dopamine release and metabolism in rat brain regions using intracerebral dialysis. J. Neurochem., 47: 113–122. Shirazi-Southall, S., Rodriguez, D.E. and Nomikos, G.G. (2002) Effects of typical and atypical antipsychotics and receptor selective compounds on acetylcholine efﬂux in the hippocampus of the rat. Neuropsychopharmacology, 26: 583–594. Shou, M., Smith, A.D., Shackman, J.G., Peris, J. and Kennedy, R.T. (2004) In vivo monitoring of amino acids by microdialysis sampling with on-line derivatization by naphthalene-2,3-dicarboxaldehyde and rapid micellar electrokinetic capillary chromatography. J. Neurosci. Methods, 138: 189–197. Smith, B.J., Doran, A.C., McLean, S., Tingley, F.D. III, O’Neill, B.T. and Kajiji, S.M. (2001) P-glycoprotein efﬂux at the blood-brain barrier mediates differences in brain disposition and pharmacodynamics between two structurally related neurokinin1 receptor antagonists. J. Pharmacol. Exp. Ther., 298: 1252–1259. Sossi, V. and Ruth, T.J. (2005) Micropet imaging: in vivo biochemistry in small animals. J. Neural Transm., 112: 319–330. Sramek, J.J., Zarotsky, V. and Cutler, N.R. (2002) Generalised anxiety disorder: treatment options. Drugs, 62: 1635–1648. Steward, C.A., Marsden, C.A., Prior, M.J., Morris, P.G. and Shah, Y.B. (2005) Methodological considerations in rat brain BOLD contrast pharmacological MRI. Psychopharmacology, 180: 687–704. Stip, E., Chouinard, S. and Boulay, L.J. (2005) On the trail of a cognitive enhancer for the treatment of schizophrenia. Prog. Neuropsychopharmacol. Biol. Psychiatry, 29: 219–232. Stormer, E., von Moltke, L.L., Perloff, M.D. and Greenblatt, D.J. (2001) P-glycoprotein interactions of nefazodone and trazodone in cell culture. J. Clin. Pharmacol., 41: 708–714. Stouch, T.R. and Gudmundsson, O. (2002) Progress in understanding the structure-activity relationship of P-glycoprotein. Adv. Drug Deliv. Rev., 54: 315–328. Suzuki, T., Ohnuma, T., Abe, S., Yamaguchi, M., Hori, T. and Arai, H. (2004) Glutamatergic and GABAergic dysfunctions in animal models of schizophrenia – comparison of data with those from schizophrenia. Curr. Top. Pharmacol., 8: 91–120. Svensson, T.H. (2003) Alpha-adrenoceptor modulation hypothesis of antipsychotic atypicality. Prog. Neuropsychopharmacol. Biol. Psychiatry, 27: 1145–1158. Takada, Y., Yoshida, M., Sakairi, M. and Koizumi, H. (1995) Detection of gamma-aminobutyric acid in a living rat brain using in vivo microdialysis-capillary electrophoresis/mass spectrometry. Rapid Commun. Mass. Spectrom., 9: 895–896. Takano, M., Yumoto, R. and Murakami, T. (2006) Expression and function of efﬂux drug transporters in the intestine. Pharmacol. Ther., 109: 137–161.

510 Tamai, I. and Tsuji, A. (2000) Transporter-mediated permeation of drugs across the blood-brain barrier. J. Pharm. Sci., 89: 1371–1388. Tanaka, S. (2006) Dopaminergic control of working memory and its relevance to schizophrenia: a circuit dynamics perspective. Neurosciences, 139: 153–171. Tashiro, M., Mochizuki, H., Iwabuchi, K., Sakurada, Y., Itoh, M., Watanabe, T. and Yanai, K. (2002) Roles of histamine in regulation of arousal and cognition: functional neuroimaging of histamine H1 receptors in human brain. Life Sci., 72: 409–414. Tatsumi, M., Jansen, K., Blakely, R.D. and Richelson, E. (1999) Pharmacological proﬁle of neuroleptics at human monoamine transporters. Eur. J. Pharmacol., 368: 277–283. Timmerman, W. and Westerink, B.H.C. (1997). Brain microdialysis of GABA and glutamate: what does it signify? Synapse, 27: 242–261. Uhr, M. and Grauer, M.R.T. (2003) abcd1ab P-glycoprotein is involved in the uptake of citalopram and trimipramine into the brain of mice. J. Psychiatric Res., 37: 179–185. Ungerstedt, U. (1991) Microdialysis – principles and applications for studies in animals and man. J. Intern. Med., 230: 365–373. Ungerstedt, U. and Rostami, E. (2004) Microdialysis in neurointensive care. Curr. Pharm. Des., 10: 2145–2152. Uutela, P., Reinila, R., Piepponen, P., Ketola, R.A. and Kostiainen, R. (2005) Analysis of acetylcholine and choline in microdialysis samples by liquid chromatography/tandem mass spectrometry. Rapid Commun. Mass Spectrom., 19: 2950–2956. Van der Sandt, I.C., Smolders, R., Nabulsi, L., Zuideveld, K.P., de Boer, A.G. and Breimer, D.D. (2001) Active efﬂux of the 5-HT1A receptor agonist ﬂesinoxan via P-glycoprotein at the blood-brain barrier. Pharm. Sci., 14: 81–86. Van Kampen, J.M. and Robertson, H.A. (2005) A possible role for dopamine D3 receptor stimulation in the induction of neurogenesis in the adult rat substantia nigra. Neuroscience, 136: 381–386. Vayer, P., Duret, G., Junges, C., Decle`ves, X. and Bertrand, M. (2004) In silico model for prediction of in vivo brain permeability. Drug Metab. Rev., 36(Suppl 1): p251. Volkmann, J. (2004) Deep brain stimulation for the treatment of Parkinson’s disease. J. Clin. Neurophysiol., 21: 6–17. Volkow, N.D., Wang, G.J., Fowler, J.S., Learned-Coughlin, S., Yang, J., Logan, J., Schlyer, D., Gatley, J.S., Wong, C., Zhu, W., Pappas, N., Schueller, M., Jayne, M., Carter, P., Warner, D., Ding, Y.S., Shea, C. and Xu, Y. (2005) The slow and long-lasting blockade of dopamine transporters in human brain induced by the new antidepressant drug radafaxine predict poor reinforcing effects. Biol. Psychiatry, 57: 640–646. Wang, J.S., Taylor, R., Ruan, Y., Donovan, J.L., Markowitz, J.S. and Lindsay and De Vane, C. (2004) Olanzapine penetration into brain is greater in transgenic Abcb1a P-glycoprotein-deﬁcient mice than FVB1 (wild-type) animals. Neuropsychopharmacology, 29: 551–557.

Watanabe, K., Sawano, T., Endo, T., Sakata, M. and Sato, J. (2002) Studies on intestinal absorption of sulpiride (1): carrier-mediated uptake of sulpiride in the human intestinal cell line Caco-2. Biol. Pharm. Bull., 25: 885–890. Watanabe, K., Sawano, T., Jinriki, T. and Sato, J. (2004) Studies on intestinal absorption of sulpiride (3): intestinal absorption of sulpiride in rats. Biol. Pharm. Bull., 27: 77–81. Weber, C.C., Kressmann, S., Ott, M., Fricker, G. and Muller, W.E. (2005) Inhibition of P-glycoprotein function by several antidepressants may not contribute to clinical efﬁcacy. Pharmacopsychiatry, 38: 293–300. Weiner, D.M., Meltzer, H.Y., Veinbergs, I., Donohue, E.M., Spalding, T.A., Smith, T.T., Mohell, N., Harvey, S.C., Lameh, J., Nash, N., Vanover, K.E., Olsson, R., Jayathilake, K., Lee, M., Levey, A.I., Hacksell, U., Burstein, E.S., Davis, R.E. and Brann, M.R. (2004) The role of M1 muscarinic receptor agonism of N-desmethylclozapine in the unique clinical effects of clozapine. Pschopharmacology, 177: 207–216. Weiss, J., Dormann, S.M., Martin-Facklam, M., Kerpen, C.J., Ketabi-Kiyanvash, N. and Haefeli, W.E. (2003) Inhibition of P-glycoprotein by newer antidepressants. J. Pharmacol. Exp. Ther., 305: 197–204. Westerink, B.H. (1995) Brain microdialysis and its application for the study of animal behaviour. Behav. Brain Res., 70: 103–124. Westerink, B.H. (2002) Can antipsychotic drugs be classiﬁed by their effects on a particular group of dopamine neurons in the brain? Eur. J. Pharmacol., 455: 1–18. Westerink, B.H., Cremers, T.I., De Vries, J.B., Liefers, H., Tran, N. and De Boer, P. (2002) Evidence for activation of histamine H3 autoreceptors during handling stress in the prefrontal cortex of the rat. Synapse, 43: 238–243. Westerink, B.H., Damsma, G., Rollema, H., De Vries, J.B. and Horn, A.S. (1987) Scope and limitations of in vivo brain dialysis: a comparison of its application to various neurotransmitter systems. Life Sci., 41: 1763–1776. Westerink, B.H. and Tuinte, M.H. (1986) Chronic use of intracerebral dialysis of the in vivo measurement of 3,4-dihydroxyphenylethylamine and its metabolite 3,4-dihydroxyphenylacetic acid. J. Neurochem., 46: 181–185. Williams, J.B., Mallorga, P.J., Jeffrey Conn, P., Pettibone, D.J. and Sur, C. (2004) Effects of typical and atypical antipsychotics on human glycine transporters. Schizophr. Res., 71: 103–112. Wood, P.L. and Altar, A. (1988) Dopamine release in vivo from nigrostriatal, mesolimbic, and mesocortical neurone: utility of 3-methoxytyramine measurements. Pharmacol. Rev., 40: 163–187. Yasui-Furukori, N., Mihara, K., Takahata, T., Suzuki, A., Nakagami, T., De Vries, R., Tateishi, T., Kondo, T. and Kaneko, S. (2004) Effects of various factors on steady-state plasma concentration of risperidone and 9-hydroxyrisperidone: lack of impact of MDR-1 genotypes. Br. J. Clin. Pharmacol., 57: 569–575.

511 Zetterstrom, T., Sharp, T., Marsden, C.A. and Ungerstedt, U. (1983) In vivo measurement of dopamine and its metabolites by intracerebral dialysis: changes after D-amphetamine. J. Neurochem., 41: 1769–1773. Zhang, M.Y. and Beyer, C.E. (2006) Measurement of neurotransmitters from extracellular ﬂuid in brain by in vivo

microdialysis and chromatography-mass spectrometry. J. Pharm. Biomed. Anal., 40: 492–499. Zocchi, A., Fabbri, D. and Heidbreder, C.A. (2005) Aripiprazole increases dopamine but not noradrenaline and serotonin levels in the mouse prefrontal cortex. Neurosci. Lett., 387: 157–161.

CHAPTER 6.2

Use of microdialysis in drug discovery and development: industry and regulatory perspectives Hans Rollema1, and Chandra S. Chaurasia2 1

Department of Neuroscience, Pfizer Global Research and Development, Groton Laboratories, CT, USA 2 Division of Bioequivalence, Food and Drug Administration, Rockville, MD, USA

Abstract: This chapter discusses the role of microdialysis in the discovery and development of drugs that act on the central nervous system (CNS). The focus is on strategies that will help to select high-quality compounds for development and that contribute to reducing the current high attrition rate in the pharmaceutical industry. Pharmacodynamic and pharmacokinetic microdialysis applications are discussed to illustrate how and when microdialysis can be used in the different stages of the discovery and development process. Since microdialysis data are likely to become an important part of new drug submissions, and thus may potentially contribute to the FDA Critical Path Initiative to facilitate innovation in drug development, some regulatory aspects will be discussed as well.

surprising that unforeseen toxicity, unrelated to the mechanism of action of the drug, is one of the main causes of drug candidate failure, it is somewhat less expected that poor biopharmaceutical properties and lack of in vivo efﬁcacy are other main reasons for the considerable loss of early lead candidates. Early identiﬁcation of compounds or chemical series that are likely to have good biopharmaceutical properties has thus a signiﬁcant impact on the attrition rate, since this selection process determines the potential success of a drug for a large part (Lipinski, 2001). From a drug development perspective, microdialysis is an important tool that is not only used to determine in vivo pharmacodynamic and pharmacokinetic properties of candidate drugs, but also to improve optimal candidate selection from a large number of compounds. Compared with highthroughput methods, microdialysis is a slow, labor-intensive technique that is not suitable for screening a large number of compounds. However, it can successfully address speciﬁc questions for selected compounds in each stage of the

I. The drug discovery process and attrition The great majority of newly synthesized compounds that are selected for further development will not become marketed drugs, since their development is halted in one of the development stages. This results in a very high attrition rate, which is one of the major contributing factors to the spiraling costs of drug development. Recent calculations, which take into account the costs of failed candidates, estimate that the overall costs to discover, develop, and launch a single new drug, that is, the investment per successful compound, is currently nearly US$ 1.7 billion, a more than 50% increase over the estimated average costs over the 1995–2000 period (DiMasi et al., 2003; Gilbert et al., 2003; FDA, 2004). One of the major challenges facing the pharmaceutical industry is thus to analyze why attrition occurs and how to reduce attrition to an acceptable rate. Whereas it is not Corresponding author: E-mail: hans.rollema@pﬁzer.com

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

513

DOI: 10.1016/S1569-7339(06)16027-0 Copyright 2007 Elsevier B.V. All rights reserved

514

discovery and development process. With regard to its impact on the three main attrition factors, that is, toxicity, poor bioavailability, and lack of efﬁcacy, microdialysis holds the greatest potential in the assessment of in vivo bioavailability and in vivo activity, and thus proof of concept, at any point in the discovery and development process (Fig. 1). The tremendous progress in combinatorial chemistry, computational methods, and highthroughput screening techniques has made it possible to rapidly synthesize and characterize a large number of compounds and to focus on the identiﬁcation of highly potent and selective compounds based on in vitro binding afﬁnities for the target receptor. However, increased in vitro potencies of candidates can be offset by poor druglike properties, such as high lipophilicity, low solubility, low permeability, and inadequate absorption, which together result in poor bioavailability and unacceptably low in vivo potency, ultimately leading to compound attrition. Recent advances to assess the absorption, distribution, metabolism, and excretion (ADME assays) characteristics of a compound in in silico computational methods and

in vitro assays have greatly improved the quality of predicting the desired biopharmaceutical proﬁle of a compound. Lipinski and colleagues have pioneered the use of computational and in vitro methods for the estimation of solubility and permeability parameters that determine their biopharmaceutical properties (Lipinski et al., 1997; Lipinski and Hopkins, 2004). Based on an analysis of the properties of orally active drugs, Lipinski et al. (1997) formulated the ‘Rule of Five’, a guideline for minimal solubility and permeability properties that aid in the prediction of compounds to have drug-like absorption and permeation properties: molecular weight o500, cLog po5, H-bond donors o5, H-bond acceptors o10 (i.e., all properties are multiples of 5). The rule poses that if only molecules that meet most of these criteria are selected to advance to the next phase, compounds will have a much better survival rate through the later development stages. These approaches have been proven successful, given the fact that compound attrition due to failures in ADME has signiﬁcantly declined in recent years. However, this success is counteracted by the very high clinical failure rate, in particular, among CNS

In Vitro/In Silico Potency,, Selectivity,, Functionality - PSA, LogP, Rule of 5, MW, CACO-2, MCDK, ADME, In Vivo Protein Binding CSF, Brain tissue Fu, KO mice, PK-Microdialysis, Behavior, Transmitter Release PD-Microdialysis

Discovery compounds

Target identification ? Hits ⇒ ? Leads Screening ⇒ Lead Optimization Candidate Nomination

Preclinical Preclinical

Development Compounds

First in Human (Phase I) First in Patient (Phase II) Phase II Phase III

Launched drugs

Market introduction Phase IV g Surveillance Post-Marketing Other Indications

Clinical Clinical

Fig. 1. Role of microdialysis at all stages of the drug discovery and development process: to select and characterize candidate compounds, to analyze PK–PD relationship for development compounds, and to further characterize or differentiate marketed drugs.

515

drugs. Therefore, extending predictions beyond in vitro bioavailability and especially efﬁcacy estimates by taking into account actual in vivo measures early in the discovery stage would obviously greatly facilitate the selection process. An early readout of actual in vivo activities by including assessments of pharmacokinetic and/or pharmacodynamic properties of representatives of a chemical class can thus be extremely advantageous and cost effective, since compounds with optimal potency, selectivity, and pharmacokinetic proﬁle will greatly reduce the attrition risk. In practice, potency and biopharmaceutical properties have to be weighed against the novelty and medical need of a treatment, since suboptimal properties may be acceptable for innovative approaches to meet a high medical need. II. Microdialysis in drug discovery and development Intracerebral microdialysis was originally developed by Ungerstedt in the 1980s as a sampling method to study in vivo neurotransmitter release and the majority of microdialysis applications have been pharmacological studies that explore neurochemical dynamics and the mechanism of action of drugs. Microdialysis has signiﬁcant advantages over blood sampling or post mortem techniques in that it can sample from the target area of the live animal to measure both the biochemical effects (in vivo pharmacodynamics) and the free concentrations of the drug (in vivo pharmacokinetics). II.A. Pharmacodynamic applications The ability to measure changes in the extracellular levels of a wide variety of endogenous compounds, including neurotransmitters, has made microdialysis the preferred method for examining neurochemical responses to compounds that act on the CNS. Since the effect on transmitters is one of the key in vivo endpoints in the evaluation of centrally acting drugs, intracerebral microdialysis is increasingly used in the pharmaceutical industry for evaluating novel CNS compounds. Its primary applications are the in vivo characterization of

mechanism of action, the estimation of in vivo potency and selectivity in comparison with pharmacological standards or marketed compounds, and demonstration of the presence or absence of certain pharmacological properties. Each application will require a different level of microdialysis support. For instance, compounds with a known effect on transmitters, thought to be related to their efﬁcacy, are more easily examined than compounds with an unprecedented mechanism of action. In the latter case, an extensive neurochemical proﬁling, in which effects on different neurotransmitters in different brain areas are assessed, may be necessary. Numerous applications are described in detail in this volume and the chapter by Millan et al. (2006) is an excellent overview of the use of microdialysis for measuring effects of CNS drugs on various transmitters, in particular, antipsychotics. Here we illustrate how microdialysis has been used in the discovery and development of varenicline, a partial agonist at a4b2 nicotinic acetylcholine receptors (nAChRs), as a novel aid for smoking cessation (Coe et al., 2005a, b). There is a high medical need for such a drug, since most smokers repeatedly fail to quit despite the availability of smoking cessation aids. The rationale for developing an a4b2 nAChR partial agonist was based on the knowledge that a4b2 nAChRs mediate the ﬁrst step in the reinforcing effects of nicotine. A partial agonist is expected to have agonist and antagonist activity at the high-afﬁnity a4b2 nAChRs in the mesolimbic dopamine system, which mediates the pharmacological and behavioral effects of nicotine. As an agonist, it relieves the craving and withdrawal symptoms in smokers who try to quit by partially activating a4b2 nAChRs, whereas as an antagonist, it reduces the reinforcing effects of tobacco use by blocking the effects of nicotine at a4b2 nAChRs. The screening strategy (Coe et al., 2005a, b) involved the identiﬁcation of compounds with potent in vitro binding afﬁnity at a4b2 nAChRs and assessment of their functional activity by electrophysiological methods in oocytes or cell lines expressing a4b2 nAChRs, followed by in vitro and in vivo studies of selected compounds (Table 1). A starting point for the synthetic program was ()-cytisine, a natural compound found in

516 Table 1. Key parameters in the discovery of varenicline from in vitro potency and functional activity to in vivo pharmacodynamic and pharmacokinetic assessment a4b2 receptor

Dopamine release

In vitro electrophysiology In vitro [3H]-DA Agonist activity 0.45 Antagonist activity 0.54 (10+10 mM Nic)

Ex vivo+PKturnover

In vivo+PKmicrodialysis

0.51 0.40 0.60 0.47 (10+10 mM Nic) 0.47 (1.8+1 mg/kg nicotine) 0.41 (1+0.3 mg/kg nicotine)

Relative agonist and antagonist efﬁcacies of varenicline are expressed as the fraction of the maximal nicotine response ( ¼ 1). Antagonist efﬁcacy was assessed at the indicated concentrations or doses of varenicline and nicotine.

H N

H N

H N

=

N O (-)-cytisine

1

3 4 2 nicotinic

OH

N H

H N

O

N

varenicline

HN

HN ~ =

N

~ =

OH morphine

2 2a antinociceptive

Fig. 2. Structural similarities between morphine, cytisine, benzazepine 3, and varenicline (Coe et al., 2005a, b).

numerous plant species, known to have partial agonist activity at a4b2 nAChRs (Papke and Heinemann, 1994). Interestingly, cytisine was studied in the 1960s for smoking cessation, but did not exhibit robust efﬁcacy, most likely due to unfavorable biopharmaceutical properties, in particular, poor brain penetration. We focused our attention therefore on compounds that combine partial agonist activity with optimal physicochemical properties to identify potent, orally bioavailable compounds. Fig. 2 shows the striking resemblance between the carbon framework of ()-cytisine and derivatives such as 1 and substructures of morphine 2, 2a, as well as benzazapine 3, which had been reported to be devoid of morphine-like antinociceptive activity. Recognizing the relationship of 3 to ()-cytisine and 1, Coe et al. (2005a) examined these compounds and found that 3

has nicotinic-like activity. Further exploration of benzazepine (3) derivatives ultimately led to compounds with the desired partial agonist proﬁle, for example, varenicline (Fig. 2). Since a4b2 nAChR activation stimulates the mesolimbic dopaminergic system, dopamine release was chosen as the key pharmacodynamic endpoint, initially using an in vitro model, [3H]-dopamine release from rat brain slices (Table 1). Keeping in mind the importance of biopharmaceutical properties for the success of a compound later in development, we set out in an early stage to examine in vivo properties of those compounds that showed the desired in vitro proﬁle, using oral administration of the compounds. In this way, we obtained early readouts on oral bioavailability and in vivo activity and efﬁcacy, representing an early proof of concept at the same time. As mentioned above, conducting microdialysis studies for every compound within a discovery series is impractical; therefore, we used an ex vivo model with a higher throughput, dopamine utilization, or turnover, that is, the ratio of the tissue levels of dopamine and its metabolites dihydroxyphenylacetic acid (DOPAC) and homovanillic acid (HVA). Turnover can be used to compare potencies of compounds with the same mechanism of action and to obtain information on their oral bioavailability. Whereas dopamine turnover measures the change in tissue levels of the metabolites DOPAC and HVA vs. DA at a certain time point, microdialysis measures the extracellular levels of transmitter and metabolites over time. For partial agonists, these two measures correspond reasonably well, since the changes in tissue metabolites parallel the changes in the transmitter release, as illustrated in Fig. 3.

200

180

175

DA

160 140 120 100

1

nicotine

0.056 0.178 0.56 1.78 3.2 varenicline

5.6

1.0 nic + 3.2 var

Dose (mg/kg sc)

Extracellular levels in n. accumbens % of baseline + SEM

C. DA microdialysis

Nicotine (0.3 mg/kg sc) Varenicline po

200

Varenicline po + Nicotine (0.3 mg/kg sc)

175 150 125

HVA

DOPAC

150

125 100

75 -120

0

120

240

360 -120

0

120

240

360 -120

0

120

240

360

Time (min)

0.32 mg/kg sc NICOTINE 200

DA

HVA

DOPAC

175 150

125 100

100 75

0.1mg/kg po VARENICLINE

B. DA turnover

200

Extracellular levels in rat n. accumbens (% of basal + SEM)

DA turnrover inrat n. accumbens % of control + SEM

A. DA turnover

vehicle 0.01

0.1 Dose (mg/kg p.o.)

1

10

75 -120

0

120

240

360 -120

0

120

240

360 -120

0

120

240

360

Time (min)

Fig. 3. Effects of nicotine and varenicline on mesolimbic dopamine (rat nucleus accumbens). (A) Turnover: dose-dependent effects of varenicline and nicotine alone and combined, on the ratio of the tissue levels of dopamine and its metabolites DOPAC and HVA (top left panel). (B and C) Microdialysis: time courses for effects of varenicline and nicotine on extracellular levels of dopamine and metabolites (B, right panel), and dose–response curves for effects of varenicline and nicotine alone and combined (bottom left panel). Relative agonist efﬁcacies of varenicline vs. nicotine ranged from 40% (turnover) to 60% (release).

517

518

Both dopamine turnover and microdialysis studies demonstrated that varenicline increases dopamine release in the mesolimbic system, but does so with a signiﬁcantly lower efﬁcacy than nicotine. Maximally effective oral doses of varenicline increase dopamine release to only 40–60% of the maximal nicotine response. When combined with nicotine, varenicline dose-dependently inhibits nicotine’s effect on dopamine release to the maximum effect of varenicline alone, consistent with partial agonist properties. In addition, the time course for the effects on dopamine in the nucleus accumbens indicated that varenicline has a much slower onset and a longer lasting effect than nicotine. This difference in dopamine time courses is in agreement with the much reduced abuse liability potential of varenicline compared with nicotine, since fast rates of dopamine increase and decline have been associated with drugs of abuse (Volkow et al., 1997; Spencer et al., 2006). These data thus provided in vivo evidence for the desired proﬁle of a smoking cessation agent, that is, it produced a moderate dopamine increase to provide relief of craving and withdrawal associated with quitting smoking, and prevented full activation of a4b2 nAChRs by nicotine to prevent reinforcement by nicotine when smoking. Studies in animal models of nicotine reinforcement, such as nicotine self-administration under ﬁxed and progressive ratios, conﬁrmed the potential of varenicline as a smoking cessation agent. Results of clinical studies in smokers who had previously averaged 21 cigarettes a day for 25 years demonstrated that varenicline was superior to placebo in helping smokers to quit. In two placebo-controlled studies that included bupropion (Zybans) as a comparator, vareniclinetreated patients were more successful in giving up smoking than those treated with bupropion (Gonzales et al., 2006; Jorenby et al., 2006). Varenicline was approved by the FDA and the EMEA in 2006 as an aid to smoking cessation treatment and is marketed as the prescription drug Chantixs in the USA and as Champixs in Europe. In conclusion, this example illustrates how microdialysis can be used for the assessment of

oral bioavailability, in vivo potency, in vivo time course, and proof of concept early on in the discovery process, as well as for the further characterization of a selected candidate. The use of microdialysis was a logical choice for this program, since the mechanism of action and clinical efﬁcacy are thought to be related to changes in dopamine release. The microdialysis data were therefore also included in the pre-clinical pharmacology and abuse potential package of the new drug application (NDA) submission. Microdialysis is obviously the preferred method for the preclinical pharmacological evaluation of compounds when efﬁcacy is related to effects on transmitter release, for example, reuptake inhibitor, autoreceptor antagonists, etc.

II.B. Pharmacokinetic applications To improve the prediction of clinical outcomes from pre-clinical studies of novel compounds, knowledge of the exposure of a drug candidate in the appropriate biophase would greatly facilitate selection of compounds as well as doses for clinical studies. Since microdialysis allows sampling the drug in almost any compartment, it can help to assess the concentrations at or near the site of action. Microdialysis samples only unbound compound in the extracellular space in blood or the target tissue and is thus, in principle, the method of choice for assessing tissue pharmacokinetics. The free drug levels that reﬂect the fraction of the drug that interacts with the target are essential for correlating in vivo and in vitro data and for selecting compounds for further development based on pharmacokinetic proﬁles. In addition, microdialysis provides detailed information on the time course of drug exposure and kinetics in a single animal and as a data-rich and animalsparing in vivo method; it is increasingly applied in the pharmaceutical industry to investigate the pharmacokinetic proﬁles of drugs (Welty, 2005). For the discovery and development of centrally acting drugs, it is important to verify that a compound reaches the target in the brain and at a concentration that elicits a pharmacological response. Brain penetration is often determined by

519

measuring levels in brain homogenate in relation to plasma levels. This ratio has to be interpreted with caution, since it does not provide information on the free drug levels, especially if the molecule is highly bound to brain tissue. The measurement of drug concentrations in the CSF is commonly used as a surrogate for free fractions in the brain, but CSF levels are a function of drug transport across the blood–CSF barrier (choroid plexus), whereas free drug in the brain is a function of drug transport across the blood-brain barrier (capillary endothelium), which can have a very different permeability proﬁle. Transporters play a major role in brain uptake of CNS drugs, especially for high potency compounds, since very low doses are most susceptible to the low capacity transporters. In addition, the expression and function of transporters on the blood-brain and blood–CSF barriers are diverse and still largely unknown (Pardridge, 2005). A direct, but labor-intensive measure of blood-brain permeability is the in vivo quantiﬁcation of the permeability surface by carotid arterial injection or quantitative intravenous injection methods (Pardridge, 2003). Microdialysis sampling is thus an excellent alternative for estimating free drug levels in the target brain compartment, with the caveat that practical issues can limit its applicability. First of all, sufﬁcient time must be allowed after probe implantation to make sure that free drug

Compound B

Compound A % Recovery or % Delivery

concentrations are not measured under conditions of increased blood-brain barrier permeability due to damage caused by probe insertion. Second, for an accurate assessment of true extracellular levels, the dialysate concentrations need to be corrected for the in vivo recovery of the compound through the membrane of the probe used. Several methods have been described to estimate in vivo recovery that are discussed in detail elsewhere in this volume, but the feasibility of these correction methods depends for a large part on the properties of the compound. Highly bound, lipophilic molecules can have such a poor in vivo recovery or can stick to such an extent to tubings, probes, and syringes that it is impossible to either determine free brain levels by microdialysis or correct dialysate levels for recovery. In particular, retrodialysis methods can lead to incorrect results, as compounds can be ‘lost’ during the retrodialysis procedure and give seemingly high, but erroneous, in vivo recoveries. This is illustrated in Fig. 4 showing in vitro and in vivo recoveries of two CNS discovery compounds A and B and the migraine drug eletriptan. Clearly, in vivo recovery (‘delivery’) determinations of A and B by retrodialysis are impossible because of signiﬁcant loss of the compound when perfused through the microdialysis probe implanted in rat brain or an extremely slow equilibrium. This is not due to adsorption at tubings or other materials, since in vitro recoveries determined with

Eletriptan

100

100

100

75

75

75

50

50

50

25

25

25

0

In Vitro Recovery

0 0

1

2

3

4

In Vivo Delivery

0 0

1 2 3 Time (hours)

4

0

1

2

3

4

Fig. 4. In vitro recovery vs. in vivo delivery (retrodialysis) for compounds A, B, and eletriptan. In vitro recovery and in vivo delivery are the drug fractions collected in the dialysate from the drug in the vial and lost from the perfused drug, respectively.

520

wild-type vs. transgenic mice that lack the transporter will conﬁrm that a candidate compound is a substrate. However, compensatory pathways that do not exist in wild-type animals may confound results from knockout models. Microdialysis in wild-type animals using speciﬁc inhibitors and substrates of transporters is a promising alternative approach and several microdialysis studies have been performed to address this issue, for instance, by comparing dialysate levels in wild-type and P-gp KO mice. An elegant rat model that was originally developed by Burgio et al. (1998) to study the disposition of chemotherapeutic agents and applied by others to CNS drugs (Potschka et al., 2004, 2002; Yeo, 2005; De Lange, 2006) utilizes dual probes to detect differences in microdialysate levels collected in the presence and absence of a transport inhibitor (Fig. 5). In this model, two microdialysis probes are implanted in the left and right sides of the brain: one is perfused with aCSF and the other with the transporter inhibitor in aCSF, and drug levels in microdialysate samples from both probes are compared after drug administration. If a compound is a substrate for Pgp, its dialysate levels collected in the presence of a P-gp inhibitor will be much higher than those without inhibitor (Fig. 5). A variety of compounds can be used to disrupt the ATP-dependent transport activity, such as P-gp inhibitors (verapamil and cyclosporine), metabolic inhibitors (NaCN, 2,4-DNP, and NaN3), or

aCSF

aCSF + P-gp inhibitor

Dialysate concentration

the same equipment are normal and reach a rapid equilibrium. A comparison with in vitro and in vivo recoveries of eletriptan (Johnson et al., 2001) shows that the retrodialysis method is feasible for more hydrophilic molecules: both in vivo delivery and in vitro recovery reach a rapid equilibrium and the recovery is, as expected, lower in vivo than in vitro. Other examples of failures to estimate in vivo recoveries of high molecular weight, lipophilic compounds by reversed dialysis were shown by Yeo (2005). It seems therefore that for most CNS compounds, recovery methods that avoid reversed dialysis procedures, such as ultra-slow ﬂow methods (Cremers et al., 2006), are preferred, but it is clear that microdialysis sampling may not be applicable for certain compounds. Since other chapters in this volume discuss pharmacokinetic applications in detail, just one example will be mentioned here to illustrate the use of microdialysis for an issue that is extremely important for the development of CNS drugs, that is, assessing the impact of transporters on drug exposure. It is well recognized that carrier-mediated transport can profoundly affect drug uptake and inﬂux and that transporters play thus an important role in the disposition of CNS drugs. Efﬂux carriers that are present in the brain capillary endothelial cells of the functional bloodbrain barrier and the choroid plexus can ‘export’ CNS compounds of different chemical classes from the brain. The organic anion-transporting polypeptide (OATP) and the P-glycoprotein encoded by the MDR1 gene (P-gp) are transporters that have been most widely studied so far. These transporters can prevent CNS drugs from reaching their target tissues by effectively secreting drugs from the brain or CSF into the blood. This is even more important in view of genetic polymorphism that can affect therapeutic drug levels and thus directly impact the therapeutic safety and efﬁcacy of a drug. Therefore, it is important to know to what extent development candidates are substrates for transporters, which will affect brain levels, or inhibitors of transporters, which can cause drug– drug interactions. Several in vitro methods can predict whether compounds are likely to be substrates for the efﬂux transporter, whereas differences in brain levels in

Fig. 5. Theoretical effect of local perfusion with a P-gp inhibitor on microdialysate levels of a CNS compound that is a substrate for P-gp. Perfusion with a P-gp inhibitor (right probe) signiﬁcantly increases the extracellular levels of the compound compared with aCSF perfusion (left probe).

521

calmodulin inhibitors (triﬂuoperazine). Obviously, the availability of very selective inhibitors for speciﬁc transporter would greatly facilitate the identiﬁcation of the transporter for which the compound is a substrate. An important consideration for the interpretation of drug levels obtained with reversed dialysis of transport or metabolic inhibitors is the amount of inhibitor actually delivered via the probe and the distance it diffuses in relation to the location of the transporter. It is unlikely that perfusions with one concentration of different inhibitors will give the same concentration or spatial distribution in the extracellular space, which will thus result in a differential effect on transport activity. In addition, despite the fact that the same drug is sampled via identical probes in the same animal, the presence of inhibitors that modulate transport or uptake activity can inﬂuence the recovery of the drug, so that different in vivo recoveries in each probe can confound the results (De Lange et al., 1998).

II.C. Clinical applications There are few attempts to use microdialysis clinically on a more regular basis in drug industry, but several applications from academic research suggest that the method may be helpful in drug development. Technically and scientiﬁcally, there are no substantial differences between clinical and pre-clinical microdialysis, except for the size of the subjects, their degree of inter-individual variability, and the regulatory and ethical aspects (Benfeldt, 2005; Stahle, 2005). Pharmacokinetic microdialysis’ applications are becoming a routine for peripherally acting drugs, for example, for tissue pharmacokinetics of antibiotics (Derendorf, 2005) and for drug distribution across the skin after dermal application (Benfeldt, 2005; Stahle, 2005). However, clinical applications of microdialysis for routine questions about CNS penetration are seemingly not possible at this time due to ethical reasons. The most important clinical CNS application of microdialysis is the monitoring of biomarkers for cerebral metabolism in traumatic brain injury or stroke or patients in the neurointensive care setting

that are discussed in detail in this volume (Chapter 7.3) and elsewhere (Tolias and Bullock, 2004; Hillered et al., 2006). The combination of microdialysis with other non-invasive monitoring methods such as tomography and magnetic resonance imaging is a very promising approach that is increasingly used and will contribute to a better understanding of processes in stroke and traumatic brain injury (see also Section III.C.). III. FDA Critical Path Initiative and regulatory aspects As discussed above, the current drug discovery and development process has become lengthy, inefﬁcient, and very costly, while the vast majority of investigational drugs that enter clinical trials have failed. This high attrition rate is in part due to the limited predictive validity of several techniques used in critical phases of drug development to reliably evaluate novel drug candidates. A concerted effort is required to explore the application of new techniques and improve the drug development ‘toolkit’ and the FDA has taken the initiative for attempts to identify and prioritize the most pressing development problems and the greatest opportunities for rapid improvement. In this ‘Critical Path Initiative’, the FDA outlines the need for new scientiﬁc and technical tools to facilitate the selection of the most promising compounds earlier in the process and thus to make drug development more efﬁcient. The goal of this initiative is thus to ‘‘ensure that basic discoveries turn into new and better medical treatments by developing robust development pathways that are efﬁcient and predictable and result in products that are safe, effective, and rapidly available to patients’’ (FDA, 2004) (Fig. 6). III.A. Pharmacodynamic perspectives The contributions in this volume make it abundantly clear that microdialysis plays a crucial role in the evaluation of CNS drugs and that it is increasingly used for the pharmacokinetic and pharmacodynamic assessment of candidate compounds in drug discovery and development. Therefore, it

522

Critical Path Research

Basic Research Target Selection

Discovery Screening Hits to Leads

Preclinical Development of Candidate

Pre-IND meeting

Clinical trials Phase 1

Phase 2

Phase 3

Safety Update

End of Phase 2a and 2 Meeting IND Submission Ongoing Submission

FDA filing Approval Launch

Market Application Submission Pre-NDA Meeting

Fig. 6. Stages of the drug discovery and development process with points of interaction between FDA and sponsors. The ‘Critical Path Initiative’ coordinates efforts of researchers from FDA, industry, and academia to improve the drug development process by establishing and implementing new evaluation tools. Drug developers often meet with the agency before submitting a new investigational drug (IND) application to discuss early development plans. An IND must be ﬁled and cleared by the FDA before human testing can commence, after which a new drug application (NDA) can be ﬁled. During the clinical phase, new protocols and results of testing continue to be submitted (adapted from FDA, 2004).

can be expected that pre-clinical microdialysis data will become a regular part of submission packages for CNS drugs. Although these data have as of yet limited regulatory signiﬁcance, the agencies are prepared to consider microdialysis data as part of the regulatory ﬁling. This is deﬁnitely the case when providing pre-clinical in vivo evidence that a drug candidate is either efﬁcacious as an addiction treatment or is devoid of abuse potential. Examples of the ﬁrst are a novel drug for smoking cessation, varenicline, described in detail in this chapter and a novel drug for cocaine addiction, CPP-109 (Schiffer et al., 2003; Chaurasia, 2005). Data on the effect of these compounds on mesolimbic dopamine release are likely to play an important role in the reviews by the agencies. In addition, the use of microdialysis data that support the absence of abuse liability in novel drugs was recently recommended in a European guideline for pre-clinical studies on dependence potential, issued by the Committee on Proprietary Medicinal Products (CPMP) of the European Medicines Agency (EMEA). In this document, microdialysis is speciﬁcally included in a paragraph on in vivo studies: ‘‘y use of neuropharmacological models, e.g. microdialysis (for example

dopamine release in nucleus accumbens), neurotransmitter turnover, y’’ (EMEA, 2006). Finally, the fact that pre-clinical microdialysis data have been accepted in support of the mechanism of action of the anti-seizure drug zonisamide (ZonegranTM FDA-approved label, Physicians’ Desk Reference, 2006) is further evidence that the FDA considers microdialysis as an important tool in drug development.

III.B. Pharmacokinetic perspectives While microdialysis is not speciﬁcally mentioned in the 2004 FDA document, it is clear that this technique represents a good example of an innovative technique discussed in the Critical Path Initiative. Microdialysis, in combination with extremely sensitive and speciﬁc analytical methodologies, has the potential to signiﬁcantly improve the assessment of drug levels at the site of action over currently used bioavailability assessments, which are based on plasma sample analysis. It should be kept in mind that the Food and Drug title of the Code of Federal Regulations deﬁnes bioavailability as ‘the rate and extent to which the active

523

ingredient or active moiety is absorbed from a drug product and becomes available at the site of action’ (21. CFR Section 320.1). Time courses of free drug levels at the site of action are already being determined for drugs that act at targets that are easily accessible by microdialysis, such as in peripheral tissue (e.g., antibiotics; Derendorf, 2005) or in the skin (e.g., dermatology; Bashaw, 2005; Benfeldt, 2005; Stahle, 2005). In view of the considerable experience in this area and the emphasis of the CFR deﬁnition on the importance of knowing drug levels at the site of action, microdialysis seems a good candidate to become accepted as a novel tool for assessing drug availability. It should be noted that while microdialysis has been used for quite some time as a pharmacokinetic sampling method in the drug discovery and selection process, it will need further development and considerations by the pharmaceutical industry, the US, and other regulatory bodies before it is recognized as a novel technique with regulatory signiﬁcance. Most importantly, validation of the method by comparison and calibration of probes and correlating results from microdialysis with clinical responses will be required. Obviously, the technique will gain acceptance as it is more widely and frequently applied, using standardized protocols, so that more experience is obtained in reviewing the results of these studies (Bashaw, 2005; Chaurasia, 2005). Finally, pharmacokinetic applications of microdialysis for drugs that act upon CNS targets are limited to preclinical animal studies, where free drug levels can be measured in the brain, as discussed above.

III.C. Clinical perspectives The use of microdialysis for human clinical studies on CNS drugs at the site of action is very limited for obvious reasons and is not accepted by the FDA as a replacement for in vivo studies or for in vivo clinical equivalence studies. The required evidence (Code of Federal Regulations 21. CFR Section 320.1) for demonstrating in vivo biological activity for a NDA or bioequivalency to the reference listed drug for an abbreviated new drug application (ANDA) has thus to be provided by

traditional clinical efﬁcacy and bioavailability studies. Brain microdialysis is however increasingly applied in clinical studies in human patients with ischemia, hypoxia, or epilepsy and is described in detail in this volume and elsewhere (Tolias and Bullock, 2004; Hillered et al., 2006; Ungerstedt and Rostami, 2006). It should be mentioned here that the FDA has so far approved two microdialysis devices for clinical studies: the intracerebral microdialysis probe to collect brain ﬂuid in conjunction with a cerebral tissue monitoring system (CMA) for the measurement of intracranial glucose, lactate, and pyruvate levels, and the GlucoWatchTMBiographer, a microdialysis device that acts as a glucose sensor and alerts for high or low glucose levels (Robert, 2002; Chaurasia, 2005). Microdialysis is obviously not as generally applicable in brain studies as a non-invasive technique like imaging, which is speciﬁcally mentioned in the Critical Path Initiative as an example of an innovative technique. Brain imaging is a ‘key technology for assessing, accelerating the development of, and guiding the use of new therapeutic options’ since the ‘synergy between drug development programs and current imaging techniques can help to make drug development more cost effective’ (Mills, 2005). The combination of microdialysis sampling with an imaging technique is obviously a very powerful technology, which was recently shown in a combined PET imaging and microdialysis study in an aneurysmal subarachnoid hemorrhage patient, to measure glutamate and lactate as biomarkers in a neurointensive care setting (Sarrafzadeh et al., 2004). In summary, microdialysis data will be increasingly used in drug submissions, especially for the assessment of bioavailability in clinical studies, whereas the application for CNS drugs will be limited for now to demonstrate in vivo pharmacological activity in pre-clinical models. Further applications of microdialysis are being explored, for example, to address speciﬁc safety issues related to systemic drug delivery (Anti-Infective Drug Advisory Committee Meeting, 1998), to optimize formulations, as an adjunct to in vivo bioavailability trials (Chaurasia, 2005). While the FDA does not require microdialysis studies at this point of time, the agency is receptive to

524

microdialysis data as part of an overall pre-clinical and clinical pharmacology package in an NDA, whereas the European agencies have speciﬁed the use of microdialysis for pre-clinical abuse liability studies. IV. Conclusions Microdialysis is a practical, data-rich, animalsparing in vivo tool that was originally developed as a sampling method for neurotransmitters and allows the determination of free drug levels and pharmacological effects at the target tissue, thereby improving the interpretation of pre-clinical data and prediction of clinical dose. It is a labor-intensive, and thus not a high-throughput technique, but is used for selected compounds at different stages in the drug discovery and development process. As part of the screening strategy, a balance between the resource intensiveness of microdialysis and the pertinent information it provides must be considered to decide when and how microdialysis should be used in drug discovery and development. Microdialysis is of great value to improve the candidate selection from compounds with the required in vitro potency for the target by taking in vivo efﬁcacy and bioavailability into account. Although the clinical applications of microdialysis for compounds in development or already marketed are still limited, microdialysis is now emerging as a clinical tool, primarily for biomarkers and pharmacokinetics. In addition, the availability of different probes and the possibility to combine microdialysis with imaging or behavioral techniques offer the potential to correlate targettissue concentration with responses in a clinical setting. Microdialysis could therefore be one of the key technologies referred to in the ‘FDA Critical Path Initiative’ to accelerate the development process and make it more cost effective. Increasing the number of high-quality microdialysis data that are submitted for regulatory review and initiating discussions between academic, industrial, and regulatory agencies on how microdialysis studies will ﬁt into drug development plans, will help to develop some practical guidance.

Acknowledgments The authors thank Helen Yeo (Roche, Palo Alto, CA), Devin Welty (Allergan, Irvine, CA), Lars Sta˚hle (Astra-Zeneca, So¨derta¨lje, Sweden), Jotham Coe (Pﬁzer, Groton, CT), and Christopher Lipinski (Pﬁzer, Groton, CT) for insightful comments and suggestions. References Anti-Infective Drug Advisory Committee Meeting. (1998) Guidance documents on developing antimicrobial drugs: general considerations and individual indications. Gaitherburg, MD, July 31, http://0-www.fda.gov.lilac.une.edu/ ohrms/dockets/ac/cder98t.htm#Anti-InfectiveDrugs. Bashaw, E.D. (2005) Use of microdialysis in topical drug evaluation: regulatory outlook. In: AAPS Workshop: Microdialysis Principles, Applications and Regulatory Perspective, Nashville, TN, http://www.aapspharmaceutica.com/workshops/Microdialysis110405/bashaw.pdf. Benfeldt, E.M. (2005) Application in bioequivalency. In: AAPS Workshop: Microdialysis Principles, Applications and Regulatory Perspective, Nashville, TN (http://www.aapspharma ceutica.com/workshops/Microdialysis110405/benfeldt.pdf). Burgio, D.E., Gosland, M.P. and McNamara, P.J. (1998) Effects of P-glycoprotein modulators on etoposide elimination and central nervous system distribution. J. Pharmacol. Exp. Ther., 287: 911–917. Chaurasia, C.S. (2005) Potential role of microdialysis in drug development. In: AAPS Workshop: Microdialysis Principles, Applications and Regulatory Perspective, Nashville, TN, http: // www.aapspharmaceutica.com/workshops/Microdialsis 110405/chaurasia.pdf. Coe, J.W., Brooks, P.R., Vetelino, M.G., Wirtz, M.C., Arnold, E.P., Huang, J., Sands, S.B., Davis, T.I., Lebel, L.A., Fox, C.B., Shrikhande, A., Heym, J.H., Schaeffer, E., Rollema, H., Lu, Y., Mansbach, R.S., Chambers, L.J., Rovetti, C.C., Schulz, D.W., Tingley, F.D. and O’Neill, B.T. (2005a) Varenicline: an a4b2 nicotinic receptor partial agonist for smoking cessation. J. Med. Chem., 48: 3474–3477. Coe, J.W., Vetelino, M.G., Bashore, C.G., Wirtz, M.C., Brooks, P.R., Arnold, E.P., Lebel, L.A., Fox, C.B., Sands, S.B., Davis, T.I., Schulz, D.W., Rollema, H., Tingley, F.D. and O’Neill, B.T. (2005b) In pursuit of alpha4 beta2 nicotinic receptor partial agonists for smoking cessation: carbon analogs of ()-cytisine. Bioorg. Med. Chem. Lett., 15: 2974–2979. Cremers, T., de Vries, M., Wientjes, K.J., Westerink, B. and de Lange, E. (2006) Improvement in quantitative microdialysis; modiﬁed ultraslow versus dynamic-no-net-ﬂux microdialysis methodology in pharmacokinetic studies. In: DiChiara, G. (Ed.), Proceedings of the 11th International Conference on In Vivo Methods: Monitoring Molecules in Neuroscience. Mafﬁa Press, Cagliari, Italy, pp. 230–233.

525 De Lange, E.C.M. (2006) Microdialysis as a method to study blood-brain barrier transport mechanisms. Chapter 6.4, p. 549, this volume. De Lange, E.C.M., deBock, G., Schinkel, A.H., deBoer, A.G. and Breimer, D.D. (1998) BBB transport and P-glycoprotein functionality using MDR1A (/) and wild-type mice. Total brain versus microdialysis concentration proﬁles of rhodamine-123. Pharm. Res., 15: 1657–1665. Derendorf, H. (2005) Application in pharmacokinetic exposure response: development of new anti-infective agents. In: AAPS Workshop: Microdialysis Principles, Applications and Regulatory Perspective, Nashville, TN, http://www.aapspharmaceutica.com/workshops/Microdialysis 110405/derendorf.pdf. DiMasi, J.A., Hansen, R.W. and Grabowski, H.G. (2003) The price of innovation: new estimates of drug development costs. J. Health Econ., 22: 151–185. EMEA. (2006) In vivo studies. In: Guideline on the Non-Clinical Investigation of the Dependence Potential of Medicinal Products. European Medicines Agency, Evaluation of Medicines for Human Use, http://www.emea.eu.int/pdfs/human/ swp/9422704en.pdf, Chapter 4.1.2. FDA. (2004) Innovation or stagnation? Challenge and opportunity on the Critical Path to new medical products. US Department of Health and Human Services, Food and Drug Administration, May 2004, http://www.fda.gov/oc/ initiatives/criticalpath/whitepaper.pdf. Gilbert, J., Henske, P. and Singh, A. (2003) Rebuilding Big Pharma’s Business Model. In Vivo, the Business and Medicine Report, Windhover Information, 21: No. 10. Gonzales, D., Rennard, S.I., Nides, M., Oncken, C., Azoulay, S., Billing, C.B., Watsky, E.J., Gong, J., Williams, K.E. and Reeves, K.R. (2006) Varenicline, an a4b2 nicotinic acetylcholine receptor partial agonist, vs sustained-release bupropion and placebo for smoking cessation: a randomized controlled trial. J. Am. Med. Ass., 296: 47–55. Hillered, L., Persson, L., Nilsson, P., Ronne-Engstrom, E. and Enblad, P. (2006) Continuous monitoring of cerebral metabolism in traumatic brain injury: a focus on cerebral microdialysis. Curr. Opin. Crit. Care, 12: 112–118. Johnson, D.E., Rollema, H., Schmidt, A.W. and McHarg, A.D. (2001) Serotonergic effects and extracellular brain levels of eletriptan, zolmitriptan and sumatriptan in rat brain. Eur. J. Pharmacol., 425: 203–210. Jorenby, D.E., Hays, J.T., Rigotti, N.A., Azoulay, S., Watsky, E.J., Williams, K.E., Billing, C.B., Gong, J. and Reeves, K.R. (2006) Efﬁcacy of Varenicline, an a4b2 nicotinic acetylcholine receptor partial agonist, vs placebo or sustained-release bupropion for smoking cessation: a randomized controlled trial. J. Am. Med. Ass., 296: 56–63. Lipinski, C.A. (2001) Drug-like properties and the causes of poor solubility and poor permeability. J. Pharmacol. Toxicol. Methods, 44: 235–249, see also http:// www.pharmalabauto.com/Solubility/Presentations/presentations. html. Lipinski, C. and Hopkins, A. (2004) Navigating chemical space for biology and medicine. Nature, 432: 855–861.

Lipinski, C.A., Lombardo, F., Dominy, B.W. and Feeney, P.J. (1997) Experimental and computational approaches to estimate solubility and permeability in drug discovery and developmental settings. Adv. Drug Deliv. Rev., 23: 3–29. Millan, M.J., Panayi, F., Rivet, J.M., Di Cara, B., Cistarelli, L., Billiras, R., Girardon, S. and Gobert, S. (2006) The Role of Microdialysis in Drug Discovery: Focus on Antipsychotic Agents, Chapter 6.1, this volume. Mills, G. (2005) Regulatory opportunities and challenges of imaging as a drug development tool. CDER/FDA, May 5. Papke, R.L. and Heinemann, S.F. (1994) Partial agonist properties of cytisine on neuronal nicotinic receptors containing the b2 subunit. Mol. Pharmacol., 45: 142–149. Pardridge, W.M. (2003) Blood-brain barrier genomics and the use of endogenous transporters to cause drug penetration into the brain. Curr. Opin. Drug Discov. Dev., 6: 683–691. Pardridge, W.M. (2005) The blood-brain barrier: bottleneck in brain drug development. NeuroRx, 2: 3–14. Physicians’ Desk Reference. (2006) Zonegrans Clinical Pharmacology FDA-Approved Label, http://www.thom sonhc.com /pdrel/ librarian/PFDefaultActionId/pdrcommon. IndexSearchTranslator. Potschka, H., Baltes, S. and Lo¨scher, W. (2004) Inhibition of multidrug transporters by verapamil or probenecid does not alter blood-brain barrier penetration of levetiracetam in rats. Epilepsy Res., 58: 85–91. Potschka, H., Fedrowitz, M. and Lo¨scher, W. (2002) Pglycoprotein-mediated efﬂux of phenobarbital, lamotrigine, and felbamate at the blood-brain barrier: evidence from microdialysis experiments in rats. Neurosci. Lett., 327: 173–176. Robert, J.-J. (2002) Continuous monitoring of blood glucose. Horm. Res., 57(Suppl. 1): 81–84. Sarrafzadeh, A.S., Haux, D., Lu¨demann, L., Amthauer, H., Plotkin, M., Ku¨chler, I. and Unterberg, A.W. (2004) Cerebral ischemia in aneurysmal subarachnoid hemorrhage: a correlative microdialysis-PET study. Stroke, 35: 638–643. Schiffer, W.K., Marsteller, D. and Dewey, S.L. (2003) Subchronic low dose gamma-vinyl GABA (vigabatrin) inhibits cocaine-induced increases in nucleus accumbens dopamine. Psychopharmacology, 168: 339–343, see also http:// www.bnl.gov/CTN/GVG/CPP.asp. Spencer, T.J., Biederman, J., Ciccone, P.E., Madras, B.K., Dougherty, D.D., Bonab, A.A., Livni, E., Parasrampuria, D.A. and Fischman, A.J. (2006) PET study examining pharmacokinetics, detection and likeability, and dopamine transporter receptor occupancy of short- and long-acting oral methylphenidate. Am. J. Psychiatry, 163: 387–395. Stahle, L. (2005) Use of microdialysis in drug development and drug pproval: Industry Outlook. In: AAPS Workshop: Microdialysis Principles, Applications and Regulatory Perspective, Nashville, TN, http://www.aapspharmaceutica.com/ workshops/Microdialysis110405/Stahle.pdf.

526 Tolias, C.M. and Bullock, M.R. (2004) Critical appraisal of neuroprotection T1 injury: what have we learned? NeuroRx, 1: 71–79. Ungerstedt, U. and Rostami, E. (2006) Microdialysis in the Human Brain: Clinical Applications, Chapter 7.3, this volume. Volkow, N., Wang, G., Fischman, M., Foltin, R., Fowler, J., Abumrad, N., Vitkun, S., Logan, J., Gatley, S., Pappas, N., Hitzemann, R. and Shea, C. (1997) Relationship between subjective effects of cocaine and dopamine transporter occupancy. Nature, 386: 827–830.

Welty, D. (2005) Microdialysis: preclinical applications and development strategies. In: AAPS Workshop: Microdialysis Principles, Applications and Regulatory Perspective, Nashville, TN, http://www.aapspharmaceutica.com/workshops/ Microdialysis110405/welty.pdf. Yeo, H. (2005) Preclinical microdialysis in drug discovery – pharmacokinetic perspective. In: AAPS Workshop: Microdialysis Principles, Applications and Regulatory Perspective, Nashville, TN, http://www.aapspharmaceutica.com/workshops/Microdialysis110405/yeo.pdf.

CHAPTER 6.3

The use of brain microdialysis in antidepressant drug research Francesc Artigas and Albert Adell Department of Neurochemistry and Neuropharmacology, Institut d’Investigacions Biome`diques de Barcelona (CSIC), IDIBAPS, Barcelona, Spain

Abstract: The appropriate treatment of depression is a major challenge in health policies, given the large prevalence of this psychiatric condition. Current antidepressant treatments have two main problems: slowness of action and limited efﬁcacy. Ideally, antidepressants should exert most their clinical action in a relatively short time (e.g., 2 weeks) and be effective in the majority of treated patients. However, the most used antidepressants such as the selective serotonin reuptake inhibitors (SSRIs) induce a clinical response (reduction to half of the initial severity) in only 60% of the patients after 6 weeks of treatment. Microdialysis studies over the last 15 years have helped to determine that the limited action of antidepressant drugs is partly due to negative feedback mechanisms involving 5-HT autoreceptors. The activation of such receptors by the excess 5-HT in the extracellular brain space reduces serotonergic cell ﬁring and 5-HT release, thus attenuating the increase produced by reuptake inhibition. Chronic antidepressant treatment results in a progressive desensitization of these negative feedback mechanisms, enabling 5-HT neurons to recover their ﬁring and release activities. This process is thought to play an important role in the delayed therapeutic action of antidepressants. This chapter reviews the use of the microdialysis technique to study the mode of action of SSRIs and other marketed (noradrenaline reuptake inhibitors, serotonin, and noradrenaline reuptake inhibitors) or potential antidepressant drugs (NK1 and CRF antagonists). Overall, microdialysis has largely contributed to the current knowledge on the mode of action of antidepressant drugs and to the development of potential new therapeutic strategies in the ﬁeld. 1992). A more recent multinational study conducted in several European countries reported 6month prevalence rates of 6.9% (Lepine et al., 1997). Likewise, a study supported by the World Health Organization predicts that major depression will be the second leading cause of illnessinduced disability in 2020, after ischemic heart disease (Murray and Lopez, 1997). Moreover, a recent study of the European Brain Council shows that brain illnesses of all kind have an annual cost of h386,000 million, of which almost one-third are due to affective disorders (Andlin-Sobocki et al, 2005). In keeping with all the above, it is not surprising that antidepressant drugs rank third in the largest sales of medicines worldwide in recent

I. Introduction I.A. The challenge of major depression A large number of epidemiological studies in Europe and North America underlie the impact of major depression in modern societies. It is generally accepted that lifetime prevalence for men and women is 10% and 20%, respectively. One-year prevalence for major depressive disorder (DSMIII) varies between 2.6% and 6.2%, with lower rates for dysthymia and bipolar depression (Angst,

Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

527

DOI: 10.1016/S1569-7339(06)16028-2 Copyright 2007 Elsevier B.V. All rights reserved

528

years, with annual sales worldwide of >$20,000 million. I.B. Therapeutic action of antidepressant drugs Antidepressant drugs have been available for over 40 years, and treatment has evolved during this time with the regular introduction of new drugs that are progressively better tolerated and simpler to use. However, the therapeutic action of these agents is far from optimal. The ﬁrst generation of reuptake blockers (tricyclic antidepressants), in addition to blocking serotonin (5-HT) and/or noradrenaline (NA) reuptake into the corresponding neurons (the pharmacological property underlying their therapeutic action), interact with acetylcholine muscarinic receptors, histamine receptors and a1-adrenoceptors. These unwanted pharmacological activities are responsible for the wide range of side effects of tricyclic drugs (dry mouth, constipation, blurred vision, sedation, and orthostatic hypotension), which severely compromise the adherence to treatments and reduce the overall therapeutic effectiveness, despite some of them, like clomipramine, have been shown to be more effective in severely depressed patients than newer drugs, such as the selective serotonin reuptake inhibitors (SSRIs) citalopram or paroxetine (Danish University Antidepressant Group, 1986, 1990). The SSRIs (citalopram, ﬂuoxetine, ﬂuvoxamine, paroxetine, and sertraline) and the selective NA reuptake inhibitors (such as reboxetine) represent a conceptual advance over tricyclic drugs in that they selectively target the 5-HT or NA transporters and are devoid of afﬁnity for the receptors responsible for the side effects associated to the use of tricyclic drugs. More recently, the selective serotonin and noradrenaline reuptake inhibitors (SNRI), such as venlafaxine, milnacipran, and duloxetine (see Section IV) are drugs that block the reuptake of both amines, yet to a different extent, and therefore are more similar to some of the ﬁrst generation tricyclics, such as imipramine, amitryptyline, or clomipramine. In common with SSRIs and unlike tricyclic drugs, the SNRIs lack afﬁnity for aminergic receptors, which make these drugs safer, better tolerated, and easier to

handle than tricyclic drugs. Another class of antidepressants, illustrated by mianserin, mirtazapine, trazodone, and nefazodone, possess a therapeutic activity based upon their ability to interact with presynaptic autoregulatory (e.g., a2-adrenoceptors) or postsynaptic (e.g., 5-HT2A/2C) monoamine receptors. Yet, most antidepressant drugs (including monoamine oxidase inhibitors) share the ability to potentiate central neurotransmission mediated by NA, 5-HT, or both (Artigas, 1995; Nutt, 2002). Although the efﬁcacy of these antidepressant drugs is unquestionable, current treatment of major depression is far from being optimal. Hence, the onset of the clinical antidepressant action is slow, taking several weeks to achieve a signiﬁcant symptom reduction (Tollefson and Holman, 1994). Typically, the proportion of patients who experience a clinical response (deﬁned as a reduction to 50% of the initial severity) is 60% and the rates of symptom remission (reduction to a score in the Hamilton scale to 8 points or lower) are 35–40% after 6 weeks of effective treatment (Bech et al., 2000). Clinical improvement increases with treatment time so that the World Health Organization recommends continued treatment for 6–9 months to prevent relapses of recovered patients but frequently, a relatively high percentage of patients experience partial or incomplete responses. This increases the chances of future relapse and recurrence and worsens quality of life (Cornwall and Scott, 1997). Hence, the suboptimal efﬁcacy of existing drugs creates a medical need for more efﬁcacious and rapid antidepressant treatments. Since most antidepressant drugs block 5-HT and/or NA reuptake and therefore increase their extracellular or ‘‘synaptic’’ concentrations in brain, the use of the microdialysis technique has largely contributed to a better understanding of their mechanism of action, and particularly of the factors delaying its clinical action (Artigas, 1993). Likewise, microdialysis has become a routine technique in drug development inasmuch as a potential antidepressant action is typically associated with an increase in the extracellular 5-HT and/or NA concentrations in forebrain areas such as frontal cortex or hippocampus.

529

The present chapter reviews the evidence accumulated over the last 15 years or so in the understanding of the neurobiological basis of the clinical antidepressant action and on the development of new treatment strategies that can overcome the existing limitations.

II. SSRIs: selective but complex actions on 5-HT neurons II.A. Effects of SSRIs on extracellular 5-HT The term SSRI encompasses several chemical agents that share their ability to inhibit selectively the function of the serotonin (5-hydroxytryptamine, 5-HT) transporter. These drugs are commonly used as the ﬁrst choice treatment for major depression and other psychiatric conditions in most countries worldwide due to their efﬁcacy and better side effect proﬁle than ﬁrst generation antidepressants. The 5-HT transporter is located on the membranes of serotonergic and glial cells of the brain and other cells outside the central nervous system (CNS), such as platelets, enterochromafﬁn cells of the gut, endothelial cells, and mastocytes. The 5-HT transporter was cloned in 1991 from different cellular sources (Blakely et al., 1991; Hoffman et al., 1991) and belongs to the same family as dopamine or noradrenaline transporters

characterized by 12 transmembrane domains and intracellular N- and C-terminals (Uhl and Hartig, 1992). The cloned transporter displays the same pharmacological proﬁle as the native protein (Hyttel, 1994; Tatsumi et al., 1997) and the identity of the CNS and peripheral 5-HT transporters was soon recognized (Lesch et al., 1993). By interfering with the process of internalization of 5-HT via the 5-HT transporter, the SSRIs enhance the ratio between the extra- and intracellular compartments of 5-HT. This property has been fundamental for the use of the microdialysis technique in assessing the mechanism of action of antidepressant drugs inasmuch as 5-HT reuptake inhibition should theoretically be monitored by sampling the extracellular brain compartment with microdialysis probes. Hence, the administration of SSRIs by reverse dialysis (i.e., dissolved in the physiological ﬂuid used to perfuse the microdialysis probes) in a forebrain area such as frontal cortex increases in a concentration-dependent manner the extracellular concentration of 5-HT, making it reach a maximal level of 6- to 10-fold the basal value (Fig. 1A). However, the systemic administration of large doses of the same agents produced more moderate effects (Bel and Artigas, 1992; Malagie´ et al., 1995; Herva´s et al., 1998, 2000). In contrast, the increase in the extracellular 5-HT concentration produced by SSRIs is larger in the dorsal and median raphe nuclei of the midbrain (DR and MnR, respectively), where cell

Fig. 1. Differential effects of the SSRIs on dialysate 5-HT concentration depending on the administration route. (A) The local application of the SSRI ﬂuoxetine by reverse dialysis in the prefrontal cortex increased the dialysate 5-HT concentration in a concentration-dependent manner up to sixfold. (B) However, the administration of 10 mg/kg ﬂuoxetine i.p., a dose that fully blocks the 5-HT transporter in vivo, only doubled dialysate 5-HT. Redrawn from data in Herva´s et al. (2000).

530

bodies and dendrites of 5-HT neurons are located, than in any other forebrain area tested so far (Bel and Artigas, 1992; Malagie´ et al., 1995; Herva´s and Artigas, 1998). These two characteristics are also common to nonselective antidepressants such as the tricyclic drug clomipramine and MAO inhibitors (Adell and Artigas, 1991; Celada and Artigas, 1993), which suggested a common mechanism for all 5-HT-enhancing drugs, irrespective of their selectivity for the 5-HT transporter. The clariﬁcation of the neurobiological mechanisms involved in these differences was made possible through the use of two in vivo techniques: single unit recordings and brain microdialysis. Indeed, the systemic administration of single doses of selective and nonselective 5-HT reuptake inhibitors decreases the ﬁring frequency of serotonergic neurons of the DR (Scuve´e-Moreau and Dresse, 1979; Blier et al., 1987). Similarly, the administration of tricyclic drugs that block NA reuptake reduces the ﬁring of noradrenergic neurons of the locus coeruleus (Scuve´e-Moreau and Dresse, 1979; Quinaux et al., 1982), indicating that similar mechanisms occur in serotonergic and noradrenergic neurons after the inhibition of the respective transporters. Microdialysis studies showed that the inhibition of serotonergic cell ﬁring was due to a remarkable increase of extracellular 5-HT in the cell body area and the subsequent activation of somatodendritic 5-HT1A receptors (Adell and Artigas, 1991; Artigas et al., 1996). In pioneering dual-probe experiments, it was shown that the local application of the nonselective reuptake inhibitor clomipramine in a midbrain area sampling the DR and the MnR markedly increased extracellular 5-HT in this area and reduced simultaneously that in frontal cortex (Adell and Artigas, 1991). Subsequent dual-probe experiments revealed the relative contribution of the DR and MnR to this effect (Romero and Artigas, 1997). Hence, the local application of citalopram in the DR reduced the release of 5-HT in frontal cortex to approximately 50% of the basal value. In contrast, the local application of citalopram in the MnR had a more modest effect on hippocampal 5-HT release (reduction to 70% of baseline), suggesting that the ascending DR pathway was more sensitive to the inhibitory effect of

reuptake blockade in the cell body area (Romero and Artigas, 1997). This negative feedback triggered by the 5-HT reuptake blockers was also shown by using another experimental approach involving only one microdialysis probe in forebrain. In this experimental paradigm, the microdialysis ﬂuid contained an SSRI so as to block (at least partially) the 5-HT transporter in the area sampled by the probe. Then, the systemic administration of an SSRI reduced rather than increasing the dialysate 5-HT concentration due to the activation of the midbrain-based negative feedback (Rutter and Auerbach, 1993; Hjorth and Auerbach, 1994; Romero and Artigas, 1997). Contrary to the results obtained in dualprobe experiments, this approach did not allow the determination of the brain area involved in the negative feedback but was useful in pharmacological experiments due to its lower complexity. The large elevation of the 5-HT concentration in the extracellular raphe space is most likely due to two main factors: (1) a high density of 5-HT reuptake sites in the raphe nuclei, common to rodents and humans (Corte´s et al, 1988; Hrdina et al., 1990), and (2) the presence of a substantial physiological release of 5-HT within the raphe nuclei, which is greater than in forebrain (He´ry et al., 1982; Adell and Artigas, 1991; Adell et al., 1993; Matos et al., 1996). Moreover, while the 5-HT elevation induced by reuptake blockade in a terminal area soon reaches a concentration plateau (e.g., Fig. 1A), this does not seem to take place in the DR or MnR, which suggests the existence of differential release mechanisms between somatodendritic and terminal regions of 5-HT neurons (Tao et al., 2000). Indeed, the precise origin of the extracellular 5-HT found in the DR and MnR is not known, but likely reﬂects the release by dendrites and by the proximal segments of efferent axons within the boundaries of the nuclei since immunostaining of the DR and MnR reveals the presence of a high density of serotonergic ﬁbers in these locations (Halliday et al., 1995). II.B. Involvement of 5-HT autoreceptors The presence of 5-HT1A autoreceptors on serotonergic neurons is a key element in the regional

531

selectivity of SSRIs. 5-HT1A receptors are located in the somatodendritic region of serotonergic neurons (presynaptic) and on other neuronal types (postsynaptic, mostly on pyramidal neurons in the cortex and hippocampus) (Pazos and Palacios, 1985; Sotelo et al., 1990; Pompeiano et al., 1992; Santana et al., 2004). 5-HT1A receptors are coupled to a K+ channel via a pertussis toxin-sensitive G protein (Andrade et al, 1986; Innis and Aghajanian, 1987). In the hippocampus, inhibition and activation of adenylate cyclase have also been reported as effector systems, but these are apparently lacking in the DR (De Vivo and Maayani, 1986; Markstein et al., 1986; Clarke et al., 1996). The activation of raphe 5-HT1A receptors by selective 5-HT1A receptor agonists leads to a ﬁring-dependent reduction in 5-HT synthesis and release in the forebrain (Hutson et al., 1989). Likewise, the excess extracellular 5-HT produced by SSRIs in the DR and MnR activates somatodendritic 5-HT1A autoreceptors and reduces 5HT release in the projection areas of these two nuclei, such as frontal cortex, striatum, and, to a lesser extent, the hippocampus (Romero et al., 1994; Romero and Artigas, 1997). The involvement of 5-HT1A receptors in this effect was demonstrated by the use of selective 5-HT1A receptor antagonists, such as WAY-100635 or UH-301. Hence, the administration of these agents (a) reversed the inhibition of serotonergic cell ﬁring and 5-HT release induced by SSRIs, and (b) potentiated the increase of extracellular 5-HT produced by SSRIs (Arborelius et al., 1995, 1996; Gartside et al., 1995; Malagie´ et al., 1996; Romero et al., 1996b; Gobert et al., 1997b, 2000; Invernizzi et al., 1997; Romero and Artigas, 1997). Likewise, several b-adrenoceptor blockers that are also 5-HT1A receptor antagonists, such as ( )pindolol or ( )tertatolol, are capable of preventing the fall in 5-HT release induced in a DR-innervated area by SSRIs, either after local application in the DR or after their systemic administration in conditions of local reuptake blockade (Hjorth and Auerbach, 1994; Romero et al., 1994, 1996a; Auerbach et al., 1995). Likewise, these nonselective antagonists potentiated the increase in extracellular 5-HT produced by the systemic administration of SSRIs in standard conditions (e.g., no reuptake inhibitor in

Fig. 2. Maximal increase in dialysate 5-HT in the frontal cortex of rats treated with antiderpesant drugs (AD) alone or in combination with WAY-100635. The doses used were: clomipramine (CIM, 10 mg/kg i.p.), ﬂuoxetine (FLX, 10 mg/kg i.p.), ﬂuvoxamine (FVX, 10 mg/kg i.p.), and paroxetine (PAR, 3 mg/ kg i.p.). Way-100635 was administered at 1 mg/kg s.c., a dose that did not alter 5-HT release by itself. Redrawn from data in Romero et al. (1996a) and Romero and Artigas (1997).

the dialysis ﬂuid) (Hjorth, 1993; Dreshﬁeld et al., 1996; Romero et al., 1996a; Gobert and Millan, 1999). Fig. 2 shows the potentiation of the effect of selective (SSRIs) and nonselective 5-HT reuptake inhibitors (clomipramine) induced by the co-administration of the selective 5-HT1A receptor antagonist WAY-100635, which by itself does not change dialysate 5-HT concentration at the dose used (typically 0.1–1 mg/kg s.c.). A second type of serotonergic autoreceptors (5-HT1B) are also involved in the limitation of the effect of SSRIs. Whereas 5-HT1A autoreceptors are located in the soma and dendrites of 5-HT neurons, 5-HT1B autoreceptors are located in 5-HT axons (Riad et al., 2000), and they inhibit neurotransmitter release upon activation by 5-HT in a ﬁring-independent manner. They also function as terminal heteroreceptors, limiting the release of other neurotransmitters like GABA or glutamate (Raiteri, 2001). The combined administration of SSRIs and selective or nonselective 5-HT1B/1D receptor antagonists evoked an increase in extracellular 5-HT greater than the SSRI alone in various brain areas (Davidson and Stamford, 1996; Rollema et al., 1996; Gobert et al., 1997b, 2000; Herva´s et al., 1998, 2000; Malagie´ et al., 2001). The generation of mice lacking 5-HT1A or 5HT1B receptors has prompted a series of microdialysis studies whose results have provided additional evidence for the crucial role played by such autoreceptors in limiting the action of

532

antidepressant drugs, increasing 5-HT availability in CNS. Hence, the administration of SSRIs to these mice results invariably in greater increases of extracellular 5-HT than those achieved in wildtype mice (Knobelman et al., 2001; Malagie´ et al., 2001; Bortolozzi et al., 2004). The importance in the coupling between 5-HT neuron activity and 5-HT release is shown by the fact that SSRI administration did not reduce serotonergic cell ﬁring in 5-HT1A knockout mice, due to the absence of 5-HT1A autoreceptor-mediated negative feedback that occurs in serotonergic neurons in normal rodents (Amargo´s-Bosch et al., 2004). Fig. 3 shows schematically the various factors involved in the presynaptic modulation of extracellular 5-HT in forebrain by SSRIs and other 5HT-enhancing antidepressants (tricyclic drugs, SNRIs). On the one hand, these agents increase the extracellular 5-HT concentration in forebrain by blocking 5-HT reuptake in axon terminals. In contrast, the activation of somatodendritic (5HT1A) and terminal (5-HT1B/1D) autoreceptors offsets this effect by reducing ﬁring-dependent and ﬁring-independent 5-HT release, respectively. A large body of evidence indicates that both receptor subsets desensitize following repeated administration of antidepressant drugs (see Blier and de Montigny, 1994 for review), thus reducing the efﬁcacy of these negative feedback mechanisms, yet in a partial manner (Arborelius et al., 1996). This effect results in a progressive normalization of the discharge of DR 5-HT neurons and of the ﬁringdependent 5-HT release in brain structures innervated by the DR. Consequently, SSRIs increase more markedly extracellular 5-HT in frontal cortex after prolonged administration (Bel and Artigas, 1993; Invernizzi et al., 1994; Rutter et al., 1994; Herva´s et al., 2001). However, other researchers have not found such an effect in the dorsal hippocampus, which is mainly innervated by the MnR (Bosker et al., 1995a, b, Invernizzi et al., 1995). This may indicate the existence of differences between DR and MnR pathways with regard to the prolonged blockade of the 5-HT transporter. However, other factors, such as the dose of the antidepressant and the route used may contribute to this discrepancy given the different pharmacokinetic properties of antidepressant

Fig. 3. (a) Opposing effects of selective serotonin reuptake inhibitors (SSRIs) on the concentration of 5-HT in forebrain synapses result from inhibiting 5-HT reuptake at two distinct anatomical sites. Reuptake inhibition in forebrain nerve terminals increases the extracellular 5-HT concentration. The concurrent inhibition in the midbrain raphe also increases 5-HT (more than in forebrain), which activates 5-HT1A autoreceptors and reduces 5-HT cell ﬁring and 5-HT release by forebrain axons. The activation of terminal (5-HT1B) autoreceptors also reduces 5-HT release. Asterisks denote the possible sites of action of pindolol in human brain (unlike in rodents, pindolol lacks signiﬁcant afﬁnity for human 5-HT1B receptors). (b) Autoreceptor antagonists potentiate the effects of SSRIs. As revealed by microdialysis experiments, blockade of 5-HT1A and/ or 5-HT1B receptors with selective antagonists (0.3 mg/kg s.c. WAY100635 and 4 mg/kg i.p. SB224289, respectively) potentiates the effects of the administration of the SSRI ﬂuoxetine (FLX) (10 mg/kg i.p.) on extracellular 5-HT in frontal cortex. Results are mean7SEM values of extracellular 5-HT (expressed as percentage of baseline) in the various experimental groups. Reproduced from Artigas et al. (2001).

dugs. Moreover, the existence of a large 5-HT1A receptor reserve in the DR may also contribute to minimize the impact of the desensitization observed in individual cells on a population-based measure such as the dialysate 5-HT concentration.

533

II.C. Clinical relevance The above data may have an impact on the treatment of major depression if similar mechanisms are also occurring in the human brain. Indeed, as mentioned above, the human DR contains a very high density of the 5-HT transporter (Corte´s et al., 1988) and 5-HT1A receptors are present as well as assessed by various techniques including PET scan using 11 C-WAY-100635 as a radiotracer (Martinez et al., 2001). Early microdialysis observations indicating the presence of a raphe-based negative feedback triggered by nonselective (clomipramine, MAOIs) and selective (SSRIs) antidepressants (Adell and Artigas, 1991; Bel and Artigas, 1992; Invernizzi et al., 1992; Celada and Artigas, 1993) prompted the realization of a pilot study, which preliminarily tested the hypothesis that 5-HT1A receptor blockade might accelerate or enhance the clinical action of 5-HT-acting antidepressants using the nonselective 5-HT1A receptor (partial) antagonist pindolol (Artigas et al., 1994). This was followed by a large number of double-blind, placebo-controlled clinical trials (see Artigas et al., 2001 for review). The results of a meta-analysis of these data indicate that pindolol accelerates the clinical improvement induced by SSRIs. Yet it seems ineffective in bringing about a clinical improvement in previously unresponsive patients (Ballesteros and Callado, 2004). However, the appropriate testing of the above hypothesis requires the development of more selective 5-HT1A receptor antagonist or dual action drugs (e.g., 5-HT reuptake inhibition+5-HT1A antagonism) that can be used in clinical trials. II.D. Involvement of postsynaptic 5-HT receptors Clinical evidence indicates that atypical antipsychotics (preferential 5-HT2A/2C vs. DA D2 antagonists) and some antidepressants blocking 5-HT2 receptors enhance the therapeutic action of SSRIs (see Marek et al., 2003 for review). These 5-HT receptors are located postsynaptically, mainly in cortical and limbic areas, which raises the possibility that postsynaptic 5-HT receptors may modulate serotonergic function via descending pathways to the DR/MnR or act in local circuits to alter 5-HT release.

Indeed, evidence has accumulated over the last years indicating that postsynaptic 5-HT receptors present in forebrain areas feeding back to the raphe nuclei can also modulate serotonergic activity and terminal 5-HT release (Celada et al., 2001; Martı´ n-Ruiz et al., 2001). Particular attention has been paid to the medial prefrontal cortex since this cortical area (a) contains a large population of pyramidal neurons projecting to the DR and controlling serotonergic activity (Sesack et al., 1989; Hajo´s et al., 1998; Peyron et al., 1998; Celada et al., 2001), and (b) it contains a large density of cells expressing 5-HT1A, 5-HT2A, and – to a lesser extent – 5-HT2C and 5-HT3 receptors (Pompeiano et al., 1992, 1994; Puig et al., 2004; Santana et al., 2004). Hence, the local activation of 5-HT1A and 5-HT2A receptors in medial prefrontal cortex by agonists decreases and increases, respectively, the 5-HT release in rodent brain (Celada et al., 2001; Martı´ n-Ruiz et al., 2001; Amargo´s-Bosch et al., 2004). Likewise, 5-HT1A receptors in the amygdala also contribute to the modulation of the 5-HT release in the caudal linear raphe nucleus (Bosker et al., 1997). However, while the role of 5-HT autoreceptors in the SSRI-mediated inhibition of serotonergic activity and 5-HT release is ﬁrmly established, it is still uncertain whether SSRIs can modulate 5-HT neurons through these long loops. One study suggested a role for postsynaptic 5-HT1A receptors in the amygdala (Bosker et al., 2001). Similarly, postsynaptic 5-HT2C receptors may be involved in the enhancement of the action of SSRIs by the 5-HT2C receptor antagonists SB 242084 and RS 102221. These observations are in accordance with the fact that ﬂuoxetine increases cortical extracellular 5-HT levels more in 5-HT2C receptor knockout than in wild-type mice (Cremers et al., 2004). Blockade of 5-HT2A receptors with a low dose of the selective 5-HT2A receptor antagonist M100907 enhanced the effect of ﬂuoxetine in a behavioral test of depression (DRL 72 s reinforcement schedule). However, this effect does not seem to be accounted for by an increased presynaptic 5-HT function, since this agent did not potentiate the increase in extracellular 5-HT (Marek et al., 2005) although it might modulate NA or dopamine (DA) release in prefrontal cortex.

534

II.E. Involvement of catecholamine pathways in SSRI actions Due to the complex interrelationships between monoaminergic systems in the brain, the SSRIs also interact with noradrenaline and dopamine systems. Hence, ﬂuoxetine administration increases dopamine and noradrenaline levels in prefrontal cortex (Tanda et al., 1994, 1996; Gobert et al., 1999) and citalopram modulates noradrenergic transmission by an action in the locus coeruleus (Mateo et al., 2000). Likewise, the action of the nonselective 5-HT1A receptor antagonist pindolol may also involve an increase in cortical DA and NA release (Gobert and Millan, 1999; Millan and Gobert, 1999). Indeed, these additional actions on other aminergic systems may contribute to the therapeutic action of these agents. In contrast, since a2-adrenoceptors are present as terminal auto- and heteroreceptors in monoaminergic axons (both 5-HT and catecholamine) their blockade with selective and nonselective antagonists also enhances the increase in extracellular 5-HT, NA, and DA produced by SSRIs (Gobert et al., 1997a; Millan et al., 2000a, b).

III. Noradrenaline reuptake inhibitors (NRIs) The demonstration that blockade of the NA transporter by drugs like desipramine, nortryptiline, and maprotiline conferred antidepressant activity leads to the development of more selective and potent noradrenaline reuptake inhibitors (NRIs). The ﬁrst compound of this series was reboxetine, which depicted low interactions with muscarinic, histaminergic, and adrenergic receptors, thereby causing fewer side effects. Despite this good pharmacological proﬁle, its use as an antidepressant is currently limited to Europe. From a neurochemical point of view, NRIs with different degrees of selectivity raise extracellular NA in vivo in different areas of the brain (see Invernizzi and Garattini, 2004 for review). Similar to the serotonergic systems, a2-adrenoceptors regulate the somatodendritic release of NA in the cell body area of locus coeruleus (Mateo et al., 1998). Thus, chronic desipramine treatment increases NA in noradrenergic

terminal areas and desensitizes a2-adrenoceptors modulating NA release at somatodendritic and terminal levels. However, some differences may exist in comparison to the more selective NRI reboxetine. Thus, chronic reboxetine desensitizes terminal, but not somatodendritic a2-adrenoceptors controlling NA release in the dorsal hippocampus (Parini et al., 2005). Furthermore, somatodendritic a2-adrenoceptors that control locus coeruleus ﬁring activity are not apparently desensitized (Be´ı¨ que et al., 2000; Mateo et al., 2001). This is at variance with what occurs in the serotonergic system in which 5-HT1A autoreceptors controlling raphe cell ﬁring are indeed desensitized after chronic SSRI treatment (Pin˜eyro and Blier, 1999). In addition, chronic desipramine treatment markedly decreases the density of brain NA transporter, which was associated with a parallel decrease in NA uptake (Benmansour et al., 2004). Although the selectivity of reboxetine for the NA transporter over DA and 5-HT transporters is very high (Millan et al., 2001a), the close interrelationship between these monoaminergic systems makes it possible that changes in one of them may affect the others, as summarized above. To complicate the picture, it has been shown that, under some circumstances, NA can be taken up by serotonergic terminals (Vizi et al., 2004), and noradrenergic terminals also can take up 5-HT (Daws et al., 1998) and DA (Devoto et al., 2005). Although the precise mechanism is not known, reboxetine appeared to enhance extracellular 5-HT (Linner et al., 2004) and DA (Valentini et al., 2004) in the mPFC. However, reboxetine failed to alter extracellular concentrations of DA and 5-HT after a chronic regimen (Sacchetti et al., 1999), which suggests that its clinical antidepressant efﬁcacy may be independent of changes in dopaminergic or serotonergic systems.

IV. Serotonin and noradrenaline reuptake inhibitors The combination of selective blockade of 5-HT and NA has been suggested to confer a superior efﬁcacy in alleviating depressive symptoms in comparison with either SSRIs or NRIs (Thase et al.,

535 Table 1. Inhibition of human monoamine transporter binding in vitro by dual uptake inhibitors

b

Duloxetine Venlafaxinea Milnacipranc

Serotonin

Noradrenaline

Dopamine

0.870.04 8273 123711

7.570.3 2,483743 20072

240723 7,6477793 >10,000

Note: All values of Ki (nM) are expressed as mean7SEM, determined from three or more independent experiments with at least six concentrations of drug in triplicate. a

Data from Tatsumi et al. (1997). bData from Bymaster et al. (2001). cData from Koch et al. (2003).

2001). Medications in this group are known as dual reuptake inhibitors and include venlafaxine (Effexors), duloxetine (Cymbaltas), and milnacipran (Ixels) launched by Wyeth (Madison, NJ, USA), Eli Lilly and Co. (Indianapolis, IN, USA) and Pierre Fabre (Castres, France), respectively. They block 5-HT and NA reuptake with much lower selectivity toward the DA transporter (see Table 1). As for SSRIs, SNRIs do not seem to interact with cholinergic, adrenergic, or histaminergic receptors (Frazer, 1997; Sa´nchez and Hyttel, 1999). Acute administration of SNRIs to rats increases extracellular concentrations of 5-HT, NA, and, in some cases, DA (Moret and Briley, 1997; Bel and Artigas, 1999; Dawson et al., 1999; Millan et al., 2001a; Koch et al., 2003; Weikop et al., 2004; Kitaichi et al., 2005). However, compared with SSRIs, simultaneous inhibition of 5-HT and NA reuptake does not have a larger effect on extracellular 5-HT in forebrain regions (Felton et al., 2003). In addition, the increase of extracellular 5-HT, but not NA, induced by SNRIs is attenuated by the activation of 5-HT1A autoreceptors (Dawson et al., 1999; Beyer et al., 2002). Conversely, a2-adrenoceptor antagonists facilitate the increase in dialysate 5-HT and NA induced by SNRIs (Gobert et al., 1996a; Invernizzi and Garattini, 2004). The chronic administration of SNRIs produced controversial effects on brain monoamines depending on the compound, route of administration, and duration of the treatment. Thus, using osmotic minipumps, a 14-day treatment with venlafaxine increased dialysate 5-HT and NA in the frontal cortex without altering dopamine (Wikell et al.,

2001). However, this effect was not observed when venlafaxine was administered with a repeated injection schedule (Gur et al., 1999; Millan et al., 2001a). In a similar way, chronic administration of duloxetine (6.25 mg/kg, p.o.) for 14 days failed to alter basal NA and 5-HT levels in the frontal cortex, but augmented the duloxetine-induced increase in output of both transmitters (Kihara and Ikeda, 1995). It therefore seems that a sustained, high, steady-state level of drug attained with osmotic minipumps is needed to perceive the effects of this drug on monoamine release in the brain.

V. Peptide antagonists V.A. NK1 receptor antagonists Since the ﬁrst thrilling report in 1998 that a neurokinin1 (NK1) receptor antagonist, MK-809 (aprepitant), possessed antidepressant properties (Kramer et al., 1998), a possible role of substance P (SP) and its preferred receptor in the pathophysiology of depression came into play. Soon afterward, it was found that SP was able to interact with monoamines via NK1 receptors, thereby modulating the activity of such transmitter systems in the brain. In fact, acute systemic administration of the NK1 receptor antagonist, GR 205171, did not change dialysate 5-HT in the frontal cortex and hippocampus of either wildtype (Zocchi et al., 2003) or NK1 knockout mice (Froger et al., 2001; Guiard et al., 2004). However, when combined with SSRIs, GR 205171 increased dialysate 5-HT through an action in the dorsal raphe nucleus (Guiard et al., 2004). In addition, 21-day treatment with GR 205171 failed to alter the extracellular concentration of 5-HT in the frontal cortex, but enhanced that in the dorsal raphe nucleus. This latter effect was shown to result in an attenuation of 5-HT1A autoreceptor responsiveness (Haddjeri and Blier, 2001; Guiard et al., 2005). In line with this ﬁnding, functional desensitization of 5-HT1A receptors has been consistently demonstrated in NK1 receptor knockout mice (Froger et al., 2001; Santarelli et al., 2001). In contrast to what occurs in the serotonergic system, dialysate NA was found to be increased in

536

the frontal cortex and hippocampus of rats (Millan et al., 2001b) and NK1 receptor knockout mice (Herpfer et al., 2005), an effect coincident with an enhanced ﬁring rate of noradrenergic cells in the locus coeruleus (Millan et al., 2001b). A direct inﬂuence on noradrenergic neurons of the locus coeruleus appears unlikely in view of the facilitatory effect of SP upon noradrenergic cell bodies (Maubach et al., 2002; Steinberg et al., 2002). However, the location of these NK1 receptors inhibitory to noradrenergic neurones remains to be determined. With regard to the effects of NK1 receptor antagonists on the brain dopaminergic systems much less information has been gathered. In a preliminary study, it was shown that locally applied CP 96345 enhanced DA release in the striatum and potentiated that elicited by a low dose of methamphetamine (Gygi et al., 1993). However, dialysate levels of DA were increased in the frontal cortex, but not in the striatum and nucleus accumbens of conscious rats following acute systemic GR 205171 (Lejeune et al., 2002). This suggests that the inﬂuence of NK1 receptors on the release of DA may be dependent upon the brain region and route of administration. In summary, the observation that NK1 receptor antagonists potentiated the neurochemical effects of SSRIs (Guiard et al., 2004) introduced a promising new therapeutic approach to the treatment of depression in which NK1 receptor antagonists could be administered in combination with ‘classical’ antidepressant drugs to produce a greater antidepressant response. Nonetheless, a recent study has shown an absence of effect of aprepitant in the treatment of depression (Keller et al., 2006) and, therefore, the concept of NK1 receptor antagonism as an antidepressant mechanism has not been supported. As a result, several drug companies have discontinued their research programs on NK1 receptor antagonists.

(Nemeroff, 1996). This neuroendocrine axis is under the control of the hypothalamic secretion of the neuropeptide corticotropin-releasing factor (CRF), which is also associated with the integration of the physiological and behavioral responses to stress. CRF is hypersecreted in some depressed patients, which led to the hypothesis that CRF receptor antagonists might possess antidepressant properties. Two CRF receptor subtypes have been described: CRF1 and CRF2. High densities of CRF1 receptors are found in corticolimbic areas and in the pituitary. CRF2 receptors are mainly peripheral, although they are also expressed in the brain. Several CRF1 receptor antagonists have shown antidepressant-like activity in animal models (Mansbach et al., 1997). CRF neural pathways interact extensively with serotonergic and noradrenergic systems in the brain, where it seems to have a positive action. Thus, intracerebroventricularly administered CRF increases dialysate 5-HT, NA, and acetylcholine in the hippocampus (Desvignes et al., 2003; Kagamiishi et al., 2003; de Groote et al., 2005), a region known by its key role in the coordination of neuroendocrine and behavioral responses to stress. Consistent with these ﬁndings, the systemic administration of the CRF1 antagonist, CP-154,526, reduced hippocampal 5-HT and NA, although this effect depends on the brain region considered (Isogawa et al., 2000). Also, the microinfusion of CRF into the locus coeruleus enhances extracellular NA in cortex and hippocampus (Page and Abercrombie, 1999; Palamarchouk et al., 2002). Overall, it appears that CRF enhances serotonergic and noradrenergic transmission, which may seem contradictory with the purported therapeutic role of CRF antagonists. It is possible, however, that this effect may be independent of their actions on both monoaminergic systems in the brain.

V.B. Corticotropin-releasing factor antagonists VI. Concluding remarks Abnormal hypothalamic-pituitary-adrenal (HPA) axis activity has been implicated in the pathophysiology of depression (Barden, 2004). In fact, depressed patients frequently show excessive secretion of cortisol from the adrenal glands

Intracerebral microdialysis has been a fundamental tool in basic and oriented research in the ﬁeld of antidepressant drugs. The above studies have enabled to precisely determine the effects of

537

marketed and experimental drugs on monoaminergic brain systems. These studies have provided new views on the mechanism of action of antidepressants in vivo, which in many instances could not be anticipated from the in vitro properties of these compounds. Likewise, microdialysis has been extensively used in the development of new therapeutic strategies. The greatest advantage of microdialysis over other neurobiological techniques in the ﬁeld is that it can reliably monitor changes in the extracellular compartment of monoamines. Since these neurotransmitters act mostly in a paracrine, nonsynaptic manner, their concentrations in brain dialysates are an indirect estimate of the activation of postsynaptic monoaminergic receptors. Indeed, it should be recalled that, despite many research efforts, the monoamine receptors responsible for the therapeutic action remain largely unknown and the great majority of antidepressant treatments are based on a presynaptic enhancement of monoaminergic function, mainly by blocking reuptake. Hence, not surprisingly, the development of new antidepressant drugs often takes the increase in extracellular monoamine concentration as a key requirement in preclinical research. In summary, the microdialysis technique has largely contributed to what we presently know about the actions of antidepressant drugs in the living brain, and hopefully will continue to do so in the forthcoming years.

Acknowledgment This work was supported by grants from La Marato´ TV3 and CICYT (SAF 2004-05525).

References Adell, A. and Artigas, F. (1991) Differential effects of clomipramine given locally or systemically on extracellular 5-HT in raphe nuclei and frontal cortex. Naunyn Schmiedebergs Arch. Pharmacol., 343: 237–244. Adell, A., Carceller, A. and Artigas, F. (1993) In vivo brain dialysis study of the somatodendritic release of serotonin in the raphe nuclei of the rat. Effects of 8-hydroxy-2-(di-npropylamino)tetralin. J. Neurochem., 60: 1673–1681.

Amargo´s-Bosch, M., Bortolozzi, A., Puig, M.V., Serrats, J., Adell, A., Celada, P., Toth, M., Mengod, G. and Artigas, F. (2004) Co-expression and in vivo interaction of serotonin1A and serotonin2A receptors in pyramidal neurons of prefrontal cortex. Cereb. Cortex, 14: 281–299. Andlin-Sobocki, P., Jo¨nsson, B., Wittchen, H.U. and Olesen, J. (2005) Cost of disorders of the brain in Europe. Eur. J. Neurol., 12(Suppl. 1): 1–27. Andrade, R., Malenka, R.C. and Nicoll, R.A. (1986) A G protein couples serotonin and GABAB receptors to the same channel in hippocampus. Science, 234: 1261–1265. Angst, J. (1992) Epidemiology of depression. Psychopharmacology, 106: S71–S74. Arborelius, L., Nomikos, G.G., Grillner, P., Hertel, P., Ho¨o¨k, B.B., Hacksell, U. and Svensson, T.H. (1995) 5-HT1A receptor antagonists increase the activity of serotonergic cells in the dorsal raphe nucleus in rats treated acutely or chronically with citalopram. Naunyn Schmiedebergs Arch. Pharmacol., 352: 157–165. Arborelius, L., Nomikos, G.G., Hertel, P., Salmi, P., Grillner, P., Ho¨o¨k, B.B., Hacksell, U. and Svensson, T.H. (1996) The 5-HT1A receptor antagonist (S)-UH-301 augments the increase in extracellular concentrations of 5-HT in the frontal cortex produced by both acute and chronic treatment with citalopram. Naunyn Schmiedebergs Arch. Pharmacol., 353: 630–640. Artigas, F. (1993) 5–HT and antidepressants: new views from microdialysis studies. Trends Pharmacol. Sci., 14: 262. Artigas, F. (1995) Selective serotonin noradrenaline reuptake inhibitors (SNRIs) – pharmacology and therapeutic potential in the treatment of depressive-disorders. CNS Drugs, 4: 79–89. Artigas, F., Celada, P., Laruelle, M. and Adell, A. (2001) How does pindolol improve antidepressant action? Trends Pharmacol. Sci., 22: 224–228. Artigas, F., Perez, V. and Alvarez, E. (1994) Pindolol induces a rapid improvement of depressed patients treated with serotonin reuptake inhibitors. Arch. Gen. Psychiatry, 51: 248–251. Artigas, F., Romero, L., de Montigny, C. and Blier, P. (1996) Acceleration of the effect of selected antidepressant drugs in major depression by 5-HT1A antagonists. Trends Neurosci., 19: 378–383. Auerbach, S.B., Lundberg, J.F. and Hjorth, S. (1995) Differential inhibition of serotonin release by 5-HT and NA reuptake blockers after systemic administration. Neuropharmacology, 34: 89–96. Ballesteros, J. and Callado, L.F. (2004) Effectiveness of pindolol plus serotonin uptake inhibitors in depression: a meta-analysis of early and late outcomes from randomised controlled trials. J. Affect. Disord., 79: 137–147. Barden, N. (2004) Implication of the hypothalamic-pituitaryadrenal axis in the physiopathology of depression. J. Psychiatry Neurosci., 29: 185–193. Bech, P., Cialdella, P., Haugh, M.C., Birkett, M.A., Hours, A., Boissel, J.P. and Tollefson, G.D. (2000) Meta-analysis of randomised controlled trials of ﬂuoxetine vs. placebo and tricyclic antidepressants in the short-term treatment of major depression. Br. J. Psychiatry, 176: 421–428.

538 Be´ı¨ que, J.C., de Montigny, C., Blier, P. and Debonnel, G. (2000) Effects of sustained administration of the serotonin and norepinephrine reuptake inhibitor venlafaxine: I. In vivo electrophysiological studies in the rat. Neuropharmacology, 39: 1800–1812. Bel, N. and Artigas, F. (1992) Fluvoxamine increases preferentially extracellular 5-HT in the raphe nuclei of the awake rat. An in vivo microdialysis study. Eur. J. Pharmacol., 229: 101–103. Bel, N. and Artigas, F. (1993) Chronic treatment with ﬂuvoxamine increases extracellular serotonin in frontal cortex but not in raphe nuclei. Synapse, 15: 243–245. Bel, N. and Artigas, F. (1999) Modulation of the extracellular 5-hydroxytryptamine brain concentrations by the serotonin and noradrenaline reuptake inhibitor, milnacipran. Microdialysis studies in rats. Neuropsychopharmacology, 21: 745–754. Benmansour, S., Altamirano, A.V., Jones, D.J., Sanchez, T.A., Gould, G.G., Pardon, M.C., Morilak, D.A. and Frazer, A. (2004) Regulation of the norepinephrine transporter by chronic administration of antidepressants. Biol. Psychiatry, 55: 313–316. Beyer, C.E., Boikess, S., Luo, B. and Dawson, L.A. (2002) Comparison of the effects of antidepressants on norepinephrine and serotonin concentrations in the rat frontal cortex: an in-vivo microdialysis study. J. Psychopharmacol., 16: 297–304. Blakely, R.D., Berson, H.E., Fremeau, R.T., Caron, M.G., Peek, M.M., Prince, H.K. and Bradley, C.C. (1991) Cloning and expression of a functional serotonin transporter from rat brain. Nature, 354: 66–70. Blier, P. and de Montigny, C. (1994) Current advances and trends in the treatment of depression. Trends Pharmacol. Sci., 15: 220–226. Blier, P., de Montigny, C. and Chaput, Y. (1987) Modiﬁcations of the serotonin system by antidepressant treatment: implications for the therapeutic response in major depression. J. Clin. Psychopharmacol., 7: 24S–35S. Bortolozzi, A., Amargo´s-Bosch, M., Toth, M., Artigas, F. and Adell, A. (2004) In vivo efﬂux of serotonin in the dorsal raphe nucleus of 5-HT1A receptor knockout mice. J. Neurochem., 88: 1373–1379. Bosker, F.J., Cremers, T.I., Jongsma, M.E., Westerink, B.H., Wikstrom, H.V. and den Boer, J.A. (2001) Acute and chronic effects of citalopram on postsynaptic 5-hydroxytryptamine(1A) receptor-mediated feedback: a microdialysis study in the amygdala. J. Neurochem., 76: 1645–1653. Bosker, F.J., Klompmakers, A.A. and Westenberg, H.G. (1995a) Effects of single and repeated oral administration of ﬂuvoxamine on extracellular serotonin in the median raphe nucleus and dorsal hippocampus of the rat. Neuropharmacology, 34: 501–508. Bosker, F.J., Klompmakers, A. and Westenberg, H.G. (1997) Postsynaptic 5-HT1A receptors mediate 5-hydroxytryptamine release in the amygdala through a feedback to the caudal linear raphe. Eur. J. Pharmacol., 333: 147–157.

Bosker, F.J., van Esseveldt, K.E., Klompmakers, A.A. and Westenberg, H.G. (1995b) Chronic treatment with ﬂuvoxamine by osmotic minipumps fails to induce persistent functional changes in central 5-HT1A and 5-HT1B receptors, as measured by in vivo microdialysis in dorsal hippocampus of conscious rats. Psychopharmacology, 117: 358–363. Bymaster, F.P., Dreshﬁeld-Ahmad, L.J., Threlkeld, P.G., Shaw, J.L., Thompson, L., Nelson, D.L., Hemrick-Luecke, S.K. and Wong, D.T. (2001) Comparative afﬁnity of duloxetine and venlafaxine for serotonin and norepinephrine transporters in vitro and in vivo, human serotonin receptor subtypes, and other neuronal receptors. Neuropsychopharmacology, 25: 871–880. Celada, P. and Artigas, F. (1993) Monoamine oxidase inhibitors increase preferentially extracellular 5-hydroxytryptamine in the midbrain raphe nuclei. A brain microdialysis study in the awake rat. Naunyn Schmiedebergs Arch. Pharmacol., 347: 583–590. Celada, P., Puig, M.V., Casanovas, J.M., Guillazo, G. and Artigas, F. (2001) Control of dorsal raphe serotonergic neurons by the medial prefrontal cortex: involvement of serotonin-1A, GABA(A), and glutamate receptors. J. Neurosci., 21: 9917–9929. Clarke, W.P., Yocca, F.D. and Maayani, S. (1996) Lack of 5hydroxytryptamine(1A)-mediated inhibition of adenylyl cyclase in dorsal raphe of male and female rats. J. Pharmacol. Exp. Ther., 277: 1259–1266. Cornwall, P.L. and Scott, J. (1997) Partial remission in depressive disorders. Acta Psychiatr. Scand., 95: 265–271. Corte´s, R., Soriano, E., Pazos, A., Probst, A. and Palacios, J.M. (1988) Autoradiography of antidepressant binding sites in the human brain: localization using [3H]imipramine and [3H]paroxetine. Neuroscience, 27: 473–496. Cremers, T.I., Giorgetti, M., Bosker, F.J., Hogg, S., Arnt, J., Mork, A., Honig, G., Bogeso, K.P., Westerink, B.H., den Boer, H., Wikstrom, H.V. and Tecott, L.H. (2004) Inactivation of 5-HT(2C) receptors potentiates consequences of serotonin reuptake blockade. Neuropsychopharmacology, 29: 1782–1789. Danish University Antidepressant Group. (1986) Citalopram: clinical effect proﬁle in comparison with clomipramine. A controlled multicenter study. Psychopharmacology (Berl.), 90: 131–138. Danish University Antidepressant Group. (1990) Paroxetine: a selective serotonin reuptake inhibitor showing better tolerance, but weaker antidepressant effect than clomipramine in a controlled multicenter study. J. Affect. Disord., 18: 289–299. Davidson, C. and Stamford, J.A. (1996) Serotonin efﬂux in the rat ventral lateral geniculate nucleus assessed by fast cyclic voltammetry is modulated by 5-HT1B and 5-HT1D autoreceptors. Neuropharmacology, 35: 1627–1634. Daws, L.C., Toney, G.M., Gerhardt, G.A. and Frazer, A. (1998) In vivo chronoamperometric measures of extracellular serotonin clearance in rat dorsal hippocampus: contribution of serotonin and norepinephrine transporters. J. Pharmacol. Exp. Ther., 286: 967–976.

539 Dawson, L.A., Nguyen, H.Q. and Geiger, A. (1999) Effects of venlafaxine on extracellular concentrations of 5-HT and noradrenaline in the rat frontal cortex: augmentation via 5-HT1A receptor antagonism. Neuropharmacology, 38: 1153–1163. de Groote, L., Pen˜alva, R.G., Flachskamm, C., Reul, J.M. and Linthorst, A.C. (2005) Differential monoaminergic, neuroendocrine and behavioural responses after central administration of corticotropin-releasing factor receptor type 1 and type 2 agonists. J. Neurochem., 94: 45–56. De Vivo, M. and Maayani, S. (1986) Characterization of the 5hydroxytryptamine 1A receptor-mediated inhibition of forskolin-stimulated adenylate cyclase activity in guinea-pig and rat hippocampal membranes. J. Pharmacol. Exp. Ther., 238: 248–253. Desvignes, C., Rouquier, L., Souilhac, J., Mons, G., Rodier, D., Soubrie´, P. and Steinberg, R. (2003) Control by tachykinin NK2 receptors of CRF1 receptor-mediated activation of hippocampal acetylcholine release in the rat and guinea-pig. Neuropeptides, 37: 89–97. Devoto, P., Flore, G., Saba, P., Fa, M. and Gessa, G.L. (2005) Stimulation of the locus coeruleus elicits noradrenaline and dopamine release in the medial prefrontal and parietal cortex. J. Neurochem., 92: 368–374. Dreshﬁeld, L.J., Wong, D.T., Perry, K.W. and Engleman, E.A. (1996) Enhancement of ﬂuoxetine-dependent increase of extracellular serotonin (5-HT) levels by (–)-pindolol, an antagonist at 5-HT1A receptors. Neurochem. Res., 21: 557–562. Felton, T.M., Kang, T.B., Hjorth, S. and Auerbach, S.B. (2003) Effects of selective serotonin and serotonin/noradrenaline reuptake inhibitors on extracellular serotonin in rat diencephalon and frontal cortex. Naunyn Schmiedebergs Arch. Pharmacol., 367: 297–305. Frazer, A. (1997) Pharmacology of antidepressants. J. Clin. Psychopharmacol., 17(Suppl 1): 2S–18S. Froger, N., Gardier, A.M., Moratalla, R., Alberti, I., Lena, I., Boni, C., De Felipe, C., Rupniak, N.M.J., Hunt, S.P., Jacquot, C., Hamon, M. and Lanfumey, L. (2001) 5-Hydroxytryptamine (5-HT1A) autoreceptor adaptive changes in substance P (neurokinin 1) receptor knock-out mice mimic antidepressant-induced desensitization. J. Neurosci., 21: 8188–8197. Gartside, S.E., Umbers, V., Hajos, M. and Sharp, T. (1995) Interaction between a selective 5-HT1A receptor antagonist and an SSRI in vivo: effects on 5-HT cell ﬁring and extracellular 5-HT. Br. J. Pharmacol., 115: 1064–1070. Gobert, A. and Millan, M.J. (1999) Modulation of dialysate levels of dopamine, noradrenaline, and serotonin (5-HT) in the frontal cortex of freely-moving rats by ( )-pindolol alone and in association with 5-HT reuptake inhibitors: comparative roles of beta-adrenergic, 5-HT1A, and 5-HT1B receptors. Neuropsychopharmacology, 21: 268–284. Gobert, A., Dekeyne, A. and Millan, M.J. (2000) The ability of WAY100,635 to potentiate the neurochemical and functional actions of ﬂuoxetine is enhanced by co-administration of SB224,289, but not BRL15572. Neuropharmacology, 39: 1608–1616.

Gobert, A., Rivet, J.M., Cistarelli, L., Melon, C. and Millan, M.J. (1997a) Alpha2-adrenergic receptor blockade markedly potentiates duloxetine- and ﬂuoxetine-induced increases in noradrenaline, dopamine, and serotonin levels in the frontal cortex of freely moving rats. J. Neurochem., 69: 2616–2619. Gobert, A., Rivet, J.M., Cistarelli, L., Melon, C. and Millan, M.J. (1999) Buspirone modulates basal and ﬂuoxetinestimulated dialysate levels of dopamine, noradrenaline and serotonin in the frontal cortex of freely moving rats: activation of serotonin1A receptors and blockade of alpha2adrenergic receptors underlie its actions. Neuroscience, 93: 1251–1262. Gobert, A., Rivet, J.M., Cistarelli, L. and Millan, M.J. (1997b) Potentiation of the ﬂuoxetine-induced increase in dialysate levels of serotonin (5-HT) in the frontal cortex of freely moving rats by combined blockade of 5-HT1A and 5-HT1B receptors with WAY 100,635 and GR 127,935. J. Neurochem., 68: 1159–1163. Guiard, B.P., Froger, N., Hamon, M., Gardier, A.M. and Lanfumey, L. (2005) Sustained pharmacological blockade of NK1 substance P receptors causes functional desensitization of dorsal raphe 5-HT1A autoreceptors in mice. J. Neurochem., 95: 1713–1723. Guiard, B.P., Przybylski, C., Guilloux, J.-P., Seif, I., Froger, N., De Felipe, C., Hunt, S.P., Lanfumey, L. and Gardier, A.M. (2004) Blockade of substance P (neurokinin 1) receptors enhances extracellular serotonin when combined with a selective serotonin reuptake inhibitor: an in vivo microdialysis study in mice. J. Neurochem., 89: 54–63. Gur, E., Dremencov, E., Lerer, B. and Newman, M.E. (1999) Venlafaxine: acute and chronic effects on 5-hydroxytryptamine levels in rat brain in vivo. Eur. J. Pharmacol., 372: 17–24. Gygi, S.P., Gibb, J.W., Johnson, M. and Hanson, G.R. (1993) Blockade of tachykinin NK1 receptors by CP-96345 enhances dopamine release and the striatal dopamine effects of methamphetamine in rats. Eur. J. Pharmacol., 250: 177–180. Haddjeri, N. and Blier, P. (2001) Sustained blockade of neurokinin-1 receptors enhances serotonin neurotransmission. Biol. Psychiatry, 50: 191–199. Hajo´s, M., Richards, C.D., Szekely, A.D. and Sharp, T. (1998) An electrophysiological and neuroanatomical study of the medial prefrontal cortical projection to the midbrain raphe nuclei in the rat. Neuroscience, 87: 95–108. Halliday, G., Harding, A. and Paxinos, G. (1995) Serotonin and tachykinin systems. In: Paxinos, G. (Ed.), The Rat Nervous System (2nd ed). Academic Press, Sydney, pp. 929–974. Herpfer, I., Hunt, S.P. and Stanford, S.C. (2005) A comparison of neurokinin 1 receptor knock-out (NK1 / ) and wildtype mice: exploratory behaviour and extracellular noradrenaline concentration in the cerebral cortex of anaesthetised subjects. Neuropharmacology, 48: 706–719. Herva´s, I. and Artigas, F. (1998) Effect of ﬂuoxetine on extracellular 5-hydroxytryptamine in rat brain. Role of 5-HT autoreceptors. Eur. J. Pharmacol., 358: 9–18.

540 Herva´s, I., Queiroz, C.M., Adell, A. and Artigas, F. (2000) Role of uptake inhibition and autoreceptor activation in the control of 5-HT release in the frontal cortex and dorsal hippocampus of the rat. Br. J. Pharmacol., 130: 160–166. Herva´s, I., Vilaro´, M.T., Romero, L., Scorza, M.C., Mengod, G. and Artigas, F. (2001) Desensitization of 5-HT(1A) autoreceptors by a low chronic ﬂuoxetine dose effect of the concurrent administration of WAY-100635. Neuropsychopharmacology, 24: 11–20. He´ry, F., Faudon, M. and Ternaux, J.P. (1982) In vivo release of serotonin in two raphe nuclei (raphe dorsalis and magnus) of the cat. Brain. Res. Bull., 8: 123–129. Hjorth, S. (1993) Serotonin 5-HT1A autoreceptor blockade potentiates the ability of the 5-HT reuptake inhibitor citalopram to increase nerve terminal output of 5-HT in vivo: a microdialysis study. J. Neurochem., 60: 776–779. Hjorth, S. and Auerbach, S.B. (1994) Further evidence for the importance of 5-HT1A autoreceptors in the action of selective serotonin reuptake inhibitors. Eur. J. Pharmacol., 260: 251–255. Hoffman, B.J., Mezey, E. and Brownstein, M. (1991) Cloning of a serotonin transporter affected by antidepressants. Science, 249: 1303–1306. Hrdina, P.D., Foy, B., Hepner, A. and Summers, R.J. (1990) Antidepressant binding sites in brain: autoradiographic comparison of [3H]paroxetine and [3H]imipramine, localization and relationship to serotonin transporter. J. Pharmacol. Exp. Ther., 252: 410–418. Hutson, P.H., Sarna, G.S., O’Connell, M.T. and Curzon, G. (1989) Hippocampal 5-HT synthesis and release in vivo is decreased by infusion of 8-OH-DPAT into the nucleus raphe dorsalis. Neurosci. Lett., 100: 276–280. Hyttel, J. (1994) Pharmacological characterization of selective serotonin reuptake inhibitors (SSRIs). Int. Clin. Psychopharmacol., 9: 19–26. Innis, R.B. and Aghajanian, G.K. (1987) Pertussis toxin blocks 5-HT1A and GABAB receptor-mediated inhibition of serotonergic neurons. Eur. J. Pharmacol., 143: 195–204. Invernizzi, R., Belli, S. and Samanin, R. (1992) Citalopram’s ability to increase the extracellular concentration of serotonin in the dorsal raphe prevents the drug’s effect in frontal cortex. Brain Res., 584: 322–324. Invernizzi, R., Bramante, M. and Samanin, R. (1994) Chronic treatment with citalopram facilitates the effect of a challenge dose on cortical serotonin output: role of presynaptic 5-HT1A receptors. Eur. J. Pharmacol., 260: 243–246. Invernizzi, R., Bramante, M. and Samanin, R. (1995) Extracellular concentrations of serotonin in the dorsal hippocampus after acute and chronic treatment with citalopram. Brain. Res., 696: 62–66. Invernizzi, R., Velasco, C., Bramante, M., Longo, A. and Samanin, R. (1997) Effect of 5-HT1A receptor antagonists on citalopram-induced increase in extracellular serotonin in the frontal cortex, striatum and dorsal hippocampus. Neuropharmacology, 36: 467–473.

Invernizzi, R.W. and Garattini, S. (2004) Role of presynaptic a2-adrenoceptors in antidepressant action: recent ﬁndings from microdialysis studies. Prog. Neuropsychopharmacol. Biol. Psychiatry, 28: 819–827. Isogawa, K., Akiyoshi, J., Hikichi, T., Yamamoto, Y., Tsutsumi, T. and Nagayama, H. (2000) Effect of corticotropin releasing factor receptor 1 antagonist on extracellular norepinephrine, dopamine and serotonin in hippocampus and prefrontal cortex of rats in vivo. Neuropeptides, 34: 234–239. Kagamiishi, Y., Yamamoto, T. and Watanabe, S. (2003) Hippocampal serotonergic system is involved in anxiety-like behavior induced by corticotropin-releasing factor. Brain Res., 991: 212–221. Keller, M., Montgomery, Ball, W., Morrison, M., Snavely, D., Liu, G., Hargreaves, R., Hietala, J., Lines, C., Beebe, K. and Reines, S. (2006) Lack of efﬁcacy of the substance P (neurokinin1 receptor) antagonist aprepitant in the treatment of major depressive disorder. Biol. Psychiatry, 59: 216–223. Kihara, T. and Ikeda, M. (1995) Effects of duloxetine, a new serotonin and norepinephrine uptake inhibitor, on extracellular monoamine levels in rat frontal cortex. J. Pharmacol. Exp. Ther., 272: 177–183. Kitaichi, Y., Inoue, T., Izumi, T., Nakagawa, S., Kato, A. and Koyama, T. (2005) Subchronic milnacipran treatment increases basal extracellular noradrenaline concentrations in the medial prefrontal cortex of rats. Eur. J. Pharmacol., 520: 37–42. Knobelman, D.A., Hen, R. and Lucki, I. (2001) Genetic regulation of extracellular serotonin by 5-hydroxytryptamine(1A) and 5-hydroxytryptamine(1B) autoreceptors in different brain regions of the mouse. J. Pharmacol. Exp. Ther., 298: 1083–1091. Koch, S., Hemrick-Luecke, S.K., Thompson, L.K., Evans, D.C., Threlkeld, P.G., Nelson, D.L., Perry, K.W. and Bymaster, F.P. (2003) Comparison of effects of dual transporter inhibitors on monoamine transporters and extracellular levels in rats. Neuropharmacology, 45: 935–944. Kramer, M.S., Cutler, N., Feighner, J., Shrivastava, R., Carman, J., Sramek, J.J., Reines, S.A., Liu, G., Snavely, D., Wyatt-Knowles, E., Hale, J.J., Mills, S.G., MacCoss, M., Swain, C.J., Harrison, T., Hill, R.G., Hefti, F., Scolnick, E.M., Cascieri, M.A., Chicchi, G.G., Sadowski, S., Williams, A.R., Hewson, L., Smith, D., Carlson, E.J., Hardgreaves, R.J. and Rupniak, N.M.J. (1998) Distinct mechanism for antidepressant activity by blockade of central substance P receptors. Science, 281: 1640–1645. Lejeune, F., Gobert, A. and Millan, M.J. (2002) The selective neurokinin (NK)1 antagonist, GR205,171, stereospeciﬁcally enhances mesocortical dopaminergic transmission in the rat: a combined dialysis and electrophysiological study. Brain Res., 935: 134–139. Lepine, J.P., Gastpar, M., Mendlewicz, J. and Tylee, A. (1997) Depression in the community: the ﬁrst pan-European study DEPRES (Depression Research in European Society). Int. Clin. Psychopharmacol., 12: 19–29.

541 Lesch, K.P., Wolozin, B.L., Murphy, D.L. and Riederer, P. (1993) Primary structure of the human platelet serotonin uptake site – identity with the brain serotonin transporter. J. Neurochem., 60: 2319–2322. Linner, L., Wiker, C., Arborelius, L., Schalling, M. and Svensson, T.H. (2004) Selective noradrenaline reuptake inhibition enhances serotonergic neuronal activity and transmitter release in the rat forebrain. J. Neural Transm., 111: 127–139. Malagie´, I., Trillat, A.C., Bourin, M., Jacquot, C., Hen, R. and Gardier, A.M. (2001) 5-HT1B Autoreceptors limit the effects of selective serotonin re-uptake inhibitors in mouse hippocampus and frontal cortex. J. Neurochem., 76: 865–871. Malagie´, I., Trillat, A.C., Douvier, E., Anmella, M.C., Dessalles, M.C., Jacquot, C. and Gardier, A.M. (1996) Regional differences in the effect of the combined treatment of WAY 100635 and ﬂuoxetine: an in vivo microdialysis study. Naunyn Schmiedebergs Arch. Pharmacol., 354: 785–790. Malagie´, I., Trillat, A.C., Jacquot, C. and Gardier, A.M. (1995) Effects of acute ﬂuoxetine on extracellular serotonin levels in the raphe: an in vivo microdialysis study. Eur. J. Pharmacol., 286: 213–217. Mansbach, R.S., Brooks, E.N. and Chen, Y.L. (1997) Antidepressant-like effects of CP-154,526, a selective CRF1 receptor antagonist. Eur. J. Pharmacol., 323: 21–26. Marek, G.J., Carpenter, L.L., McDougle, C.J. and Price, L.H. (2003) Synergistic action of 5-HT2A antagonists and selective serotonin reuptake inhibitors in neuropsychiatric disorders. Neuropsychopharmacology, 28: 402–412. Marek, G.J., Martin-Ruiz, R., Abo, A. and Artigas, F. (2005) The selective 5-HT2A receptor antagonist M100907 enhances antidepressant-like behavioral effects of the SSRI ﬂuoxetine. Neuropsychopharmacology, 30: 2205–2215. Markstein, R., Hoyer, D. and Engel, G. (1986) 5-HT1A-receptors mediate stimulation of adenylate cyclase in rat hippocampus. Naunyn Schmiedebergs Arch. Pharmacol., 333: 335–341. Martinez, D., Hwang, D., Mawlawi, O., Slifstein, M., Kent, J., Simpson, N., Parsey, R.V., Hashimoto, T., Huang, Y., Shinn, A., Van Heertum, R., Abi-Dargham, A., Caltabiano, S., Malizia, A., Cowley, H., Mann, J.J. and Laruelle, M. (2001) Differential occupancy of somatodendritic and postsynaptic 5HT(1A) receptors by pindolol: a dose-occupancy study with [11C]WAY 100635 and positron emission tomography in humans. Neuropsychopharmacology, 24: 209–229. Martı´ n-Ruiz, R., Puig, M.V., Celada, P., Shapiro, D.A., Roth, B.L., Mengod, G. and Artigas, F. (2001) Control of serotonergic function in medial prefrontal cortex by serotonin2A receptors through a glutamate-dependent mechanism. J. Neurosci., 21: 9856–9866. Mateo, Y., Ferna´ndez-Pastor, B. and Meana, J.J. (2001) Acute and chronic effects of desipramine and clorgyline on a2-adrenoceptors regulating noradrenergic transmission in the rat brain: a dual-probe microdialysis study. Br. J. Pharmacol., 133: 1362–1370.

Mateo, Y., Pineda, J. and Meana, J.J. (1998) Somatodendritic a2-adrenoceptors in the locus coeruleus are involved in the in vivo modulation of cortical noradrenaline release by the antidepressant desipramine. J. Neurochem., 71: 790–798. Mateo, Y., Ruiz-Ortega, J.A., Pineda, J., Ugedo, L. and Meana, J.J. (2000) Inhibition of 5-hydroxytryptamine reuptake by the antidepressant citalopram in the locus coeruleus modulates the rat brain noradrenergic transmission in vivo. Neuropharmacology, 39: 2036–2043. Matos, F.F., Urban, C. and Yocca, F.D. (1996) Serotonin (5HT) release in the dorsal raphe´ and ventral hippocampus: raphe´ control of somatodendritic and terminal 5-HT release. J. Neural Transm., 103: 173–190. Maubach, K.A., Martin, K., Chicchi, G., Harrison, T., Wheeldon, A., Swain, C.J., Cumberbatch, M.J., Rupniak, N.M.J. and Seabrook, G.R. (2002) Chronic substance P (NK1) receptor antagonist and conventional antidepressant treatment increases burst ﬁring of monoamine neurones in the locus coeruleus. Neuroscience, 109: 609–617. Millan, M.J. and Gobert, A. (1999) ( )-Pindolol increases dialysate concentrations of dopamine and noradrenaline, but not serotonin, in the frontal cortex of freely-moving rats. Neuropharmacology, 38: 909–912. Millan, M.J., Gobert, A., Lejeune, F., Newman-Tancredi, A., Rivet, J.M., Auclair, A. and Peglion, J.-L. (2001a) S33005, a novel ligand at both serotonin and norepinephrine transporters: I. Receptor binding, electrophysiological, and neurochemical proﬁle in comparison with venlafaxine, reboxetine, citalopram, and clomipramine. J. Pharmacol. Exp. Ther., 298: 565–580. Millan, M.J., Gobert, A., Rivet, J.M., Adhumeau-Auclair, A., Cussac, D., Newman-Tancredi, A., Dekeyne, A., Nicolas, J.P. and Lejeune, F. (2000a) Mirtazapine enhances frontocortical dopaminergic and corticolimbic adrenergic, but not serotonergic, transmission by blockade of alpha2-adrenergic and serotonin2C receptors: a comparison with citalopram. Eur. J. Neurosci., 12: 1079–1095. Millan, M.J., Lejeune, F., De Nanteuil, G. and Gobert, A. (2001b) Selective blockade of neurokinin (NK)1 receptors facilitates the activity of adrenergic pathways projecting to frontal cortex and dorsal hippocampus in rats. J. Neurochem., 76: 1949–1954. Millan, M.J., Lejeune, F. and Gobert, A. (2000b) Reciprocal autoreceptor and heteroreceptor control of serotonergic, dopaminergic and noradrenergic transmission in the frontal cortex: relevance to the actions of antidepressant agents. J. Psychopharmacol., 14: 114–138. Moret, C. and Briley, M. (1997) Effects of milnacipran and pindolol on extracellular noradrenaline and serotonin levels in guinea pig hypothalamus. J. Neurochem., 69: 815–822. Murray, C.J.L. and Lopez, A.D. (1997) Alternative projections of mortality and disability by cause 1990–2020: Global Burden of Disease Study. Lancet, 349: 1498–1504. Nemeroff, C.B. (1996) The corticotrophin-releasing factor (CRF) hypothesis of depression: new ﬁndings and new directions. Mol. Psychiatry, 1: 336–342.

542 Nutt, D.J. (2002) The neuropharmacology of serotonin and noradrenaline in depression. Int. Clin. Psychopharmacol., 17: S1–S12. Page, M.E. and Abercrombie, E.D. (1999) Discrete local application of corticotropin-releasing factor increases locus coeruleus discharge and extracellular norepinephrine in rat hippocampus. Synapse, 33: 304–313. Palamarchouk, V.S., Swiergiel, A.H. and Dunn, A.J. (2002) Hippocampal noradrenergic responses to CRF injected into the locus coeruleus of unanesthetized rats. Brain Res., 950: 31–38. Parini, S., Renoldi, G., Battaglia, A. and Invernizzi, R.W. (2005) Chronic reboxetine desensitizes terminal but not somatodendritic a2-adrenoceptors controlling noradrenaline release in the rat dorsal hippocampus. Neuropsychopharmacology, 30: 1048–1055. Pazos, A. and Palacios, J.M. (1985) Quantitative autoradiographic mapping of serotonin receptors in the rat brain. I. Serotonin-1 receptors. Brain Res., 346: 205–230. Peyron, C., Petit, J.M., Rampon, C., Jouvet, M. and Luppi, P.H. (1998) Forebrain afferents to the rat dorsal raphe nucleus demonstrated by retrograde and anterograde tracing methods. Neuroscience, 82: 443–468. Pin˜eyro, G. and Blier, P. (1999) Autoregulation of serotonin neurons: role in antidepressant drug action. Pharmacol. Rev., 51: 533–591. Pompeiano, M., Palacios, J.M. and Mengod, G. (1992) Distribution and cellular localization of MnRA coding for 5-HT1A receptor in the rat brain: correlation with receptor binding. J. Neurosci., 12: 440–453. Pompeiano, M., Palacios, J.M. and Mengod, G. (1994) Distribution of the serotonin 5-HT2 receptor family mRNAs: comparison between 5-HT2A and 5-HT2C receptors. Brain Res. Mol. Brain Res., 23: 163–178. Puig, M.V., Santana, N., Celada, P., Mengod, G. and Artigas, F. (2004) In vivo excitation of GABA interneurons in the medial prefrontal cortex through 5-HT3 receptors. Cereb. Cortex, 14: 1365–1375. Quinaux, N., Scuve´e-Moreau, J. and Dresse, A. (1982) Inhibition of in vitro and ex vivo uptake of noradrenaline and 5-hydroxytryptamine by ﬁve antidepressants; correlation with reduction of spontaneous ﬁring rate of central monoaminergic neurones. Naunyn Schmiedebergs Arch. Pharmacol., 319: 66–70. Raiteri, M. (2001) Presynaptic autoreceptors. J. Neurochem., 78: 673–675. Riad, M., Garcia, S., Watkins, K.C., Jodoin, N., Doucet, E., Langlois, X., El Mestikawy, S., Hamon, M. and Descarries, L. (2000) Somatodendritic localization of 5HT1A and preterminal axonal localization of 5-HT1B serotonin receptors in adult rat brain. J. Comp. Neurol., 417: 181–194. Rollema, H., Clarke, T., Sprouse, J.S. and Schulz, D.W. (1996) Combined administration of a 5-hydroxytryptamine (5HT)1D antagonist and a 5-HT reuptake inhibitor synergistically increases 5-HT release in guinea pig hypothalamus in vivo. J. Neurochem., 67: 2204–2207.

Romero, L. and Artigas, F. (1997) Preferential potentiation of the effects of serotonin uptake inhibitors by 5-HT1A receptor antagonists in the dorsal raphe pathway: role of somatodendritic autoreceptors. J. Neurochem., 68: 2593–2603. Romero, L., Bel, N., Artigas, F., deMontigny, C. and Blier, P. (1996a) Effect of pindolol on the function of pre- and postsynaptic 5-HT1A receptors: in vivo microdialysis and electrophysiological studies in the rat brain. Neuropsychopharmacology, 15: 349–360. Romero, L., Celada, P. and Artigas, F. (1994) Reduction of in vivo striatal 5-hydroxytryptamine release by 8-OH-DPAT after inactivation of Gi/Go proteins in dorsal raphe nucleus. Eur. J. Pharmacol., 265: 103–106. Romero, L., Herva´s, I. and Artigas, F. (1996b) The 5-HT1A antagonist WAY-100635 selectively potentiates the presynaptic effects of serotonergic antidepressants in rat brain. Neurosci. Lett., 219: 123–126. Rutter, J.J. and Auerbach, S.B. (1993) Acute uptake inhibition increases extracellular serotonin in the rat forebrain. J. Pharmacol. Exp. Ther., 265: 1319–1324. Rutter, J.J., Gundlah, C. and Auerbach, S.B. (1994) Increase in extracellular serotonin produced by uptake inhibitors is enhanced after chronic treatment with ﬂuoxetine. Neurosci. Lett., 171: 183–186. Sacchetti, G., Bernini, M., Bianchetti, A., Parini, S., Invernizzi, R.W. and Samanin, R. (1999) Studies on the acute and chronic effects of reboxetine on extracellular noradrenaline and other monoamines in the rat brain. Br. J. Pharmacol., 128: 1332–1338. Sa´nchez, C. and Hyttel, J. (1999) Comparison of the effects of antidepressants and their metabolites on reuptake of biogenic amines and on receptor binding. Cell. Mol. Neurobiol., 19: 467–489. Santana, N., Bortolozzi, A., Serrats, J., Mengod, G. and Artigas, F. (2004) Expression of serotonin1A and serotonin2A receptors in pyramidal and GABAergic neurons of the rat prefrontal cortex. Cereb. Cortex, 14: 1100–1109. Santarelli, L., Gobbi, G., Debs, P.C., Sibille, E.L., Blier, P., Hen, R. and Heath, M.J.S. (2001) Genetic and pharmacological disruption of neurokinin 1 receptor function decreases anxiety-related behaviours and increases serotonergic function. Proc. Natl. Acad. Sci. U.S.A., 98: 1912–1917. Scuve´e-Moreau, J. and Dresse, A. (1979) Effect of various antidepressant drugs on the spontaneous ﬁring rate of locur coeruleus and dorsal raphe neurons of the rat. Eur. J. Pharmacol., 57: 219–225. Sesack, S.R., Deutch, A.Y., Roth, R.H. and Bunney, B.S. (1989) Topographical organization of the efferent projections of the medial prefrontal cortex in the rat: an anterograde tract–tracing study with Phaseolus vulgaris leucoagglutinin. J. Comp. Neurol., 290: 213–242. Sotelo, C., Cholley, B., El Mestikawy, S., Gozlan, H. and Hamon, M. (1990) Direct immunohistochemical evidence of the existence of 5-HT autoreceptors on serotonergic neurons in the midbrain raphe nuclei. Eur. J. Neurosci., 2: 1144–1154.

543 Steinberg, R., Alonso, R., Rouquier, L., Desvignes, C., Michaud, J.-C., Cudennec, A., Jung, M., Simiand, J., Griebel, G., Edmons-Alt, X., Le Fur, G. and Soubrie´, P. (2002) SSR240600 [(R)-2(1-{2-[4-{2-[3,5-bis(triﬂuoromethyl)phenyl]acetyl}-2-(3,4-dicholorophenyl)-2-morpholinyl]ethyl}4-piperidinyl)-2-methylpropanamide], a centrally active nonpeptide antagonist of the tachykinin neurokinin 1 receptor: II. Neurochemical and behavioral characterization. J. Pharmacol. Exp. Ther., 303: 1180–1188. Tanda, G., Carboni, E., Frau, R. and Di Chiara, G. (1994) Increase of extracellular dopamine in the prefrontal cortex: a trait of drugs with antidepressant potential? Psychopharmacology (Berl.), 115: 285–288. Tanda, G., Frau, R. and Di Chiara, G. (1996) Chronic desipramine and ﬂuoxetine differentially affect extracellular dopamine in the rat prefrontal cortex. Psychopharmacology (Berl.), 127: 83–87. Tao, R., Ma, Z. and Auerbach, S.B. (2000) Differential effect of local infusion of serotonin reuptake inhibitors in the raphe versus forebrain and the role of depolarization-induced release in increased extracellular serotonin. J. Pharmacol. Exp. Ther., 294: 571–579. Tatsumi, M., Groshan, K., Blakely, R.D. and Richelson, E. (1997) Pharmacological proﬁle of antidepressants and related compounds at human monoamine transporters. Eur. J. Pharmacol., 340: 249–258. Thase, M.E., Entsuah, A.R. and Rudolph, R.L. (2001) Remission rates during treatment with venlafaxine or selective serotonin reuptake inhibitors. Br. J. Psychiatry, 178: 234–241. Tollefson, G.D. and Holman, S.L. (1994) How long to onset of antidepressant action – a metaanalysis of patients treated

with ﬂuoxetine or placebo. Int. Clin. Psychopharmacol., 9: 245–250. Uhl, G.R. and Hartig, P.R. (1992) Transporter explosion: update on uptake. Trends Pharmacol. Sci., 13: 421–425. Valentini, V., Frau, R. and Di Chiara, G. (2004) Noradrenaline transporter blockers raise extracellular dopamine in medial prefrontal but not parietal and occipital cortex: differences with mianserin and clozapine. J. Neurochem., 88: 917–927. Vizi, E.S., Zsilla, G., Caron, M.G. and Kiss, J.P. (2004) Uptake and release of norepinephrine by serotonergic terminals in norepinephrine transporter knock-out mice: implications for the action of selective serotonin reuptake inhibitors. J. Neurosci., 24: 7888–7894. Weikop, P., Kehr, J. and Scheel-Kruger, J. (2004) The role of a1- and a2-adrenoreceptors on venlafaxine-induced elevation of extracellular serotonin, noradrenaline and dopamine levels in the rat prefrontal cortex and hippocampus. J. Psychopharmacol., 18: 395–403. Wikell, C., Hjorth, S., Apelqvist, G., Kullingsjo, J., Lundmark, J., Bergqvist, P.B. and Bengtsson, F. (2001) Sustained administration of the antidepressant venlafaxine in rats: pharmacokinetic and pharmacodynamic ﬁndings. Naunyn Schmiedebergs Arch. Pharmacol., 363: 448–455. Zocchi, A., Varnier, G., Arban, R., Griffante, C., Zanetti, L., Bettelini, L., Marchi, M., Gerrard, P.A. and Corsi, M. (2003) Effects of antidepressant drugs and GR 205171, an neurokinin-1 (NK1) receptor antagonist, on the response in the forced swim test and on monoamine extracellular levels in the frontal cortex of the mouse. Neurosci. Lett., 345: 73–76.

CHAPTER 6.4

Microdialysis as a method to study blood-brain barrier transport mechanisms Elizabeth C.M. de Lange Division of Pharmacology, Leiden/Amsterdam Center for Drug Research, Leiden University, Gorleaus Laboratories, Leiden, The Netherlands

Abstract: Often, blood-brain barrier (BBB) transport is still considered qualitatively, in terms of ‘‘limited’’ or ‘‘readily’’. However, for a given compound, the rate and extent of BBB transport will be the net result of the contribution of many possible passive and active dynamically regulated transport mechanisms. This implies that knowledge of BBB transport mechanisms and regulation is critical for the understanding of brain homeostasis; on one hand to gain insights into how disturbances thereof may lead to CNS diseases, and on the other hand to ultimate being able to predict the relation between the kinetics and the dynamics of CNS active compounds. Among a number of in vivo techniques to study BBB transport, in vivo microdialysis has the important unique characteristic to provide data on free concentrations in the extracellular (EC) spaces in the body. This characteristic makes microdialysis especially valuable as a technique to study the kinetics of BBB membrane transport. In this chapter, the focus is on the use of intracerebral microdialysis technique to study BBB transport and (within) brain distribution. A total of 282 references are used. Following introduction, in II, the determinants in BBB functionality are presented with circumstances that affect BBB functionality. Then, part III deals with in vivo techniques to study BBB transport, including microdialysis. Subsequently part IV presents crucial methodological factors that determine the validity of microdialysis outcomes, approaches in pharmacokinetic modelling of BBB transport using microdialysis, brain tissue and plasma/blood data. Then, in V, the currently available reviews on BBB transport studies with microdialysis are presented, followed by the more recent original papers dealing with passive and active BBB transport mechanisms, as well as potential changes thereof in a number of disease conditions. Finally (VI), it is concluded that the use of microdialysis combined with brain tissue and plasma/blood data provides a good approach in characterising BBB transport mechanisms in many species, under a variety of conditions.

system (CNS) diseases. In addition, BBB transport mechanisms may have an important inﬂuence on the concentration–time proﬁles of CNS active drugs at different sites within the brain, including the target site (the biophase) as an important determinant of CNS drug responses (dynamics). Knowledge on these mechanisms is essential for the ultimate goal to predict the relation between the kinetics and the dynamics of CNS active compounds.

I. Introduction The blood-brain barrier (BBB) regulates the exchange of compounds between blood and brain. Knowledge on BBB transport mechanisms and regulation is critical for the understanding of brain homeostasis, as well as for insights into how disturbances thereof may lead to central nervous Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

545

DOI: 10.1016/S1569-7339(06)16029-4 Copyright 2007 Elsevier B.V. All rights reserved

546

Often, BBB transport is still considered qualitatively, in terms of ‘‘limited’’ or ‘‘readily’’, without taking into account that for a given compound the rate and extent of BBB transport will be the net result of the contribution of many possible passive and active transport mechanisms. As BBB functionality is dynamically regulated, the contribution of those mechanisms to the overall BBB transport may vary under different conditions. Such conditions include time, subject, disease state, genetic background and age. Among a number of techniques available to study BBB transport, intracerebral microdialysis is an in vivo technique that permits monitoring local concentrations of solutes at speciﬁc sites in the brain. As microdialysis is a delicate technique, several factors have to be considered at the surgical, experimental and analytical level. Provided that these factors are optimised to give valid and quantitative results, intracerebral microdialysis provides many high-resolution concentration–time data from a relatively small number of individual animals. Actually, the most important unique characteristic of microdialysis is that information is obtained on free concentrations in the extracellular (EC) spaces in the body. This characteristic makes microdialysis especially valuable as a technique to study the kinetics of transport equilibration across the BBB. This chapter will focus on the use of the intracerebral microdialysis technique to study BBB transport and (within) brain distribution. In the ﬁrst part, important aspects of the BBB will be presented; its characteristics and transport mechanisms that altogether deﬁne BBB functionality, as well as circumstances that affect BBB functionality. The second part will deal shortly with in vivo methodologies that may provide information on BBB transport and/or brain distribution, with the special value of microdialysis in this. The third part reviews important methodological considerations in using microdialysis for studying BBB transport. It then provides equations that can be used in pharmacokinetic modelling of microdialysis, brain tissue and plasma/blood data to reveal pharmacokinetic parameters of BBB transport and brain distribution, in terms of clearances and volumes, or rate constants and amounts. Subsequently, a number of reviews on BBB transport

studies using microdialysis are given, followed by the presentation of the more recent original papers dealing with passive and active BBB transport mechanisms, and potential changes thereof in a number of disease conditions. Finally, the conclusion is presented.

II. The blood-brain barrier (BBB) II.A. Characteristics of the BBB The BBB is the cellular layer that separates the blood compartment from the brain. It is a selectively permeable barrier that prevents the unrestricted exchange of molecules (Davson and Welch, 1971; Fenstermacher et al., 1974; Rapoport et al., 1979; Bradbury, 1984; Cornford, 1985; Pardridge, 1988; Vorbrodt, 1988; Steward and Mikulis, 1998; Rubin and Staddon, 1999; Somjen, 2002). As it highly regulates the exchange of compounds between blood and brain, the BBB has an important role in maintaining the optimal environment for the adequate functioning of the CNS (Abbott and Revest, 1991; Banks, 1999; Mayhan, 2001; Abbott, 2002). The actual BBB is formed by an ensemble of different cell types, including brain endothelial cells, pericytes, the basal lamina, astrocytes and neurons. The main component of the BBB is the brain endothelial cell that serves as the vessel wall of the brain capillaries, thus having direct contact with the blood compartment (the lumen). A special feature of these brain endothelial cells is the presence of so-called ‘‘tight junctions’’ between adjacent cells that narrow the paracellular space (Huber et al., 2001). Besides, the brain endothelial cells express numerous active transport mechanisms, in a polarised fashion. Furthermore, the brain endothelial cells are in close contact with the pericytes that potentially serve a role in vascular contraction and may as well inﬂuence the tightness of the BBB endothelium (Minakawa et al., 1991). Both the brain endothelial cells and the pericytes are surrounded by the basal lamina that provides mechanical support for cell attachment. At the brain side of the basal lamina, the astrocytes have their end feet covering the full surface along the brain capillary endothelium. The interactions

547

of these astrocytic end feet with the brain endothelial cells are of importance for BBB morphology, protein expression and functionality (Cancilla and DeBault, 1983; Beck et al., 1984; Arthur et al., 1987; Abbott and Revest, 1991; Abbott, 2002; Abbott, et al., 2006). Finally, also the neurons in close vicinity of the BBB endothelium contribute to the regulation of BBB functionality. II.B. Transport mechanisms at the BBB The BBB has many possible transport mechanisms, generally to be divided into passive and active transport into and out of the brain (Oldendorf, 1974; Levin, 1980; Bodor and Brewster, 1983; Collins and Dedrick, 1983; Aird, 1984; Bradbury, 1984; Fenstermacher et al., 1995; Groothuis and Levy, 1997; Gloor et al., 2001; de Lange and Danhof, 2002; de Boer et al., 2003; Begley, 2004). The different transport mechanisms are presented here. II.B.1. Passive BBB transport Passive transport occurs on the basis of simple and facilitated diffusion. Simple diffusion is directly dependent on the magnitude of the concentration gradient of (exchangeable fraction of) the solute over the membrane, and may occur between cells (paracellular diffusion) and across cells (transcellular diffusion). In general, the driving force for transport across a membrane is to reach equilibrium in the concentration of the exchangeable fraction of a solute. This fraction is governed by many factors such as (protein) binding in plasma and brain. Also for charged solutes with pKa values around the physiological pH range, in this respect, a pH difference at either side of the BBB is of importance. Diffusion of hydrophilic solutes across the BBB is mostly restricted due to the presence of the ‘‘tight junctions’’, which narrow the paracellular route. More lipophilic compounds may also use the transcellular route. As a rule of thumb, a higher lipophilicity of a compound will enhance its ability to cross the cell membranes (Oldendorf, 1974; Levin, 1980). In the case of facilitated diffusion (a form of carrier-mediated endocytosis), there is a binding of a solute to a transporter on one side of the membrane that

triggers a conformational change in the protein. This conformational change results in carrying the solute from the high to the low concentration side of the BBB. Facilitated diffusion contributes to BBB transport of compounds like amino acids, monocarboxylates, hexoses, amines, nucleosides, glutathione and small peptides, and contributes to brain homeostasis. II.B.2. Active BBB transport Active transport mechanisms include pinocytosis, receptor-mediated endocytosis, absorptive- and carrier-mediated transport and the transportermediated efﬂux and inﬂux (Brightman, 1977; Brightman et al., 1983; Mooradian, 1988; Tsuji and Tamai, 1999; Cornford and Cornford, 2002; de Boer et al., 2003). Pinocytosis is non-speciﬁc ﬂuid uptake that is temperature and energy dependent, non-competitive and non-saturable. Under physiological conditions, the degree of pinocytosis in brain endothelial cells is very limited. It may increase under disturbed conditions (such as seizures). Receptor-mediated endocytosis is a highly speciﬁc type of energy dependent transport in membrane areas called coated pits. Macromolecules cannot pass the BBB under normal conditions if not for the presence of receptor-mediated endocytosis by which many different types of macromolecules are selectively taken up, like transferrin, insulin-like growth hormone, leptin and immunoglobulin molecules. Absorptive-mediated transport is triggered by an electrostatic interaction between a positively charged solute (usually a charge moiety of a peptide) and the negatively charge plasma membrane surface (i.e., glycocalyx). Carrier-mediated transport is dependent on energy and/or co-transport of another substance. Co-transport may occur in the same direction (symport) or in the opposite direction (antiport), from high to low concentration. Other signiﬁcant transport mechanisms at the BBB are transporter-mediated inﬂux and efﬂux (Greig et al., 1987; Ooie et al., 1997; Suzuki et al., 1997; Takasawa et al., 1997; Sugiyama et al., 1999; Gao et al., 2000; Kusuhara, 2001a, b; Lee et al., 2001; de Lange and Danhof, 2002; Mahar Doan et al., 2002; Kusuhara et al., 2003; Mori et al., 2003; Sun et al., 2003; Gibbs et al., 2004; Kikuchi et al.,

548

2004; Loscher and Potschka, 2005). Especially the efﬂux transporters at the BBB have gained enormous attention in research of the last decade. The presence of these efﬂux transporters explain the fact that for many drugs, despite their favourable lipophilic character, a relatively poor brain distribution was found due to the fact that these drugs are substrates for the efﬂux transporters at the BBB. Here the P-glycoprotein (Pgp, or ofﬁcially ABCB1), the multidrug resistance related proteins (MRP’s, or ofﬁcially ABCC’s,) and the breast cancer resistance protein (BCRP, or ofﬁcially ABCG2), all belonging to the ABC transporters (Dean et al., 2001; Higgins, 2001; Borst and Oude Elferink, 2002), are shortly presented below. II.B.2.a. P-glycoprotein (ABCB1). Pgp is the most well-known among the ABC transporters and was the ﬁrst to be identiﬁed in man in which it plays a critical role in drug resistance in treatment of cancers. Numerous investigations with many drugs have demonstrated that Pgp has an important role in determining the concentration–time proﬁles of Pgp-substrates in the different parts of the body, including the brain as Pgp is expressed at the luminal face of the BBB (Cordon-Cardo et al., 1989; Schinkel, 1994, 1995, 1996; Mayer et al., 1997; Kim et al., 1998; Schinkel, 2001; Bendayan et al., 2002; Kim, 2002). In general, Pgp preferentially extrudes large hydrophobic, positively charged molecules (Klopman et al., 1997; Seelig et al., 1998, 2000). In addition of having many substrates among CNS active drugs, this transporter also is involved in the transport of certain cytokines and may even play a role in the inhibition of apoptosis and in the pathogenesis of Alzheimer’s disease, by direct interaction of Pgp with Amyloid-b40 and Amyloid-b42, or by inﬂuencing the accumulation of these proteins (Lam et al., 2001; Vogelgesang et al., 2002). Also, Pgp is of importance in the neuroendocrine functioning and regulation of the hypothalamic-pituitary-adrenocortical (HPA) axis, by regulating the efﬂux of certain natural and synthetic glucocorticoids from the brain (Karssen et al., 2001, 2002). II.B.2.b. Multidrug resistance related proteins (ABCC’s). The total MRP family now includes

MRP1 to MRP9 (ABCC7 and ABCC8) (HuaiYun et al., 1998; Borst et al., 2000; Hopper et al., 2001; Belinsky et al., 2002; Wijnholds, 2002; Chen et al., 2003). It seems that MRP2 and MRP4 are expressed at the luminal face of the BBB (Miller et al., 2000, 2002), acting as efﬂux transporters, while MRP1, MRP3, MRP5 and MRP6 seem to be expressed at the albuminal face of the BBB as inﬂux transporters (Regina et al., 1998; Zhang et al., 2000; Schinkel, 2001; Kusuhara and Sugiyama, 2005). Their functionality in vivo, and the substrate speciﬁcity at the level of the BBB is still largely unknown. MRP1 was ﬁrst discovered and actually is a prototype glutathion conjugate pump that transports a variety of drugs conjugated to glutathione, sulfate or to glucuronate, but also anionic drugs and dyes, neutral-basic amphiphatic drugs and even oxyanions (Jedlitschky et al., 1996; Hipfner et al., 1999; Leslie et al., 2001; Qian et al., 2001). The MRP’s have a broad tissue distribution and are able to extrude negatively charged anionic drugs and neutral drugs conjugated to glutathione, glucuronate or sulfate. Some MRP’s are able to transport neutral drugs if cotransported with glutathione. The most important substrate of MRP1 is the endogenous leukotriene LTC4. MRP2 and MRP3 have overlapping substrate speciﬁcities. MRP4 and MRP5 broaden the spectrum of drugs to the nucleotide analogs. II.B.2.c. Breast cancer resistance protein (ABCG2). In general, the BCRP transporter preferentially transports large hydrophobic, positively charged molecules. BBB is expressed as an efﬂux transporter (Lee et al., 2001), at the luminal face of the brain endothelial cells (Cooray et al., 2002; Eisenblatter et al., 2003). II.C. Changes in BBB functionality Under normal conditions already regional differences in BBB functionality may exist (Gross et al., 1986), but as the BBB is dynamically regulated a diversity of conditions have been shown to affect BBB functionality. This may occur at different levels of mechanisms of BBB transport. While it is known that BBB functionality changes upon treatments like osmotic opening (Rapoport and

549

Robinson, 1985; de Lange et al., 1995b), it may as well result from ageing (Mooradian, 1988) pathological conditions. Just to mention a few examples BBB functionality may change by neuroinﬂammation (de Vries et al., 1997; Lossinsky and Shiver, 2004), multiple sclerosis (Brosman and Claudio, 1998), hypertension, HIV-infection (Nottet, 1999), traumatic brain injury (Povlishock, 1998; Bouw et al., 2001b; Lo et al., 2001), brain tumours (Bolwig et al., 1977; Blasberg et al., 1981; Hasegawa et al., 1983; Groothuis et al., 1984; de Lange et al., 1995a), cerebral malaria (Turner, 1998), epilepsy and/or seizures (Petito et al., 1977; Nitsch and Klatzo, 1983; Bolwig et al., 1997; Cornford, 1999; Dombrowski et al., 2001; Seegers et al., 2002; Seiffert et al., 2004), meningitis (Kim et al., 1997) and stress (Fenstermacher et al., 1995). However, the impact of a disease (and its progression) on the changes in speciﬁc mechanisms of BBB transport deserves further investigation.

III. In vivo techniques for BBB transport Blood-brain barrier transport can be assessed by a number of in vivo and ex vivo techniques (Bonate, 1995; de Lange et al., 1997a, b, 1999; Foster and Roberts, 2000; de Lange and Danhof, 2002; Feng et al., 2001; Dash and Elmquist, 2003; Smith, 2003; Bickel, 2005; Plock and Kloft, 2005). These include the more classical single time-point pharmacokinetic techniques like intravenous (i.v.) administration and tissue sampling, the brain uptake index (Hardebo et al., 1970; Oldendorf, 1979), and the brain perfusion technique (Takasato et al., 1984). In the brain uptake and brain perfusion techniques (part of), the brain tissue is used for analysis, following a short experimental period of perfusion of the brain with a perfusate containing the compound of interest, as performed in animals under anaesthesia. The brain uptake of the compound is expressed as a percentage of that of a marker with maximal BBB transport. A correction for the ‘‘vascular space contribution’’ of the perfusion concentrations to the total brain homogenate concentration is made by the incorporation of an impermeable marker in the perfusate. The brain efﬂux index technique (Kakee et al., 1996)

characterises the efﬂux (loss) of the compound of interest following its local brain injection on the basis of the ‘‘left over’’ brain tissue concentrations together relative to an impermeable marker that has been added to the injectate as a check for BBB integrity and actually administered dose. Another single-time-point method is the quantitative autoradiography (Mans et al., 1987; Namba et al., 1987; Wallace et al., 1992) in which the spatial distribution of the label over the brain is measured, supposed to represent the radiolabelled compound of interest. All the above-mentioned in vivo techniques do not discriminate between intracellular (IC) and extracellular ﬂuid (ECF) concentrations of the compound. The serial cerebrospinal ﬂuid (CSF) sampling technique (Bouman et al., 1979; van Bree et al., 1991; de Lange et al., 1997a) is able to reﬂect the brain ﬂuid concentrations as present in the ventricular space, as a function of time. Often the cisterna magna is sampled. It is more exception than a rule that CSF concentrations are equal to the ECF concentrations around the parenchymal cells in the brain tissue (de Lange and Danhof, 2002). The dimensions of time as well as spatial resolution are encompassed by the imaging techniques; positron emission tomography (PET; Agon et al., 1991, 1988; Ponto-Boles et al., 1992a, b; Yu et al., 1992; Elsinga et al., 2004), and nuclear magnetic resonance (NMR; Bartels et al., 1991; Pouremad et al., 1999; de Graaf et al., 2001). Though in principle very elegant and powerful, for PET scanning the half-life of positron emitting isotopes is the main limiting factor, while for NMR many scans are needed to obtain a timeaveraged picture with acceptable signal-to-noise ratio. Also, for PET and NMR scans, experimental animals need to have a ﬁxed position, often accomplished by anaesthesia. As these imaging technologies suffer from being very expensive, they are only available at a limited number of locations. Intracerebral microdialysis is a technique to monitor unbound concentrations in the brain ECF, as well as in plasma, in freely moving individual subjects (de Lange et al., 1997b). This technique is of special value for the determination of BBB membrane transport characteristics, as to

550

that end information on the unbound concentration–time proﬁles at either side of the BBB are needed as driving forces for membrane transport. Beside BBB transport, ECF concentrations will be determined by bulk ﬂow, EC metabolism, and intra-EC exchange. The relationship between ECF and IC concentrations can be derived from the ECF and the total brain (part) concentrations. The latter may be obtained at appropriate times of ending microdialysis experiments. Thus, BBB transport characteristics of compounds can be assessed in terms of rate and extent of inﬂux and efﬂux clearances (Hammarlund-Udenaes et al., 1997; Deguchi and Morimoto, 2001; Deguchi, 2002; Bickel, 2005).

IV. Microdialysis in BBB transport IV.A. Methodological considerations Speciﬁcally for BBB transport measurements, a number of important methodological aspects needs to be considered before and during the use of intracerebral microdialysis: (1) the implantation of the microdialysis probe is an invasive procedure and may affect brain tissue and the BBB, (2) the composition of the microdialysis perfusion solution inﬂuences the data outcomes, and (3) the in vivo concentration recovery (the relation between microdialysate concentrations and the true brain ECF concentrations) needs to be determined for quantitative data. IV.A.1. Probe implantation The implantation of a microdialysis probe is, despite its small size, invasive and could affect brain tissue and BBB functionality. Factors that will inﬂuence tissue trauma and therewith the data obtained at the time of the microdialysis experiment, include the speed of implantation (Allen et al., 1992) and the size of the probe (Coleman et al., 1974; Matlaga et al., 1976), as well as the duration of the post-surgery interval. Studies on tissue changes and BBB integrity are presented below. Following implantation of a microdialysis probe into the hippocampus, Benveniste et al. (1987) investigated local cerebral blood ﬂow (LCBF) and

local cerebral glucose metabolism (LCGM), being sensitive indicators of regional damage. After initial disturbances, at 24 h post-surgery in all rats nearly normalised LCMR and LCBF were found, while disturbances of glucose phosphorylation and blood ﬂow were absent. Histological evaluations of tissue trauma by intracerebral microdialysis have been performed by several groups (Imperato et al., 1984; Imperato and Di Chiara, 1985; L’Heureux et al., 1986; Benveniste and Diemer, 1987; Ruggeri et al., 1990; Shuaib et al., 1990; de Lange et al., 1995d). In general, only minimal reactions to the probe were found. At 1–2 days after implantation of the microdialysis probe, the histological changes were mostly conﬁned to a very thin zone around the microdialysis probe, occasionally with haemorrhages. At later times, activated astrocytes as ependyma-like glia were present in some cases, sometimes intruding the probe membrane material. A difference was found between histological tissue reactions to a non-perfused and a perfused probe, respectively silently present for 4 days or perfused on 4 consecutive days. The latter showed more hypercellularity around the probe (de Lange et al., 1995d). Using autoradiography of 14C-alpha-aminoisobutyric acid (14C-AIB) no 14C-AIB was found to be present around the microdialysis probe except for the areas close to the burr holes for the transversal probe implantation, indicating an intact BBB at the measuring site (Benveniste et al., 1984). With experiments starting at 24 h after probe implantation, de Lange et al. (1994) investigated BBB transport of a hydrophilic and a moderately lipophilic drug following i.v. injection. In accordance with outcomes using other experimental techniques to measure BBB transport as based on AUC ratios, it was found that the hydrophilic drug atenolol was restricted in BBB transport (4%) ,while the more lipophilic drug acetaminophen appeared to penetrate the brain more easily (18%; de Lange et al., 1994). For atenolol, no real changes in BBB transport were found following i.v. administration on repetitive experiments performed at 3 consecutive days (de Lange et al., 1995d), indicating the possibility of repeated BBB transport measurements.

551

While the studies above indicate that it is possible to maintain an intact BBB under their experimental conditions, also contrasting results were found. Major et al. (1990) showed that the permeability of the BBB for Cr-51-EDTA was increased directly after probe implantation, and declined subsequently but still remained elevated up to 24 h after surgery. Westergren et al. (1995) evaluated BBB integrity by Evans Blue-albumin staining, immunohistochemistry, 3H-inulin and glutamate. They observed Evans Blue around the probe and in surrounding brain tissue, while 3Hinulin was detected in the dialysate, indicating a compromised BBB. Groothuis et al. (1998) found changes in BBB permeability to be associated with probe insertion of brain cannulas and microdialysis probes using 14C-sucrose and 14C-dextran. BBB permeability seemed to be disturbed in response to insertion of the brain cannula or probe tip. This disturbance varied with time, and could persist for at least 28 days after insertion. Morgan et al. (1996) found indications for BBB damage, by comparison of microdialysis with a classic method to measure the apparent brain uptake of two polar permeants, 3H-sucrose and 14C-urea, the latter having 20-fold higher blood-to-brain transfer constant in the classic one-point-per-animal method. During microdialysis sampling they found (1) the loss of 3H-sucrose from brain ECF on termination of an i.v. infusion to be biphasic, with the initial phase evident immediately on termination by the infusion, suggesting that a fraction of the microdialysis probe resides in a region in rapid equilibrium with plasma; (2) complete loss of selectivity in the rate constants for CNS entry of sucrose vs. urea; and (3) substantially higher AUCbrain ECF/AUCplasma ratios for both sucrose and urea generated by microdialysis than the corresponding ratios (AUCCSF/AUCplasma or AUCbrain/AUCplasma) obtained by classic methods. However, in this study, the potential inﬂuence of EC trapping and thereby higher ECF than total brain concentrations (Stain et al., 1995), has not been addressed. The integrity of the BBB has also been evaluated by assessing the change in BBB transport induced by techniques known to open the BBB. Westergren et al. (1995) found no further increase in 3H-inulin

levels in the dialysate after opening of the BBB with intracarotid injection of protamine. In contrast, Allen et al. (1992) found that the levels of 4-trimethylammonium antipyrine levels in the dialysate increased signiﬁcantly with BBB opening by oleic acid. In addition, the results of this study provided information about a possible relationship between the rate of probe implantation and the degree of BBB damage. This may in part account for the differences found in results on BBB integrity. Finally, de Lange et al. (1995b) could demonstrate a 10-fold increase in atenolol, upon i.v. administration, dialysate levels after intracarotid injection of a hypertonic mannitol solution while plasma proﬁles remained unchanged. Intruigingly, a circadianic variability in this effect was observed as the BBB permeability increases only in ‘‘afternoon’’ experiments. IV.A.2. Perfusate The microdialysate perfusion solution should mimic the periprobe ECF in the sense that the system under investigation will not be disturbed by deviations in ion ﬂuid composition of the perfusate. Moghaddam and Bunney (1989) were the ﬁrst to assess the effect of such ionic composition differences on microdialysate outcomes, based on measurements of EC dopamine (DA) levels during resting conditions and following a pharmacological manipulation. The probe perfusion with a Ringer’s solution (with the ionic composition that mimics that of plasma, as mostly used in those days) or an artiﬁcial ECF. This resulted in different data on DA turnover rate and basal release. Especially the higher calcium levels, that is 3.4 mM, relatively to those present in the ECF of the brain striatum (1.2 mM, as determined by the authors, was able to alter the pharmacological responsiveness of the nigrostriatal DA system to synthesis inhibition. Also McNay and Sherwin (2004) reported effects of small variations in the ionic composition of microdialysis perfusate. Such variations produced fourfold differences (0.53–2.18 mM) in the results obtained when measuring brain ECF glucose. These changes may be linked to concomitant alterations of local neural activity caused by the perfusate composition. In addition to perfusate composition, probe

552

type also proved to have a signiﬁcant impact on microdialysis measurements. Furthermore, signiﬁcant differences between brain ECF and CSF ionic composition as measured by microdialysis were presented by these authors. For BBB transport, de Lange et al. (1994) have shown that BBB transport can be affected by the composition of the perfusate. For 3 consecutive days, repetitive experiments in rats were performed to measure brain dialysate concentrations of the hydrophilic compound atenolol following i.v. administration, using either an isotonic or a hypotonic perfusion solution. While highly reproducible dialysate concentrations of atenolol were observed with the isotonic perfusate, the use hypotonic perfusate solution gave rise to a different dialysate concentration–time proﬁle at the ﬁrst experiment, while considerably elevated dialysate levels of atenolol were found at the second and the third day. This was presumed to be a result of increased permeability of the BBB, as in the same study no such changes were observed for the more lipophilic acetaminophen. Also, in this study (de Lange et al., 1994), the effect of a potential temperature difference between the microdialysis perfusion solution and the periprobe tissue on microdialysis data was considered, and investigated again for atenolol and acetaminophen. A comparison was made between a perfusion solution at room temperature when entering the probe (direct connection), and a perfusion solution that was led through a subcutaneous cannula before entering the probe (indirect connection), assumed to be equilibrated to rat body temperature. A temperature effect was indeed observed, but only for the use of the hypotonic perfusate. It was hypothetised that the periprobe tissue, already ‘‘stressed’’ by the hypotonic condition, looses its capability to compensate temperature effects. This indicates that perfusate temperature may be especially important in pathological circumstances. However, it is recommended to perform all microdialysis experiments with perfusion ﬂuids at body temperature. IV.A.3. Quantitative microdialysis The relation between the concentrations of the compound of interest in the dialysates and ECF (termed in vivo concentration recovery or often

also in vivo recovery) depends on the exchange of the compound between the probe perfusate and the periprobe ECF. This relation is affected by the probe characteristics, and experimental variables (Parsons and Justice, 1994). A number of approaches have been proposed to experimentally estimate in vivo recovery (Lerma et al., 1986; Lonnroth et al., 1987; Larsson, 1991; Scheller and Kolb, 1991; Olson and Justice, 1993; Yokel et al., 1992). Mathematical studies indicated that tissue processes play an important role in in vivo recovery. Bungay et al. (1990) developed a mathematical framework for in vivo concentration recovery. Assuming that (1) the probe is inserted in tissue within a normal state; (2) there is intimate contact between tissue and outer surface of the probe and that (3) the diffusion through the tissue for hydrophilic compounds takes place through the ECF with linear transport processes, it was found that in vivo concentration recovery is determined by a series of mass transfer resistances at the level of the tissue, the microdialysis membrane and the dialysate. For most experiments, the resistance for mass transfer through the tissue is by far the largest. The theoretical model indicated that under steady-state conditions all processes that contribute to elimination of a drug, will affect in vivo concentration recovery, such as efﬂux to the microvasculature, irreversible EC metabolism, and the composite of irreversible IC metabolism and EC–IC exchange. Inﬂux of the compound does not affect the in vivo concentration recovery, but plays a role in determining the concentrations of the drug in brain ECF. By Morrison et al. (1991), the mathematical approach was extended to transients conditions. In line with these theoreretical implications, a number of microdialysis experiments have conﬁrmed the importance of elimination processes in the brain on the in vivo recovery values found. Olson and Justice (1993) presented their dynamicno-net-ﬂux (DNNF) microdialysis method for the quantitative determination of EC concentrations of DA under transient conditions. Following cocaine and amphetamine administration, a signiﬁcantly greater increase in extracellular DA was found than was estimated from the dialysate using conventional microdialysis methods. It was found that this discrepancy was due to in vivo recovery

553

decreasing concurrently with the increased DA ECF concentrations following cocaine and amphetamine injections. In a study by de Lange et al. (1998), a 4.5-fold difference in vivo recovery of the Pgp substrate rhodamine-123 (R123) from the brain was found between the mdr1a(–/–) and the wild-type mice, due to the absence of the efﬂux of R123 by Pgp at the BBB of the mdr1a(–/–) mice. Similar ﬁndings in mdr1a(–/–) compared with wild-type mice were reported for morphine (Xie et al., 1999) and for sparﬂoxacin (de Lange et al., 2000b). Then, Sun et al. (2001a) investigated the effects of BBB transporter inhibitors on the in vivo recovery. Two combinations of transporter substrate and inhibitor were studied (ﬂuorescein and probenecid, and quinidine and LY335979), and in vivo recovery was investigated for the substrate either with or without inhibitor. Probenecid decreased the in vivo recovery of ﬂuorescein in the frontal cortex, while the Pgp inhibitor LY335979 did not change the in vivo recovery of quinidine. This indicates that only in certain cases, changes in BBB efﬂux may affect actual in vivo recovery values of the substrate. Another example is provided by Linden et al. (2003). Alovudine blood and brain ECF proﬁles were determined with and without the presence of the transport inhibitors quinidine or probenecid. The ratio of brain ECF over plasma AUC values were 0.24 under control conditions. This ratio declined to 0.17 with probe perfusion of probenecid. With quinidine, this ratio was even lower, 0.085, however without concomitant changes in total brain tissue concentrations. In this study, the in vivo recovery of alovudine was increased by quinidine, while no changes were observed by probenecid. To obtain meaningful data on BBB transport and brain distribution knowledge of actual in vivo concentration recovery is needed. If information on in vivo recovery is properly obtained, that is not locally inﬂuencing the tissue characteristics, it is of high interest to use it to gain insight in periprobe elimination mechanisms of the compound of interest.

play a role in the relation between plasma concentrations and CNS drug effects, in other words to contribute to distinguish between biophase equilibration processes, and signal transduction processes. For the determination of the kinetics of BBB transport characteristics, the actual unbound concentrations at either side of the BBB are the driving forces for BBB transport (HammarlundUdenaes et al., 1997). However, other factors apart from only BBB transport processes may be important determinants for the actual ECF concentrations. These factors may include EC metabolism, extra-IC distribution and exchange of the drug between ECF and CSF (Cserr, 1984; Williams et al., 1995). As an example, for a drug with low BBB permeability but fast accumulation into brain cells, the ECF concentrations will be lower than in case no IC accumulation takes place. Therefore, it is a pre-requisite to take total brain concentrations into account, because the extent and rate of transport into the brain will be underestimated. Moreover, intra-EC exchange may include active transport mechanisms (Lee et al., 2001), and potential change in this transport by co-administration of transport inhibitors, intended to modify BBB transport, could as well modify extra-IC exchange. Pharmacokinetic equations for the exchange of amounts between different compartments at either side of the BBB or blood–CSF-barrier (BCSFB; de Lange et al., 2005) can be described by the change in the amount per unit of time for a certain compartment, which equals the change in the concentration times the volume of that compartment (C V). This will be the result of transport into and transport out of that compartment. Assuming that the compartments and their transport connections can be considered as depicted in Fig. 1, the derivatives in Table 1 will describe the change in the amount within the compartments.

V. BBB transport studies using microdialysis IV.A.4. Compartmental modelling of BBB transport An important aspect of determining BBB transport is that will help to reveal the mechanisms that

Over the years many studies on BBB transport using microdialysis have been reviewed (de Lange et al., 1997a, b, 1999, 2000a, 2002, 2005; Groothuis

554

Fig. 1. Compartmental view of the exchange of a compound between brain extracellular ﬂuid (ECF), as the compartment for microdialysis sampling, and other body parts. C, concentration; V, volume; CSF, cerebrospinal ﬂuid; parenchyma, brain parenchymal cells; tissue, peripheral tissues; u, unbound; b, bound. In this scheme, it is assumed that: binding of the compound in brain ECF is negligible; binding of the compound in CSF is negligible; exchange of the compound occurs between compartments as indicated by the lines; the unbound fraction of the compound may be eliminated and the arterial and venous blood concentrations are equal.

and Levy, 1997; Sawchuk and Yang, 1999; Hammarlund-Udenaes, 2000; Sawchuk and Elmquist, 2000; Deguchi and Morimoto, 2001; Deguchi, 2002). Below presented are the more recent studies on the use of microdialysis in BBB transport studies. These have been performed in rats, unless stated otherwise. V.A. Passive BBB transport Xie et al. (1998) investigated the distribution of codeine across the BBB following i.v. infusion of three dosage schemes of codeine. Microdialysis

was used to determine codeine unbound concentrations in blood and brain ECF, with nalorphine used as a calibrator for measurement of in vivo recovery. Total brain tissue and plasma concentrations were also determined. It was found that codeine was rapidly transported into the brain ECF with identical CLin and CLout values for transport across the BBB. Total brain tissue concentrations were 3.671.2-fold higher than the brain ECF concentrations. A rapid distributional equilibrium with equal unbound concentrations in blood and brain, and the absence of any dosedependency of codeine BBB transport, indicate that codeine has a purely passive and fast transport across the BBB. Walker et al. (2000) compared serum, CSF (by direct sampling) and brain ECF pharmacokinetics of lamotrigine following systemic administration of lamotrigine. The serum pharmacokinetics were biphasic (initial distribution phase half-life 3 h and a prolonged elimination phase of 30 h). The serum pharmacokinetics were linear over the range 10 – 40 mg/kg. The calculated penetration halftime into CSF was 0.4270.15 h. At equilibrium, the CSF to total serum concentration ratio (0.6170.02) was greater than the ratio of free to total serum concentration (0.3970.01). Microdialysis information was obtained from the frontal cortex and hippocampus. Using in vivo recovery corrected microdialysis data, the calculated penetration half-time of lamotrigine into brain ECF was 0.5170.11 h, similar to that for CSF, and was not AUC or dose dependent. At equilibrium, the brain ECF to total serum concentration ratio (0.4070.04) was similar to the free to total serum concentration (0.3970.01). No differences were found between the frontal cortex and the hippocampus. It was concluded that lamotrigine has a relatively slow penetration into both CSF and brain ECF compartments compared with antiepileptic drugs used in acute seizures. Furthermore, the free-serum drug concentration is not the sole contributor to the CSF compartment, and the CSF concentration is an overestimate of the brain ECF concentration of lamotrigine. Tong and Patsalos (2001) studied the BBB transport of the novel antiepileptic drug levetiracetam by the combination of serial blood sampling

555 Table 1. The exchange of the compound between the brain extracellular ﬂuid compartment and other compartments of the body, with assumptions as presented in Fig. 1 Compartment

Change in amount/time

CSF

dCCSF/dt VCSF ¼ CLplasma to CSF Cplasma,u – CLCSF CECF – CLCSF to ECF CCSF – CLmetabolism CCSF

Parenchyma,b

dCparenchyma,b/dt Vparenchyma,b ¼ CLparenchyma,u to parenchyma,b Cparenchyma,u – CLparenchyma,b to parenchyma,u Cparenchyma,b

Parenchyma,u

dCparenchyma,u/dt Vparenchyma,u ¼ CLECF to Cparenchyma,u – CLmetabolism Cparenchyma,u

ECF

dCECF/dt VECF ¼ CLplasma,u to ECF Cplasma – CLECF to plasma,u CECF+CCSF to ECF CCSF – CLECF to CSF CECF+CLparenchyma,u to ECF Cparenchyma,u – CLECF to parenchyma,u Cparenchyma,u – CLmetabolism CECF

Plasma,b

dCplasma,b/dt Vplasma,b ¼ CLplasma,u

Plasma,u

dCplasma,u /dt Vplasma,u ¼ +CLECF to plasma,u CECF – CLplasma,u to ECF Cplasma,u+CLCSF CCSF – CLplasma,u to CSF Cplasma,u+CLtissue,u to plasma,u Ctissue,u – CLplasma,u to tissue,u Cplasma,u – CLelimination Cplasma,u

CNS

parenchyma,u

to plasma,u

CCSF+CLECF

CECF – CLparenchyma,u

to CSF

to ECF

Periphery to plasma,b

Vplasma,u – CLplasma,b

to plasma,u

Vplasma,b

to plasma,u

Tissue,b

dCtissue,b/dt Vtissue,b ¼ +CLtissue,u

Tissue,u

dCtissue,u/dt Vtissue,u ¼ +CLplasma,u to tissue,u Cplasma,u – CLtissue,u to plasma,u Ctissue,u+CLtissue,b to tissue,u Ctissue,b – CLtissue,u to tissue,b Ctissue,u – CLmetabolism Ctissue,u

to tissue,b

Ctissue,u – CLtissue,b

to tissue,u

Ctissue,b

Note: The exchange is expressed in terms of change in amount per unit of time (d/dt), using the primary pharmacokinetic parameters, clearance (CL) and volume of distribution (V), with C, concentration; V, volume; ECF, brain extracellular ﬂuid; CSF, cerebrospinal ﬂuid; parenchyma, brain parenchymal cells; tissue, peripheral tissues; u, unbound; b, bound. The equations can easily be converted from the macroconstants (CL, V) to the parameters of k (ﬁrst-order rate constants) and A (amount), by using C V ¼ k A.

and concomitant microdialysis brain ECF sampling in frontal cortex and hippocampus was used. Levetiracetam was administered intravenously at two dosages. No serum protein binding was found and it was revealed that brain distribution was region-independent, while the half-lives in brain were 1.5-fold larger than in serum, and no equilibrium with serum levetiracetam levels was obtained within the experimental period. This is indicative for passive and slow BBB transport equilibration. Schaddelee et al. (2004) studied the clearance for transport from blood to the brain by simultaneous analysis of the blood and brain ECF concentrations using a compartmental pharmacokinetic model for two synthetic A1 receptor agonists 8-methylamino-N6-cyclopentyladenosine (MCPA) and 20 -deoxyribose-N6-cyclo-pentyladenosine (20 dCPA), following i.v. infusions. Total brain tissue

and plasma concentrations were also determined. The in vivo microdialysis recoveries were determined by the DNNF. No time-dependency was found. The values of the intercompartmental clearance for the distribution from blood to brain were very low, 1.970.4 mL/min for MCPA and 1.670.3 mL/min for 20 -dCPA, in accordance with in vitro tests. Furthermore, a much slower elimination from the brain compartment was observed relative to that in plasma. These data indicates a passive and highly restricted BBB transport of those adenosine A1 analogs. V.B. Active BBB transport Actually the number of studies that indicate purely passive BBB transport are scarce relatively to those dealing with active efﬂux or inﬂux BBB transport mechanisms. Microdialysis research on

556

active BBB transport have included the antiretroviral drugs, the antiepileptic drugs and the opioids. V.B.1. Antiretroviral drugs The group of antiretroviral drugs for the treatment of human immunodeﬁciency virus (HIV) received a lot of attention, as due to the BBB, the brain is considered a protected ‘‘sanctuary site’’, and the role of the brain remains particularly important when it comes to both active and latent HIV infection and related syndromes. Therefore, information of antiretroviral drug transport into the brain is of importance. Stahle and Borg (2000) investigated the transport of alovudine into the brain and CSF, using microdialysis. Following i.v. administration, the AUCCSF/AUCblood ratio of alovudine was higher than the corresponding AUCbrain ECF/AUCblood ratio, without any effects of thymidine neither acetazolamide on the AUCbrain ECF/AUCblood ratio. It was concluded that alovudine reaches brain ECF, not via the CSF, but via the BBB. The existing concentration gradient of alovudine over the BBB as well as BCSFB might be due to the presence of active efﬂux transport mechanisms. For alovudine also active efﬂux mechanisms were indicated in another study by Linden et al. (2003) in which microdialysis was used to sample brain ECF concentrations in order to study the inﬂuence of the well-known transport inhibitors probenecid and quinidine on the transport of alovudine between the blood and the brain ECF or whole brain tissue. No differences were found for the AUCbrain ECF/AUCserum of 0.2 following administration of alovudine alone, or combined with probenecid. Perfusion through the microdialysis probe with probenecid also had no effect on the AUCbrain ECF/AUCserum for alovudine. However, the AUCbrain ECF/AUCserum for alovudine was signiﬁcantly reduced to 0.09 by co-administration of quinidine, without apparent changes in whole brain tissue concentration. Finally, the microdialysis recovery of alovudine increased with increasing concentrations of alovudine in the perfusion ﬂuid. The explanations to these ﬁndings are pending further research. For zidovudine Fox et al. (2002) studied in non-human primates the concentrations in blood,

muscle and brain ECF, by microdialysis, in serum ultraﬁltrate, and in CSF samples during a continuous i.v. infusion and after bolus dosing. The goal was to determine whether CSF drug penetration is a valid surrogate for BBB penetration. Recovery was estimated by in vivo by zero net ﬂux for the continuous infusion, and by retrodialysis for the bolus dosing. In vivo recovery was tissuedependent and was lower in brain than in blood or muscle. At steady state, the brain ECF penetration and CSF penetration were 0.1370.06 and 0.1770.02, respectively. These data indicated that CSF and brain ECF zidovudine concentrations were comparable at steady state, while also the corresponding AUCs were comparable after bolus injection. It was concluded that in non-human primates the zidovudine penetration in brain ECF and CSF is limited to a similar extent as in other species, presumably by active transport. V.B.2. Antiepileptic drugs In epilepsy, pharmacoresistance is a major problem, and occurs in 25% of the epilepsy patients. One of the reasons for pharmacoresistance may be restricted BBB transport and/or unfavourable brain distribution. It is, therefore, important to learn about BBB transport mechanisms of antiepileptic drugs. Luer et al. (1999) studied whether the fraction of gabapentin crossing the BBB is linear over a broad range of doses, using the microdialysis technique for measuring gabapentin concentrations in the brain ECF, combined with plasma sampling. Although higher AUCbrain ECF values were obtained with higher AUCplasma values, changes in AUCbrain ECF were less than proportional to observed changes in AUCplasma. It seemed that BBB transport of gabapentin was saturable. In the report of Feng et al. (2001), the BBB inﬂux and efﬂux of pregabalin were investigated with microdialysis. BBB inﬂux (CLin) and efﬂux (CLout) permeability for pregabalin were 4.8 and 37.2 mL/ min/g brain, respectively, following i.v. infusion. The results indicate that pregabalin is able to enter the brain. A signiﬁcant delay in anticonvulsant action of pregabalin was found relatively to the estimated ECF drug concentrations. Using a PK/ PD link model, the counter-clockwise delay in the

557

relationship between pregabalin brain ECF concentration and the anticonvulsant effect showed that the concentration in the hypothetical effect compartment (Ce) versus effect (PD) proﬁle exhibits a sigmoidal curve and the calculated ECe50 and Keo values were 95.3 ng/mL and 0.0092/min, respectively. The small value for the Keo indicates that the effect is not directly proportional to the amount of pregabalin in the ECF compartment possibly due to inherent delay at other levels than BBB transport. The effects of brain ECF–parenchymal exchange has been clearly demonstrated by the study on valproate by Scism et al. (2000), using rabbits. It was shown that the unfavourable brain-toplasma gradient was the result of the coupled efﬂux transport processes at both the parenchymal cells and the BBB. BBB transport and brain distribution of valproic acid were investigated in absence and presence of probenecid, using microdialysis and total tissue sampling during steadystate i.v. infusion of valproic acid. In control conditions, the intracellular brain concentration (ICC) was 2.8 times higher than the corresponding ECF concentrations. Co-infusion of probenecid elevated the ratio of ICC over ECF concentrations to 4.2. This indicated the presence of a probenecid-sensitive efﬂux transporter at the brain parenchymal cells. The EFC to unbound plasma concentration ratio was 0.3, and was not signiﬁcantly inﬂuenced by probenecid. Herewith the presence of distinctly different organic anion transporters for the efﬂux of valproic acid at the parenchymal cells and capillary endothelium in the brain was indicated. As increased expression of multidrug transporters (e.g., ABCB1/Pgp) has been reported in epileptogenic brain tissue from pharmacoresistant patients undergoing epilepsy surgery, the inﬂuence of Pgp inhibition on BBB transport of phenobarbital, lamotrigine and felbamate was investigated by in vivo microdialysis by Potschka et al. (2002). Local perfusion of verapamil via the microdialysis probe increased the concentration of the three antiepileptic drugs in the cortical brain ECF. This indicates that overexpression of Pgp in epileptic tissue might limit brain access of phenobarbital, lamotrigine and felbamate.

In another study, Potschka et al. (2003) investigated further whether ABCC2 (/MRP2) is functionally involved in transport of carbamazepine, lamotrigine and felbamate across the BBB. The distribution of these drugs into the brain was determined using ABCC2-deﬁcient TR-rats. The microdialysis results gave no evidence that ABCC2 function modulates entry of carbamazepine, lamotrigine or felbamate into the CNS. However, ABCC2 deﬁciency was associated with an increased anticonvulsant response of carbamazepine in the amygdala-kindling model of epilepsy. For levitiracetam, as a new antiepileptic drug, the expectations were quite high as it seemed to be an effective and well-tolerated drug in many patients with otherwise pharmacoresitant epilepsy. Thus, Potschka et al. (2004) investigated whether the concentration of levitiracetam in the cortical brain ECF could be modulated by inhibition of Pgp or MRPs, using the Pgp inhibitor verapamil and the MRP1/2 inhibitor probenecid. Local perfusion with verapamil or probenecid via the microdialysis probe did not increase the brain ECF concentration of levitiracetam. This indicates that brain uptake of levitiracetam is not affected by Pgp or MRP1/2. This could explain its antiepileptic efﬁcacy in patients whose seizures are poorly controlled by other antiepileptic drugs. The relation between brain ECF concencentrations following systemic administration of oxcarbazepine and its effects on local ECF levels of DA and serotonin was investigated by Clinckers et al. (2005), including modulation of oxcarbazepine BBB transport. The intrahippocampal perfusion of verapamil, a ABCB1/Pgp inhibitor, and probenecid, a MRP inhibitor, on the BBB passage of oxcarbazepine were investigated. Simultaneously, the effects on hippocampal monoamines were studied as pharmacodynamic markers for the anticonvulsant activity of oxcarbazepine, in the focal pilocarpine model for limbic seizures. Systemic oxcarbazepine administration alone did not prevent the rats from developing seizures. The co-administration of verapamil or probenecid offered complete protection, while in parallel, signiﬁcant increases in brain ECF hippocampal DA and serotonin levels were observed. These data indicated that oxcarbazepine is a substrate for

558

multidrug transporters at the BBB, and that co-administration of multidrug transporter inhibitors signiﬁcantly potentiates the anticonvulsant activity of oxcarbazepine, indicating the impact of BBB transport for the CNS effects of this antiepileptic drug. V.B.3. Opioids and metabolites Relatively many studies have been performed on opioids and their metabolites, which remain to be intruiging in their PK/PD relationships. As there were data available to indicate that the anaesthetic effects of morphine could modulated by Pgp inhibition, Letrent et al. (1999) studied the brain distribution and antinociceptive effects of morphine with microdialysis. Upon co-administration of the ABCB1/Pgp inhibitor GF120918, the half-life of morphine in the brain increased about threefold, while the antinociceptive effect by morphine was increased. These effects appeared to be associated with modulation of the BBB transport of morphine. The contribution of BBB transport of morphine to the delay in antinociceptive effect was assessed in a quantitative manner by Bouw et al. (2000). The PK of unbound morphine was determined by microdialysis in venous blood and striatal brain ECF, while serial arterial sampling was used for the PK in arterial blood. Retrodialysis by drug was used for in vivo calibration of the MD probes. Following a short infusion of either 10 or 40 mg/ kg, the BBB equilibration of morphine, expressed as AUCbrain ECF/AUCblood, was less than unity (0.28 7 0.09 and 0.22 7 0.17 for 10 and 40 mg/kg, respectively), indicating active efﬂux of morphine across the BBB. The concentration–effect relationship showed a clear hysteresis. The effect delay half-life of 32 min based on arterial concentrations, while being only 5 min based on striatal ECF concentrations. This means that the BBB transport of morphine has a signiﬁcant impact on the effect delay towards the antinociceptive effect. Xie et al. (2000) studied the BBB transport characteristics of the morphine metabolite morphine-3-glucuronide (M3G) in the absence and presence of probenecid, using a cross-over experimental design. Normal BBB transport of M3G was 8%, and increased about twofold upon

co-administration of probenecid. NONMEM modelling of BBB transport indicated that this effect was due to changes in CLin rather than changes in CLout for morphine. Morphine-6-glucuronide (M6G) is more potent than morphine and Stain-Texier et al. (1999) addressed the mechanisms afterwards, by combining transcortical microdialysis with classical tissue sampling. Thus, the distribution of morphine and M6G in the brain cortex and total brain was determined, and thereby the distribution into brain ECF. It was found that in the brain, morphine IC levels were approximately four times higher than ECF levels, while for M6G the ECF levels were 125-fold higher than IC levels. Thus, it was shown that M6G is almost exclusively distributed into brain ECF, which is the compartment of the opioid receptors. In conclusion, M6G has a favourable distribution into the EC biophase of the m-receptor, in comparison with morphine. The quantitative PK/PD relationship of M6G was investigated by Bouw et al. (2001a), for its antinociceptive action. On 2 consecutive days, an exponential infusion of M6G was administered for 4 h aiming at a target concentration of 3,000 ng/ mL in blood. The concentrations of unbound M6G were determined in brain ECF and venous blood using microdialysis and in arterial blood by regular sampling. The calibration of the microdialysis probes for the in vivo recovery of M6G was by using retrodialysis by drug prior to drug administration. The BBB transport equilibration of M6G was determined using unbound brain ECF and blood concentrations. The half-life of M6G was 2375 min in arterial blood, 26710 min in venous blood and 58717 min in brain ECF. The BBB equilibration, expressed as the unbound steady-state concentration ratio, was 0.2270.09, indicating active efﬂux in the BBB transport of M6G. A two-compartment model best described the brain distribution of M6G. The unbound volume of distribution was 0.2070.02 mL/g brain. The concentration–antinociceptive effect relationships exhibited a clear hysteresis, resulting in an effect delay half-life of 103 min in relation to blood concentrations and a remaining effect delay halflife of 53 min in relation to brain ECF concentrations. It could, therefore, be concluded that the

559

very long half-life of M6G antinociceptive effects was in part (50%) due to slow transport across the BBB. The remaining part should be the result of drug distribution within the brain tissue or ratelimiting mechanisms at the receptor level. The possible inﬂuence of probenecid on morphine transport across the BBB was studied by Tunblad et al. (2003). The microdialysis probes were placed into the striatum and jugular vein and calibrated using retrodialysis by drug. Morphine was administered as a 4-h exponential infusion on 2 consecutive days, the ﬁrst day without and the second day with the addition of probenecid, administered as a bolus dose (20 mg/kg) followed by a constant infusion (20 mg/kg/h). The steady-state ratio of unbound morphine concentration in brain to that in blood was 0.2970.07. This indicates that morphine is actively efﬂuxed at the BBB. Probenecid co-administration increased the ratio to 0.3970.04 (po0.05). Models in which probenecid inﬂuenced the brain efﬂux clearance rather than the inﬂux clearance, well described the data. The half-life in brain increased from 5879 min to 115725 min when probenecid was co-administered. Systemic clearance of morphine also decreased upon probenecid co-administration. This study indicated that morphine is a substrate for the probenecid-sensitive transporters at the BBB. Also for M6G, the impact of probenecid on BBB transport was investigated (Tunblad et al., 2005). Microdialysis probes were placed in the striatum and into the jugular vein. Probe calibration was performed in vivo by retrodialysis by drug. M6G was administered as a 4-h exponential i.v. infusion, on 2 consecutive days, respectively without and with the addition of probenecid. An integrated model including the total arterial concentrations, the dialysate concentrations in brain and blood, and the in vivo recovery measurements, was developed. The extent of BBB transport, expressed as the ratio between CLin/CLout was estimated as 0.29 for both days. This indicated the presence of efﬂux transport of M6G at the level of the BBB. However, the probenecid-sensitive transporters are not involved in the brain efﬂux. The half-life of M6G was longer in the brain than in blood on both experimental days (po0.05). The systemic elimination of M6G decreased by 22%

(po0.05) upon probenecid co-administration, however, the ratio CLin/CLout was unaltered upon probenecid co-administration. It was concluded that probenecid decreased the systemic elimination of M6G, but had no effect on the BBB transport of M6G. V.B.4. Transporter knockout mice Apart from studies performed in rats using ratio’s like AUCbrain ECF/AUCplasma, unbound or CLin and CLout values, either without or with the use of transport inhibitors to indication transporter mediated transport, the use of transporter knockout animals could be used. The example of the ABCC2 deﬁcient rats has been mentioned previously in the section of the antiepileptic drugs. Using ABCB1(–/–) and wild-type mice, de Lange et al. (1998) studied the BBB transport and Pgp functionality using total brain and brain cortical microdialysis concentration proﬁles of the model Pgp substrate R123. Maintenance of BBB integrity was indicated by equal total brain/blood ratios of Flu and FD-4 in both mice types. The Pgp substrate rhodamine-123 was infused and total brain, blood and brain microdialysate concentrations in ABCB1(–/–) mice and wild-type mice were compared. The no-net-ﬂux method was used to estimate in vivo recovery in the ABCB1(–/–) and wild-type mice, which revealed a fourfold lower in vivo recovery values in the ABCB1(–/–) mice, due to the absence of active efﬂux of R123 by the BBB. R123 concentrations in brain and in brain ECF after i.v. infusion were about fourfold higher in ABCB1(–/–) than in wild-type mice without changes in blood levels. It was concluded that Pgp plays an important role in R123 distribution into the brain, while it was stressed that changes in in vivo recovery by changes in active efﬂux from brain ECF in using intracerebral microdialysis should be considered carefully. In these mice, Xie et al. (1999) studied the effect of Pgp functionality at the BBB for morphine. ABCB1(–/–) and wild-type mice received a constant infusion of morphine for 1, 2, or 4 h, at a rate of 9 nmol/min/mouse. Total brain and blood samples were obtained at these time-points. Microdialysis was used to estimate morphine unbound concentrations in brain ECF during the 4 h

560

infusion. Two methods of estimating in vivo recovery were used, retrodialysis with nalorphine as a calibrator, and the DNNF method, and comparable results were obtained (including a lower in vivo recovery of morphine for the ABCB1(–/–) mice). The brain ECF concentration ratios of (–/–)/wild-type were 2.7 for the retrodialysis and 3.6 for the DNNF method at 4 h, with corresponding total brain concentration ratios of ABCB1(–/–)/wild-type of 2.3 and 2.6, respectively. The total concentration ratios of brain/plasma were 1.1 and 0.5 for ABCB1(–/–) and wild-type mice, respectively. By comparison of the ABCB1 (–/–) and wild-type mice, it was concluded that ABCB1/Pgp participates in regulating the amount of morphine transport across the BBB. Also, these mice were used to study the impact of ABCB1/Pgp on BBB transport of ﬂuoroquinolones (de Lange et al., 2000b). A series of ﬂuoroquinolones were investigated on their interaction with the Pgp transporter. In vitro, inhibition of Pgp mediated transport of R123 was assessed in monolayers of cells transfected with Pgp, and for those ﬂuoroquinolones that appeared to have an effect, also the direct transport modiﬁcation by Pgp was investigated. Thus, for peﬂoxacin, norﬂoxacin, ciproﬂoxacin, ﬂeroﬂoxacin and sparﬂoxacin an interaction was observed. Using ABCB1(–/–) and wild-type mice, the BBB transport characteristics of these ﬂuoroquinolones was determined on the basis of total brain over plasma concentrations. For sparﬂoxacin clear effect of Pgp functionality on BBB transport was found. BBB transport characteristics were investigated in detail by microdialysis, using the DNNF method for quantiﬁcation, indicating a ﬁvefold increase in brain ECF distribution in the absence of Pgp efﬂux. Again, differences in in vivo probe recovery of sparﬂoxacin were found, being lower in ABCB1(–/–) relative to that in wild-type mice. Inano et al. (2003) studied the impact of the carnitine transporter (SLC22A5) functionality by comparison of wild-type mice and jvs mice with a defective SLC22A5 gene. The no-net-ﬂux method revealed a higher in vivo recovery of acetyl-L-carnitine (ALCAR) and lower physiological ALCAR concentration in the ECF of the thalamus in jvs versus wild-type mice. Administration of ALCAR

resulted in a signiﬁcant lower initial uptake of ALCAR across the BBB in the jvs mouse. These results indicated that the SLC22A5 transporter is functionally involved in ALCAR transport across BBB. V.B.5. Miscellaneous compounds Using a multiple isotope method, Beagles et al. (1998) studied the in vivo synthesis rates, EC concentrations, and intercompartmental distributions of quinolinic acid in normal and immune-activated brain. Quinolic acid is able to kill neurons by activation of the NMDA receptors, facing the EC space in the brain. Microdialysis was employed to quantify the brain ECF quinolinic acid levels. Different carbon isotopes forms of quinolinic acid (13C7-quinolonic acid and 2HC7-quinolonic acid) were used to be administrated via either the subcutaneous route, the probe perfusion, while endogenous quinolonic acid with the natural isotope was determined by mass spectrometry. 13C7quinolonic acid was perfused through the probe for in vivo calibration to accurately quantify brain ECF quinolonic acid concentrations. In normal brain, 85% of brain ECF quinolonic acid levels (110 nM) originated from blood, whereas 59% of tissue homogenate quinolonic acid (130 pmol/g) originated from local de novo synthesis. During systemic immune activation (intraperitoneal injection of endotoxin), blood quinolonic acid levels increased (10-fold) and caused a rise in homogenate (11-fold) and brain ECF (19-fold) quinolonic acid levels with an increase in the proportions of quinolonic acid derived from blood. During CNS inﬂammation (local infusion of endotoxin), increases in brain homogenate (250-fold) and brain ECF (65-fold) quinolonic acid levels occurred because of an increase in local synthesis rate (146-fold) and a reduction in efﬂux/inﬂux ratio (by 53%). These results demonstrate the use of microdialysis in combination with other techniques, to reveal the dynamic regulation of endogenous compounds within the brain and by transport across the BBB. Taylor et al. (2000) found that AIT-082, a cognitive enhancer, is transported into the brain by a non-saturable inﬂux mechanism and out of the brain by a saturable efﬂux mechanism. The

561

authors combined in their study the use of brain perfusion with capillary depletion and microdialysis. The brain inﬂux of radiolabelled AIT-082 was measured, alone, and in the presence of 110–600fold excess of cold AIT-082. No changes were observed in the inﬂux, indicating a non-saturable inﬂux mechanism of BBB transport. The 600-fold cold excess of AIT-082 decreased the efﬂux of radiolabelled AIT-082, as shown by microdialysis investigations, with an increase in elimination half-life from 20 to 35 min. Using a quantitative microdialysis technique the efﬂux of the 6-mercaptourine across the BBB was studied by Deguchi et al. (2000). The brain tissue, CSF and hippocampal ECF concentrations of 6-mercaptourine were very low relatively to the unbound plasma concentration. This suggests that 6-mercaptourine distribution into the brain is highly restricted. Kinetic analysis showed that efﬂux clearance from brain ECF to plasma is 20 times higher that the inﬂux clearance from plasma to brain. N-ethylmaleimide, probenecid, p-amminohippuric acid, benzoate and salicylate signiﬁcantly reduced the efﬂux clearance, while choline, nor tetraetylammonium had any effect. These data suggested that the restricted brain distribution of 6-mercaptourine may be ascribed to efﬁcient efﬂux from the brain, by an organic anion transport system, as well as by the monocarboxylic acid transport system. Fluorescein is a model compound often used to indicate changes in BBB permeability. However, its transport across the BBB is not solely determined by passive diffusion. Sun et al. (2001b) studied the BBB transport kinetics of ﬂuorescein by microdialysis in brain ECF and CSF, following a constant rate infusion of 6 mg/kg/h with and without i.v. co-administration of probenecid (loading dose 100 mg/kg and maintenance infusion rate of 30 mg/kg) in a cross-over design. Probenecid decreased the probe recovery of ﬂuorescein in frontal cortex, from 0.2170.017 to 0.1770.020. The distribution of ﬂuorescein across the BBB and BCSFB was enhanced by 2.2- and 1.9-fold, respectively, when probenecid was co-administered, corrected for increased ﬂuorescein plasma concentrations and plasma-free fraction. These results demonstrate that ABCC’s (MRPs) or ABCC-like

transport system(s) may play an important role in ﬂuorescein distribution across both BBB and BCSFB. Tsai et al. (2001) studied the brain distribution of camptothecin, and the effects of the Pgp transport modulators, cyclosporin A, berberine, quercetin, naringin and naringenin. In the presence of cyclosporin A, the AUCbrain ECF of camptothecin was signiﬁcantly elevated. No changes in camptothecin AUCbrain ECF was observed in the presence of berberine, quercetin, naringin and naringenin. Shimizu et al. (2001) studied the BBB transport of paraquat that is structurally similar to Nmethyl-4-phenyl- pyridium (MPP+) and could, therefore, induce dopaminergic toxicity, like MPP+ does. It was found that paraquat appeared in the brain striatum dialysate, in a dose-dependent manner, while MPP+ was found to be unable to penetrate the BBB, either in control or paraquat pre-treated rats. Penetration of paraquat is, therefore, not the result of changes in BBB that might be induced by paraquat or a paraquat radical. Pre-treatment of L-valine, high substrate afﬁnity for the neutral amino-acid transporter, markedly reduced the BBB penetration of paraquat. These and other data indicate that paraquat is transported into the brain by the neutral amino acid transporter. The BBB is of importance for the homeostatis of the brain tissue environment. Also ionic exchange is subjected to regulation via BBB transport, as described for aluminium by Yokel (2002). Microdialysis probes were implanted in the jugular vein as well as the left and right frontal cortex. At steady-state the brain ECF to blood ratio was smaller than 1, indicating a potential active efﬂux mechanism. No changes were observed in this ratio by probe perfusion with the metabolic inhibitor 2,4-dinitrophenol (10 mM). The addition of valproic to the perfusate (10 and 100 mM) was ineffective. For pyruvic acid the addition (100 mM) signiﬁcantly increased the aluminium citrate brainto-blood ratio from 0.19 to 0.31. Pyruvic acid (1 M in the dialysate) increased the aluminium citrate brain-to-blood ratio to a value not different from unity, suggesting competition between aluminium citrate and pyruvic acid for transport.

562

Based on these data combined with other knowledge, it was proposed that the monocarboxylic acid transporter at the BBB maintains a steadystate aluminium citrate brain-to-blood ratio much less than 1. Mano et al. (2002) studied the kinetics of YM992, a novel antidepressant, in plasma, CSF, brain ECF and in brain at the steady-state after i.v. infusion. The concentration ratio of total brain to total plasma was 70, while those of the free concentrations in brain ECF to plasma and CSF to plasma were comparable. The carotid artery injection study showed that the brain uptake index of YM992 was 140%. Furthermore, the uptake clearance into brain after i.v. dosing was 0.6 mL/ min/g brain, indicating a high permeability at the BBB. The microdialysis data were used to proof the very high distribution volume of YM992 in brain (375 mL/g brain), suggesting high binding in the brain cells. These ﬁndings suggested that the high partition of YM992 to the brain is attributed to its high level of binding in the brain as well as its high permeability at the BBB. The impact of BBB transport on the fact that the plasma and brain ECF pharmacokinetics of norﬂoxacin are linearly related, while dramatic hysteresis between brain ECF and EEG proﬁles have been observed, was the scope of the study of Chenel et al. (2004). Simultaneous CNS distribution and PK/PD modelling of the electroencephalogram (EEG) effect of norﬂoxacin administered at a i.v. convulsant bolus dose (150 mg/kg). Blood samples were collected for total norﬂoxacin plasma concentration measurements, the corresponding unbound levels were determined in brain ECF using microdialysis, while quantitative EEG recording was conducted during 9 h post-dose. Brain ECF norﬂoxacin concentrations were much lower than plasma levels (AUC ratio of 10%) but peaked very early, and PK proﬁles were parallel in both biological ﬂuids. The best PK model was obtained by considering that brain ECF concentrations were part of the central compartment, with a proportionality factor. The Tmax of the EEG effect was delayed and the EEG effect versus plasma concentration curves exhibited a dramatic hysteresis. A PK/PD compartment model with a spline function to describe

the relationship between the EEG effect and concentration at the effect site successfully described the data. Comparisons of PK/PD parameters estimated from plasma and brain ECF concentrations showed that most of the delayed norﬂoxacin EEG effect was not due to BBB transport. Bourasset et al. (2005) focussed on the BBB uptake and brain intra- and brain ICF–ECF partitioning of S18986 [(S)-2,3-dihydro-[3,4]cyclopentano-1,2,4-benzothiadiazine-1,1-dioxide], a new positive allosteric modulator of the AMPA receptor. BBB transport of S18986 was measured using the in situ brain perfusion technique, while brain ECF concentrations were determined by microdialysis in the two effector areas, that is, frontal cortex and dorsal hippocampus. Blood samples were collected simultaneously. CSF and brain tissue concentrations were determined using a conventional PK approach. PK modelling indicated that the brain uptake clearance of S18986 was high, 20 mL/s/g. Terminal half-lives were similar in plasma and brain, at 1 h. The ratios of AUCbrain ECF/AUCplasma,u were 0.25 in the frontal cortex and dorsal hippocampus, whereas ratios of AUCbrain ICF/AUCplasma were 1 and 1.5, respectively. The authors concluded that despite the ratio of AUCbrain ECF/AUCplasma,u below 1, there is nevertheless a relatively high BBB uptake of S18986, to be explained by the non-homogenous brain ICF–ECF partitioning of S18986, favouring the brain ICF. This is a an illustration of the importance of taking brain ICF–ECF partitioning into account to determine BBB transport characteristics. Gupta et al. (2006) studied the brain distribution of levocetirizine and dextrocetirizine, and compared three different tissue-to-plasma partition coefﬁcients, K(p), K(p,u) and K(p,uu). The objective of this study was to compare the BBB transport and brain distribution of levocetirizine and dextrocetirizine. Microdialysis probes, calibrated using retrodialysis by drug, were placed into the frontal cortex and right jugular vein of eight guinea pigs. Racemic cetirizine (2.7 mg/kg) was administered as a 60-min i.v. infusion. Unbound and total concentrations of the enantiomers were measured in blood and brain. The brain distribution of the cetirizine enantiomers were compared

563

using the parameters K(p), K(p,u), K(p,uu) and V(u,br). K(p) compared the total brain concentration to total plasma concentration, K(p,u) compensated for binding to plasma components, whereas K(p,uu) also compensated for binding within the brain tissue and directly quantiﬁes the transport across the BBB. V(u,br) described the binding within the brain. The stereoselective brain distribution indicated by the K(p) of 0.22 and 0.04 for dextrocetirizine and levocetirizine, respectively, was caused by different binding to plasma proteins. The transport of the cetirizine enantiomers across the BBB was not stereoselective, since the K(p,uu) was 0.17 and 0.14 (not signiﬁcant) for dextrocetirizine and levocetirizine, respectively. The K(p,uu) values show that the enantiomers are efﬂuxed to a large extent across the BBB. The V(u,br) of 2.5 mL/g brain was also similar for both the enantiomers, and the value indicates high binding to brain tissue. Thus, when determining stereoselectivity in brain distribution, it is important to study all factors governing this distribution, binding in blood and brain, and the BBB equilibrium. V.C. Disease conditions and BBB transport Many conditions may affect BBB functionality and thereby BBB transport characteristics of compounds. Using microdialysis, Diserbo et al. (2002) studied the effects of total body irradiation of 4.5 Gy on BBB permeability to 3H-alpha-aminoisobutyric acid and 14C-sucrose in brain striatum. At 1, 3 and 6 weeks and at 3, 5 and 8 month, no changes were observed in BBB transport of these compounds. However, initially, between 3 and 17 h after irradiation, a signiﬁcant but transient increase in the BBB permeability to the two markers was found, while a secondary transient increased BBB permeability for only 14C-sucrose was observed at 28 h post-irradiation, with a decrease in BBB permeability to 3H-alpha-aminoisobutyric acid between 33 and 47 h post-irradiation. It was concluded that early modiﬁcations of BBB permeability may result following a moderate-dose whole-body irradiation. In diabetes, the plasma glucose levels are poorly controlled. As brain function may be impaired by

prolonged elevations of blood glucose the question was if the BBB would adapt to glucose transport. Jacob et al. (2002) used microdialysis to measure brain ECF levels of glucose, lactate and betahydroxybutyrate in the inferior colliculus. Results in chronically hyperglycemic BB/wor diabetic rats were compared with those obtained in control (Sprague–Dawley) rats, during euglycemia and acute hyperglycemia. The brain ECF/plasma glucose ratio of 0.3 was remarkably similar for all three groups, with proportionally higher brain ECF glucose levels in the hyperglycemic groups. It was also found that the brain ECF levels of lactate and beta-hydroxybutyrate were increased in diabetic rats as compared with controls. These results suggested that the BBB does not have a protective adaptation of the transfer of glucose occurs in chronic hyperglycemia. It was concluded that brain tissue, during poorly controlled diabetes, may be chronically exposed to markedly elevated levels of glucose and other metabolic fuels, and on the long-term may suffer from adverse effects of hyperglycemia as well. Swelling of cerebral glial cells, is a typical complication in patients with acute liver failure, which often results in high intracranial pressure and brain herniation before or during liver transplantation. Tofteng et al. (2002) were interested in metabolic alterations that cause the high intracranial pressure in patients with acute liver failure. Using microdialysis in a young man with severe acute liver failure and cerebral oedema, these authors found that the brain ECF content of lactate gradually increased during the surgery. These brain ECF lactate levels correlated in this patient with arterial lactate concentrations during and after grafting, but not with arterial glucose concentrations. Also, brain ECF glutamate and glycerol levels were severely elevated before liver transplantation, but tended to decrease in the hours after grafting. Apart from disturbances in glutamate neurotransmission, arachidonic acid metabolism, these ﬁndings indicate distortions in lactate ﬂux across the BBB in patients with acute liver failure. Ederoth et al. (2004) hypothesised in this clinical study that the active efﬂux of morphine from brain to blood also occurs in the human brain, and that

564

brain injury would alter cerebral morphine pharmacokinetics. These authors investigated BBB transport characteristics of morphine in patients with traumatic brain injury. Three microdialysis probes were inserted; two in the brain and one in abdominal subcutaneous fat for metabolic monitoring. The cerebral probe locations were termed as ‘‘better brain tissue’’ and ‘‘worse brain tissue’’, referring to the degree of traumatic injury. Morphine (10 mg) was infused intravenously over a 10min period in seven patients in the intensive care setting. Tissue and plasma morphine concentrations were obtained during the subsequent 3-h period with microdialysis and regular blood sampling. It was found that the ratio of AUCbrain ECF/AUCplasma,u was 0.65 in the ‘‘better brain tissue’’, 0.78 in the ‘‘worse brain tissue’’, and 1.00 in subcutaneous fat. These data indicated that in ‘‘worse brain tissue’’ in trauma patients an increased BBB permeability for morphine might be found. Furthermore, the terminal half-life and Tmax values were longer in the brain when compared with plasma and fat, respectively, while within the brain the Tmax value tended to be shorter in the ‘‘worse brain tissue’’. Also, it was found that the relative in vivo recovery for morphine in ‘‘better brain tissue’’ was higher than in the ‘‘worse brain tissue’’. As the AUCbrain ECF/ AUCplasma,u was below unity in the ‘‘better brain tissue’’, the presence of an active efﬂux of morphine at the level of the BBB was concluded. BBB transport of morphine was also studied in diseased brain following experimentally induced meningitis in piglets (Tunblad et al., 2004). In the generally anaesthetised piglets, one occipital and two frontal microdialysis brain probes and one pressure transducer were inserted into the brain tissue. Another probe was placed into the jugularis interna. Morphine 1 mg/kg was administered as a 10-min infusion, and morphine concentrations were subsequently measured for 3 h. Then, meningitis was induced by lipopolysaccharide injection in cisterna magna. When meningitis was established, the morphine experiment was repeated. The AUCbrain ECF/AUCblood,u of morphine was 0.47 during the control period, and 0.95 during meningitis. The increase in the AUCbrain ECF/ AUCblood,u ratio during meningitis implied a

decreased active efﬂux and an increased passive diffusion of morphine over the BBB. The half-life of morphine in brain was longer than in blood during both periods, and was unaffected by meningitis. These data showed that the morphine distribution into the brain is signiﬁcantly increased during meningitis as compared with the control situation. Cefoselis is a widely used beta-lactam antibiotic, but occasionally induces seizures and convulsions in elder and renal failure patients. However, betalactams are known not to pass the BBB. Ohtaki et al. (2004) examined the BBB penetration of cefoselis in normal and renal failure rats using microdialysis. Cefoselis was found to appear in AUCbrain ECF, in a dose-dependent fashion, linearly related to its blood level. The apparent elimination constant from brain ECF was slightly lower than that from blood. These results indicated that cefoselis is able to penetrate the BBB either by diffusion or by active transport. In renal failure rats, the elimination half-lives of cefoselis from both blood and brain were extensively prolonged. It was concluded that this might contribute to the occurrence of seizures seen in patients, but probably is not the only reason, as an extremely high dose was required to induce seizures in renal failure rats. The authors indicated that additional factors, such as a decreased brain function in the elder patients, would be involved in seizures in patients who received cefoselis.

VI. Discussion and conclusions It is concluded that the use of microdialysis combined with brain tissue and blood concentrations, for example to obtained by brain tissue and serial blood sampling, provides very useful data to determine the kinetics of transport equilibration across the BBB under a variety of conditions, in the species of choice, such as mice, rats, rabbits, piglets and monkeys, as well as in intensive care patients. Such information on the rate and extent of BBB transport, and modulations thereof, will be useful to further distinguish the contribution of many possible passive and active dynamically regulated

565

transport mechanisms. The knowledge thus provided on BBB transport mechanisms and regulation is critical for the understanding of brain homeostasis, and how disturbances thereof may lead to CNS diseases. Also it will be critical in ultimate being able to predict the relation between the kinetics and the dynamics of CNS active compounds.

References Abbott, N.J. (2002) Astrocyte-endothelial interactions and blood-brain barrier permeability. J. Anat., 200(6): 629–638. Abbott, N.J. and Revest, P.A. (1991) Control of brain endothelial permeability. Cerebrovasc. Brain Met. Rev., 3: 39–72. Abbott, N.J., Ronnback, L. and Hansson, E. (2006) Astrocyteendothelial interactions at the blood-brain barrier. Nat. Rev. Neurosci., 7(1): 41–53. Agon, P., Braeckman, R., Van Haver, D., Denutte, H., Goethals, P., Donche, -H., Vermeullen, F., Deman, J. and Kaufman, J.M. (1988) Drug distribution in dog brain studied by positron emission tomography. Biopharm. Drug. Dispos., 9: 567–577. Agon, P., Goethals, P., van Haver, D.K. and Kaufman, J.M. (1991) Permeability of BBB for atenolol studies by PET. J. Pharm. Pharmacol., 43: 597–600. Aird, R.B. (1984) A study of intrathecal, cerebrospinal ﬂuid-tobrain exchange. Exp. Neurol., 86: 342–358. Allen, D.D., Crooks, P.A. and Yokel, R.A. (1992) 4-Trimethylammonium antipyrine: a quaternary ammonium nonradionuclide marker for blood-brain barrier integrity during in vivo microdialysis. J. Pharmacol. Toxicol. Methods, 28(3): 129–135. Arthur, F.E., Shivers, R.R. and Bowman, P.D. (1987) Astrocyte-mediated induction of tight junctions in brain capillary endothelium, an efﬁcient in vitro model. Brain Res., 433(1): 155–159. Banks, W.A. (1999) Physiology and pathology of the bloodbrain barrier, implications for microbial pathogenesis, drug delivery and neurodegenerative disorders. J. Neurovirol., 5: 538–555. Bartels, M., Gunther, U., Albert, K., Mann, K., Schuff, N. and Stuckstedte, H. (1991) 19F nuclear magnetic resonance spectroscopy of neuroleptics, the ﬁrst in vivo pharmacokinetics of triﬂuoperazine in the rat brain and the ﬁrst in vivo spectrum of ﬂuphenazine in the human brain. Biol. Psychiatry., 30: 656–662. Beagles, K.E., Morrison, P.F. and Heyes, M.P. (1998) Quinolinic acid in vivo synthesis rates, extracellular concentrations, and intercompartmental distributions in normal and immune-activated brain as determined by multiple-isotope microdialysis. J. Neurochem., 70(1): 281–291.

Beck, D.W., Vinters, H.V., Hart, M.N. and Cancilla, P.A. (1984) Glial cells inﬂuence polarity of the blood-brain barrier. J. Neuropathol. Exp. Neurol., 43(3): 219–224. Begley, D.J. (2004) ABC transporters and the blood-brain barrier. Curr. Pharm. Des., 10(12): 1295–1312. Belinsky, M.G., Chen, Z-.S., Shchaveleva, I., Zeng, H. and Kruh, G.D. (2002) Characterization of the drug resistance and transport properties of multidrug resistance protein 6 (MRP6, ABCC6). Cancer Res., 62: 6172–6177. Bendayan, R., Lee, G. and Bendayan, M. (2002) Functional expression and localization of P-glycoprotein at the blood brain barrier. Microsc. Res. Tech., 57(5): 365–380. Benveniste, H. and Diemer, N.H. (1987) Cellular reactions to implantation of a microdialysis tube in the rat hippocampus. Acta Neuropath. (Berl.), 74: 234–238. Benveniste, H., Drejer, J., Schoesboe, A. and Diemer, N.H. (1984) Elevation of the extracellular concentrations of glutamate and aspartate in rat hippocampus during transient cerebral ischemia monitored by intracerebral microdialysis. J. Neurochem., 43: 1369–1374. Benveniste, H., Drejer, J., Schoesboe, A. and Diemer, N.H. (1987) Regional cerebral glucose phosphorylation and blood ﬂow after insertion of a microdialyis ﬁber through the dorsal hippocampus in the rat. J. Neurochem., 49: 729–734. Bickel, U. (2005) How to measure drug transport across the blood-brain barrier. NeuroRx, 2(1): 15–26. Blasberg, R.G., Kobayashi, T., Patlak, C.S., Shinohara, M., Miyoaka, M., Rice, J.M. and Shapiro, W.R. (1981) Regional blood ﬂow, capillary permeability, and glucose utilization in two brain tumor models, preliminary observations and pharmacokinetic implications. Cancer Treat. Rep., 65(Suppl 2): 3–12. Bodor, N. and Brewster, M.E. (1983) Problems of drug delivery of drugs to the brain. Pharma. Ther., 19: 337–386. Bolwig, T.G., Hertz, M.M., Paulson, O.B., Spotoft, H. and Rafaelsen, O.J. (1977) The permeability of the blood-brain barrier during electrically induced seizures in man. Eur. J. Clin. Invest., 7: 87–93. Bonate, P.L. (1995) Animal models for studying transport across the blood-brain barrier. J. Neurosci. Methods, 56(1): 1–15. Borst, P., Evers, R., Kool, M. and Wijnholds, J. (2000) A family of drug transporters, the multidrug resistance-associated proteins. J. Natl. Cancer Inst., 2(16): 1295–1302. Borst, P. and Oude Elferink, R. (2002) Mammalian ABC transporters in health and disease. Annu. Rev. Biochem., 71: 537–592. Bouman, H.J. and Van Wimersma Greidanus, T.B. (1979) A rapid and simple cannulation technique for repeated sampling of cerebrospinal ﬂuid in freely moving rats. Brain Res. Bull., 4(4): 575–577. Bourasset, F., Bernard, K., Munoz, C., Genissel, P. and Scherrmann, J.M. (2005) Neuropharmacokinetics of a new alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) modulator, S18986 [(S)-2,3-dihydro-[3,4]cyclopentano-1,2,4-benzothiadiazine-1,1-dioxide], in the rat. Drug Metab. Dispos., 33(8): 1137–1143.

566 Bouw, R., Ederoth, P., Lundberg, J., Ungerstedt, U., Nordstrom, C.H. and Hammarlund-Udenaes, M. (2001b) Increased blood-brain barrier permeability of morphine in a patient with severe brain lesions as determined by microdialysis. Acta Anaesthesiol. Scand., 45: 390–392. Bouw, M.R., Gardmark, M. and Hammarlund-Udenaes., M. (2000) Pharmacokinetic-pharmacodynamic modelling of morphine transport across the blood-brain barrier as a cause of the antinociceptive effect delay in rats – a microdialysis study. Pharm. Res., 17(10): 1220–1227. Bouw, M.R., Xie, R., Tunblad, K. and Hammarlund-Udenaes, M. (2001a) Blood-brain barrier transport and brain distribution of morphine-6-glucuronide in relation to the antinociceptive effect in rats – pharmacokinetic/pharmacodynamic modelling. Br. J. Pharmacol., 134(8): 1796–1804. Bradbury, M.W. (1984) The structure and function of the blood-brain barrier. Fed. Proc., 43(2): 186–190. Brightman, M.W. (1977) Morphology of blood-brain interfaces. Exp. Eye Res., 25(Suppl.): 1–25. Brightman, M.W., Zis, K. and Anders, J. (1983) Morphology of cerebral endothelium and astrocytes as determinants of the neuronal microenvironment. Acta Neuropathol. Suppl. (Berl.), 8: 21–33. Brosman, C.F. and Claudio, L. (1998) Brain microvasculature in multiple sclerosis. In: Pardridge, W.M. (Ed.), Introduction to the Blood-Brain Barrier; Methodology, Biology and Pathology. Cambridge University Press, Cambridge, UK, pp. 386–400. Bungay, P.M., Morrison, P.F. and Dedrick, R.L. (1990) Steady-state theory for quantitative microdialysis of solutes and water in vivo and in vitro. Life Sci., 46: 105–119. Cancilla, P.A. and DeBault, L.E. (1983) Neutral amino acid transport properties of cerebral endothelial cells in vitro. J. Neuropathol. Exp. Neurol., 42(2): 191–199. Chen, Z-.S., Hopper Borge, E., Belinsky, M.G., Shchaveleva, I., Kotova, E. and Kruh, G.D. (2003) Characterization of the transport properties of human multidrug resistance protein 7 (MRP7, ABCC10). Mol. Pharmacol., 63: 351–358. Chenel, M., Marchand, S., Dupuis, A., Lamarche, I., Paquereau, J., Pariat, C. and Couet, W. (2004) Simultaneous central nervous system distribution and pharmacokinetic-pharmacodynamic modelling of the electroencephalogram effect of norﬂoxacin administered at a convulsant dose in rats. Br. J. Pharmacol., 142(2): 323–330. Clinckers, R., Smolders, I., Meurs, A., Ebinger, G. and Michotte, Y. (2005) Quantitative in vivo microdialysis study on the inﬂuence of multidrug transporters on the blood-brain barrier passage of oxcarbazepine, concomitant use of hippocampal monoamines as pharmacodynamic markers for the anticonvulsant activity. J. Pharmacol. Exp. Ther., 314(2): 725–731. Coleman, D.L., King, R.N. and Anrade, J.D. (1974) The foreign body reaction, a chronic inﬂammatory response. J. Biomed. Mater. Res., 8: 199–211. Collins, J.M. and Dedrick, L.D. (1983) Distributed model for drug delivery to CSF and brain tissue. J. Am. Physiol., 14: R303–R310.

Cooray, H.C., Blackmore, C.G., Maskell, L. and Barrand, M.A. (2002) Localisation of breast cancer resistance protein in microvessel endothelium of human brain. Neuroreport, 13(16): 2059–2063. Cordon-Cardo, B., O’Brien, J.P., Casals, D., Rittman-Grauer, L., Biedler, J.L., Melamed, M.R. and Bertino, J.R. (1989) Multidrug-resistance gene (P-glycoprotein) is expressed by endothelial cells at blood-brain barrier sites. Proc. Natl. Acad. Sci. U.S.A., 86: 695–689. Cornford, E.M. (1985) The blood-brain barrier: a dynamic regulatory interface. Mol. Physiol., 7: 219–260. Cornford, E.M. (1999) Epilepsy and the blood-brain barrier, endothelial cell responses to seizures. Adv. Neurol., 79: 845–862. Cornford, E.M. and Cornford, M.E. (2002) New systems for delivery of drugs to the brain in neurological disease. Lancet Neurol., 1(5): 306–315. Cserr, H.F. (1984) Convection of brain interstitial ﬂuid. In: Shapiro, K., Marmarou, A. and Portnoy, H. (Eds.), Hydrocephalus. Raven Press, New York, pp. 59–68. Dash, A.K. and Elmquist, W.F. (2003) Separation methods that are capable of revealing blood-brain barrier permeability. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 797(1–2): 241–254. Davson, H. and Welch, K. (1971) The permeation of several materials into the ﬂuids of the rabbit’s brain. J. Physiol., 218(2): 337–351. de Boer, A.G., van der Sandt, I. and Gaillard, P.J. (2003) The role of drug transporters at the blood-brain barrier. Ann. Rev. Pharmacol. Toxicol., 43: 629–656. de Graaf, R.A., Pan, J.W., Telang, F., Lee, J.H., Brown, P., Novotny, E.J., Hetherington, H.P. and Rothman, D.L. (2001) Differentiation of glucose transport in human brain gray and white matter. J. Cereb. Blood Flow Metab., 21(5): 483–492. de Lange, E.C. and Danhof, M. (2002) Considerations in the use of cerebrospinal ﬂuid pharmacokinetics to predict brain target concentrations in the clinical setting, implications of the barriers between blood and brain. Clin. Pharmacokinet., 41(10): 691–703. de Lange, E.C., Danhof, M., de Boer, A.G. and Breimer, D.D. (1997b) Methodological considerations of intracerebral microdialysis in pharmacokinetic studies on drug transport across the blood-brain barrier. Brain Res. Rev., 25(1): 27–49. de Lange, E.C., de Bock, G., Schinkel, A.H., de Boer, A.G. and Breimer, D.D. (1998) BBB transport and P-glycoprotein functionality using MDR1A (–/–) and wild-type mice. Total brain versus microdialysis concentration proﬁles of rhodamine-123. Pharm. Res., 15: 1657–1665. de Lange, E.C., de Boer, B.A. and Breimer, D.D. (1999) Microdialysis for pharmacokinetic analysis of drug transport to the brain. Adv. Drug Deliv. Rev., 36(2–3): 211–227. de Lange, E.C., de Boer, A.G. and Breimer, D.D. (2000a) Methodological issues in microdialysis sampling for pharmacokinetic studies. Adv. Drug Deliv. Rev., 45(2–3): 125–148.

567 de Lange, E.C., Marchand, S., van den Berg, D., van der Sandt, I.C., de Boer, A.G., Delon, A., Bouquet, S. and Couet, W. (2000b) In vitro and in vivo investigations on ﬂuoroquinolones; effects of the P-glycoprotein efﬂux transporter on brain distribution of sparﬂoxacin. Eur. J. Pharm. Sci., 12(2): 85–93. de Lange, E.C., Ravenstijn, P.G., Groenendaal, D. and van Steeg, T.J. (2005) Toward the prediction of CNS drug-effect proﬁles in physiological and pathological conditions using microdialysis and mechanism-based pharmacokinetic-pharmacodynamic modeling. AAPS J., 7(3): E532–E543. de Lange, E.C.M., Danhof, M., De Boer, A.G. and Breimer, D.D. (1994) Critical factors of intracerebral microdialysis as a technique to determine the pharmacokinetics of drugs in rat brain. Brain Res., 666: 1–8. de Lange, E.C.M., de Boer, A.G. and Breimer, D.D. (1997a) Monitoring in vivo blood-brain barrier drug transport, cerebrospinal ﬂuid sampling, the unit impulse response method and, with special reference, intracerebral microdialysis. STP Pharma. Sci., 7: 17–27. de Lange, E.C.M., De Vries, J.D., Zurcher, C., Danhof, M., De Boer, A.G. and Breimer, D.D. (1995a) The use of intracerebral microdialysis for the determination of pharmacokinetic proﬁles of anticancer drugs in tumor-bearing rat brain. Pharm. Res., 12: 1924–1931. de Lange, E.C.M., Hesselink, M.B., Danhof, M., De Boer, A.G. and Breimer, D.D. (1995b) The use of intracerebral microdialysis to determine changes in blood-brain barrier transport. Pharm. Res., 12(12): 1924–1931. de Lange, E.C.M., Zurcher, C., Danhof, M., De Boer, A.G. and Breimer, D.D. (1995c) Repeated microdialysis perfusions, periprobe tissue reactions and BBB permeability. Brain Res., 702: 261–265. de Vries, H.E., Kuiper, J., de Boer, A.G., van Berkel, T.J.C. and Breimer, D.D. (1997) The role of the blood-brain barrier in neuro-inﬂammatory diseases. Pharmacol. Rev., 49: 143–156. Dean, M., Rzhetsky, A. and Allikmets, R. (2001) The human ATP-binding cassette (ABC) transporter superfamily. Genome Res., 11(7): 1156–1166. Deguchi, Y. (2002) Application of in vivo brain microdialysis to the study of blood-brain barrier transport of drugs. Drug Metab. Pharmacokinet., 17(5): 395–407. Deguchi, Y. and Morimoto, K. (2001) Application of an in vivo brain microdialysis technique to studies of drug transport across the blood-brain barrier. Curr. Drug Metab., 2(4): 411–423. Deguchi, Y., Yokoyama, Y., Sakamoto, T., Hayashi, H., Naito, T., Yamada, S. and Kimura, R. (2000) Brain distribution of 6-mercaptopurine is regulated by the efﬂux transport system in the blood-brain barrier. Life Sci., 66(7): 649–662. Diserbo, M., Agin, A., Lamproglou, I., Mauris, J., Staali, F., Multon, E. and Amourette, C. (2002) Blood-brain barrier permeability after gamma whole-body irradiation, an in vivo microdialysis study. Can. J. Physiol. Pharmacol., 80(7): 670–678.

Dombrowski, S.M., Desai, S.Y., Marroni, M., Cucullo, L., Goodrich, K., Bingaman, W., Mayberg, M.R., Bengez, L. and Janigro, D. (2001) Overexpression of multiple drug resistance genes in endothelial cells from patients with refractory epilepsy. Epilepsia, 42(12): 1501–1506. Ederoth, P., Tunblad, K., Bouw, R., Lundberg, C.J., Ungerstedt, U., Nordstrom, C.H. and Hammarlund-Udenaes, M. (2004) Blood-brain barrier transport of morphine in patients with severe brain trauma. Br. J. Clin. Pharmacol., 57(4): 427–435. Eisenblatter, T., Huwel, S. and Galla, H.J. (2003) Characterisation of the brain multidrug resistance protein (BMDP/ ABCG2/BCRP) expressed at the blood-brain barrier. Brain Res., 971(2): 221–231. Elsinga, P.H., Hendrikse, N.H., Bart, J., Vaalburg, W. and van Waarde, A. (2004) PET studies on P-glycoprotein function in the blood-brain barrier: how it affects uptake and binding of drugs within the CNS. Curr. Pharma. Des., 10(13): 1493–1503. Feng, M.R., Turluck, D., Burleigh, J., Lister, R., Fan, C., Middlebrook, A., Taylor, C. and Su, T. (2001) Brain microdialysis and PK/PD correlation of pregabalin in rats. Eur. J. Drug Metab. Pharmacokinet., 26(1–2): 123–128. Fenstermacher, J.D., Patlak, C.S. and Blasberg, R.G. (1974) Transport of material between brain extracellular ﬂuid, brain cells and blood. Fed. Proc., 33: 2070–2074. Fenstermacher, J.D., Wei, L., Acuff, V., Lin, S.Z., Chen, J.L., Bereczki, D., Otsuka, T., Nakata, H., Tajima, A., Hans, F.J., Ghersi-Egea, J.F., Finnegan, W., Richardson, G., Haspel, H. and Patlak, C. (1995) The dependency of inﬂux across the blood-brain barrier on blood ﬂow and the apparent ﬂowindependence of glucose inﬂux during stress. In: Greenwood, J., Begley, D.J. and Segal, M. (Eds.), New Concepts of a Blood-Brain Barrier. Plenum Press, New York and London, pp. 89–101. Foster, K.A. and Roberts, M.S. (2000) Experimental methods for studying drug uptake in the head and brain. Curr. Drug Metab., 1(4): 333–356. Fox, E., Bungay, P.M., Bacher, J., McCully, C.L., Dedrick, R.L. and Balis, F.M. (2002) Zidovudine concentration in brain extracellular ﬂuid measured by microdialysis, steadystate and transient results in rhesus monkey. J. Pharmacol. Exp. Ther., 301(3): 1003–1011. Gao, B., Hagenbuch, B., Kullak-Ublick, G.A., Benke, D., Aguzzi, A. and Meier, P.J. (2000) Organic anion-transporting polypeptides mediate transport of opioid peptides across blood-brain barrier. J. Pharmacol. Exp. Ther., 294(1): 73–79. Gibbs, J.P., Adeyeye, M.C., Yang, Z. and Shen, D.D. (2004) Valproic acid uptake by bovine brain microvessel endothelial cells, role of active efﬂux transport. Epilepsy Res., 58(1): 53–66. Gloor, S.M., Wachtel, M., Bolliger, M.F., Ishihara, H., Landmann, R. and Frei, K. (2001) Molecular and cellular permeability control at the blood-brain barrier. Brain Res. Brain Res. Rev., 36(2–3): 258–264. Greig, N.H., Momma, S., Sweeney, D.J., Smith, Q.R. and Rapoport, S.I. (1987) Facilitated transport of melphalan at the rat blood-brain barrier by the large neutral amino acid carrier system. Cancer Res., 47: 1571–1576.

568 Groothuis, D.R. and Levy, R.M. (1997) The entry of antiviral and antiretroviral drugs into the central nervous system. J. Neurovirol., 3(6): 387–400. Groothuis, D.R., Molnar, P. and Blasberg, R.G. (1984) Regional blood ﬂow and blood-to-tissue transport in ﬁve brain tumor models. Prog. Exp. Tumor Res., 27: 132–153. Groothuis, D.R., Ward, S., Schlageter, K.E., Itskovich, A.C., Schwerin, S.C., Allen, C.V., Dills, C. and Levy, R.M. (1998) Changes in blood-brain barrier permeability associated with insertion of brain cannulas and microdialysis probes. Brain Res., 803(1–2): 218–230. Gross, P.M., Sposito, N.M., Pettersen, S.E. and Fenstermacher, J.D. (1986) Differences in function and structure of the capillary endothelium in gray matter, white matter, and a circumventricular organ of rat brain. Blood Vessels, 23: 261–270. Gupta, A., Chatelain, P., Massingham, R., Jonsson, E.N. and Hammarlund-Udenaes, M. (2006) Brain distribution of cetirizine enantiomers, comparison of three different tissueto-plasma partition coefﬁcients, K(p), K(p,u), and K(p,uu). Drug Metab. Dispos., 34(2): 318–323. Hammarlund-Udenaes, M. (2000) The use of microdialysis in CNS drug delivery studies. Pharmacokinetic perspectives and results with analgesics and antiepileptics. Adv. Drug Deliv. Rev., 45(2–3): 283–294. Hammarlund-Udenaes, M., Paalzow, L.N. and De Lange, E.C.M. (1997) Drug equilibration across the blood-brainbarrier – pharmacokinetic considerations based on the microdialysis method. Pharm. Res., 14: 128–134. Hardebo, J.E., Nilsson, B. and Oldendorf, W.H. (1970) Measurement of brain uptake of radiolabelled substances using a tritiated water internal standard. Brain Res., 24: 372–376. Hasegawa, H., Ushio, Y., Hayakawa, T., Yamada, K. and Mogami, H. (1983) Changes of the blood-brain barrier in experimental metastatic brain tumors. J. Neurosurg., 59: 304–310. Hipfner, D.R., Deeley, R.G. and Cole, S.P. (1999) Structural, mechanistic and clinical aspects of MRP1. Biochim. Biophys. Acta, 1461(2): 359–376. Higgins, C.F. (2001) ABC transporters: physiology, structure and mechanisms – an overview. Res. Microbiol, 152(3–4): 205–210. Hopper, E., Belinsky, M.G., Zeng, H., Tosolini, A., Testa, J.R. and Kruh, G.D. (2001) Analysis of the structure and expression pattern of MRP7 (ABCC10): a new member of the MRP subfamily. Cancer Lett., 162: 181–191. Huai-Yun, H., Secrest, D.T., Mark, K.S., Carney, D., Brandquist, C., Elmquist, W.F. and Miller, D.W. (1998) Expression of multidrug resistance-associated protein (MRP) in brain microvessel endothelial cells. Biochem. Biophys. Res. Commun., 243(3): 816–820. Huber, J.D., Egleton, R.D. and Davis, T.P. (2001) Molecular physiology and pathophysiology of tight junctions in the blood-brain barrier. Trends Neurosci., 24: 719–725. Imperato, A. and Di Chiara, G. (1984) Trans-striatal dialysis coupled to reverse phase high performance liquid chromatography with electrochemical detection: a new method for the study of in vivo release of endogenous dopamine and metabolites. J. Neurosci., 4: 966–977.

Imperato, A. and Di Chiara, G. (1985) Dopamine release and metabolism in awake rats after systemic neuroleptics as studied by trans-striatal dialysis. J. Neurosci., 5: 297–306. Inano, A., Sai, Y., Nikaido, H., Hasimoto, N., Asano, M., Tsuji, A. and Tamai, I. (2003) Acetyl-L-carnitine permeability across the blood-brain barrier and involvement of carnitine transporter OCTN2. Biopharm. Drug Dispos., 24(8): 357–365. Jacob, R.J., Fan, X., Evans, M.L., Dziura, J. and Sherwin, R.S. (2002) Brain glucose levels are elevated in chronically hyperglycemic diabetic rats, no evidence for protective adaptation by the blood brain barrier. Metabolism, 51(12): 1522–1524. Jedlitschky, G., Leier, I., Buchholz, U., Barnouin, K., Kurz, G. and Keppler, D. (1996) Transport of glutathione, and sulfur conjugates by the MRP gene-encoded conjugate export pump. Cancer Res., 56: 988–994. Kakee, A., Terasaki, T. and Sugiyama, Y. (1996) Brain efﬂux index as a novel method of analyzing efﬂux transport at the blood-brain barrier. J. Pharmacol. Exp. Ther., 277(3): 1550–1559. Karssen, A.M., Meijer, O.C., van der Sandt, I.C., De Boer, A.G., De Lange, E.C. and De Kloet, E.R. (2002) The role of the efﬂux transporter P-glycoprotein in brain penetration of prednisolone. J. Endocrinol., 175(1): 251–260. Karssen, A.M., Meijer, O.C., van der Sandt, I.C., Lucassen, P.J., de Lange, E.C., de Boer, A.G. and de Kloet, E.R. (2001) Multidrug resistance P-glycoprotein hampers the access of cortisol but not of corticosterone to mouse and human brain. Endocrinology, 142(6): 2686–2694. Kikuchi, R., Kusuhara, H., Sugiyama, D. and Sugiyama, Y. (2004) Contribution of organic anion transporter 3 (Slc22a8) to the elimination of p-aminohippuric acid and benzylpenicillin across the blood-brain barrier. J. Pharmacol. Exp. Ther., 311(3): 1147–1153. Kim, K.S., Wass, C.A. and Cross, A.S. (1997) Blood-brain barrier permeability during the development of experimental bacterial meningitis in the rat. Exp. Neurol., 145: 253–257. Kim, R.B. (2002) Drugs as P-glycoprotein substrates, inhibitors, and inducers. Drug Metab. Rev., 34(1–2): 47–54. Kim, R.B., Fromm, M.F., Wandel, C., Leake, B., Wood, A.J.J., Roden, D.M. and Wilkinson, G.R. (1998) The drug transporter P-glycoprotein limits oral absorption and brain entry of HIV1 protease inhibitors. J. Clin. Invest., 101: 289–294. Klopman, G., Leming, M.S. and Avner, R. (1997) Quantitative structure-activity relationship of multidrug resistance reversal agents. Mol. Pharmacol., 52: 323–334. Kusuhara, H., He, Z., Nagata, Y., Nozaki, Y., Ito, T., Masuda, H., Meier, P.J., Abe, T. and Sugiyama, Y. (2003) Expression and functional involvement of organic anion transporting polypeptide subtype 3 (Slc21a7) in rat choroid plexus. Pharm. Res., 20(5): 720–727. Kusuhara, H. and Sugiyama, Y. (2001a) Efﬂux transport systems for drugs at the blood-brain barrier and bloodcerebrospinal ﬂuid barrier (Part 1). Drug Discov. Today, 6(3): 150–156.

569 Kusuhara, H. and Sugiyama, Y. (2001b) Efﬂux transport systems for drugs at the blood-brain barrier and blood-cerebrospinal ﬂuid barrier (Part 2). Drug Discov. Today, 6(4): 206–212. Kusuhara, H. and Sugiyama, Y. (2005) Active efﬂux across the blood-brain barrier, role of the solute carrier family. NeuroRx, 2(1): 73–85. Lam, F.C., Liu, R., Lu, P., Shapiro, A.B., Renoir, J.M., Sharom, F.J. and Reiner, P.B. (2001) Beta-amyloid efﬂux mediated by p-glycoprotein. J. Neurochem., 76(4): 1121–1128. Larsson, C.I. (1991) The use of an ‘‘internal standard’’ for control of the recovery in microdialysis. Life Sci., 49: 73–78. Lee, G., Dallas, S., Hong, M. and Bendayan, R. (2001) Drug transporters in the central nervous system, brain barriers and brain parenchyma considerations. Pharmacol. Rev., 53: 569–596. Lerma, J., Herranz, A.S., Herreras, O., Abraira, V. and Del Rio, R. (1986) In vivo determination of extracellular concentrations of amino acids in the rat hippocampus. A method based on brain dialysis and computerized analysis. Brain Res., 384: 145–155. Leslie, E.M., Deeley, R.G. and Cole, S.P. (2001) Toxicological relevance of the multidrug resistance protein 1, MRP1 (ABCC1) and related transporters. Toxicology, 167(1): 3–23. Letrent, S.P., Pollack, G.M., Brouwer, K.R. and Brouwer, K.L. (1999) Effects of a potent and speciﬁc P-glycoprotein inhibitor on the blood-brain barrier distribution and antinociceptive effect of morphine in the rat. Drug Metab. Dispos., 27(7): 827–834. Levin, V.A. (1980) Relationship of octanol/water partition coefﬁcient and molecular weight to rat brain capillary permeability. J. Med. Chem., 23: 682–684. L’Heureux, R., Dennis, T., Curet, O. and Scatton, B. (1986) Measurement of endogenous noradrenaline release in the rat cerebral cortex in vivo by transcortical dialysis: effects of drugs affecting noradrenergic transmission. J. Neurochem., 46: 1794–1801. Linden, K., Stahle, L., Ljungdahl-Stahle, E. and Borg, N. (2003) Effect of probenecid and quinidine on the transport of alovudine (30 -ﬂuorothymidine) to the rat brain studied by microdialysis. Pharmacol. Toxicol., 93(5): 226–232. Lo, E.H., Singhal, A.B., Torchilin, V.P. and Abbott, N.J. (2001) Drug delivery to damaged brain. Brain Res. Brain Res. Rev., 38: 140–148. Lonnroth, P., Jansson, P.A. and Smith, U. (1987) A microdialysis method allowing characterization of intercellular water space in humans. Am. J. Physiol., 253(Endocrinol.Metab. 16): E228–E231. Loscher, W. and Potschka, H. (2005) Blood-brain barrier active efﬂux transporters, ATP-binding cassette gene family. NeuroRx, 2(1): 86–98. Lossinsky, A.S. and Shivers, R.R. (2004) Structural pathways for macromolecular and cellular transport across the bloodbrain barrier during inﬂammatory conditions. Histol. Histopathol., 19(2): 535–564. Luer, M.S., Hamani, C., Dujovny, M., Gidal, B., Cwik, M., Deyo, K. and Fischer, J.H. (1999) Saturable transport of gabapentin at the blood-brain barrier. Neurol. Res., 21(6): 559–562.

Mahar Doan, K.M., Humphreys, J.E., Webster, L.O., Wring, S.A., Shampine, L.J., Serabjit-Singh, C.J., Adkison, K.K. and Polli, J.W. (2002) Passive permeability and P-glycoprotein-mediated efﬂux differentiate central nervous system (CNS) and non-CNS marketed drugs. J. Pharmacol. Exp. Ther., 303(3): 1029–1037. Major, O., Shdanova, T., Duffek, L. and Nagy, Z. (1990) Continuous monitoring of blood-brain barrier opening to Cr51-EDTA by microdialysis following probe injury. Acta Neurochir. Suppl. (Wien), 51: 46–48. Mano, Y., Higuchi, S. and Kamimura, H. (2002) Investigation of the high partition of YM992, a novel antidepressant, in rat brain – in vitro and in vivo evidence for the high binding in brain and the high permeability at the BBB. Biopharm. Drug Dispos., 23(9): 351–360. Mans, A.M., DeJoseph, M.R., Davis, D.W. and Hawkins, R.A. (1987) Regional amino acid transport into brain during diabetes: effect of plasma amino acids. Am. J. Physiol., 253(5 Pt 1): E575–E583. Matlaga, B.F., Yasenchak, L.P. and Salthouse, T.H. (1976) Tissue response to implanted polymers: the signiﬁcance of sample shape. J. Biomed. Mater. Res., 10: 391–397. Mayer, U., Wagenaar, E., Dorobek, B., Beijnen, J.H., Borst, P. and Schinkel, A.H. (1997) Full blockade of intestinal Pglycoprotein and extensive inhibition of blood-brain barrier P-glycoprotein by oral treatment of mice with PSC833. J. Clin. Invest., 100: 2430–2436. Mayhan, W.G. (2001) Regulation of blood-brain barrier permeability. Microcirculation, 8: 89–104. McNay, E.C. and Sherwin, R.S. (2004) From artiﬁcial cerebrospinal ﬂuid (aCSF) to artiﬁcial extracellular ﬂuid (aECF): microdialysis perfusate composition effects on in vivo brain ECF glucose measurements. J. Neurosci. Methods, 132(1): 35–43. Miller, D.S., Graeff, C., Droulle, L., Fricker, S. and Fricker, G. (2002) Xenobiotic efﬂux pumps in isolated ﬁsh brain capillaries. Am. J. Physiol. Regul. Integr. Comp. Physiol., 282(1): R191–R198. Miller, D.S., Nobmann, S.N., Gutmann, H., Toeroek, M., Drewe, J. and Fricker, G. (2000) Xenobiotic transport across isolated brain microvessels studied by confocal microscopy. Mol. Pharmacol., 58(6): 1357–1367. Minakawa, T., Bready, J., Berliner, J., Fisher, M. and Cancilla, P.A. (1991) In vitro interaction of astrocytes and pericytes with capillary-like structures of brain microvessel endothelium. Lab. Invest., 65(1): 32–40. Moghaddam, B. and Bunney, B.S. (1989) Ionic composition of microdialysis perfusing solution alters the pharmacological responsiveness and basal outﬂow of striatal dopamine. J. Neurochem., 53: 652–654. Mooradian, A.D. (1988) Effect of aging on the blood-brain barrier. Neurobiol. Aging, 9(1): 31–39. Morgan, M.E., Singhal, D. and Anderson, B.D. (1996) Quantitative assessment of blood-brain barrier damage during microdialysis. J. Pharmacol. Exp. Ther., 277(2): 1167–1176.

570 Mori, S., Takanaga, H., Ohtsuki, S., Deguchi, T., Kang, Y.S., Hosoya, K. and Terasaki, T. (2003) Rat organic anion transporter 3 (rOAT3) is responsible for brain-to-blood efﬂux of homovanillic acid at the abluminal membrane of brain capillary endothelial cells. J. Cereb. Blood Flow Metab., 23: 432–440. Morrison, P.F., Bungay, P.M., Hsiao, J.K., Ball, B.A., Mefford, I.N. and Dedrick, R.L. (1991) Quantitative microdialysis, analysis of transients and application to pharmacokinetics in brain. J. Neurochem., 57: 103–119. Namba, H., Lucignani, G., Nehlig, A., Patlak, C., Pettigrew, K., Kennedy, C. and Sokoloff, L. (1987) Effects of insulin on hexose transport across blood-brain barrier in normoglycemia. Am. J. Physiol., 252(3 Pt 1): E299–E303. Nitsch, C. and Klatzo, I. (1983) Regional patterns of bloodbrain barrier breakdown during epileptiform seizures induced by various convulsive agents. J. Neurol. Sci., 59: 305–322. Nottet, H.S.L.M. (1999) Interactions between macrophages and brain microvascular endothelial cells, role in pathogenesis of HIV-1 infection and blood-brain barrier function. J. Neurovirol., 5: 659–669. Ohtaki, K., Matsubara, K., Fujimaru, S., Shimizu, K., Awaya, T., Suno, M., Chiba, K., Hayase, N. and Shiono, H. (2004) Cefoselis, a beta-lactam antibiotic, easily penetrates the blood-brain barrier and causes seizure independently by glutamate release. J. Neural Transm., 111(12): 1523–1535. Oldendorf, W.H. (1974) Lipid solubility and drug penetration of the blood-brain barrier. Proc. Exp. Biol. Med., 14: 813–816. Oldendorf, W.H. (1979) Estimation of cerebral extraction of circulating compounds by the brain uptake index method, inﬂuence of circulation time, volume injection, and cerebral blood ﬂow. Acta Physiol. Scand., 107(2): 153–159. Olson, R.J. and Justice, J.B. (1993) Quantitative microdialysis under transient conditions. Anal. Chem., 65: 1017–1022. Ooie, T., Suzuki, H., Terasaki, T. and Sugiyama, Y. (1997) Kinetic evidence for active efﬂux transport across the bloodbrain barrier of quinolone antibiotics. J. Pharmacol. Exp. Ther., 283: 293–304. Pardridge, W.M. (1988) Recent advances in blood-brain barrier transport. Ann. Rev. Pharmacol. Toxicol., 28: 25–39. Parsons, L.H. and Justice, J.B. Jr. (1994) Quantitative approaches to in vivo brain microdialysis. Crit. Rev. Neurobiol., 8(3): 189–220. Petito, C.K., Schaefer, J.A. and Plum, F. (1977) Ultrastructural characteristics of the brain and blood-brain barrier in experimental seizures. Brain Res., 127: 251–267. Plock, N. and Kloft, C. (2005) Microdialysis-theoretical background and recent implementation in applied life-sciences. Eur. J. Pharm. Sci., 25(1): 1–24. Ponto-Boles, L.L. and Ponto, J.A. (1992a) Uses and limitations of positron emission tomography in clinical pharmacokinetics/dynamics. Part I. Clin. Pharmacokinet., 22: 211–222. Ponto-Boles, L.L. and Ponto, J.A. (1992b) Uses and limitations of positron emission tomography in clinical pharmacokinetics/dynamics. Part II. Clin. Pharmacokinet., 22: 274–283.

Potschka, H., Baltes, S. and Loscher, W. (2004) Inhibition of multidrug transporters by verapamil or probenecid does not alter blood-brain barrier penetration of levetiracetam in rats. Epilepsy Res., 58(2–3): 85–91. Potschka, H., Fedrowitz, M. and Loscher, W. (2002) P-glycoprotein-mediated efﬂux of phenobarbital, lamotrigine, and felbamate at the blood-brain barrier, evidence from microdialysis experiments in rats. Neurosci. Lett., 327(3): 173–176. Potschka, H., Fedrowitz, M. and Loscher, W. (2003) Brain access and anticonvulsant efﬁcacy of carbamazepine, lamotrigine, and felbamate in ABCC2/MRP2-deﬁcient TR-rats. Epilepsia, 44(12): 1479–1486. Pouremad, R., Bahk, K.D., Shen, Y.J., Knop, R.H. and Wyrwicz, A.M. (1999) Quantitative 19F NMR study of triﬂuorothymidine metabolism in rat brain. NMR Biomed., 12(6): 373–380. Povlishock, J.T. (1998) The pathophysiology of blood-brain barrier dysfunction due to traumatic brain injury. In: Pardridge, W.M. (Ed.), Introduction to the Blood-Brain Barrier; Methodology, Biology and Pathology. Cambridge University Press, Cambridge, UK, pp. 441–453. Qian, Y.M., Qiu, W., Gao, M., Westlake, C.J., Cole, S.P. and Deeley, R.G. (2001) Characterization of binding of leukotriene C4 by human multidrug resistance protein 1, evidence of differential interactions with NH2- and COOH-proximal halves of the protein. J. Biol. Chem., 276(42): 38636–38644. Rapoport, S.I., Ohno, K. and Pettigrew, K.D. (1979) Drug entry into the brain. Brain. Res., 172: 354–359. Rapoport, S.I. and Robinson, P.J. (1985) Tight junctional modiﬁcation as the basis of osmotic opening of the bloodbrain barrier. Ann. N.Y. Acad. Sci., 36: 251–267. Regina, A., Koman, A., Piciotti, M., El Hafny, B., Center, M.S., Bergmann, R., Couraud, P.O. and Roux, F. (1998) Mrp1 multidrug resistance-associated protein and P-glycoprotein expression in rat brain microvessel endothelial cells. J. Neurochem., 71(2): 705–715. Rubin, L.L. and Staddon, J.M. (1999) The cell biology of the blood-brain barrier. Annu. Rev. Neurosci., 22: 11–28. Ruggeri, M., Zoli, M., Grimaldi, R., Ungerstedt, U., Eliasson, A., Agnati, L.F. and Fuxe, K. (1990) Aspects of neural plasticity in the central nervous system-III, methodological studies on the microdialysis technique. Neurochem. Int., 16: 427–443. Sawchuk, R.J. and Elmquist, W.F. (2000) Microdialysis in the study of drug transporters in the CNS. Adv. Drug Deliv. Rev., 45: 295–306. Sawchuk, R.J. and Yang, Z. (1999) Investigation of distribution, transport and uptake of anti-HIV drugs to the central nervous system. Adv. Drug Deliv. Rev., 39: 5–31. Schaddelee, M.P., Groenendaal, D., DeJongh, J., Cleypool, C.G., Ijzerman, A.P., De Boer, A.G. and Danhof, M. (2004) Population pharmacokinetic modeling of blood-brain barrier transport of synthetic adenosine A1 receptor agonists. J. Pharmacol. Exp. Ther., 311(3): 1138–1146. Scheller, D. and Kolb, J. (1991) The internal reference technique in microdialysis: a practical approach to monitoring dialysis efﬁciency and to calculating tissue concentration from dialysate samples. J. Neurosci. Methods, 40: 31–38.

571 Schinkel, A.H. (2001) The roles of P-glycoprotein and MRP1 in the blood-brain and blood-cerebrospinal ﬂuid barriers. Adv. Exp. Med. Biol., 500: 365–372. Schinkel, A.H., Smit, J.J.M., Van Tellingen, O., Beijnen, J.H., Wagenaar, E., Van Deemter, L., Mol, C.A.A.M., Van der Valk, M.A., Robanus-Maandag, E.C., Te Riele, H.P.J., Berns, A.J.M. and Borst, P. (1994) Disruption of the mouse mdr1a P-glycoprotein gene leads to a deﬁciency in the blood-brain barrier and to increased sensitivity to drugs. Cell, 77: 491–502. Schinkel, A.H., Wagenaar, E., Mol, C.A.A.M., Van Deemter, L. and Borst, P. (1996) P-glycoprotein in the blood-brain barrier of mice inﬂuences the brain penetration and pharmacological activity of many drugs. J. Clin. Invest., 97: 2517–2524. Schinkel, A.H., Wagenaar, E., Van Deemter, L., Mol, C.A.A.M. and Borst, P. (1995) Absence of the mdr1a p-glycoprotein in mice affects tissue distribution and pharmacokinetics of dexamethasone, digoxin, and cyclosporin A. J. Clin. Invest., 96: 1698–1705. Scism, J.L., Powers, K.M., Artru, A.A., Lewis, L. and Shen, D.D. (2000) Probenecid-inhibitable efﬂux transport of valproic acid in the brain parenchymal cells of rabbits: a microdialysis study. Brain Res., 884(1–2): 77–86. Shimizu, K., Ohtaki, K., Matsubara, K., Aoyama, K., Uezono, T., Saito, O., Suno, M., Ogawa, K., Hayase, N., Kimura, K. and Shiono, H. (2001) Carrier-mediated processes in bloodbrain barrier penetration and neural uptake of paraquat. Brain Res., 906(1–2): 135–142. Shuaib, A.X., Crain, B., Siren, A.-L., Feuerstein, G., Hallenbeck, J. and Davis, J.N. (1990) Assessment of damage from implantation of microdialysis probes in the rat hippocampus with silver degeneration staining. Neurosci. Lett., 112: 149–154. Smith, Q.R. (2003) A review of blood-brain barrier transport techniques. Methods Mol. Med., 89: 193–208. Somjen, G.G. (2002) Ion regulation in the brain, implications for pathophysiology. Neuroscientist, 8(3): 254–267. Stahle, L. and Borg, N. (2000) Transport of alovudine (30 -ﬂuorothymidine) into the brain and the cerebrospinal ﬂuid of the rat, studied by microdialysis. Life Sci., 66(19): 1805–1816. Stain, F., Barjavel, M.J., Sandouk, P., Plotkine, M., Scherrmann, J-.M. and Bhargava, H.N. (1995) Analgesic response and plasma and brain extracellular ﬂuid pharmacokinetics of morphine and morpine-6-b D-glucuronide in the rat. J. Pharmacol. Exp. Ther., 274: 852–857. Stain-Texier, F., Boschi, G., Sandouk, P. and Scherrmann, J.M. (1999) Elevated concentrations of morphine 6-beta-D-glucuronide in brain extracellular ﬂuid despite low blood-brain barrier permeability. Br. J. Pharmacol., 128(4): 917–924. Steward, P.A. and Mikulis, D. (1998) The blood-brain barrier in brain tumours. In: Pardridge, W.M. (Ed.), Introduction to the Blood-Brain Barrier; Methodology, Biology and Pathology. Cambridge University Press, pp. 434–440. Sugiyama, Y., Kusuhara, H. and Suzuki, H. (1999) Kinetic and biochemical analysis of carrier-mediated efﬂux of drugs through the blood-brain and blood-cerebrospinal ﬂuid barriers, importance in the drug delivery to the brain. J. Control. Release, 62(1–2): 179–186.

Sun, H., Bungay, P.M. and Elmquist, W.F. (2001a) Effect of capillary efﬂux transport inhibition on the determination of probe recovery during in vivo microdialysis in the brain. J. Pharmacol. Exp. Ther., 297(3): 991–1000. Sun, H., Dai, H., Shaik, N. and Elmquist, W.F. (2003) Drug efﬂux transporters in the CNS. Adv. Drug Deliv. Rev., 55(1): 83–105. Sun, H., Miller, D.W. and Elmquist, W.F. (2001b) Effect of probenecid on ﬂuorescein transport in the central nervous system using in vitro and in vivo models. Pharm. Res., 18(11): 1542–1549. Suzuki, H., Terasaki, T. and Sugiyama, Y. (1997) Role of efﬂux transport across the blood-brain barrier and bloodcerebrospinal ﬂuid barrier on the disposition of xenobiotics in the central nervous system. Adv. Drug Deliv. Rev., 25: 257–285. Takasato, Y., Rapoport, S.I. and Smith, Q.R. (1984) An in situ brain perfusion technique to study cerebrovascular transport in the rat. Am. J. Physiol., 247(3 Pt 2): H484–H493. Takasawa, K., Terasaki, T., Suzuki, H. and Sugiyama, Y. (1997) In vivo evidence for carrier-mediated efﬂux transport of 30 -azido-30 -deoxythymidine and 20 ,30 -dideoxyinosine across the blood-brain barrier via a probenecidsensitive transport system. J. Pharmacol. Exp. Ther., 281: 369–375. Taylor, E.M., Yan, R., Hauptmann, N., Maher, T.J., Djahandideh, D. and Glasky, A.J. (2000) AIT-082, a cognitive enhancer, is transported into brain by a nonsaturable inﬂux mechanism and out of brain by a saturable efﬂux mechanism. J. Pharmacol. Exp. Ther., 293(3): 813–821. Tofteng, F. and Larsen, F.S. (2002) Monitoring extracellular concentrations of lactate, glutamate, and glycerol by in vivo microdialysis in the brain during liver transplantation in acute liver failure. Liver Transpl., 8(3): 302–305. Tong, X. and Patsalos, P.N. (2001) A microdialysis study of the novel antiepileptic drug levetiracetam, extracellular pharmacokinetics and effect on taurine in rat brain. Br. J. Pharmacol., 133(6): 867–874. Tsai, T.H., Lee, C.H. and Yeh, P.H. (2001) Effect of P-glycoprotein modulators on the pharmacokinetics of camptothecin using microdialysis. Br. J. Pharmacol., 134(6): 1245–1252. Tsuji, A. and Tamai, I. (1999) Carrier-mediated or specialized transport of drugs across the blood-brain barrier. Adv. Drug Deliv. Rev., 36: 277–290. Tunblad, K., Ederoth, P., Gardenfors, A., HammarlundUdenaes, M. and Nordstrom, C.H. (2004) Altered brain exposure of morphine in experimental meningitis studied with microdialysis. Acta Anaesthesiol. Scand., 48(3): 294–301. Tunblad, K., Hammarlund-Udenaes, M. and Jonsson, E.N. (2005) Inﬂuence of probenecid on the delivery of morphine-6glucuronide to the brain. Eur. J. Pharm. Sci., 24(1): 49–57. Tunblad, K., Jonsson, E.N. and Hammarlund-Udenaes, M. (2003) Morphine blood-brain barrier transport is inﬂuenced by probenecid co-administration. Pharm. Res., 20(4): 618–623.

572 Turner, G. (1998) Cerebral malaria and brain microvasculature. In: Pardridge, W.M. (Ed.), Introduction to the Blood-Brain Barrier; Methodology, Biology and Pathology. Cambridge University Press, Cambridge, UK, pp. 454–461. Van Bree, J.B., Heijligers-Feijen, C.D., de Boer, A.G., Danhof, M. and Breimer, D.D. (1991) Stereoselective transport of baclofen across the blood-brain barrier in rats as determined by the unit impulse response methodology. Pharm. Res., 8(2): 259–262. Vogelgesang, S., Cascorbi, I., Schroeder, E., Pahnke, J., Kroemer, H.K., Siegmund, W., Kunert-Keil, C., Walker, L.C. and Warzok, R.W. (2002) Deposition of Alzheimer’s beta-amyloid is inversely correlated with P-glycoprotein expression in the brains of elderly non-demented humans. Pharmacogenetics, 12(7): 535–541. Vorbrodt, A.W. (1988) Ultrastrucutral cytochemistry of bloodbrain barrier endothelia. Prog. Histochem. Cytochem., 18: 1–99. Walker, M.C., Tong, X., Perry, H., Alavijeh, M.S. and Patsalos, P.N. (2000) Comparison of serum, cerebrospinal ﬂuid and brain extracellular ﬂuid pharmacokinetics of lamotrigine. Br. J. Pharmacol., 130: 242–248. Wallace, M.C., Teasdale, G.M. and McCullough, J. (1992) Autoradiographic analysis of 3H-MK-801 (dizocilpine) in vivo uptake and in vitro binding after focal cerebral ischemia in the rat. J. Neurosurg., 76: 127–133. Westergren, I., Nystrom, B., Hamberger, A. and Johansson, B.B. (1995) Intracerebral microdialysis and the blood-brain barrier. J. Neurochem., 64: 229–234. Wijnholds, J. (2002) Drug resistance caused by multidrug resistance-associated proteins. Novartis Found Symp., 243: 69–79 discussion 80–82, 180–185. Williams, S.A., Davson, H. and Segal, M.B. (1995) Transport of the nucleoside thymidine, in the central nervous system,

the blood-cerebrospinal ﬂuid and blood-brain barriers. In: Greenwood, J., Begley, D.J. and Segal, M.B. (Eds.), New Concepts of a Blood-Brain Barrier. Plenum Press, New York and London, pp. 175–188. Xie, R. and Hammarlund-Udenaes, M. (1998) Blood-brain barrier equilibration of codeine in rats studied with microdialysis. Pharm. Res., 15(4): 570–575. Xie, R., Bouw, M.R. and Hammarlund-Udenaes, M. (2000) Modelling of the blood-brain barrier transport of morphine3-glucuronide studied using microdialysis in the rat, involvement of probenecid-sensitive transport. Br. J. Pharmacol., 131(8): 1784–1792. Xie, R., Hammarlund-Udenaes, M., de Boer, A.G. and de Lange, E.C.M. (1999) The role of P-glycoprotein in blood-brain barrier transport of morphine, transcortical microdialysis studies in mdr1a (–/–) and mdr1a (+/+) mice. Br. J. Pharmacol., 128: 563–568. Yokel, R.A. (2002) Brain uptake, retention, and efﬂux of aluminum and manganese. Environ. Health Perspect., 110(Suppl 5): 699–704. Yokel, R.A., Allen, D.D., Burgio, D.E. and McNamara, P.J. (1992) Antipyrine as a dialyzable reference to correct differences in efﬁciency among and within sampling devices during in vivo microdialysis. J. Pharmacol. Toxicol. Methods, 27: 135–142. Yu, D.-W., Gatley, S.J., Wolf, A.P., MacGregor, R.R., Dewey, S.L., Fowle, J.S. and Schlyer, D.J. (1992) Synthesis of carbon-11 labeled ionidated cocaine derivatives and their distribution in baboon brain measured using positron emission tomography. J. Med. Chem., 35: 2178–2183. Zhang, Y., Han, H., Elmquist, W.F. and Miller, D.W. (2000) Expression of various multidrug resistance-associated protein (MRP) homologues in brain microvessel endothelial cells. Brain Res., 876(1–2): 148–153.

CHAPTER 6.5

Assaying protein-unbound drugs using microdialysis techniques Tung-Hu Tsai Institute of Traditional Medicine, School of Medicine, National Yang-Ming University, Taipei, Taiwan and Department of Education and Research, Taipei City Hospital, Taipei, Taiwan

Abstract: This article provides a concise overview of available methodological approaches for sampling protein-unbound drugs using microdialysis. Microdialysis is a technique for protein-unbound drug sampling without withdrawal of biological ﬂuids, and thus involving minimal disturbance of physiological functions. Conventional total drug sampling includes both protein-bound and -unbound drugs. However, only the unbound fraction of a drug is available for absorption, distribution, metabolism, and elimination, as well as for delivery to the target sites for pharmacological action. Although several techniques have been used to determine protein-unbound drugs from biological ﬂuids, including ultraﬁltration, equilibrium dialysis, and microdialysis, only microdialysis allows simultaneous in vivo online sampling of protein-unbound chemicals from blood, tissues, and body ﬂuids, such as bile juice and cerebral spinal ﬂuid for pharmacokinetic and pharmacodynamic studies. This chapter describes the technique of microdialysis and its application in pharmacokinetic studies. Furthermore, the advantages and limitations of microdialysis are discussed, including the detailed surgical techniques for animal experiments with rat blood, brain, muscle, subcutaneous adipose tissue, liver, and bile duct to evaluate hepatobiliary excretion and enterohepatic circulation of unbound drug sampling and in vitro cell culture applications. investigate pharmacokinetic mechanisms of protein-unbound drugs. Although most drugs are found in the blood as both protein-unbound and -bound forms in equilibrium, only the protein-unbound drugs are able to penetrate cell membranes either from the gastrointestinal tract to blood vessels or from the blood to the extravascular sites of drug action. This unbound concentration provides direct interaction at the site of action for pharmacological and toxicological effects. It would seem reasonable that concentration–response relationships be based on protein-unbound concentration rather than on total drug concentration in blood or tissue. With microdialysis, the sampling probe is implanted directly into the tissue of research interest as microdialysis collects only the protein-unbound

I. Introduction Generally speaking, this method allows the sampling of unbound endogenous and exogenous substances, which surround a microdialysis probe (de Lange et al., 2000; Verbeeck, 2000; Bourne, 2003; Tsai, 2003; Plock and Kloft, 2005). The ﬁrst paper involving microdialysis, where it was applied to the study of dopamine neurotransmission, was published by Ungerstedt and Pycock (1974). By August 2006, a search of database PubMed revealed 10,685 articles listed under the keyword ‘‘microdialysis’’ and 1,094 articles for ‘‘microdialysis and pharmacokinetics’’. Thus, microdialysis has rapidly become a widely established technique to Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

573

DOI: 10.1016/S1569-7339(06)16030-0 Copyright 2007 Elsevier B.V. All rights reserved

574

fraction of a drug. For this reason, a microdialysis sampling system excludes the fraction of protein binding (50–90%) and recovery by the microdialysis membrane (10–50%). The ﬁnal concentration of dialysate is approximately 1–25% of the total concentration of analyte. Therefore, to have a sensitive assay system capable of monitoring these, very low concentrations of analyte from the dialysate is a critical issue for this technique. According to classical pharmacokinetics, once a drug molecule enters the bloodstream, the drug is distributed to the body tissues of its target site and has its particular pharmacological effect. However, only the protein-unbound fraction of drugs is available for distribution to the target organ or tissue, and for binding to cell-membrane receptors to thereby activate cellular responses. Traditional biological ﬂuid sampling methods measure the total drug concentration, which includes both protein-bound and -unbound drug concentrations, and therefore, does not reﬂect the true pharmacologically active concentration of the drug at the cellular level. Several methods have been used to measure the protein-unbound fraction from the total fraction of drug. Techniques such as ultraﬁltration, equilibrium dialysis, and microdialysis have been commonly applied for sampling the unbound fraction of drugs from biological samples.

II. Principles of microdialysis A permeable membrane on the tip of two concentric silica tubes is the basic structure of the microdialysis probe (Fig. 1). The complexity of microdialysis is because of the interactions between the dialysis membrane, perfusion solution, and surrounding environment of living tissue. When the probe is implanted into the region of interest (either a tissue or a medium), a suitable solution (perfusate) that enters through the inner tube is perfused at a constant ﬂow rate and ﬂows to its distal end. The perfusate exits by the inner tube and enters the space between the inner tube and the outer dialysis membrane, so the molecular exchange occurs at the tip of the probe. By its molecular weight cut-off, the membrane excludes larger molecules from the surrounding

Fig. 1. Structure of a microdialysis probe. (Adapted from Tsai, 2003.)

environment into the perfusate, so only small hydrophilic molecules in their protein-unbound form can penetrate the membrane (Ungerstedt, 1991). As enzymes are also excluded, no further enzymatic biodegradation occurs in the dialysate. The dialysate leaving the probe is collected and analyzed by analytical systems. The basic principle of microdialysis can be explained in terms of Fick’s ﬁrst law of passive diffusion. Accordingly, the driving force to penetrate the membrane is the concentration gradient. The diffusion rate of analytes, including endogenous substances and delivery of drugs, is dependent on the surface area of the dialysis membrane, the ﬂow of the perfusion solution, and the substance’s speed of diffusion through the extracellular ﬂuid. The permeation rate of a drug through microdialysis membrane depends on the following factors: drug concentration, oil/water partition coefﬁcient of the drug, and surface area of the dialysis membrane. In addition, temperature and pH nearby the microdialysis probe, as well as molecular weight, shape, and charge are also factors that inﬂuence elements of recovery. If the substance concentration in the probe lumen is higher than that in the extracellular space, the substance will cross the membrane into the extracellular space according to the concentration gradient. This is the process for administration. In contrast, if the perfusate is a blank artiﬁcial physiological ﬂuid, it will perfuse into the body to

575

collect substances from the body. In this situation, the substance concentration in the surrounding environment of the microdialysis probe is higher than that in the probe lumen, and the substance will be collected into the microdialysis probe according to the concentration gradient and be continuously removed by the perfusates. This is the process for sampling. With the exception of administration and sampling, if the half-life of the substance is very short, for example, 109 s for hydroxyl free radicals, a trapping reagent such as salicylic acid can be perfused into the probe lumen to capture the hydroxyl free radicals (Fig. 2). In this case, the salicylic acid concentration in the perfusate of the probe lumen is higher than the surrounding environment of the probe. Therefore, salicylic acid will diffuse out of the dialysis membrane and then hydroxyl radicals may have a chance to attach to the para- and metapositions of salicylic acid, thereby generating 2,3and 2,5-dihydroxyl benzoic acids (2,3-DHBA and 2,5-DHBA). By passive diffusion, these two products 2,3-DHBA and 2,5-DHBA can be collected from the dialysates according to the concentration gradient (Tsai, 2003). As the microdialysis technique causes no biological ﬂuid loss due to sampling from the body, in contrast to conventional sampling methods, microdialysis provides the major advantages of

Fig. 2. Enlarged view of the tip of a microdialysis probe for the reaction of hydroxyl radicals with salicylic acid through the membrane of the microdialysis probe to 2,3- and 2,5-dihydroxyl benzoic acids. (Adapted from Tsai, 2003.)

higher temporal resolution in the sampling interval and continuous sampling over longer periods of time. This is especially useful to increase data points from a relatively small animal without disturbing the physiological conditions. Following the drug administration, the drug molecules in the extracellular space of plasma or tissue diffuse into the dialysate and may be recovered to determine their concentrations. Hence, the level of drug concentration can be detected for further pharmacokinetic study. The drug concentration data from sampling of extracellular ﬂuid are comparable with the data from ordinary blood sampling, and this microdialysis pharmacokinetic characteristic has been evaluated by a multicompartmental model (Stahle, 1993). Regular microdialysis probes are 1–4 mm membrane long and 200 mm in diameter, which gives poor spatial resolution, and this diameter limits their use for small nuclei in the brain. To improve spatial resolution for in vivo sampling, a miniaturized probe using a fused silica capillary tube with 18–40 mm inner diameter (without dialysis membrane) and an outer diameter of 90 mm is used to measure amino acids in striata of anesthetized rats (Kennedy et al., 2002). Multiple probes implanted at the same time in different tissues or regions of interest of an experimental animal are also used. For example, Davies and Lunte (1996) used three microdialysis probes implanted into separate liver lobes in the same rat to study the metabolism of phenol, as different enzymes in different lobes may have individual activity. In another study, three microdialysis probes have been used to investigate blood, brain, and bile of anesthetized rat to study the pharmacokinetics of peﬂoxacin (Tsai, 2001). These studies provide information on brain distribution and hepatobiliary excretion of various drugs. Four microdialysis probes simultaneously implanted into the blood, brain, liver, and bile of the same anesthetized rat have been used to investigate the biotransformation of drugs in the liver, brain, distribution in the brain, and hepatobiliary excretion in the bile (Tseng and Tsai, 2004). As described above, microdialysis provides protein-unbound samples, which permits the dialysate to be injected directly into the analytical

576

binding can be represented as: unbound drug þ free protein ¼ drug2protein complex

Fig. 3. Microdialysis hyphenation for (A) ofﬂine and (B) online systems. (Adapted from Tsai, 2003.)

instrument via an ofﬂine or an online injection system (Fig. 3) (Tsai, 2003). Online microdialysis coupled with an analytical system provides the advantages of automatic sampling and detection, time and labor savings, suitability for unstable samples, and no sample contamination. One limitation of the online system is the requirement that the sampling period should be the same as the injection interval of the analytical system, which constrains the temporal resolution. Although each sample can be injected only once using the online system, the advantage of the ofﬂine system is that the sample can be further concentrated to enhance the sensitivity or it can be desalted prior to the injection into a mass spectrometer.

III. Protein binding and equilibrium dialysis The major drug binding proteins in plasma are albumin, alpha-1 acid glycoprotein, and lipoproteins. Drugs are loosely bound to plasma proteins, such as albumin for acidic drugs and alpha-1 acid glycoprotein for basic drugs, thereby forming an equilibrium ratio between bound and unbound drugs. Drug protein binding is a reversible interaction of drugs with proteins in plasma, and the rates of drug binding and release are very fast, occurring in the milliseconds range. Drugs can also bind reversibly to red blood cells, tissue membranes as well as other blood and tissue constituents. Protein

The drug–protein complex can rapidly dissociate for drugs that are initially bound to proteins. For example, liver cells very efﬁciently extract free drugs from the blood, and these can be extracted in one pass through the liver. Compounds, such as acrylonitrile can highly react alkylate with cysteine of protein (Campian et al., 2002). Binding of a drug to plasma proteins limits its concentration in tissues and its pharmacological action as only the unbound drug is in equilibrium across membranes. Accordingly, after distribution equilibrium is achieved, the intracellular concentration of active, unbound drug is the same as that in plasma, except when carrier-mediated transport is involved. Drug transport, metabolism, and excretion are also limited by plasma binding. In vitro and in vivo protein binding of methotrexate has been assessed by microdialysis, with results indicating that the relative recovery was independent of methotrexate concentrations. However, in that study, recovery was a factor that was assessed in the presence of proteins surrounding the dialysis membrane, and this was related to the physiological buffer, human serum albumin, and human plasma. The study revealed that the protein content can directly affect microdialysis probe recovery (Maia et al., 1996). Comparing microdialysis sampling with blood sample withdrawal for the in vivo pharmacokinetics of ﬂurbiprofen revealed that simultaneous sampling of blood and intravenous (i.v.) microdialysis can be used to study the pharmacokinetics of ﬂurbiprofen in an individual rat (Evrard et al., 1996). To investigate the variation in microdialysis and ultraﬁltration unbound concentrations, one early study used both microdialysis sampling via the rabbit femoral vein and collection of whole blood via the rabbit ear vein after valproate injection. As a result, concentrations of free valproate in plasma were determined by ultraﬁltration method that could be compared with the microdialysis method. The results indicated that there is no difference in the elimination half-life of valproate determined

577

by microdialysis and by ultraﬁltration. With correction of dialysate recovery, the area under the concentration versus time curve (AUC) of valproate in the dialysate was also in good agreement with the value obtained from plasma ultraﬁltration (Nakashima et al., 1994). For unbound drug sampling, the conventional equilibrium dialysis method has the limitation of excessive dialysis time, which can cause drug and protein degradation and allow bacterial growth. Its alternative, microdialysis, is based on the same basic principle and uses a semipermeable membrane for collecting unbound drug. Some in vitro experiments have already shown good agreement from the evaluation of unbound drug by equilibrium dialysis and by ultraﬁltration (Herrera et al., 1990; Eckblom et al., 1992). Additional studies have demonstrated that there is no apparent bias in the correlation of the microdialysis and the equilibrium dialysis data, thus supporting the use of microdialysis in pharmacokinetic studies (Eckblom et al., 1992; Sarre et al., 1992).

IV. Microdialysis experiments IV.A. Blood microdialysis For implantation of a microdialysis probe into a rat blood vessel, a ﬂexible microdialysis probe is implanted in the jugular vein via the superior vena cava toward the right atrium of the heart, and it is then perfused with the anticoagulant, citrate dextrose (ACD solution: citric acid 3.5 mM; sodium citrate 7.5 mM; dextrose 13.6 mM), to avoid blood clotting around the dialysis ﬁber. This probe tip extends about 3 cm to reach the right superior vena cava/atrial junction, and the probe has an active dialysis membrane 1 cm long for drug sampling. The major reason locating the probe in the right atrial junction is to obtain a larger blood pool surrounding the probe (Fig. 4). Although blood ﬂow is impeded when a microdialysis probe is inserted into the jugular vein, the afﬂuent blood from the inferior vena cava provides adequate blood supply for effective blood microdialysis. In contrast, if the microdialysis probe is implanted into other veins, such as the femoral vein, the

Fig. 4. Position of a microdialysis probe inserted into the jugular vein toward the heart. (Adapted from Tsai, 2003.)

blood vessel would be occluded, resulting in poor blood circulation and poor dialysis efﬁciency. These considerations make the jugular vein the most suitable location for blood sampling for microdialysis. In vivo blood microdialysis sampling has widely been employed to study the pharmacokinetics and metabolism of xenobiotics. A naturally occurring antioxidant, the unbound fraction of caffeic acid has been sampled by online microdialysis and applied to pharmacokinetic study. To avoid auto-oxidation of analyte, the dialysate was automatically injected into a liquid chromatographic system (Tsai et al., 1999). Similarly, the unbound fraction of ()-epigallocatechin-3-gallate (EGCG), isolated from green tea, has been measured in rat blood and ﬁtted best by two-compartmental pharmacokinetic model (Lin et al., 2004b). For radiolabel analysis, Haaparanta et al. (2004) used a microdialysis system coupled to planar chromatography and digital autoradiography with a phosphoimager plate to investigate pharmacokinetics of various radiopharmaceuticals. This highly sensitive method provided simultaneous analysis of PET tracers and their metabolites in blood. 2-[18F]Fluoro-2-deoxy-D-glucose (FDG) is actively transported into the cells by the glucose transporters, and microdialysis has been used to measure its extracellular concentrations to study

578

glucose metabolism. Furthermore, the kinetics of [18F]-radioactivity in the blood, muscle, liver, and adipose tissue as well as analyses of the metabolites of FDG have been measured by HPLC in muscle, liver, and adipose tissue homogenates (Haaparanta et al., 2003). To reduce stress on the experimental animal, an automated blood sampler for simultaneous sampling of systemic blood and brain microdialysates has been developed to investigate the systemic pharmacokinetics and its neurotransmitters release in an awake and freely moving rat. Dialysates could be divided into two parts for the evaluation of drug level and neurotransmitter release. After oral administration of olanzapine, the dialysate collected from the brain indicated that the drug elevates dopamine release from the striatum. Comparison of the drug levels in the blood and brain indicated that the olanzapine level remained elevated in the brain after plasma concentration had started to decline (Gunaratna et al., 2004). Microdialysis has been used to demonstrate the connection of liquid chromatography (LC) and chemiluminescence detection. The unbound level of levodopa was measured by online microdialysis in the rabbit auricle vein, and the detection limit of levodopa was 3.0 ng/mL with a linear range of 10–1,000 ng/mL. This detection system can also be applied to the determination of levodopa concentration in human serum and urine (Funan et al., 2005). Furthermore, the microdialysis technique has been used to assess herb–drug interactions. For example, a herbal ingredient, both as a single herbal extract and as a complex herbal preparation using that ingredient, was investigated to determine the interaction of theophylline and caffeine (Jan et al., 2005; Tsai et al., 2005). IV.B. Brain microdialysis In the past, push–pull perfusion (Fig. 5) was one of the major in vivo methods for direct measurements of neurotransmitters. This was modiﬁed by Ungerstedt and Pycock (1974) with using microdialysis in the neurosciences for the measurement of neurotransmitters. The central nervous system (CNS) is protected from the peripheral circulation

PULL

PUSH

DENTAL CEMENT GUIDE TUBE SKULL

Fig. 5. Push–pull perfusion for tissue ﬂuid sampling.

system by the blood-brain barrier (BBB) and the blood–cerebrospinal ﬂuid (BCF) barrier. A drug molecule must penetrate these barriers surrounding the CNS by passive diffusion via endothelial cells or the tight junctions. It is also possible to penetrate the BBB by active transport. Several speciﬁc transporter mechanisms have been identiﬁed as efﬂux pumps on the luminal side of the BBB, for example, P-glycoprotein (HammarlundUdenaes, 2000; Sawchuk and Elmquist, 2000). Brain penetration is deﬁned as the brainto-blood distribution, which is calculated by dividing the analyte AUC in brain by its AUC in blood (k ¼ AUCbrain/AUCblood) (de Lange et al., 1997, 1999). Peﬂoxacin has been found to penetrate the BBB, and it has been shown that the pharmacokinetic proﬁles of peﬂoxacin in rat blood and brain are not altered by the treatment of P-glycoprotein modulator (Tsai, 2001). Unbound morphine and its metabolite morphine-3-glucuronide concentrations have been monitored in the rat brain extracellular ﬂuid of the striatum using microdialysis (Bouw et al., 2000; Xie et al., 2000). The elimination half-life of unbound morphine and morphine-3-glucuronide in the rat brain was 44 and 81 min, respectively. In addition, this research group has presented clinical research using intracerebral microdialysis to obtain data on morphine penetration of the BBB in a patient with severe brain injury (Bouw et al., 2001). The elimination half-lives of morphine in uninjured and injured brain tissue were

579

found to be 178 and 169 min, respectively, indicating that morphine is retained in brain tissue longer than in the peripheral circulation system and adipose tissue. Direct evidence of the effect of P-glycoprotein on the brain distribution for the mdr1a (/) gene knockout mice was used to investigate the mechanism for penetration of the BBB by morphine (Xie et al., 1999), rhodamine-123 (de Lange et al., 1998), and sparﬂoxacin (de Lange et al., 2000b). The results indicate that both morphine and rhodamine-123 levels in brain of mdr1a (/) mice were higher than those in wild-type mdr1a (+/+) mice after i.v. infusion and without changes in blood levels. These studies demonstrate in general that microdialysis is suitable to investigate drug distribution within the CNS; and speciﬁcally, that comparison between mdr1a (/) and (+/+) mice indicates that P-glycoprotein participates in regulating the amount of morphine and rhodamine-123 transported across the BBB (de Lange et al., 1998, 2000a; Xie et al., 1999). To eliminate the need for gene knockout, certain chemicals, such as cyclosporine or verapamil, can be used to partially inhibit the transporter of P-glycoprotein (Sakata et al., 1994; de Lange and Danhof, 2002). Cyclosporine has also been investigated for its ability to increase brain concentrations of cefepime (Chang et al., 2001), camptothecin (Tsai et al., 2001), and baicalein (Tsai et al., 2002). GF120918 is also a P-glycoprotein modulator, and it has been demonstrated to increase the concentration of unbound amprenavir in the CNS of rats, signiﬁcantly increasing the brain-to-blood distribution ratio (Edwards et al., 2002). It has been hypothesized that active efﬂux of morphine occurs in the human brain, and brain injury could alter the pharmacokinetics of cerebral morphine. The AUC ratio of unbound morphine in brain tissue to plasma was found to be 0.64 in normal brain tissue and 0.78 in similar brain tissue that had been injured. With active efﬂux of unbound morphine across the BBB, injury to brain tissue can increase the permeability for morphine to cross the BBB (Ederoth et al., 2004). Microdialysis allows the dynamic monitoring of biochemical analysis in the blood of gerbils

subjected to cerebral ischemia/reperfusion study. Thus, during cerebral ischemia and reperfusion, levels of magnesium, glucose, and lactate can be monitored. It was found that glucose levels continued to have no signiﬁcant difference during cerebral ischemia, and had gradually fallen to 50% of basal levels by the end of reperfusion. Magnesium levels gradually rose during cerebral ischemia and returned to the basal levels within 30 min after reperfusion (Lin et al., 2004a). IV.C. Microdialysis in subcutaneous adipose tissue The principal barrier of the skin is the stratum corneum. Hydrophilic drugs, such as aciclovir and penciclovir may not be absorbed in topical applications because of the lipid matrix of this barrier layer. Morgan et al. (2003) developed a cutaneous microdialysis system to measure dermal drug levels of acyclovir and penciclovir in the forearm of healthy volunteers. After removal of the stratum corneum by tape stripping, the results indicated that the absorption of penciclovir and aciclovir were increased by 1,300- and 440-fold, respectively. These results conﬁrm that the stratum corneum is the major barrier to hydrophilic drug absorption through the skin. To measure unbound ﬂurbiprofen levels in the dermis and in subcutaneous tissue, two linear microdialysis probes were implanted in the dermis and in subcutaneous tissue at depths of 398.3 and 1,878 mm, respectively. The results indicated that unbound ﬂurbiprofen concentration in the dermis was about 13-fold higher than that in the subcutaneous tissue, and thus iontophoresis delivery was more efﬁcient for delivery of ﬂurbiprofen through the dermis (Mathy et al., 2005a). To investigate ﬂuconazole, which is a water-soluble antifungal agent, two microdialysis probes were implanted in the jugular vein and dermis of rats, which then awoke and were free to move around. This study shows that cutaneous microdialysis is an effective and minimally invasive tool to evaluate the dermal pharmacokinetics of ﬂuconazole following i.v. or topical administration (Mathy et al., 2005b). An online in vivo microdialysis and glucose sensor was designed to monitor glucose levels in diabetic patients in the abdominal subcutaneous

580

interstitium (Schoonen and Wientjes, 2003). As the analyte concentration in the perfusate is substantially lower than that in the interstitial compartment, an ultra-slow microdialysis procedure with perfusion rates below 0.1 mL/min was developed by Kaptein et al. (1998). Recently, a LC–tandem mass spectrometry was used to measure scopolamine concentrations in blood sample and dialysate collected from subcutaneous adipose tissue. The pharmacokinetic data indicate that the elimination half-life of scopolamine in the serum and dialysate were not signiﬁcantly different in healthy volunteers (Stetina, et al., 2005). IV.D. Muscle microdialysis Based on the concentration gradient, a microdialysis system also allows the monitoring of unbound fractions of a drug within interstitial levels of tissue. This can investigate how the protein-unbound level of antibiotics at an infection site, such as muscle can be reﬂected in the bactericidal effect. Freddo and Dalla Costa (2002) observed unbound norﬂoxacin in skeletal muscle using microdialysis to investigate the pharmacokinetics of norﬂoxacin in rats. In comparison with the plasma concentration of norﬂoxacin and its distribution in the skeletal muscle, the muscle penetration of norﬂoxacin was correlated to total plasma and free tissue levels of the drug when the protein binding and tissue penetration factor were known. Using microdialysis, the concentration of SDZ ICM 567 was found to be 10 times lower in the muscle than in the blood (Ofner et al., 1997). The unbound metronidazole level in blood of patients achieved a value of 16.574.6 mg/L at 30 min, while in muscle, a maximum level of 7.871.5 mg/L was achieved at 114 min (Karjagin et al., 2004). The unbound level of a neuromuscular blocker, rocuronium, in the muscle tissue of dogs was measured by a microdialysis system. The results indicated that the drug concentrations in the site of action were better explained by the concentration–effect relation of muscle relaxants (Ezzine et al., 2004). Similarly, the muscle interstitial levels of glucose, lactate, pyruvate, and urea have been monitored by microdialysis in healthy subjects during

exercise. It was found that the interstitial glucose concentration in human skeletal muscle decreases for a prolonged period following a single short period of exercise, even after the blood ﬂow has returned to resting levels (Henriksson and Knol, 2005). Tissue distributions of unbound imipenem in blood, skeletal muscle, and lung extracellular ﬂuids of rats have also been investigated by microdialysis. The AUC ratios of tissue (muscle or lung) to the AUC for blood (AUCmuscle or lung/ AUCblood) were virtually equal to 1. This result demonstrates that unbound imipenem concentrations were virtually identical in blood, muscle, and lung (Marchand et al., 2005). Similar results have been found in the tissue distribution of cefpodoxime into skeletal muscle and lung in rats, indicating that the interstitial levels of unbound cefpodoxime in muscle and lung tissue are very similar (Liu et al., 2005). IV.E. Liver microdialysis Since 1991, microdialysis probes have been utilized to monitor blood and liver extracellular ﬂuid after i.v. aluminum lactate or aluminum citrate injection compared with multiple blood withdrawals. The results indicate that for toxicokinetic studies, metals can be repetitively sampled in the extracellular space of the liver using microdialysis (Yokel et al., 1991a, b). Subsequently, several reports have demonstrated that concurrent multiple-site microdialysis sampling is a useful tool for pharmacokinetic and drug metabolism studies in the liver (Scott and Lunte, 1993; Davies and Lunte 1995). Two types of microdialysis probes have been regularly used in liver sampling. One, the ﬂexible concentric probe, is identical to the probe used for sampling in the blood vessel. The other is a simple linear probe, constructed with fused silica as both the inlet and outlet tubing, and containing a 4–5 mm long dialysis membrane between two sections of the silica tubing. However, long-term implantation of either type of probe in liver tissue runs the danger of necrosis appearing at the implantation site after 12 h (Davies and Lunte, 1996). For liver probe implantation, the liver of the anesthetized rat is exposed by making an incision

581

on the midline at the xiphoid and extending it approximately 2–4 cm posteriorly. The dialysis probe is inserted into the median lobe of liver toward the head, parallel to the midline and then ﬁxed to the abdominal muscle with suture. After probe insertion, a wash-out period of 1 h is used to clear the extracellular ﬂuid space from substances released because of the cellular damage caused by the implantation procedure. It has been concluded that the microdialysis recovery of a substance from the liver is not generally affected by liver metabolism (Stenken et al., 1997). IV.F. Bile microdialysis Scott and Lunte (1993) described a ﬂow-through microdialysis probe for sampling bile while preserving normal bile ﬂow to investigate the hepatic metabolism of phenol. After i.v. infusion of phenol, phenol and its major hepatic metabolites were analyzed by LC. The results indicated that the bile concentrations of phenol and its metabolites are higher than those in the liver, demonstrating that the metabolites are actively excreted into the bile. The bile duct in the rat is about 1 mm wide and runs from the hilum of the liver through the pancreatic tissue to the duodenum. The pancreatic ducts merge into the posterior part of the bile duct. There is no gall bladder in the rat, and bile juice together with pancreatic secretions enters the duodenum. Therefore, catheterization of the duct near the hilum of the liver will allow the collection of pure bile (Fig. 6) (Scott and Lunte, 1993). For bile duct cannulation, a midline abdominal incision of about 2 cm is made in an anesthetized rat. The duodenum and a small part of the intestine are pulled out to the right and kept moist by covering with a gauze pad soaked with physiological saline. The bile duct can be seen around the region of duodenum, especially if it is traced back from the hilum of the liver (Fig. 6A). As 0.5–1 cm of the bile duct near the hilum of the liver is free of pancreatic tissue, this region is used for catheterization. One section of polyethylene (PE-10) tube is inserted into the anterior region of the bile duct (Fig. 6B), and the other section of PE10 tubing is inserted into the posterior region of the bile duct (Fig. 6C). Shunt microdialysis probe

Fig. 6. Schematic procedures for the implantation of a microdialysis probe into the bile duct: (A) ﬁnd common bile duct through duodenum end, (B) insert a PE-10 tube into common bile duct toward liver site, (C) insert another PE-10 tube into common bile duct toward duodenum site, and (D) a ﬂowthrough microdialysis probe was connected into these two end of the PE-10 tubes. (Adapted from Tsai, 2003.)

implantation is connected to the anterior region of PE-10 tubing to conduct bile ﬂow. After bile ﬂows through the chamber of the dialysis probe, the smaller molecules contained in the bile may penetrate the dialysis membrane into the dialysate according to the concentration gradient. The other end of the shunt probe is inserted into the posterior region of PE-10 toward the duodenum, allowing bile to ﬂow into the small intestine (Fig. 6D). A diagram of the shunt microdialysis probe is shown in Fig. 7. This is a valuable tool for proﬁling analytes in the rat bile duct, while preserving enterohepatic circulation without disruption of the bile ﬂow. This technique also provides continuous sampling with no net loss in ﬂuid volume and has high temporal resolution during the experimental period. A simultaneous measurement of unbound camptothecin dialysates in a study of rat blood and bile indicates that the amount of camptothecin, as estimated from the AUC, in bile versus concentration gradient signiﬁcantly exceeds that in blood, suggesting that camptothecin might be actively excreted into the bile. Treatment with P-glycoprotein modulator results in a decrease in

582

Fig. 7. Animal model for bile sampling: (A) detailed description of a custom-made bile microdialysis probe, and (B) ﬂow-through microdialysis probe used for rat bile sampling. (Adapted from Tsai, 2003.)

the concentration versus time curve of camptothecin in the bile. These results imply that the P-glycoprotein might regulate the hepatobiliary excretion of camptothecin (Tsai et al., 2001). Sinomenine, an herbal ingredient isolated from Sinomenium acutum, is used for the amelioration of arthritis. The biliary distribution ratios (k ¼ AUCbile/AUCblood) being 3.8570.29 and 3.5270.28 at 10 and 30 mg/kg, respectively, indicating active hepatobiliary excretion. The bile-toblood distribution ratio was signiﬁcantly reduced to 0.4770.05 and 0.4970.05, respectively, when cyclosporine (a P-glycoprotein inhibitor) was coadministered. These results suggest that sinomenine underwent active hepatobiliary elimination, which might be regulated by the P-glycoprotein (Tsai and Wu, 2003). Another herbal ingredient, berberine was isolated from the roots and bark of Berberis aristata or Coptis chinensis. The elimination mechanism of berberine was processed through hepatobiliary excretion, which was against a concentration gradient. Based on the bile-to-blood distribution ratios (k ¼ AUCbile/AUCblood), the active berberine efﬂux might be affected by P-glycoprotein and

organic cation transporter as coadministration of berberine and cyclosporine or quinidine signiﬁcantly decreased the berberine amount in bile (Tsai and Tsai, 2004a). Genistein, one of the major isoﬂavone in soybeans, has been shown to have a wide range of effects. Microdialysis investigation of this phytoestrogen has shown that it penetrates the BBB and goes through hepatobiliary excretion. The brain-to-blood (AUCbrain/AUCblood) and bile-toblood (AUCbile/AUCblood) distribution ratios were found to be 0.0470.01 and 1.8570.42, respectively for a 30 mg/kg dosage of genistein. After coadministration of cyclosporine, the distribution ratios of genistein in brain and bile were not signiﬁcantly altered. These results suggest that the BBB penetration and hepatobiliary excretion of genistein may not be regulated by P-glycoprotein (Tsai, 2005). In addition, peﬂoxacin (Tsai, 2001), ﬂuconazole (Lee et al., 2002), and metronidazole (Tsai and Chen, 2003) all demonstrated rapid exchange and equilibration between the blood and hepatobiliary system. However, the hepatobiliary excretion mechanism may not be related to the P-glycoprotein transporter.

583

In contrast, ranitidine (Huang et al., 2005) and baicalin (Tsai and Tsai, 2004b) undergoes hepatobiliary excretion against the concentration gradient from bile-to-blood. The ranitidine levels in bile decreased with the treatment of cyclosporine or quinidine, which suggests that the hepatobiliary excretion of ranitidine might be partially regulated by P-glycoprotein or organic cation transporter. IV.G. Enterohepatic circulation The processes of enterohepatic circulation may be comprised of sequential routes of hepatic uptake from blood, excretion from the liver into the bile, transport of the bile to the duodenum, reabsorption from the intestine, and returning to the liver via portal circulation. The secretory transporters are expressed on the apical membrane of enterocytes, so direct secretion of drugs and metabolites may occur from the systemic circulation into the intestinal lumen. Such enterohepatic recycling may signiﬁcantly prolong the presence of a drug (or toxin) and its effects within the body prior to the elimination by other pathways. Transporter-mediated hepatic uptake can be the cause of drug–drug interactions involving drugs that are actively taken up into the liver and metabolized and/or excreted in the bile. P-glycoprotein plays a very important role in hepatobiliary excretion of drugs from the bile-to-blood. For animal experiments, using a hepato-duodenal shunt connecting a drug-treated donor to another untreated recipient rat with simultaneous blood sampling through microdialysis permits multiple sampling without undue stress and biological ﬂuid consumption. The bile duct of the donor is cannulated proximal to the liver with a 20-cm section of PE-10 tubing, the other end of which is inserted through the bile duct into the duodenum of the recipient rat. To balance the ﬂuid losses and gains in the donor and recipient rats, the bile duct of the recipient rat is also cannulated to channel the bile back to the donor rat (Fig. 8). For this technique, after about 2 h of surgical stabilization period, drug may be administered to the donor rat through a femoral vein. Dialysates from the blood of the donor and recipient rats will

Fig. 8. Hepato-duodenal shunt model for the investigation of enterohepatic circulation. To balance the ﬂuid losses and gains in the donor (drug treatment) and recipient (no drug treatment) rats, the bile duct of the recipient rat was also cannulated to channel the bile back to the donor rat. (Adapted from Tsai, 2003.)

be collected for later analyses. To estimate quantitatively the degree of enterohepatic circulation taking place in the paired rats, the AUC in the recipient rat is compared with the AUC in the donor rats (AUCrecipient/AUCdonor). After chloramphenicol administration (100 mg/kg, i.v.) into the recipient rat, pharmacokinetic parameters calculated from the AUCs of unbound chloramphenicol and chloramphenicol glucuronide show that the extent of recycling (AUCrecipient/AUCdonor) is 1.8 and 4.9% for chloramphenicol and chloramphenicol glucuronide, respectively. The in vivo pairedrat animal model was demonstrated to be potentially useful for studying the pharmacokinetics and enterohepatic circulation of analytes in rats (Tsai et al., 2000). IV.H. Cell culture microdialysis Microdialysis has also been used to measure intracellular catecholamines in PC-12 cells (Cheng et al, 2000). For this, the catecholamine concentrations in the medium were analyzed by microbore liquid chromatography with electrochemical detection (LC–ED). Conventional medium sampling requires time-consuming pretreatments or complicated extraction prior to analysis by conventional LC-ED assays, which

584

may result in sample loss and may require more time to complete the experiments. Alternatively, the microdialysis technique has the advantages of requiring no sample pretreatment, and sample can be directly injected for analysis. This online method speeds up the experimental procedure, provides high sensitivity, minimizes the required sample volume, enhances the detection limits, and decreases degradation of analyzed compounds. An online microdialysis device has also been developed for the direct measurement of pyruvate and lactate in primary liver cell culture medium. This novel sampling device is constructed within a hypoxia chamber. Using these results, dynamic changes in pyruvate and lactate levels in primary liver cell culture medium under hypoxia and reperfusion have been discussed (Wu et al., 2001). In this technique, the sampling device is composed of a Petri dish, two transmission tubes, and a dialysis membrane, as illustrated in Fig. 9. The Petri dish has a receiving space with an open top. The cover has two holes separated by a predetermined distance. This microdialysis system is similar to those reported by Maas et al. (1992) and Miyamato and Chams (1991). Cell culture coupled to microdialysis technique is relatively efﬁcient, cost-effective, and less vulnerable to human error than conventional studies, in which a number of Petri dishes are used.

Fig. 9. Schematic diagram of a Petri dish used for microdialysis. (Adapted from Tsai, 2003.)

A previous study minimized pretreatment procedures for sample preparation, decreased possible contamination from sampling of culture medium, and enhanced the detection sensitivity of catecholamines in PC-12 cell culture medium. Furthermore, this novel microdialysis device can also be applied to the measurement of chemical substances in other culture systems (Cheng et al., 2000).

V. Conclusion Microdialysis provides several advantages for pharmacokinetic and pharmacodynamic studies by in vivo sampling of extracellular ﬂuid in many kinds of tissues and ﬂuids. In contrast to other methods of sampling biological ﬂuids and tissues, microdialysis provides a very clean dialysate, which requires no further clean-up procedure. Microdialysis also allows continuous monitoring of drug absorption, distribution, metabolism, and elimination at various tissue sites and various ﬂuids. The fact that there is no biological ﬂuid loss makes microdialysis sampling available for highly temporal and spatial resolutions. As no biological ﬂuid is removed from or introduced into the body during the process of microdialysis, minimal perturbation can be achieved. For pharmacokinetic and pharmacodynamic investigation, the dialysate is obtained before, during and after the drug treatment, so each animal serves as its own control. This permits a cross-over experiment performed in a single animal. Using several microdialysis probes, simultaneous sampling from various sites in a single animal avoids the problems associated with intra-animal variability for pharmacokinetic studies. The AUC data from blood could be directly compared with the AUC from the brain for brain distribution and BBB penetration studies. Multiple site sampling is also utilized in the investigation of hepatobiliary excretion and enterohepatic circulation. The brainto-blood and bile-to-blood distribution ratios are calculated by AUCbrain/AUCblood and AUCbile/ AUCblood, respectively. Furthermore, these designs have been using to investigate the mechanism of P-glycoprotein.

585

References Bourne, J.A. (2003) Intracerebral microdialysis: 30 years as a tool for the neuroscientist. Clin. Exp. Pharmacol. Physiol., 30: 16–24. Bouw, M.R., Gardmark, M. and Hammarlund-Udenaes, M. (2000) Pharmacokinetic-pharmacodynamic modelling of morphine transport across the blood-brain barrier as a cause of the antinociceptive effect delay in rats – a microdialysis study. Pharm. Res., 17: 1220–1227. Bouw, R., Ederoth, P., Lundberg, J., Ungerstedt, U., Nordstrom, C.H. and Hammarlund-Udenaes, M. (2001) Increased blood-brain barrier permeability of morphine in a patient with severe brain lesions as determined by microdialysis. Acta Anaesthesiol. Scand., 45: 390–392. Campian, E.C., Cai, J. and Benz, F.W. (2002) Acrylonitrile irreversibly inactivates glyceraldehyde-3-phosphate dehydrogenase by alkylating the catalytically active cysteine 149. Chem. Biol. Interact., 140: 279–291. Chang, Y.L., Chou, M.H., Lin, M.F., Chen, C.F. and Tsai, T.H. (2001) Effect of cyclosporine, a P-glycoprotein inhibitor, on the pharmacokinetics of cefepime in rat blood and brain: a microdialysis study. Life Sci., 69: 191–199. Cheng, F.C., Kuo, J.S., Huang, H.M., Yang, D.Y., Wu, T.F. and Tsai, T.H. (2000) Determination of catecholamines in pheochromocytoma cell (PC-12) culture medium by microdialysis-microbore liquid chromatography. J. Chromatogr. A, 870: 405–411. Davies, M.I. and Lunte, C.E. (1995) Microdialysis sampling for hepatic metabolism studies. Impact of microdialysis probe design and implantation technique on liver tissue. Drug Metab. Dispos., 23: 1072–1079. Davies, M.I. and Lunte, C.E. (1996) Simultaneous microdialysis sampling from multiple sites in the liver for the study of phenol metabolism. Life Sci., 59: 1001–1013. de Lange, E.C.M. and Danhof, M. (2002) Considerations in the use of cerebrospinal ﬂuid pharmacokinetics to predict brain target concentrations in the clinical setting: implications of the barriers between blood and brain. Clin. Pharmacokinet., 41: 691–703. de Lange, E.C.M., Danhof, M., de Boer, A.G. and Breimer, D.D. (1997) Methodological considerations of intracerebral microdialysis in pharmacokinetic studies on drug transport across the blood-brain barrier. Brain Res. Rev., 25: 27–49. de Lange, E.C.M., de Bock, G., Schinkel, A.H., de Boer, A.G. and Breimer, D.D. (1998) BBB transport and P-glycoprotein functionality using MDR1A (/) and wild-type mice. Total brain versus microdialysis concentration proﬁles of rhodamine-123. Pharm. Res., 15: 1657–1665. de Lange, E.C.M., de Boer, B.A. and Breimer, D.D. (1999) Microdialysis for pharmacokinetic analysis of drug transport to the brain. Adv. Drug Deliv. Rev., 36: 211–227. de Lange, E.C.M., de Boer, A.G. and Breimer, D.D. (2000a) Methodological issues in microdialysis sampling for pharmacokinetic studies. Adv. Drug Deliv. Rev., 45: 125–148.

de Lange, E.C.M., Marchand, S., van den Berg, D., van der Sandt, I.C., de Boer, A.G., Delon, A., Bouquet, S. and Couet, W. (2000b) In vitro and in vivo investigations on ﬂuoroquinolones; effects of the P-glycoprotein efﬂux transporter on brain distribution of sparﬂoxacin. Eur. J. Pharm. Sci., 12: 85–93. Eckblom, M., Hammarlund-Undenas, M., Lundqvist, T. and Sjoberg, P. (1992) Potential use of microdialysis in pharmacokinetics: a protein binding study. Pharm. Res., 9: 155–158. Ederoth, P., Tunblad, K., Bouw, R., Lundberg, C.J., Ungerstedt, U., Nordstrom, C.H. and Hammarlund-Udenaes, M. (2004) Blood-brain barrier transport of morphine in patients with severe brain trauma. Br. J. Clin. Pharmacol., 57: 427–435. Edwards, J.E., Brouwer, K.R. and McNamara, P.J. (2002) GF120918, a P-glycoprotein modulator, increases the concentration of unbound amprenavir in the central nervous system in rats. Antimicrob. Agents Chemother., 46: 2284–2286. Evrard, P.A., Cumps, J. and Verbeeck, R.K. (1996) Concentration-dependent plasma protein binding of ﬂurbiprofen in the rat: an in vivo microdialysis study. Pharm. Res., 13: 18–22. Ezzine, S., Yamaguchi, N. and Varin, F. (2004) Determination of interstitial rocuronium concentrations in the muscle tissue of anesthetized dogs by microdialysis. J. Pharmacol. Toxicol. Methods, 49: 121–129. Freddo, R.J. and Dalla Costa, T. (2002) Determination of norﬂoxacin free interstitial levels in skeletal muscle by microdialysis. J. Pharm. Sci., 91: 2433–2440. Funan, C., Zhujun, Z., Yingxue, Z. and Deyong, H. (2005) Microdialysis sampling and high-performance liquid chromatography with chemiluminescence detection for in-vivo on-line determination and study of the pharmacokinetics of levodopa in blood. Anal. Bioanal. Chem., 382: 211–215. Gunaratna, P.C., Kissinger, P.T., Kissinger, C.B. and Gitzen, J.F. (2004) An automated blood sampler for simultaneous sampling of systemic blood and brain microdialysates for drug absorption, distribution, metabolism, and elimination studies. J. Pharmacol. Toxicol. Methods, 49: 57–64. Haaparanta, M., Gronroos, T., Marjamaki, P., Eskola, O., Bergman, J., Paul, R. and Solin, O. (2004) In vivo sampling for pharmacokinetic studies in small experimental animals: a combination of microdialysis, planar chromatography and digital autoradiography. Mol. Imaging Biol., 6: 27–33. Haaparanta, M., Paul, R., Gronroos, T., Bergman, J., Kamarainen, E.L. and Solin, O. (2003) Microdialysis and 2-[18F]ﬂuoro-2-deoxy-D-glucose (FDG): a study on insulin action on FDG transport, uptake and metabolism in rat muscle, liver and adipose tissue. Life Sci., 73: 1437–1451. Hammarlund-Udenaes, M. (2000) The use of microdialysis in CNS drug delivery studies. Pharmacokinetic perspectives and results with analgesics and antiepileptics. Adv. Drug Deliv. Rev., 45: 283–294. Henriksson, J. and Knol, M. (2005) A single bout of exercise is followed by a prolonged decrease in the interstitial glucose concentration in skeletal muscle. Acta Physiol. Scand., 185: 313–320.

586 Herrera, A.M., Scott, D.O. and Lunte, C.E. (1990) Microdialysis sampling for determination of plasma protein binding of drugs. Pharm. Res., 7: 1077–1081. Huang, S.M., Tsai, T.R., Yeh, P.H. and Tsai, T.H. (2005) Measurement of unbound ranitidine in blood and bile of anesthetized rats using microdialysis coupled to liquid chromatography and its pharmacokinetic application. J. Chromatogr. A, 1073: 297–302. Jan, W.C., Lin, L.C., Chen, C.F. and Tsai, T.H. (2005) Herb–drug interaction of Evodia rutaecarpa extract on the pharmacokinetics of theophylline in rats. J. Ethnopharmacol., 102: 440–445. Kaptein, W.A., Zwaagstra, J.J., Venema, K. and Korf, J. (1998) Continuous ultraslow microdialysis and ultraﬁltration for subcutaneous sampling as demonstrated by glucose and lactate measurements in rats. Anal. Chem., 70: 4696–4700. Karjagin, J., Pahkla, R. and Starkopf, J. (2004) Perioperative penetration of metronidazole into muscle tissue: a microdialysis study. Eur. J. Clin. Pharmacol., 59: 809–813. Kennedy, R.T., Thompson, J.E. and Vickroy, T.W. (2002) In vivo monitoring of amino acids by direct sampling of brain extracellular ﬂuid at ultralow ﬂow rates and capillary electrophoresis. J. Neurosci. Methods, 114: 39–49. Lee, C.H., Yeh, P.H. and Tsai, T.H. (2002) Hepatobiliary excretion of ﬂuconazole and its interaction with cyclosporin A in rat blood and bile using microdialysis. Int. J. Pharm., 241: 367–373. Lin, M.C., Huang, Y.L., Liu, H.W., Yang, D.Y., Lee, J.B. and Cheng, F.C. (2004a) Microdialysis analyzer and ﬂame atomic absorption spectrometry in the determination of blood glucose, lactate and magnesium in gerbils subjected to cerebral ischemia/reperfusion. J. Am. Coll. Nutr., 23: 556S–560S. Lin, L.C., Hung, L.C. and Tsai, T.H. (2004b) Determination of ()-epigallocatechin gallate in rat blood by microdialysis coupled with liquid chromatography. J. Chromatogr. A, 1032: 125–128. Liu, P., Fuhrherr, R., Webb, A.I., Obermann, B. and Derendorf, H. (2005) Tissue penetration of cefpodoxime into the skeletal muscle and lung in rats. Eur. J. Pharm. Sci., 25: 439–444. Maas, S., Jarry, H., Teichmann, A., Rath, W., Kuhn, W. and Wuttke, W. (1992) Paracrine actions of oxytocin, prostaglandin F2 alpha, and estradiol within the human corpus luteum. J. Clin. Endocrinol. Metab., 74: 306–312. Maia, M.B., Saivin, S., Chatelut, E., Malmary, M.F. and Houin, G. (1996) In vitro and in vivo protein binding of methotrexate assessed by microdialysis. Int. J. Clin. Pharmacol. Ther., 34: 335–341. Marchand, S., Dahyot, C., Lamarche, I., Mimoz, O. and Couet, W. (2005) Microdialysis study of imipenem distribution in skeletal muscle and lung extracellular ﬂuids of noninfected rats. Antimicrob. Agents Chemother., 49: 2356–2361. Mathy, F.X., Lombry, C., Verbeeck, R.K. and Preat, V. (2005a) Study of the percutaneous penetration of ﬂurbiprofen by cutaneous and subcutaneous microdialysis after iontophoretic delivery in rat. J. Pharm. Sci., 94: 144–152.

Mathy, F.X., Ntivunwa, D., Verbeeck, R.K. and Preat, V. (2005b) Fluconazole distribution in rat dermis following intravenous and topical application: a microdialysis study. J. Pharm. Sci., 94: 770–780. Miyamoto, A. and Schams, D. (1991) Oxytocin stimulates progesterone release from microdialyzed bovine corpus luteum in vitro. Biol. Reprod., 44: 1163–1170. Morgan, C.J., Renwick, A.G. and Friedmann, P.S. (2003) The role of stratum corneum and dermal microvascular perfusion in penetration and tissue levels of water-soluble drugs investigated by microdialysis. Br. J. Dermatol., 148: 434–443. Nakashima, M., Takeuchi, N., Hamada, M., Matsuyama, K., Ichikawa, M. and Goto, S. (1994) In vivo microdialysis for pharmacokinetic investigations: a plasma protein binding study of valproate in rabbits. Biol. Pharm. Bull., 17: 1630–1634. Ofner, B., Boukhabza, A., Pancha, W., Amsterdam, C.V. and Winterseiger, R. (1997) Determination of SDZ ICM 567 in blood and muscle microdialysis samples by microbore liquid chromatography with ultraviolet and ﬂuorescence detection. J. Chromatogr. B, 700: 191–200. Plock, N. and Kloft, C. (2005) Microdialysis – theoretical background and recent implementation in applied life sciences. Eur. J. Pharm. Sci., 25: 1–24. Sakata, A., Tamai, I., Kawazu, K., Deguchi, Y., Ohnishi, T., Saheki, A. and Tsuji, A. (1994) In vivo evidence for ATPdependent and P-glycoprotein-mediated transport of cyclosporin A at the blood-brain barrier. Biochem. Pharmacol., 48: 1989–1992. Sarre, S., Van Belle, L., Smolders, I., Krieken, G. and Michotte, Y. (1992) The use of microdialysis for the determination of plasma protein binding of drugs. J. Pharm. Biomed. Anal., 10: 735–739. Sawchuk, R.J. and Elmquist, W.F. (2000) Microdialysis in the study of drug transporters in the CNS. Adv. Drug Deliv. Rev., 45: 295–307. Schoonen, A.J.M. and Wientjes, K.J.C. (2003) A model for transport of glucose in adipose tissue to a microdialysis probe. Diabetes Technol. Ther., 5: 589–598. Scott, D.O. and Lunte, C.E. (1993) In vivo microdialysis sampling in the bile, blood, and liver of rats to study the disposition of phenol. Pharm. Res., 10: 335–342. Stahle, L. (1993) Zero and ﬁrst moment area estimation from microdialysis data. Eur. J. Clin. Pharmacol., 45: 477–481. Stenken, J.A., Lunte, C.E., Southard, M.Z. and Stahle, L. (1997) Factors that inﬂuence microdialysis recovery. Comparison of experimental and theoretical microdialysis recoveries in rat liver. J. Pharm. Sci., 86: 958–966. Stetina, P.M., Madai, B., Kulemann, V., Kirch, W. and Joukhadar, C. (2005) Pharmacokinetics of scopolamine in serum and subcutaneous adipose tissue in healthy volunteers. Int. J. Clin. Pharmacol. Ther., 43: 134–139. Tsai, T.H. (2001) Pharmacokinetics of peﬂoxacin and its interaction with cyclosporin A, a P-glycoprotein modulator, in rat blood, brain and bile, using simultaneous microdialysis. Br. J. Pharmacol., 132: 1310–1316.

587 Tsai, T.H. (2003) Assaying protein-unbound drugs using microdialysis techniques. J. Chromatogr. B, 797: 161–173. Tsai, T.H. (2005) Concurrent measurement of unbound genistein in the blood, brain and bile of anesthetized rats using microdialysis and its pharmacokinetic application. J. Chromatogr. A, 1073: 317–322. Tsai, T.H. and Chen, Y.F. (2003) Pharmacokinetics of metronidazole in rat blood, brain and bile studied by microdialysis coupled to microbore liquid chromatography. J. Chromatogr. A, 987: 277–282. Tsai, T.H. and Wu, J.W. (2003) Regulation of hepatobiliary excretion of sinomenine by P-glycoprotein in SpragueDawley rats. Life Sci., 72: 2413–2426. Tsai, T.H., Chang, C.H. and Lin, L.C. (2005) Effects of Evodia rutaecarpa and rutaecarpine on the pharmacokinetics of caffeine in rats. Plant. Med., 71: 640–645. Tsai, T.H., Chen, Y.F., Chen, I.F. and Chen, C.F. (1999) Measurement of unbound caffeic acid in rat blood by on-line microdialysis coupled with liquid chromatography and its application to pharmacokinetic study. J. Chromatogr. B, 729: 119–125. Tsai, T.H., Lee, C.H. and Yeh, P.H. (2001) Effect of P-glycoprotein modulators on the pharmacokinetics of camptothecin using microdialysis. Br. J. Pharmacol., 134: 1245–1252. Tsai, T.H., Liu, S.C., Tsai, P.L., Ho, L.K., Shum, A.Y. and Chen, C.F. (2002) The effects of the cyclosporin A, a P-glycoprotein inhibitor, on the pharmacokinetics of baicalein in the rat: a microdialysis study. Br. J. Pharmacol., 137: 1314–1320. Tsai, T.H., Shum, A.Y. and Chen, C.F. (2000) Enterohepatic circulation of chloramphenicol and its glucuronide in the rat by microdialysis using a hepato-duodenal shunt. Life Sci., 66: 363–370. Tsai, P.L. and Tsai, T.H. (2004a) Hepatobiliary excretion of berberine. Drug Metab. Dispos., 32: 405–412. Tsai, P.L. and Tsai, T.H. (2004b) Pharmacokinetics of baicalin in rats and its interactions with cyclosporin A, quinidine and

SKF-525A: a microdialysis study. Planta Med., 70: 1069–1074. Tseng, T.Y. and Tsai, T.H. (2004) Measurement of unbound geniposide in blood, liver, brain and bile of anesthetized rats: an application of pharmacokinetic study and its inﬂuence by the acupuncture. Anal. Chim. Acta, 517: 47–52. Ungerstedt, U. (1991) Microdialysis – principles and applications for studies in animals and man. J. Intern. Med., 230: 365–373. Ungerstedt, U. and Pycock, C. (1974) Functional correlates of dopamine neurotransmission. Bull. Schweiz. Akad. Med. Wiss., 30: 44–55. Verbeeck, R.K. (2000) Blood microdialysis in pharmacokinetic and drug metabolism studies. Adv. Drug Deliv. Rev., 45: 217–228. Wu, Y.S., Tsai, T.H., Wu, T.F. and Cheng, F.C. (2001) Determination of pyruvate and lactate in primary liver cell culture medium during hypoxia by on-line microdialysisliquid chromatography. J. Chromatogr. A, 913: 341–347. Xie, R., Bouw, M.R. and Hammarlund-Udenaes, M. (2000) Modelling of the blood-brain barrier transport of morphine3-glucuronide studied using microdialysis in the rat: involvement of probenecid-sensitive transport. Br. J. Pharmacol., 131: 1784–1792. Xie, R., Hammarlund-Udenaes, M., de Boer, A.G. and de Lange, E.C. (1999) The role of P-glycoprotein in blood-brain barrier transport of morphine: transcortical microdialysis studies in mdr1a (/) and mdr1a (+/+) mice. Br. J. Pharmacol., 128: 563–568. Yokel, R.A., Lidums, V., McNamara, P.J. and Ungerstedt, U. (1991a) Aluminum distribution into brain and liver of rats and rabbits following intravenous aluminum lactate or citrate: a microdialysis study. Toxicol. Appl. Pharmacol., 107: 153–163. Yokel, R.A., Lidums, V. and Ungerstedt, U. (1991b) Aluminum mobilization by desferrioxamine assessed by microdialysis of the blood, liver and brain. Toxicology, 66: 313–324.

CHAPTER 6.6

Microdialysis for characterization of PK/PD relationships Margareta Hammarlund-Udenaes Division of Pharmacokinetics and Drug Therapy, Department of Pharmaceutical Biosciences, Faculty of Pharmacy, Uppsala University, Uppsala, Sweden

Abstract: The added value of using microdialysis in PK/PD research is that it gives an opportunity for direct measurement of drug concentrations close to the site of action and PD effects in parallel, thus furthering our understanding of relationships at the local target site. This area has a large potential that is not yet fully utilized. With the use of large pore size probes it may also be possible to measure larger molecules as biomarkers of drug action. Drug PK studies with microdialysis in tissues can also be combined with other types of PD measurements to obtain PK/PD relationships. Areas that beneﬁt from this approach are early drug development, where drug selection based on both PK and PD is crucial, and studies of the inﬂuence of disease states on drug distribution and effects. The brain as a target for drug action is in special focus due to the blood-brain barrier hindering drug transport. The main target areas studied so far are PK/PD relationships in brain of opioids and antiepileptics, and antibiotic distribution in muscle and adipose tissue during different conditions.

in the body. However, it is to date rare that the two measurements are combined in the same study. Combining PK and PD measurements provides valuable information on both quantitative and time aspects of drug action, and can provide a means for increasing the mechanistic understanding of the relationship. This is especially relevant in the ﬁeld of drug discovery and development where several compounds may need to be compared, and where it is important to understand the relationship between the distribution and action of the drug in question. The reason that progress in using this methodology has been slow might be that the traditional modes of working of PK researchers differ from those of researchers focusing on PD and pharmacology. This chapter covers the use of microdialysis in PK/PD studies and provides examples where appropriate.

I. Introduction The main advantage of microdialysis is that it allows measurement of unbound, pharmacologically active drug concentrations and pharmacodynamic (PD) biomarkers at a speciﬁc site such as the target organ for drug action. The technique is therefore of great value for pharmacokinetic/pharmacodynamic (PK/PD) research since it can give detailed time proﬁles of the analyte of interest within one animal/human. Microdialysis is extensively used for PD measurements of CNS transmitter activity after drug interventions. The method is also used for quantitative drug PK measurements at several locations

Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

589

DOI: 10.1016/S1569-7339(06)16031-2 Copyright 2007 Elsevier B.V. All rights reserved

590

II. Prerequisites for PK/PD measurements using microdialysis II.A. Quantitative aspects of microdialysis sampling Only a fraction of the drug or endogenous compound is captured from the tissue in the dialysate during microdialysis. The size of this fraction depends on the ﬂow rate through the probe (the lower the ﬂow rate, the higher the fraction) and the length of the probe (the longer the probe, the higher the fraction). It is also dependent on the tissue distribution properties of the solute of interest. Therefore, control or baseline recovery (extraction fraction) measurements are required to quantify the true unbound tissue concentration. Quantitative microdialysis is generally used in PK studies, while PD studies generally use relative dialysate concentrations. This may seem a hurdle if microdialysis techniques are to be used for combination PK and PD work, but is surmountable. Even if recovery measurements are desirable, the aim of the work may not actually require quantitative data. Mapping the change in dialysate concentrations from baseline may sufﬁce in transmitter studies. In vitro recovery measurements may be enough in early drug development and candidate selection studies. For academic publications, however, quantiﬁcation of the true extracellular concentrations of drugs is the goal. Justice and coworkers have made important contributions to the understanding and method development of quantitative neurotransmitter microdialysis (Justice, 1993; Olson and Justice, 1993; Cosford et al., 1994, 1996; Parsons and Justice, 1994; Smith and Justice, 1994; Thompson et al., 1995). They showed that certain interventions in the biological system alter the recovery of the analyte while other types of intervention do not. For example, the recovery of dopamine (DA) decreased extensively after the administration of an uptake inhibitor, but not after inhibition of synthesis, or intracellular or extracellular metabolism (Smith and Justice, 1994). This was also shown for serotonin and norepinephrine (Cosford et al., 1996). The authors concluded that only processes inﬂuencing the clearance of the solute of interest

from the site of measurement will result in changes in recovery. Recovery estimations can therefore be used to describe such processes. The recovery of drugs has also been shown to decrease after administration of compounds that block active efﬂux transport, such as probenecid or P-glycoprotein inhibitors (Wang et al., 1995; de Lange et al., 1998, 2000; Xie et al., 2000; Sun et al., 2001, 2003). Thus, dialysate concentrations, in vitro recovery, or recovery performed in vivo prior to an intervention might not provide the full picture, but might provide sufﬁcient information for some scenarios, depending on the question to be answered.

II.B. Site of measurement Plasma is the default site of measurement in PK studies. If tissue concentrations are required, whole tissue is generally sampled. Both of these types of samples include both unbound and bound drugs and, in whole tissue samples or PET measurements, include the average of extracellular, intracellular, bound, and unbound concentrations (Fig. 1). Microdialysis can add value to PK/PD studies by measuring the unbound concentrations Microdialysis blood sample

Microdialysis tissue sample

unbound

unbound

bound

bound

Plasma

Tissue Brain ISF ISF

Whole plasma sample

unbound

bound

Tissue cell

Whole tissue sample l

Fig. 1. Microdialysis sampling of unbound drug concentrations in blood and tissues makes this method suitable for PK/PD studies. Whole tissue and plasma samples also allow for inclusion of binding to proteins and cellular components in the models.

591

in blood or in the tissue of interest. This is suitable when the target is present in the extracellular compartment. Microdialysis does not measure intracellular concentrations and is therefore not suitable for PK/PD studies of drugs with intracellular targets. It can, however, be used to acquire information from closer to the site of action for this type of drug, and can also be combined with other methods such as PET (Langer et al., 2005). II.C. Size of molecules to be measured The cutoff point for the microdialysis probe with respect to molecular size will limit the molecular size of the analyte to be sampled. The larger the analyte molecules, the smaller is the resultant recovery. Most probes have a cutoff point of 20 kDa or less, which means that small molecules and most drugs are dialyzable. There are also probes with a cutoff of 100 kDa, which allow sampling of larger molecules. Experience with microdialysis detection of large molecules is increasing, with subsequent improvement in the possibility of measuring PD markers in tissues (Ao et al., 2005; Clough, 2005; Dabrosin, 2005; Hillman et al., 2005; Hutchinson et al., 2005). III. Concentration and time aspects of drug presence at the site of action in relation to PD III.A. Concentration aspects in tissue vs. blood It is usually assumed in PK studies that the unbound concentration of the molecule is the same throughout the body at steady state. This assumption has to be modiﬁed after microdialysis studies of unbound tissue concentrations. The discrepancies mainly concern tissues, such as the brain, with tight junctions and active transporters in their capillary walls. Microdialysis has the advantage of measuring the in vivo relationship between drug concentrations in different bodily ﬂuids. This is especially valuable as only a few transporters acting on drugs have been identiﬁed, and the importance of different efﬂux or inﬂux processes is not yet fully understood. The ratio of unbound concentration in brain interstitial ﬂuid (ISF) to

unbound concentration in blood at steady state (Kp,uu) quantiﬁes the combined effects of inﬂux and efﬂux system(s) at the blood-brain barrier (BBB) acting on the drug of interest (Hammarlund-Udenaes, 2000; Hammarlund-Udenaes et al., 1997; Gupta et al., 2006). If an understanding of which transporter is active is required, blockers of individual transporters need to be coadministered or the transporter needs to be studied separately in cell cultures. For some drugs, the concentrations in brain ISF are similar to the unbound concentration in blood (Kp,uu ¼ 1); for example, caffeine (Stahle et al., 1991), codeine (Xie and Hammarlund-Udenaes, 1998), and diazepam (Dubey et al., 1989). However, the majority of drugs studied quantitatively to date show lower or much lower concentrations in brain ISF than in blood. Examples are AZT (Wong et al., 1992), cetirizine (Gupta et al., 2006), ﬂuconazole (Yang et al., 1996), gabapentin (Welty et al., 1993), morphine (Bouw et al., 2000; Tunblad et al., 2003), morphine-6-glucuronide (Bouw et al., 2001; Tunblad et al., 2005), morphine-3-glucuronide (Xie et al., 2000), norﬂoxacin (de Lange et al., 2000; Chenel et al., 2004), and pemetrexed (Dai et al., 2005). The same relationship probably holds for other membranes with tight junctions and active efﬂux mechanisms, like the placenta and the testis barriers, although drug transport across these membranes has not yet been studied using microdialysis. The concentrations of drugs in other tissues have also been shown to differ from those in blood. For example, the unbound concentration of amoxicillin in the middle ear ﬂuid of the chinchilla is lower than in the blood (Huang et al., 2001). Perhaps less easily explained is the ﬁnding that antibiotic concentrations are lower in muscle tissue during sepsis than in healthy subjects (Brunner et al., 2000, 2002; Joukhadar et al., 2002) (see Section IV.C). Other pathologic conditions could also change the tissue-to-blood relationships with respect to drug concentrations. This will naturally have an impact on the resulting PD effects or side effects, and on the dose recommendations for successfully treating a certain disease. Drug interactions that block transporter function will increase the concentrations of the drug that is blocked. The interaction potential, and

592

Microdialysis has shown that the half-lives of several unbound drugs are longer in the brain than in blood. The half-life in the tissue will have an impact on the timing of drug effect in that tissue, and may explain time delays in PD results compared with results using plasma concentrations. It is, however, less likely that the half-lives of unbound drug will be longer in tissues with leaky capillaries, such as muscle, than in blood. The half-life in blood will ultimately be the same as that in the ‘‘slowest tissue’’. However, within a reasonable concentration range, there is a clear difference between brain and blood half-lives for several drugs. For example, half-lives are longer in the brain ISF than in blood for morphine in rats, pigs, and humans (Bouw et al., 2000; Tunblad et al., 2003, 2004; Ederoth et al., 2004), M6G (Bouw et al., 2001; Tunblad et al., 2005), M3G (Xie et al., 2000), cetirizine (Gupta et al., 2006), and tiagabine (Wang et al., 2004). The half-lives of other drugs [e.g., AZT (Wong et al., 1992) and norﬂoxacin (Chenel et al., 2004)] are similar in spite of differing absolute unbound concentrations. The local brain half-life is mainly determined by both the efﬂux clearance and the brain distribution volume (Syva¨nen et al., 2006).

(a)

(b)

Effect

III.B. Time aspects in tissue vs. blood

relationship with the drug effect. Slower evolvement of PD effects than those expected from the plasma concentration time proﬁle may be the result of several mechanisms; for example, transport of the drug to and from the site of action might be slow. As mentioned earlier, a longer half-life in the tissue may result in a longer lasting effect than that expected from the plasma concentrations (Wang et al., 2004). Further, the PD mechanism may result in a delayed effect once the drug is present at the site of action. A fourth possibility is that the PD effect is caused by an active metabolite rather than the parent drug (Weikop et al., 2004). The nature of the PK/PD relationship can be observed when plotting concentration vs. effect. If there is a direct relationship, that is, no time delay in effect, this can be observed as superimposable legs of the concentration–effect curve when the concentration is increasing and decreasing (Fig. 2a). A time delay in the effect shows as a counterclockwise ‘‘hysteresis,’’ that is, a curve where the two legs of increasing and decreasing concentrations are not superimposable (Fig. 2b). The development of tolerance would show as a clockwise ‘‘hysteresis’’. If the concentrations at the site of action are measured and there are no PD aspects to the delay, this should show as a collapse of the hysteresis to a direct relationship. This was studied regarding the delay in antinociceptive effect of morphine and M6G, where plasma and striatum ISF concentrations were measured simultaneously in the rat

Effect

therefore the risk of increased effects or side effects, will depend on the Kp,uu ratio. As the maximum possible Kp,uu is 1 with passive transport, a Kp,uu ratio of 0.1 gives an interaction potential of a 10-fold increase in local concentrations, while a ratio of 0.5 only gives an interaction potential of 2. A clinically relevant interaction will also depend on whether one or several transporters are acting on the drug. If P-glycoprotein is the major transporter, the risk of drug interactions with, for example, cyclosporine A or verapamil is more serious. If the drug is transported via several transporters, the risk decreases (Xie et al., 1999; Tunblad et al., 2003).

Concentration

III.C. PK/PD relationships Measurement of drug concentrations close to the site of action is expected to result in a direct

Concentration

Fig. 2. (a) Direct concentration–effect relationship where the legs of the curve are superimposable independently of whether concentrations increase or decrease. (b) Counter-clockwise hysteresis showing a delay in effect vs. measured concentrations.

593

(Bouw et al., 2000, 2001). This allowed for the separation of BBB transport processes and intrabrain transport and/or PD mechanisms as the cause of the delay in effect. The delay in morphine PD was shown to be mainly caused by the time taken for BBB transport (27 out of 32 min of the effect delay half-life). Thus, once morphine is present in the striatum ISF, there is still a delay before the effect, with a half-life of 5 min (Bouw et al., 2000). M6G is transported even more slowly across the BBB than morphine. Of the effect delay half-life of 1.4 h, 55 min was caused by BBB transport. As much as 50 min of the half-life was still unexplained after M6G reached the brain ISF (Bouw et al., 2001) (Fig. 3). The relationship between the BBB transport of norﬂoxacin and its convulsant side effects in rats has also been studied (Chenel et al., 2004). The authors found very similar time proﬁles for norﬂoxacin in brain ISF and in blood, but much lower unbound concentrations in brain ISF (9.7%). The EEG effect was signiﬁcantly delayed in relation to brain ISF concentration, and could therefore not be explained by slow BBB transport. The effect delay half-life was found to be 86 min, probably the result of intrabrain PD factors that take time to develop. If a PK/PD relationship is to be studied with respect to its timing, it is advisable that the

Antinociceptive effect (V)

7 6 5

striatum

4

blood 3 2 10

100 1000 Concentration (ng/ml)

10 000

Fig. 3. Morphine PK/PD relationship studied using microdialysis based on unbound concentrations in blood or brain (striatum). It can be observed that the wide counterclockwise hysteresis is decreased when striatum concentrations are plotted vs. the antinociceptive effect. From Bouw et al. (2000) with permission from the publisher.

administration of drug is based on both increases and decreases in concentration with time. The design could be an infusion lasting for three to four microdialysis sampling intervals combined with continued sampling after the end of the infusion.

IV. Examples of PK/PD studies with microdialysis IV.A. Method development An animal model suitable for PK/PD investigations has been developed by Pan and Hedaya (1998). The model allows drug administration via different routes, serial blood sampling, serial brain ISF sampling through a microdialysis probe, and monitoring the cardiovascular functions without touching the animal. The authors used this experimental setup to study the PK and neurochemical and cardiovascular relationships of cocaine after administration via different routes. Cocaine rapidly distributed to the brain with a Kp,uu of 2. This high ratio indicates active inﬂux of cocaine at the BBB. DA levels were in direct relationship with cocaine brain ISF concentrations. Hysteresis loops were observed when the cardiovascular effects were plotted vs. plasma cocaine concentrations, indicating an indirect relationship. The authors concluded that the model was useful for simultaneous PK/PD investigations, especially of centrally acting drugs. A combined in vivo PK plus in vitro PD approach has been suggested for antibiotic drugs (Johansen et al., 1997; Delacher et al., 2000; Brunner et al., 2005). Microdialysis can be used to measure interstitial drug concentrations at the target site for PK analysis. The concentration–time proﬁle is then simulated in an in vitro system in which PD measurements of bacterial growth are made (see also Section IV.C). Microdialysis combined with mechanism-based PK/PD modeling has been suggested by de Lange et al. (2005) as a method of predicting CNS drugeffect proﬁles. The mechanism-based PK/PD approach can separate drug-related parameters from physiologic parameters such as BBB functionality that may change during disease, which is important when studying pathologic conditions.

594

A triple-probe application has been suggested for fast PK/PD evaluation of dopamimetic activity by candidates for drug development (Weikop et al., 2004). One probe was placed in the blood of anesthetized rats, the second probe was placed in the brain for drug measurements, and the third probe was placed in the contralateral part of the brain for neurotransmitter measurement. In this manner, the rate and extent of extracellular drug accumulation could be measured in the brain and compared with DA levels in vivo. This approach is also applicable to other PK/PD relationships. Langer et al. (2005) utilized combined PET and microdialysis for in vivo assessment of muscle intracellular 18F-ciproﬂoxacin PK in humans. PET was used because microdialysis cannot be used for intracellular measurements. These workers found an intracellular-to-extracellular concentration ratio of 3, indicating sustained intracellular retention of ciproﬂoxacin. It is a drug which has the advantage for PET studies of moderate protein binding, a low degree of metabolism, and low unspeciﬁc binding. The uptake of ciproﬂoxacin was very fast, while efﬂux from the intracellular space was much slower.

IV.B. Brain PK/PD IV.B.1. Antiepileptic tissue distribution and PK/ PD Dual-probe microdialysis has been used to study the anticonvulsant activity of oxcarbazepine in relation to local brain concentrations of the drug in rats (Clinckers et al., 2005). Hippocampal monoamines DA and 5-HT were measured as PD markers. Pilocarpine-induced seizures were measured in parallel. The contralateral hippocampus was used as the control. Systemic administration of oxcarbazepine alone did not prevent the animals from developing seizures. When verapamil and probenecid were coadministered via the ipsilateral probe, oxcarbazepine was active on hippocampal DA and 5-HT levels. Without these drugs no effect was registered. The authors found that oxcarbazepine was a substrate for several transporters at the BBB, which hindered oxcarbazepine transport into the brain. Unfortunately, the brain ISF vs. effect

relationships were not studied. The oxcarbazepine hippocampal vs. plasma ratio increased with time, indirectly indicating a longer half-life in brain ISF than in plasma. In another investigation, Graumlich et al. (1999) studied carbamazepine PK-PD in controls and two strains of epilepsy-prone rats. Serotonin release from the hippocampus was used as the PD marker. As a result of microdialysis measurements of unbound carbamazepine in brain and plasma, the different responses in the three models were found to be the result of PK differences in plasma rather than PD differences. Plots of concentration vs. effect were not made in this study. The possibility of simultaneous measurement of vigabatrine and glutamate, L-aspartate, and GABA was presented by Benturquia et al. (2004), and this investigation showed rapid responses after intracerebral infusion of vigabatrine.

IV.B.2. Opioid tissue distribution and PK/PD Morphine and M6G have similar Kp,uu across the BBB in spite of a 10-fold difference in inﬂux permeability clearance in the rat (Bouw et al., 2000, 2001; Tunblad et al., 2003, 2005). This indicates that the mechanisms behind the extent of drug equilibration between brain ISF and blood are controlled in a different manner from that controlling the rate of drug delivery to the brain (Hammarlund-Udenaes, 2000; Hammarlund-Udenaes et al., 1997; Syva¨nen et al., 2006). It also shows that potency estimations for morphine and M6G will give very different results if the studies are based on total brain concentrations or unbound brain concentrations. Total brain-to-blood concentration ratios were 0.74 vs. 0.05 for morphine and M6G, while the unbound ratios were 0.29 and 0.27 (Stain-Texier et al., 1999; Bouw et al., 2000, 2001; Tunblad et al., 2003, 2005). The PK/PD relationship of morphine and M6G is discussed further in Section III.C. Letrent et al. (1999) developed a PK/PD model to describe the increased antinociceptive effect observed after coadministration of morphine and the P-glycoprotein blocking agent GF120918 to rats. They showed that the increased effect was fully explained by the increased brain ISF

595

concentrations of morphine. They also found that the half-life of unbound morphine in brain ISF was threefold longer after GF120918 administration. Other experiments found that coadministration of probenecid with morphine gave a longer half-life of morphine in the brain ISF (Tunblad et al., 2003); the metabolite M3G was inﬂuenced by GF120918 in both blood and brain (Letrent et al., 1999); in P-glycoprotein knockout mice, morphine but not M3G PK were changed (Xie et al., 1999); PSC833 increased spinal M6G and antinociceptive effects (Lotsch et al., 2002); and probenecid did not inﬂuence M6G transport into the brain, and therefore did not affect the PD (Tunblad et al., 2005).

IV.B.3. Other brain targets Raje et al. (2005) studied the PK/PD effects of two benztropine analogs, ANH-1055 and ANH-2005, in comparison with cocaine, on DA response. The DA response to these analogs was profound and long lasting. Modeling was based on plasma concentrations from an earlier study. No brain concentrations of the compounds were determined. Thus, it is still uncertain whether the delay in the effect of ANH-1055 and ANH-2005 was due to slow penetration across the BBB or to PD effects at the site of action. Heinzen and Pollack (2003) have studied nitric oxide (NO) production using a sensitive and speciﬁc amperometric sensor placed in the hippocampus in rats, while simultaneously measuring L-arginine by microdialysis in brain and blood. The NO precursor L-arginine was infused intravenously. The data were ﬁtted to a PK/PD model that included the brain uptake of L-arginine and explained its saturable uptake into the brain, and also explained the saturable formation of NO. In another study, metoprolol was administered intravenously and in the anterior hypothalamus via a microdialysis probe to distinguish its central vs. peripheral effects on blood pressure in control and aortic coarctated rats (Hocht et al., 2005a, b). The hypothalamic concentrations were similar for the two parts of the study. Dihydroxyphenyl acetic acid (DOPAC) and metoprolol were measured from the same microdialysis samples by dividing

the samples in two. The authors found that metoprolol rapidly reached the CNS. They suggested that metoprolol also blocks hypothalamic betaadrenoceptors as part of its antihypertensive effect, and that there is a reduction of DA turnover in hypertensive animals. A combination of two PD measurements, locomotor activity, and DA levels in brain, was combined with measurements of methylphenidate enantiomers in another study (Aoyama et al., 1996). A clockwise hysteresis was observed between methylphenidate concentrations and DA in brain ISF. The locomotor activity induced by methylphenidate was related to the increased DA levels. It was also found that (+)-methylphenidate was more active than the (–)-isomer.

IV.C. Antibiotic tissue distribution and PK/PD It has been assumed that peripheral tissue concentration–time proﬁles of antibiotics closely follow the unbound plasma concentrations with time. Microdialysis measurements in human muscle and adipose tissue have, however, shown different behaviors for different drugs. For example, moxiﬂoxacin in healthy volunteers showed ratios of 0.8–0.9 between the drug in muscle ISF and unbound drug in blood (Muller et al., 1999b); levoﬂoxacin entered the ISF with a ratio to plasma of 0.85 (Zeitlinger et al., 2003); and the fosfomycin ratio was 0.75 in septic patient ISF compared with plasma (Joukhadar et al., 2003). Metronidazole took 2–3 h to equilibrate between muscle tissue and plasma in septic shock patients, but showed similar concentrations after that (Karjagin et al., 2005). This indicates that unbound plasma concentrations are representative of active muscle tissue concentrations. The in vivo recovery was very variable (24–82%). In other studies, the unbound imipenem ratio was 1 in rat muscle tissue compared with blood (Marchand et al., 2005b), and the amoxicillin ratio was 0.8–0.9 (Marchand et al., 2005a). Ciproﬂoxacin muscle interstitiumto-serum concentration ratios ranged from 0.55 to 0.73 (Brunner et al., 1999). In a few disease states, it has been found that some antibiotics have lower concentrations in

596

peripheral tissue than those of the unbound drug in blood. As peripheral capillaries are rather leaky, these ﬁndings are difﬁcult to explain. In spite of sometimes lower concentrations of the antibiotic or antifungal agent, in vitro PD has shown that the simulated concentration–time proﬁles provide levels that are sufﬁcient to prevent bacterial or fungal growth. The area under the curve (AUC) of cefpirome in subcutaneous tissue from septic patients was only half that in tissue from healthy volunteers (Sauermann et al., 2005). However, effective bacterial growth inhibition was observed in all in vitro simulations. This was attributed to the signiﬁcantly prolonged half-life in tissue, which kept the tissue cefpirome levels above the MICs for relevant pathogens for extended periods in the septic group. The concentrations of cefpirome that were reached in the tissue interstitium and plasma exceeded minimal inhibitory concentrations for most clinically relevant pathogens in patients with sepsis and were no different from those in healthy volunteers (Joukhadar et al., 2002). Piperacillin interstitium to serum concentration ratios ranged from 0.25 to 0.27 in patients undergoing aortic valve replacement, and from 0.43 to 1.22 in volunteers (Brunner et al., 2000). In septic shock patients, interstitial piperacillin concentrations in skeletal muscle and subcutaneous adipose tissue were ﬁve- to tenfold lower than corresponding free plasma concentrations (po0.03) (Joukhadar et al., 2001). The in vitro simulations of piperacillin showed that bacterial killing may be effective in severely ill patients despite relatively low concentrations of piperacillin at the target site (Sauermann et al., 2003). This ﬁnding is thought to be the result of impaired renal function in patients in intensive care, with subsequent prolonged tissue and plasma half-lives for piperacillin. Ciproﬂoxacin distribution was studied in inﬂamed foot lesions in non-insulin-dependent diabetes mellitus patients (Muller et al., 1999a). Interstitial ciproﬂoxacin concentrations were signiﬁcantly lower than the corresponding serum concentrations, but there was no signiﬁcant difference in the penetration of ciproﬂoxacin into inﬂamed and unaffected tissue. Together with the ﬁnding by Langer et al. (2005) of extensive

intracellular accumulation for this drug, these results provide valuable information on the clinical use of ciproﬂoxacin. Gemiﬂoxacin had signiﬁcantly higher concentrations in soft tissue than the unbound concentrations in the plasma of healthy volunteers (Islinger et al., 2004). The ratios of the mean AUC for tissue to the AUC for free gemiﬂoxacin in plasma were 1.7 for skeletal muscle and 2.4 for adipose tissue. Again, this is difﬁcult to explain based on a free movement of unbound drug molecules across capillary membranes. IV.D. Other applications IV.D.1. Ocular microdialysis Microdialysis in the vitreous humor of the eye can give valuable information on drug behavior and PD markers, including active transport, clearance mechanisms and the inﬂuence of dosage forms on the active drug concentrations (Rittenhouse and Pollack, 2000; Anand et al., 2004; Duvvuri et al., 2005). IV.D.2. Cutaneous microdialysis Cutaneous microdialysis is dramatically improving data collection after dermal or subcutaneous injections in humans. Kopacz et al. (2003) showed the impact of subcutaneous microcapsules on the release of and duration of local anesthetic effects for bupivacaine. Kreilgaard et al. (2001) also demonstrated the advantages of using microdialysis for dermal drug delivery studies including simultaneous PD measurements. IV.D.3. Mapping pathophysiology The penetration of morphine into the brain is increased in brain trauma patients (Ederoth et al., 2004). Probes placed in more damaged areas showed a clear tendency towards higher drug concentrations than those placed in more healthy sites. This has implications for drug dosing. Experimentally induced meningitis in pigs abolished the BBB function for morphine transport and the unbound brain concentrations were similar to those in blood (Tunblad et al., 2004). The reason

597

for this may have been increased permeability, decreased active efﬂux mechanisms, or both. The study of neuroprotective agents in traumatic brain injury has historically been problematic because it has not been possible to measure the actual penetration in the traumatized human brain (Alves et al., 2003). The authors of a study of the evaluation of the bioavailability and effects of topiramate used cerebral microdialysis. Topiramate crossed the BBB in neuroprotective concentrations and was demonstrated to lower glutamate levels. They concluded that the strategy is a clear improvement compared to traditional clinical trial design, and that it will reduce trial costs. PET and microdialysis were combined in studying patient response after middle cerebral artery infarction in humans (Dohmen et al., 2003). The authors concluded that neuromonitoring helped to classify the clinical course by characterizing pathophysiologic markers such as the lactate:pyruvate ratio, but that it did not predict fatal outcome as early as PET.

for brain penetration and CNS drug action or to avoid such effects. The technique has also proven valuable at other target sites. Abbreviations PK PD M6G Brain ISF BBB CNS NO 5-HT 5-HIAA DOPAC DA HVA Cu PET GABA

pharmacokinetic pharmacodynamic morphine-6-glucuronide brain interstitial ﬂuid blood-brain barrier central nervous system nitric oxide serotonin hydroxyindole acetic acid dihydroxyphenyl acetic acid dopamine homovanillic acid unbound (not protein bound) concentration positron emission tomography gamma amino butyric acid

V. Conclusions The added value of using microdialysis in PK/PD research is that there is an opportunity for direct measurement of drug concentrations close to the site of action and PD effects, thus furthering our understanding of relationships at the local target site. It allows measurements of unbound drug at sites that have not previously been possible to measure, such as the eye or local sites in the brain. It also allows for local drug administration via the probe. Further, drug transport processes can be separated from PD processes; pathologic inﬂuences on drug distribution or PD biomarkers can be mapped; and sites of drug action can be distinguished after drug delivery to different locations. The number of publications addressing PK/PD relationships using microdialysis is increasing. There is much to gain from using this technique; for example, in early drug development, the distribution and PD properties of new drug candidates can be distinguished. This is especially valuable in brain research, where it is essential either to obtain the best possible drug candidates

References Alves, O.L., Doyle, A.J., Clausen, T., Gilman, C. and Bullock, R. (2003) Evaluation of topiramate neuroprotective effect in severe TBI using microdialysis. Ann. N.Y. Acad. Sci., 993: 25–34 discussion 48–53. Anand, B.S., Atluri, H. and Mitra, A.K. (2004) Validation of an ocular microdialysis technique in rabbits with permanently implanted vitreous probes: systemic and intravitreal pharmacokinetics of ﬂuorescein. Int. J. Pharm., 281: 79–88. Ao, X., Rotundo, R.F., Loegering, D.J. and Stenken, J.A. (2005) In vivo microdialysis sampling of cytokines produced in mice given bacterial lipopolysaccharide. J. Microbiol. Methods, 62: 327–336. Aoyama, T., Kotaki, H., Sawada, Y. and Iga, T. (1996) Pharmacokinetics and pharmacodynamics of methylphenidate enantiomers in rats. Psychopharmacology (Berl.), 127: 117–122. Benturquia, N., Parrot, S., Sauvinet, V., Renaud, B. and Denoroy, L. (2004) Simultaneous determination of vigabatrin and amino acid neurotransmitters in brain microdialysates by capillary electrophoresis with laser-induced ﬂuorescence detection. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 806: 237–244. Bouw, M.R., Ga˚rdmark, M. and Hammarlund-Udenaes, M. (2000) Pharmacokinetic-pharmacodynamic modelling of

598 morphine transport across the blood-brain barrier as a cause of the antinociceptive effect delay in rats – a microdialysis study. Pharm. Res., 17: 1220–1227. Bouw, M.R., Xie, R., Tunblad, K. and HammarlundUdenaes, M. (2001) Blood-brain barrier transport and brain distribution of morphine-6-glucuronide in relation to the antinociceptive effect in rats – pharmacokinetic/ pharmacodynamic modelling. Br. J. Pharmacol., 134: 1796–1804. Brunner, M., Derendorf, H. and Muller, M. (2005) Microdialysis for in vivo pharmacokinetic/pharmacodynamic characterization of anti-infective drugs. Curr. Opin. Pharmacol., 5: 495–499. Brunner, M., Hollenstein, U., Delacher, S., Jager, D., Schmid, R., Lackner, E., Georgopoulos, A., Eichler, H.G. and Muller, M. (1999) Distribution and antimicrobial activity of ciproﬂoxacin in human soft tissues. Antimicrob. Agents Chemother., 43: 1307–1309. Brunner, M., Pernerstorfer, T., Mayer, B.X., Eichler, H.G. and Muller, M. (2000) Surgery and intensive care procedures affect the target site distribution of piperacillin. Crit. Care Med., 28: 1754–1759. Brunner, M., Stabeta, H., Moller, J.G., Schrolnberger, C., Erovic, B., Hollenstein, U., Zeitlinger, M., Eichler, H.G. and Muller, M. (2002) Target site concentrations of ciproﬂoxacin after single intravenous and oral doses. Antimicrob. Agents Chemother., 46: 3724–3730. Chenel, M., Marchand, S., Dupuis, A., Lamarche, I., Paquereau, J., Pariat, C. and Couet, W. (2004) Simultaneous central nervous system distribution and pharmacokinetic-pharmacodynamic modelling of the electroencephalogram effect of norﬂoxacin administered at a convulsant dose in rats. Br. J. Pharmacol., 142: 323–330. Clinckers, R., Smolders, I., Meurs, A., Ebinger, G. and Michotte, Y. (2005) Hippocampal dopamine and serotonin elevations as pharmacodynamic markers for the anticonvulsant efﬁcacy of oxcarbazepine and 10,11-dihydro-10-hydroxycarbamazepine. Neurosci. Lett., 390: 48–53. Clough, G.F. (2005) Microdialysis of large molecules. AAPS J., 7: E686–E692. Cosford, R.J., Parsons, L.H. and Justice, J.B. Jr. (1994) Effect of tetrodotoxin and potassium infusion on microdialysis extraction fraction and extracellular dopamine in the nucleus accumbens. Neurosci. Lett., 178: 175–178. Cosford, R.J., Vinson, A.P., Kukoyi, S. and Justice, J.B. Jr. (1996) Quantitative microdialysis of serotonin and norepinephrine: pharmacological inﬂuences on in vivo extraction fraction. J. Neurosci. Methods, 68: 39–47. Dabrosin, C. (2005) Microdialysis – an in vivo technique for studies of growth factors in breast cancer. Front. Biosci., 10: 1329–1335. Dai, H., Chen, Y., Elmquist, W.F., Yang, H., Wang, Q. and Elmquist, W.F. (2005) Distribution of the novel antifolate pemetrexed to the brain. J. Pharmacol. Exp. Ther., 315: 222–229. de Lange, E.C., de Bock, G., Schinkel, A.H., de Boer, A.G. and Breimer, D.D. (1998) BBB transport and P-glycoprotein

functionality using MDR1A (–/–) and wild-type mice. Total brain versus microdialysis concentration proﬁles of rhodamine-123. Pharm. Res., 15: 1657–1665. de Lange, E.C., Marchand, S., van den Berg, D., van der Sandt, I.C., de Boer, A.G., Delon, A., Bouquet, S. and Couet, W. (2000) In vitro and in vivo investigations on ﬂuoroquinolones; effects of the P-glycoprotein efﬂux transporter on brain distribution of sparﬂoxacin. Eur. J. Pharm. Sci., 12: 85–93. de Lange, E.C., Ravenstijn, P.G., Groenendaal, D. and van Steeg, T.J. (2005) Toward the prediction of CNS drug-effect proﬁles in physiological and pathological conditions using microdialysis and mechanism-based pharmacokinetic-pharmacodynamic modeling. AAPS J., 7: E532–E543. Delacher, S., Derendorf, H., Hollenstein, U., Brunner, M., Joukhadar, C., Hofmann, S., Georgopoulos, A., Eichler, H.G. and Muller, M. (2000) A combined in vivo pharmacokinetic–in vitro pharmacodynamic approach to simulate target site pharmacodynamics of antibiotics in humans. J. Antimicrob. Chemother., 46: 733–739. Dohmen, C., Bosche, B., Graf, R., Staub, F., Kracht, L., Sobesky, J., Neveling, M., Brinker, G. and Heiss, W.D. (2003) Prediction of malignant course in MCA infarction by PET and microdialysis. Stroke, 34: 2152–2158. Dubey, R.K., McAllister, C.B., Inoue, M. and Wilkinson, G.R. (1989) Plasma binding and transport of diazepam across the blood-brain barrier. No evidence for in vivo enhanced dissociation. J. Clin. Invest., 84: 1155–1159. Duvvuri, S., Rittenhouse, K.D. and Mitra, A.K. (2005) Microdialysis assessment of drug delivery systems for vitreoretinal targets. Adv. Drug. Deliv. Rev., 57: 2080–2091. Ederoth, P., Tunblad, K., Bouw, R., Lundberg, C.J., Ungerstedt, U., Nordstrom, C.H. and Hammarlund-Udenaes, M. (2004) Blood-brain barrier transport of morphine in patients with severe brain trauma. Br. J. Clin. Pharmacol., 57: 427–435. Graumlich, J.F., McLaughlin, R.G., Birkhahn, D., Shah, N., Burk, A., Jobe, P.C. and Dailey, J.W. (1999) Carbamazepine pharmacokinetics-pharmacodynamics in genetically epilepsyprone rats. Eur. J. Pharmacol., 369: 305–311. Gupta, A., Chatelain, P., Massingham, R., Jonsson, E.N. and Hammarlund-Udenaes, M. (2006) Brain distribution of cetirizine enantiomers: comparison of three different tissueto-plasma partition coefﬁcients: K(p), K(p,u), and K(p,uu). Drug. Metab. Dispos., 34: 318–323. Hammarlund-Udenaes, M. (2000) The use of microdialysis in CNS drug delivery studies. Pharmacokinetic perspectives and results with analgesics and antiepileptics. Adv. Drug Deliv. Rev., 45: 283–294. Hammarlund-Udenaes, M., Paalzow, L.K. and de Lange, E.C. (1997) Drug equilibration across the blood-brain barrier – pharmacokinetic considerations based on the microdialysis method. Pharm. Res., 14: 128–134. Heinzen, E.L. and Pollack, G.M. (2003) Pharmacokinetics and pharmacodynamics of L-arginine in rats: a model of stimulated neuronal nitric oxide synthesis. Brain Res., 989: 67–75. Hillman, J., Aneman, O., Anderson, C., Sjogren, F., Saberg, C. and Mellergard, P. (2005) A microdialysis technique for

599 routine measurement of macromolecules in the injured human brain. Neurosurgery, 56: 1264–1268 discussion 1268–1270. Hocht, C., DiVerniero, C., Opezzo, J.A., Taira, C.A., Hocht, C., Opezzo, J.A. and Taira, C.A. (2005a) Applicability of microdialysis as a technique for pharmacokinetic-pharmacodynamic (PK-PD) modeling of antihypertensive beta-blockers. J. Pharmacol. Toxicol. Methods, 52: 244–250. Hocht, C., Opezzo, J.A. and Taira, C.A. (2005b) Hypothalamic antihypertensive effect of metoprolol in chronic aortic coarctated rats. Clin. Exp. Pharmacol. Physiol., 32: 681–686. Huang, Y., Ji, P., Inano, A., Yang, Z., Giebink, G.S. and Sawchuk, R.J. (2001) Microdialysis studies of the middle ear distribution kinetics of amoxicillin in the awake chinchilla. J. Pharm. Sci., 90: 2088–2098. Hutchinson, P.J., O’Connell, M.T., Nortje, J., Smith, P., AlRawi, P.G., Gupta, A.K., Menon, D.K. and Pickard, J.D. (2005) Cerebral microdialysis methodology – evaluation of 20 kDa and 100 kDa catheters. Physiol. Meas., 26: 423–428. Islinger, F., Bouw, R., Stahl, M., Lackner, E., Zeleny, P., Brunner, M., Muller, M., Eichler, H.G. and Joukhadar, C. (2004) Concentrations of gemiﬂoxacin at the target site in healthy volunteers after a single oral dose. Antimicrob. Agents Chemother., 48: 4246–4249. Johansen, M.J., Newman, R.A. and Madden, T. (1997) The use of microdialysis in pharmacokinetics and pharmacodynamics. Pharmacotherapy, 17: 464–481. Joukhadar, C., Frossard, M., Mayer, B.X., Brunner, M., Klein, N., Siostrzonek, P., Eichler, H.G. and Muller, M. (2001) Impaired target site penetration of beta-lactams may account for therapeutic failure in patients with septic shock. Crit. Care Med., 29: 385–391. Joukhadar, C., Klein, N., Dittrich, P., Zeitlinger, M., Geppert, A., Skhirtladze, K., Frossard, M., Heinz, G. and Muller, M. (2003) Target site penetration of fosfomycin in critically ill patients. J. Antimicrob. Chemother., 51: 1247–1252. Joukhadar, C., Klein, N., Mayer, B.X., Kreischitz, N., DelleKarth, G., Palkovits, P., Heinz, G. and Muller, M. (2002) Plasma and tissue pharmacokinetics of cefpirome in patients with sepsis. Crit. Care Med., 30: 1478–1482. Justice, J.B. Jr. (1993) Quantitative microdialysis of neurotransmitters. J. Neurosci. Methods, 48: 263–276. Karjagin, J., Pahkla, R., Karki, T. and Starkopf, J. (2005) Distribution of metronidazole in muscle tissue of patients with septic shock and its efﬁcacy against Bacteroides fragilis in vitro. J. Antimicrob. Chemother., 55: 341–346. Kopacz, D.J., Bernards, C.M., Allen, H.W., Landau, C., Nandy, P., Wu, D. and Lacouture, P.G. (2003) A model to evaluate the pharmacokinetic and pharmacodynamic variables of extended-release products using in vivo tissue microdialysis in humans: bupivacaine-loaded microcapsules. Anesth. Analg., 97: 124–131. Kreilgaard, M., Kemme, M.J., Burggraaf, J., Schoemaker, R.C. and Cohen, A.F. (2001) Inﬂuence of a microemulsion vehicle on cutaneous bioequivalence of a lipophilic model drug assessed by microdialysis and pharmacodynamics. Pharm. Res., 18: 593–599.

Langer, O., Karch, R., Muller, U., Dobrozemsky, G., Abrahim, A., Zeitlinger, M., Lackner, E., Joukhadar, C., Dudczak, R., Kletter, K., Muller, M. and Brunner, M. (2005) Combined PET and microdialysis for in vivo assessment of intracellular drug pharmacokinetics in humans. J. Nucl. Med., 46: 1835–1841. Letrent, S.P., Pollack, G.M., Brouwer, K.R. and Brouwer, K.L. (1999) Effects of a potent and speciﬁc P-glycoprotein inhibitor on the blood-brain barrier distribution and antinociceptive effect of morphine in the rat. Drug Metab. Dispos., 27: 827–834. Lotsch, J., Schmidt, R., Vetter, G., Schmidt, H., Niederberger, E., Geisslinger, G. and Tegeder, I. (2002) Increased CNS uptake and enhanced antinociception of morphine-6-glucuronide in rats after inhibition of P-glycoprotein. J. Neurochem., 83: 241–248. Marchand, S., Dahyot, C., Lamarche, I., Mimoz, O. and Couet, W. (2005a) Microdialysis study of imipenem distribution in skeletal muscle and lung extracellular ﬂuids of noninfected rats. Antimicrob. Agents Chemother., 49: 2356–2361. Marchand, S., Chenel, M., Lamarche, I. and Couet, W. (2005b) Pharmacokinetic modeling of free amoxicillin concentrations in rat muscle extracellular ﬂuids determined by microdialysis. Antimicrob. Agents Chemother., 49: 3702–3706. Muller, M., Stass, H., Brunner, M., Moller, J.G., Lackner, E. and Eichler, H.G. (1999a) Penetration of moxiﬂoxacin into peripheral compartments in humans. Antimicrob. Agents Chemother., 43: 2345–2349. Muller, M., Brunner, M., Hollenstein, U., Joukhadar, C., Schmid, R., Minar, E., Ehringer, H. and Eichler, H.G. (1999b) Penetration of ciproﬂoxacin into the interstitial space of inﬂamed foot lesions in non-insulin-dependent diabetes mellitus patients. Antimicrob. Agents Chemother., 43: 2056–2058. Olson, R. and Justice, J. (1993) Quantitative microdialysis under transient conditions. Anal. Chem., 65: 1017–1022. Pan, W.J. and Hedaya, M.A. (1998) An animal model for simultaneous pharmacokinetic/pharmacodynamic investigations: application to cocaine. J. Pharmacol. Toxicol. Methods, 39: 1–8. Parsons, L.H. and Justice, J.B. Jr. (1994) Quantitative approaches to in vivo brain microdialysis. Crit. Rev. Neurobiol., 8: 189–220. Raje, S., Cornish, J., Newman, A.H., Cao, J., Katz, J.L. and Eddington, N.D. (2005) Pharmacodynamic assessment of the benztropine analogues AHN-1055 and AHN-2005 using intracerebral microdialysis to evaluate brain dopamine levels and pharmacokinetic/pharmacodynamic modeling. Pharm. Res., 22: 603–612. Rittenhouse, K.D. and Pollack, G.M. (2000) Microdialysis and drug delivery to the eye. Adv. Drug Deliv. Rev., 45: 229–241. Sauermann, R., Delle-Karth, G., Marsik, C., Steiner, I., Zeitlinger, M., Mayer-Helm, B.X., Georgopoulos, A., Muller, M. and Joukhadar, C. (2005) Pharmacokinetics and pharmacodynamics of cefpirome in subcutaneous adipose tissue of septic patients. Antimicrob. Agents Chemother., 49: 650–655.

600 Sauermann, R., Zeitlinger, M., Erovic, B.M., Marsik, C., Georgopoulos, A., Muller, M., Brunner, M. and Joukhadar, C. (2003) Pharmacodynamics of piperacillin in severely ill patients evaluated by using a PK/PD model. Int. J. Antimicrob. Agents, 22: 574–578. Smith, A.D. and Justice, J.B. (1994) The effect of inhibition of synthesis, release, metabolism and uptake on the microdialysis extraction fraction of dopamine. J. Neurosci. Methods, 54: 75–82. Stahle, L., Segersvard, S. and Ungerstedt, U. (1991) Drug distribution studies with microdialysis. II. Caffeine and theophylline in blood, brain and other tissues in rats. Life Sci., 49: 1843–1852. Stain-Texier, F., Boschi, G., Sandouk, P. and Scherrmann, J.M. (1999) Elevated concentrations of morphine 6-beta-D-glucuronide in brain extracellular ﬂuid despite low blood-brain barrier permeability. Br. J. Pharmacol., 128: 917–924. Sun, H., Bungay, P.M. and Elmquist, W.F. (2001) Effect of capillary efﬂux transport inhibition on the determination of probe recovery during in vivo microdialysis in the brain. J. Pharmacol. Exp. Ther., 297: 991–1000. Sun, H., Dai, H., Shaik, N., Elmquist, W.F. and Bungay, P.M. (2003) Drug efﬂux transporters in the CNS. Adv. Drug Deliv. Rev., 55: 83–105. Syva¨nen, S., Xie, R., Sahin, S. and Hammarlund-Udenaes, M. (2006) Pharmacokinetic consequences of active drug efﬂux at the blood-brain barrier. Pharm. Res, 23: 705–717. Thompson, A.C., Justice, J.B. Jr. and McDonald, J.K. (1995) Quantitative microdialysis of neuropeptide Y. J. Neurosci. Methods, 60: 189–198. Tunblad, K., Ederoth, P., Gardenfors, A., HammarlundUdenaes, M. and Nordstrom, C.H. (2004) Altered brain exposure of morphine in experimental meningitis studied with microdialysis. Acta Anaesthesiol. Scand., 48: 294–301. Tunblad, K., Hammarlund-Udenaes, M. and Jonsson, E.N. (2005) Inﬂuence of probenecid on the delivery of morphine6-glucuronide to the brain. Eur. J. Pharm. Sci., 24: 49–57. Tunblad, K., Jonsson, E.N. and Hammarlund-Udenaes, M. (2003) Morphine blood-brain barrier transport is inﬂuenced by probenecid co-administration. Pharm. Res., 20: 618–623.

Wang, Q., Yang, H., Miller, D.W. and Elmquist, W.F. (1995) Effect of the P-glycoprotein inhibitor, cyclosporin A, on the distribution of rhodamine-123 to the brain: an in vivo microdialysis study in freely moving rats. Biochem. Biophys. Res. Commun., 211: 719–726. Wang, X., Ratnaraj, N. and Patsalos, P.N. (2004) The pharmacokinetic inter-relationship of tiagabine in blood, cerebrospinal ﬂuid and brain extracellular ﬂuid (frontal cortex and hippocampus). Seizure, 13: 574–581. Weikop, P., Egestad, B. and Kehr, J. (2004) Application of triple-probe microdialysis for fast pharmacokinetic/pharmacodynamic evaluation of dopamimetic activity of drug candidates in the rat brain. J. Neurosci. Methods, 140: 59–65. Welty, D.F., Schielke, G.P., Vartanian, M.G. and Taylor, C.P. (1993) Gabapentin anticonvulsant action in rats: disequilibrium with peak drug concentrations in plasma and brain microdialysate. Epilepsy Res., 16: 175–181. Wong, S.L., Wang, Y. and Sawchuk, R.J. (1992) Analysis of zidovudine distribution to speciﬁc regions in rabbit brain using microdialysis. Pharm. Res., 9: 332–338. Xie, R., Bouw, M.R. and Hammarlund-Udenaes, M. (2000) Modelling of the blood-brain barrier transport of morphine3-glucuronide studied using microdialysis in the rat: involvement of probenecid-sensitive transport. Br. J. Pharmacol., 131: 1784–1792. Xie, R. and Hammarlund-Udenaes, M. (1998) Blood-brain barrier equilibration of codeine in rats studied with microdialysis. Pharm. Res., 15: 570–575. Xie, R., Hammarlund-Udenaes, M., de Boer, A.G. and de Lange, E.C. (1999) The role of P-glycoprotein in blood-brain barrier transport of morphine: transcortical microdialysis studies in mdr1a (–/–) and mdr1a (+/+) mice. Br. J. Pharmacol., 128: 563–568. Yang, H., Wang, Q. and Elmquist, W.F. (1996) Fluconazole distribution to the brain: a crossover study in freely-moving rats using in vivo microdialysis. Pharm. Res., 13: 1570–1575. Zeitlinger, M.A., Dehghanyar, P., Mayer, B.X., Schenk, B.S., Neckel, U., Heinz, G., Georgopoulos, A., Muller, M. and Joukhadar, C. (2003) Relevance of soft-tissue penetration by levoﬂoxacin for target site bacterial killing in patients with sepsis. Antimicrob. Agents Chemother., 47: 3548–3553.

CHAPTER 6.7

Application of microdialysis in pharmacokinetic studies Ronald J. Sawchuk and Belinda W.Y. Cheung Department of Pharmaceutics, College of Pharmacy, University of Minnesota, Minneapolis, MN, USA

Abstract: This chapter focuses on the recent advances and applications of microdialysis in pharmacokinetic studies. The ﬁrst section addresses the technical and theoretical issues regarding microdialysis experiments with an emphasis on obtaining pharmacokinetic data. The principles and methodologies for microdialysis probe calibration are examined. Various setups for microdialysis sampling and subsequent data analysis are also illustrated. The second section includes research articles since the year 2000, chosen from an extensive literature review to report and discuss the utilization and implications of microdialysis sampling in pharmacokinetics. These articles are discussed in three subsections dealing with the kinetics of drug delivery to speciﬁc targets, the kinetics of drug transport and metabolism, and the use of microdialysis to study the effects of pathophysiology or surgical intervention on pharmacokinetics. A concluding section brieﬂy underscores the critical need to characterize microdialysis recovery in pharmacokinetic investigations.

to speciﬁc targets; (2) studies focusing on the kinetics of drug transport and metabolism (including interactions); and (3) reports that utilize microdialysis sampling to study the effects of pathophysiology or surgical intervention on pharmacokinetics. The concluding section reiterates the importance of assessing microdialysis recovery in pharmacokinetic investigations, and provides caveats related to the reliability of conclusions drawn from such studies as they relate to assessment of analyte recovery.

I. Introduction This chapter is broadly divided into three sections. The ﬁrst of these addresses the characteristics of microdialysis sampling that facilitate its application in pharmacokinetic studies, in animals as well as in human subjects. This section also discusses some of the difﬁculties and limitations that may be encountered in such studies, particularly where the goal is to characterize concentration–time proﬁles into quantitative issues, that is, by extracting pharmacokinetic parameters that describe the system. The second section provides speciﬁc examples of studies that utilize microdialysis sampling in pharmacokinetic investigations reported in the literature since 2000. This section is organized according to the general goal of these studies, and represents a summary of: (1) reports dealing with the kinetics of drug delivery

II. Microdialysis sampling in pharmacokinetic studies Pharmacokinetics is the study of the absorption, distribution, metabolism, and excretion of a drug. Of the four areas, microdialysis sampling of tissues has been most widely used in the determination and understanding of drug distribution and metabolism. With the continual broadening

Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

601

DOI: 10.1016/S1569-7339(06)16032-4 Copyright 2007 Elsevier B.V. All rights reserved

602

and advances in microdialysis methodology, its application in other areas of pharmacokinetic is likely to increase. One of the earliest applications of microdialysis in the ﬁeld of pharmacokinetics was the study of drug distribution in the brain. In the 1980s, several pre-clinical brain microdialysis models were developed, primarily for the investigation of neurotransmitter dynamics (Tossman and Ungerstedt, 1986; Ungerstedt and Hallstrom, 1987; Egawa et al., 1988). Using these animal models, the protein-unbound, extracellular concentrations of exogenous drugs in different regions of the brain can be measured continuously. In early 1990s, the technique began to ﬁnd application in clinical pharmacokinetic studies (Lonnroth et al., 1987; Lonnroth, 1991). Over the subsequent two decades, microdialysis has been performed in a wide variety of tissues, including brain (Wang and Sawchuk, 1995), spinal cord (Skilling et al., 1988), cerebrospinal ﬂuid (CSF)/intrathecal (Krupp and Bernards, 2004), skeletal and heart muscle, blood, adipose tissue, lung, liver, skin (Elmquist and Sawchuk, 1997), bone (Stolle et al., 2004), peritoneum (Riese et al., 2003), kidney (Nishiyama et al., 2002), middle ear ﬂuid (MEF; Zhu et al., 2003), eye (Anand et al., 2004), synovial ﬂuid (SF; Qian et al., 2003), gut lumen (Solliga˚rd et al., 2005), breast tissue (Dabrosin, 2005), and tumor (Brunner and Muller, 2002). The majority of the pre-clinical microdialysis experiments have been conducted in animals such as rats, rhesus monkeys, rabbits, chinchillas, and dogs. More recently, miniaturization and new probe designs have seen the mouse added to this list (Boschi and Scherrmann, 2000). In studies requiring less invasive surgical procedures, especially those involving the muscle, skin, adipose, certain regions of the brain, and blood, subjects may be allowed to recover from anesthesia following probe implantation and remain conscious or even freely moving during microdialysis (Elmquist and Sawchuk, 1997; Galvan et al., 2003). These conscious models are ideal for pharmacokinetic evaluation because they are devoid of the effects of anesthetic agents, and normal physiological conditions may be maintained so that microdialysis sampling of tissues can continue for many hours, or even days.

II.A. Comparing microdialysis and discrete-point sampling Prior to the introduction of microdialysis as a sampling methodology, virtually all pharmacokinetic studies used discrete-point sampling to measure drug concentrations in blood or peripheral tissues. There are ﬁnite limits regarding the sample volume and sampling frequency that are feasible without significantly disturbing the physiology and homeostasis of the subjects. Whole tissue homogenates have often been used for a single determination of tissue drug concentration. However, to construct the tissue concentration–time proﬁle, different groups of animals are sacriﬁced at each time point, and large data variability is often observed. In contrast, microdialysis does not involve the removal or depletion of biological matrices, allowing for more frequent sampling to better characterize rapid changes in concentrations. This sampling technique reduces the total number of subjects required by permitting numerous measurements from each subject, and because it raises the possibility of crossover study designs. The magnitude of data scatter is also likely to be lower, since inter-animal variability does not contribute to the measured concentrations. However, employing microdialysis sampling in pharmacokinetic studies raises challenges that must be considered and carefully addressed.

II.B. Microdialysis probe recovery and calibration During microdialysis, a perfusion ﬂuid, or perfusate, is delivered through the probe, typically at a constant ﬂow rate. Exchange of solutes occurs along the semi-permeable membrane of the probe, resulting in a dialysate solute concentration that is a fraction of the actual tissue extracellular, unbound level. This fraction is also called the relative recovery. For studies examining changes in endogenous compound levels from their baseline values, it is usually not necessary to determine the relative recovery as long as it is reasonable to assume that recovery remains relatively constant throughout the experiment. However, in

603

pharmacokinetic studies, there are no baseline values for exogenous substances. Knowledge of the relative recovery becomes crucial for the determination of true extracellular tissue concentration of the analyte – the drug, or its metabolites. Many factors affect probe recovery including microdialysis (perfusion) ﬂow rate, temperature, probe membrane area, nature of the dialyzed tissue, physicochemical properties of the analyte of interest, and any other factors that alter the diffusion characteristics. In general, the higher the ﬂow rate through the probe, the lower the relative recovery. Higher temperatures and larger probe membrane areas usually result in better recovery. These factors are addressed in detail elsewhere in this volume. A number of methods have been developed and used to calibrate the microdialysis probes by estimating the relative recovery in vivo. They include the ﬂow rate or stop ﬂow method (Jacobson et al., 1985), the zero net ﬂux or concentration difference or equilibrium method (Lonnroth et al., 1987), and the retrodialysis or reverse dialysis method (Wang et al., 1991). Both the ﬂow rate and the zero net ﬂux methods require the analyte in the dialyzed tissue to be at steady state. These time-consuming calibration processes are typically done prior to the actual pharmacokinetic studies. However, in vivo microdialysis is a dynamic process and changes in probe recovery are often observed, especially with longer duration experiments (Benveniste, 1989). Air bubbles may become trapped and tissue components may accumulate along the probe, affecting its performance. Therefore, recovery estimated prior to or following the study may not reﬂect what happens during the course of the study. One way to monitor probe recovery throughout the experimental period is by retrodialysis using a calibrator. This method involves the use of a calibrator that exhibits a similar permeability–area product to that of the compound of interest. A known concentration of the calibrator (Cic) is added to the perfusion ﬂuid. With each dialysate sample, a residual calibrator concentration (Cec) can be measured. The relative loss of the calibrator is 1 minus the ratio of Cec to Cic (Wang et al., 1991). Ideally, the loss of the calibrator is identical to the recovery of the solute of interest. The calibrator needs to be

tested carefully under the eventual study conditions to ensure that its loss across the membrane reﬂects changes in probe function that affect the analyte in a similar manner. An important issue related to in vivo calibration of probe recovery was highlighted by Bungay et al. (2001). The authors pointed out that, with rapidly changing analyte concentration in the dialyzed tissue, there is a time-dependent change in the relative recovery of the analyte. This is especially germane for microdialysis sampling in solid tissues. The phenomenon stems from the concentration gradient from the probe membrane with radial distance into the adjacent tissue. Simulation results indicate that with conventional methods of probe calibration such as retrodialysis, the transient changes in recovery lead to a discrepancy between corrected (apparent) tissue concentration and actual level in the dialyzed site. Proper calibration of this transient change can be achieved by the administration of an appropriate calibrator simultaneously with the solute of interest using the same route and at the same location or by the transient zero net ﬂux method proposed by Olson and Justice (1993). However, these forms of calibration are often not practical for most in vivo pharmacokinetic studies. Therefore, it becomes difﬁcult to correlate individual dialysate concentrations to the actual extracellular concentrations. One way to minimize the effect of the changing recovery is to perfuse the probes at low ﬂow rates, thus keeping the recovery near equilibrium, that is, as close to unity as possible. This method is not always feasible because of assay sensitivity limits for very small sample volumes. If the goal of the study is to obtain the area under the concentration–time curve (AUC) of the analyte in the dialyzed tissue, the problem associated with time-dependent changes in recovery becomes far less important. This is because the discrepancy between apparent and actual tissue concentration changes the shape of the concentration–time proﬁle without much impact on the area under the curve to inﬁnite time (Bungay et al., 2001). For microdialysis sampling of analytes in a ﬂuid-ﬁlled space where there is convective ﬂow that minimizes spatial concentration differences (e.g., blood), the transient nature of probe recovery may be less

604

important. Here, the concentration–time course of the drug may be closer to that in the sampled space, and pharmacokinetic parameters derived from an analysis of the recovery-corrected dialysate concentrations may more accurately reﬂect the concentration–time course in the sampled tissue. II.C. Analysis of dialysate samples and concentration data Once perfusion of the microdialysis probe begins, the efﬂuent ﬂuid or dialysate can be collected either off-line or on-line. For off-line collection, a fraction collector directs the dialysate ﬂow into an appropriate sample container and switches to a new container at the end of each sampling interval. Some fraction collectors can cool the samples to minimize evaporation and degradation. These samples can then be processed and analyzed using a validated analytical assay for the drug, or stored until analysis at a later time. For on-line collection, efﬂuent from the probe is routed to a small volume sample loop, and at the end of each sampling period, the loop contents are injected directly onto an analytical instrument. No human handling of the samples is required. Dialysates from more than one tissue site can be collected either simultaneously using the off-line method or programmed to inject samples alternately using the on-line method. Concentrations of drug in each dialysate sample reﬂect the average level over the collection interval, and the mid-point of the interval is often used as the sample time (after correcting for lag time). Lag time is deﬁned as the time it takes for the dialysate to ﬂow from the probe membrane to the collection device or the analytical instrument. With very low perfusion ﬂow rates, lag times of more than 1 h are not uncommon. To correct for lag time, knowledge of the tubing length and dead volume is required. II.D. Pharmacokinetic analysis Following correction for lag time and relative recovery, a tissue concentration–time proﬁle can be

constructed and the data can be analyzed or modeled to obtain estimates of various pharmacokinetic parameters similar to those obtained from direct blood sampling. For noncompartmental analysis, area under the concentration–time curve and area under the ﬁrst moment curve (AUMC) can be determined using the trapezoidal rule. The ratio of AUMC to AUC yields mean residence time (MRT). Stahle (1993) pointed out that if the probe recovery remains constant throughout the experiment, MRT can be calculated based on uncorrected dialysate data. In other words, the ratio of AUMC to AUC is not affected by the magnitude of microdialysis recovery. Other pharmacokinetic parameters that can be estimated without correction for recovery include time to maximum concentration, terminal half-life, and the rate constants into and out of the dialyzed tissue. However, if probe recovery ﬂuctuates or drifts with time, concentrations in dialysate samples need to be corrected before any parameters can be determined. The ability to characterize tissue concentration–time proﬁle in each subject with frequent sampling provides valuable insight regarding mechanisms of drug transport. The ratio of AUC of the dialyzed tissue to that of plasma gives an estimate of the ratio of drug transport clearance into and out of the tissue. A clearance ratio of unity (where concentrations are expressed in terms of unbound values) is expected where passive diffusion of solutes exists between tissue and blood. A greater than unity clearance ratio indicates a possible involvement of inﬂux transporter(s) from blood to tissue. In contrast, a clearance ratio less than 1 suggests the existence of efﬂux transporter(s) from tissue to blood. The transfer clearances in both directions can be determined separately if the dialyzed tissue is treated as an individual entity during compartmental data analysis with inﬂux and efﬂux rate constants as ﬁtted parameters. Unlike passive diffusion, facilitated or active transporter systems are saturable. Therefore, by varying the administered dose of the solute of interest, changes in the AUC ratio (tissue to plasma) indicate that transporter system(s) may play a role in drug distribution. Further information about

605

the involvement of speciﬁc transporters can be obtained by introducing known inhibitors to the tissue site and examining the effect, if any, on the AUC or clearance ratio (Zhu et al., 2003; Dai and Elmquist, 2005). Crossover study designs are preferred for inhibition studies to reduce inter-animal variability and increase the power to detect differences between control and treated subjects. II.E. Additional resources Over the past several years, several comprehensive reviews and tutorials that address methodological issues as they relate to microdialysis sampling in pharmacokinetic studies have appeared in the literature (Boschi and Scherrmann, 2000; Chu and Gallo, 2000; de Lange et al., 2000; de la Pena et al., 2000; Muller, 2000; Verbeeck, 2000; Garrison et al., 2002; Kreilgaard, 2002; Cano-Cebrian et al., 2005; Plock and Kloft, 2005). Some of these reviews summarize publications that appeared 10–15 years ago, and therefore represent an earlier examination of the topic than those addressed in the next section of this chapter. III. Selected studies utilizing microdialysis sampling in pharmacokinetic investigations III.A. Kinetics of drug delivery to specific targets III.A.1. Brain Using serial blood sampling and simultaneous brain microdialysis in a rat model, Tong and Patsalos (2001) investigated the concentration–time course of a new antiepileptic drug, levetiracetam, in serum and brain extracellular ﬂuid (ECF). Frontal cortex and hippocampus were examined. The antiepileptic drug was administered by interperitoneal injection at doses of 40 or 80 mg/kg. Blood samples were withdrawn at 20min intervals over the ﬁrst hour and subsequently at 30-min intervals up to 8 h. Relative microdialysis recovery (13%) was determined at 371C in vitro. This recovery was used to adjust dialysate concentrations to represent ECF concentration data. The authors state that a one-compartment pharmacokinetic model was used to analyze the

sera and microdialysate data. However, they report areas under the curve calculated by the trapezoidal rule, and Tmax and Cmax values obtained by visual inspection. After administration of levetiracetam, it rapidly appeared in both serum (Tmax, 0.4–0.7 h) and ECF (Tmax, 2.0–2.5 h). Brain ECF concentrations rose in a linear fashion. The data suggest that transport across the blood–brain barrier is rapid. The range of serum-free fraction (0.93–1.05) was independent of concentration and conﬁrms that levetiracetam is not bound to plasma proteins. However, half-lives for levetiracetam disappearance from brain ECF were significantly longer than those for serum (mean range, 3.0–3.3 h vs. 2.1–2.3 h) and brain ECF concentrations in the decline phase were 60–70% of those in serum at both dose levels. This study clearly demonstrates that levetiracetam readily and rapidly enters the brain without regional speciﬁcity, although it should be noted that microdialysis recovery was assessed in vitro and assumed to be the same in both brain regions. The somewhat prolonged residence time of this drug in the brain may explain its prolonged duration of action. The pharmacokinetics of ethanol as measured in blood are known to be gender-dependent in the rat. However, gender-speciﬁc differences in the concentrations of this drug in the brain have not been carefully studied. Robinson et al. (2002) used quantitative microdialysis to determine ethanol pharmacokinetics in the nucleus accumbens following modest doses of ethanol, comparing males with females during the estrous cycle. The investigators administered ethanol intravenously or intragastrically in doses of 1 g/kg to male and female rats. Ethanol concentrations in the nucleus accumbens were determined using microdialysis with in vivo probe calibration. These concentrations were compared with those in jugular venous blood. Noncompartmental analysis of the data was conducted. Upon intravenous dosing, ethanol elimination from both brain and blood was found to be 15% faster in females than males. After intragastric delivery, blood ethanol levels peaked faster in females than males by 20 min, suggesting more rapid absorption from the gastrointestinal tract. In addition, ethanol concentrations in the nucleus accumbens peaked more rapidly in females than

606

males by 25 min, suggesting more rapid distribution to the brain. These parameters did not differ across the estrous cycle, and were not related to body water. This appears to be the ﬁrst quantitative assessment of ethanol distribution kinetics to brain ECF of conscious rats. Marchand et al. (2003) examined the effect of dose on the pharmacokinetics of norfloxacin and its distribution into brain ECF in freely moving rats. These investigators used microdialysis to measure unbound concentrations of norfloxacin in rat hippocampus following an intravenous bolus dose ranging from 12.5 to 150 mg/kg. In vivo recovery of norfloxacin was determined by retrodialysis employing ciproﬂoxacin as the calibrator. Norfloxacin and ciproﬂoxacin exhibited similar loss when perfused simultaneously, with mean ratios (norfloxacin:ciproﬂoxacin) ranging from 0.94 to 1.07. Relative recoveries in vitro for both analytes at a ﬂow rate of 1 mL/min ranged from 32 to 37%. Microdialysis samples were analyzed by high-performance liquid chromatography (HPLC) with ﬂuorescence detection. Noncompartmental analysis was performed on the data from each animal individually. Although maximum brain ECF concentrations of norfloxacin were obtained rapidly, brain:plasma AUC ratios were low and linearly related to dose, averaging 8.275.8%. Interestingly, norfloxacin systemic pharmacokinetics were nonlinear. These investigators carefully characterized the recovery of several potential calibrators, and used retrodialysis with a homolog calibrator during the entire study time course. As such, retrodialysis by calibrator avoids artifacts that may be introduced by a changing recovery during the time course of the experiment. Zhang et al. (2003) investigated the pharmacokinetics of L-3,4-dihydroxyphenylalanine (L-dopa) in the plasma and ECF of striatum in marmosets. The authors measured the concentration of L-dopa in plasma and striatal ECF in common marmosets using microdialysis and HPLC with electrochemical detection. Noncompartmental analysis of the data was performed. The mean Cmax was 20.3 mM in plasma and 442.9 nM in ECF of striatum. The L-dopa concentration in ECF was much lower than that typically observed during in vivo studies of L-dopa toxicity.

Bagger and Bechgaard (2004) examined the delivery of ﬂuorescein from the nasal cavity to the brain in rats using microdialysis. The aim of their study was to characterize the targeting of the brain by olfactory absorption using a hydrophilic model compound, sodium ﬂuorescein, an agent with limited permeability across the blood-brain barrier. Microdialysis probes were placed in the right and left sides of the brain (striatum), as well as in blood, in rats. The time course of sodium ﬂuorescein was studied for 3 h following intravenous and nasal administration. The fractional rate of elimination from brain was significantly slower after nasal administration (0.004 min1 compared with intravenous administration, 0.012 min1), suggesting slow but continued delivery to brain tissue. However, ratios of the brain to plasma area under the curve for ﬂuorescein were extremely low (2–3%) regardless of the route of administration. The authors concluded that targeted delivery via the nasal route of hydrophilic compounds such as sodium ﬂuorescein is limited. Wang et al. (2004) investigated the serum, cortical, hippocampal, and CSF kinetics of tiagabine, a new antiepileptic drug in a freely moving rat model. One group of rats was implanted with a jugular vein catheter and a cisterna magna (CM) catheter for blood and CSF sampling, respectively. A second group was ﬁtted with a blood catheter and a microdialysis probe in the hippocampus and frontal cortex. Tiagabine was dosed intraperitoneally. Blood, CSF, and ECF were collected for assay of tiagabine concentrations by HPLC. Noncompartmental analysis was performed on the data. Tiagabine concentrations in blood and CSF were linearly related to dose. Mean CSF:serum tiagabine concentration ratios (range, 0.008–0.01) were much lower than the mean serum free fraction of tiagabine (0.04570.003). The study is notable as it provides information regarding the distribution of a drug into three separate regions of the central nervous system – cortex, hippocampus, and CSF. Elimination half-life in CSF was similar to that in serum, but the half-lives in ECF were three times longer. Tiagabine demonstrated linear kinetic behavior with rapid brain penetration. CSF concentrations did not reﬂect free concentrations in serum. The observation that tiagabine elimination

607

from the brain is much slower than that seen in blood may account for the long duration of action of this antiepileptic drug. Tunblad et al. (2004b) proposed an integrated model for the interpretation of microdialysis data that considers all data, including recovery assessment, in a single model. The authors underscore that noncompartmental analysis provides estimates of certain parameters, such as clearance and volume of distribution, but does not typically give information regarding the rate of transport across the blood-brain barrier. It should be noted that the report of Bungay et al. (2001) suggests that the drug concentration–time proﬁle characterized by microdialysis in solid tissues differs in theory from the true unbound concentration–time proﬁle in the tissue ECF because of a timedependent recovery. This observation may call into question the applicability of compartmental modeling where rapid changes in tissue concentration are observed. The data in this study were derived from a crossover study in rats designed to investigate the effect of probenecid (PBD) on the blood-brain barrier transport of morphine. Recovery of morphine from femoral vein blood and striatum was determined by retrodialysis by drug prior to dosing. Noncompartmental data analysis was performed providing estimates of clearance, steady-state volume of distribution, free fraction in plasma, and microdialysis recovery. Systemic pharmacokinetic parameters were obtained from the arterial data, where regular blood sampling was performed. Incorporating the free fraction in blood, the brain concentrations were analyzed, assuming that blood-brain barrier transport could be described in terms of clearance into and out of the brain. The investigators applied a unique modeling approach in which the systemic pharmacokinetics were described using all the data collected in blood, that is, the arterial concentrations measured by traditional sampling, the venous dialysate concentrations, and the recovery measurements in venous blood. Although the analysis of brain and blood data together in a global pharmacokinetic model is not unique, the determination of recovery using both arterial blood levels of morphine and the venous dialysate concentrations appears to be novel.

Juan and Tsai (2005) studied the pharmacokinetics of the alkaloid vincamine, and investigated the potential role of the multi-drug transporter P-glycoprotein on the distribution and disposition of this alkaloid. Microdialysis probes were inserted into the jugular vein and in the hippocampus of Sprague–Dawley rats. Vincamine (10 and 30 mg/kg) was administered intravenously. Probe calibration was accomplished by microdialysis by drug prior to the experiment. The average recovery in vivo of vincamine in blood was 53.7% and in brain, 17.5%. Pharmacokinetic parameters of vincamine were determined using a two-compartment model that was selected over the one-compartment model based upon a lower Akaike Information Criterion. In the presence of cyclosporine, an inhibitor of P-glycoprotein, given concurrently in a dose of 10 mg/kg intravenously, unbound vincamine AUCs were increased significantly in both blood and brain. However, the average brain-to-blood ratio of areas under the curve did not appear to be different in treated and controlled rats at either dose level. The extent to which cyclosporine increased the brain AUC of this alkaloid may have been due in large part to the increase in the blood AUC, that is, to a reduction in total clearance of vincamine in the presence of cyclosporine. Chang et al. (2005) recently developed a method for simultaneously measuring unbound levels of nicotine and its metabolite, cotinine, in blood and brain tissue of rats. Microdialysis probes were inserted into the jugular vein/right atrium and the striatum of Sprague–Dawley rats, and nicotine (2 mg/kg) was administered intravenously. Microdialysis sampling and an HPLC–UV assay were used to deﬁne the pharmacokinetics of nicotine and cotinine at both sites. The limit of quantiﬁcation for nicotine and cotinine was 0.25 and 0.05 mg/mL, respectively. The authors used microdialysis by drug to calibrate recovery. Nicotine was measurable in the blood as well as in the brain; however, cotinine, which exhibited an extremely long half-life in the blood, could not be quantitated in striatum. Although the recovery of nicotine and cotinine from blood was 30%, the recovery from brain for both analytes was only in the range of 3–3.5%, raising questions concerning the

608

reliability of the brain level measurements. Because the rats were maintained under anesthesia throughout the experiment, the time course of the measurements, particularly for nicotine brain sampling, was limited. The blood and brain concentration–time proﬁle of nicotine suggests that nicotine rapidly equilibrates with striatum and that cotinine exhibits a long residence time in blood. The reason for the extremely high AUC of unbound nicotine in brain compared with plasma is not known.

III.A.2. Spinal region Clement et al. (2000) employed microdialysis to study the intrathecal bioavailability of a mixture of lidocaine and bupivacaine in a rabbit model of spinal anesthesia. A catheter and a microdialysis probe were inserted in either the epidural or the intrathecal space. Retrodialysis, using ropivacaine as the calibrator, was used to determine relative recovery. One interesting feature of the study was that the investigators were able to collect dialysates and inject them every 2 min onto the HPLC, providing extremely good time resolution in the data. The epidural clearance of bupivacaine was higher than that of lidocaine, suggesting a more significant uptake of bupivacaine into the systemic circulation and/or into the CSF. The intrathecal bioavailability of bupivacaine and lidocaine was 12.3 and 17.9%, respectively, whereas it was 5.5 and 17.7% when each of these anesthetics was administered separately. The investigators present a model that describes the competitive processes involved in the disposition of drugs administered into the epidural space. Based upon this, they speculate that a reduction in systemic absorption of bupivacaine might enhance its availability to spinal CSF when both anesthetics are administered together, perhaps because of a vasoconstrictor effect of lidocaine. Song et al. (2004) investigated the kinetics of disposition of two antiviral nucleosides in CSF during intrathecal administration using a novel conscious rabbit model. They also examined the possible involvement of an organic anion transporter (OAT) in the efﬂux of these drugs from CSF. Microdialysis probes were implanted in the

lumbar (L) and thoracic (T) regions of the rabbit. A CSF sampling catheter was implanted at the CM. A dosing catheter was also implanted in the lumbar region adjacent to the lumbar microdialysis probe. In vivo on-line microdialysis (L and T) and direct sampling (CM) were employed to monitor CSF concentrations of drugs. Zidovudine (N ¼ 8) or stavudine (N ¼ 5) was infused at 5 mg/min in the control phase, or at 2.5 mg/min when coinfused with PBD at a rate of 0.5 mg/min in crossover. A retrodialysis calibrator was used during the studies to correct for recovery. Zidovudine and stavudine concentration proﬁles in CSF at L were analyzed by noncompartmental analysis. Zidovudine half-life, efﬂux clearance, and AUC/dose were 1307101 min, 0.062370.0497 mL/min, and 24.7724.5 mg min/mL/mg, respectively. These values were all significantly different (po0.05) from stavudine counterparts (275752.0 min, 0.02167 0.0109 mL/min, and 70.3739.0 mg min/mL/mg, respectively). There was a significant difference in the AUC/dose of zidovudine (po0.05) in the presence of PBD. These investigators demonstrated that drug concentration is clearly dependent on the distance between the dosing and sampling site during intrathecal administration, and hypothesized that OATs may participate in zidovudine, but not stavudine, efﬂux from the CSF. Krupp and Bernards (2004) appear to be the ﬁrst to investigate the pharmacokinetics of oligonucleotides following intrathecal administration. This is of interest because it is very likely that antisense oligonucleotides will be given by this route of administration to patients in the future. These investigators implanted microdialysis probes intrathecally in pigs at L4, L1, and T11 vertebral levels and epidurally at the L4 vertebral level. The study oligonucleotides were 10-, 18-, or 30nucleotide-long sequences of the human MDR-1 gene. Microdialysis samples were obtained until 180 min after dosing for the analysis of oligodeoxynucleotide. Noncompartmental pharmacokinetic analysis was conducted. The AUC differed significantly among the oligodeoxynucleotides at all sampling sites. This study demonstrates that the intrathecal pharmacokinetics of oligodeoxynucleotides are largely dependent on oligodeoxynucleotide length, in direct contrast with

609

ﬁndings for smaller molecules, where intrathecal and epidural pharmacokinetics are mainly determined by lipophilicity, not molecular weight.

III.A.3. Lung and skeletal muscle Using continuous microdialysis sampling, Sasongko et al. (2002) studied the distribution of the neuromuscular blocker, gallamine, into muscle tissue. Rats, under urethane anesthesia, received an intravenous bolus dose (6 mg/kg). Microdialysis sampling was used to continuously determine gallamine concentrations in muscle interstitial ﬂuid (MIF). In vivo microdialysis recovery of gallamine was determined using ‘‘retrodialysis by drug’’. Post-study tissue homogenization was also performed to determine gallamine distribution within muscle tissue. Gallamine was rapidly distributed into MIF (MIF–plasma partition coefﬁcient was 0.970.1). However, the estimated gallamine concentration in muscle tissue homogenate was only 2375% of that in MIF, suggesting that gallamine is largely excluded from the cells within muscle. These studies demonstrate that microdialysis coupled with tissue homogenization provides valuable insight into drug distribution within tissues. Freddo and Dalla Costa (2002) studied norfloxacin pharmacokinetics in rats to determine if free ECF levels of this drug could be predicted from total plasma concentrations. They measured free tissue and total plasma levels of norfloxacin in Wistar rats receiving intravenous bolus doses. The plasma pharmacokinetics could be described by a two-compartment model. However, the unbound norfloxacin concentrations in muscle ECF were lower than those predicted by plasma data. Their analysis therefore incorporated a proportionality factor, estimated to be 0.2570.08. Because it may be expected that this ‘‘tissue penetration factor’’ would be different for each drug and perhaps for each tissue examined, the utility of unbound plasma levels in predicting free antibiotic levels in muscle as described here is not clear. Persky et al. (2003) examined the pharmacokinetics of creatine (Cr) in six healthy males receiving a single oral dose of Cr monohydrate and at steady state after 6 days of oral Cr administration.

Oral clearance (CL/F) after the single dose was 0.2070.066 L/h/kg. At steady state, CL/F decreased to 0.1270.016 L/h/kg. The relative AUCs (muscle:plasma) of Cr determined by microdialysis was 0.4770.09 for the single dose and 0.3770.27 at steady state. Plasma and muscle data were simultaneously ﬁtted with a model incorporating a saturable absorption (to account for the reduced oral clearance at steady state) and ﬁrst-order elimination process. The authors suggest that repeated dosing of Cr caused a reduction in clearance that could result from saturation of the skeletal muscle pool. Rojas et al. (2003) used microdialysis to compare rat plasma and muscle concentrations of triamcinolone acetonide (TA). Steady-state experiments were carried out using an initial intravenous loading bolus of the phosphate ester of TA and constant-rate infusions. In vivo recovery was measured by retrodialysis. The unbound muscle ECF concentration at steady state, corrected for recovery, was 2.7370.42 mg/mL compared with 21.972.3 mg/mL in plasma. A two-compartment model was used to describe the plasma data. The systemic clearance of TA was 0.94 L/h/kg. The measured microdialysate levels of TA in muscle, corrected for recovery, were comparable to the model-predicted free drug levels in the peripheral compartment of the two-compartment model. Protein binding in rat plasma, measured by ultraﬁltration, was 90.1%. The microdialysis recovery in vivo in muscle was similar to the in vitro recovery under stirred conditions. Liu et al. (2005) investigated the pharmacokinetics of cefpodoxime in rat skeletal muscle and lung interstitial tissue ﬂuids using microdialysis to determine whether the free plasma or muscle levels might be predictive of those in lung tissue, since the lung is much more difﬁcult to access experimentally. Cefpodoxime was given as single intravenous bolus doses of either 10 or 20 mg/kg to anesthetized male Wistar rats or as a constant infusion of 260 mg/h with a loading dose to hasten the approach to steady state. The protein-bound fraction of the cephalosporine was measured in rat plasma using ultraﬁltration and was found to average 38%. During constant-rate infusion, the unbound concentrations in the muscle and the

610

lung as determined by microdialysis were comparable, but less than those in plasma. These data were analyzed simultaneously using a twocompartmental body model. The authors concluded that because unbound ECF concentrations of cefpodoxime in muscle and lung tissue were similar but lower than those measured in plasma, unbound levels in muscle ECF may be a surrogate for unbound concentrations of this drug in lung tissue. III.A.4. Eye In an extension of a previous microdialysis study in which Macha and Mitra (2001) studied the ocular pharmacokinetics of cephalexin, cephazolin, and cephalothin in rabbits, these authors (Macha and Mitra, 2002) also characterized the ocular pharmacokinetics of ganciclovir (GCV) using microdialysis sampling of the eye (New Zealand albino rabbits) following intravitreal administration. In addition, they examined the utility of the acyl monoester prodrugs of GCV (acetate, propionate, butyrate, and valerate) for delivering GCV to the vitreous. The animals were kept under anesthesia. A concentric and a linear microdialysis probe were implanted in the vitreous and in the anterior chamber across the cornea, respectively. The drugs were administered intravitreally and the samples were collected for 10 h. The vitreal clearance of the prodrugs increased with ester chain length. A parabolic relationship was observed between the vitreal elimination rate constant and the ester chain length. MRT of GCV generated upon prodrug administration was three to four times that observed after GCV injection. III.A.5. Middle and inner ear Cefditoren (CDTR) is a new oral cephalosporin antibiotic that exhibits a broad spectrum of bactericidal activity. Little is known about the penetration of this antibiotic into MEF. Zhu et al. (2003) conducted a study to determine CDTR transport kinetics between plasma and MEF by characterizing inﬂux (CLin) and efﬂux (CLout) clearances. Simultaneous intravenous bolus and intramiddle-ear doses were administered to two groups of chinchillas – control and infected. In

vivo microdialysis was used to determine free CDTR levels in MEF. Parameters determined in both groups were compared to assess the effect of infection and inﬂammation on CDTR distribution kinetics. The unbound transport clearances, CLin and CLout, estimated by compartmental and noncompartmental analysis agreed closely. The calculated CLin:CLout ratio was 0.7670.23 and 0.5670.25 in normal (N ¼ 9) and infected (N ¼ 6) animals, respectively. The CLin:CLout ratio determined in chinchillas compared well with values estimated from data in pediatric patients. Since the 95% conﬁdence interval of this ratio did not include unity, the authors speculated that an active efﬂux mechanism in middle ear mucosa may be involved in CDTR distribution in MEF. The round window membrane (RWM) separates the inner ear from the middle ear and is permeable to many compounds. RWM instillation of gentamicin, a preferentially vestibulotoxic aminoglycoside, is used as a therapeutic treatment for patients with intractable vertigo, despite considerable variations in the incidence and severity of hearing loss associated with gentamicin. Hunter et al. (2003) employed inner ear microdialysis to measure perilymph concentrations of low molecular weight compounds applied directly on the RWM in a chinchilla model in an attempt to establish optimal dosages for the treatment of inner ear disorders. Using urea as a low molecular weight marker applied on the RWM, they extrapolated results from a concentration–time plot of dialysates. The authors concluded that inner ear microdialysis can be used to characterize the pharmacokinetics of a low molecular weight agent within the perilymphatic space without the need for repeated direct sampling. Plontke et al. (2004) point out that the local delivery of drugs to the cochlea may be an important alternative to the usual systemic treatment of inner ear disorders. In addition to careful study design and interpretation of experiments examining pharmacokinetics in the inner ear, the authors make a case for the use of continuous sampling methods such as microdialysis and the use of three-dimensional models that consider the complex geometry of the inner ear when performing interspecies scaling of pharmacokinetic behavior.

611

III.A.6. Synovial fluid Knowledge of the distribution kinetics of a drug into SF is important in developing anti-inﬂammatory agents and understanding the relationship between their physicochemical properties and antiinﬂammatory efﬁcacy. The purpose of a study described by Qian et al. (2003) using the anti-arthritic drug, methotrexate (MTX), was to establish and validate an animal model of drug distribution and disposition in SF by comparing microdialysis with direct sampling of SF. Microdialysis probes in the stiﬂe joint space of four dogs had been calibrated in dog SF in vitro using the no net ﬂux method. After an intravenous bolus of 2.5 mg/kg of MTX, unbound levels in SF, as well as total plasma concentrations, were measured by liquid chromatography tandem mass spectrometry up to 48 h after dosing. The mean probe recovery and unbound fraction of MTX in SF were 46.8 and 44.8%, respectively. The drug was observed to penetrate into the joint space rapidly, producing an unbound MTX area under the curve in SF 40% of the total area under the curve in plasma. The data agreed well with previous data obtained using arthrocentesis.

III.A.7. Skin and adipose tissue In an effort to evaluate the dermal delivery of both hydrophilic and lipophilic drugs, Kreilgaard (2001) examined the utility of microemulsion vehicles for topical delivery of lidocaine and prilocaine in rats. Following topical application of three different formulations, unbound cutaneous concentrations of the drugs were determined by in vivo microdialysis in rats. Recovery was assessed throughout the studies using retrodialysis by calibrator. Pharmacokinetic modeling provided an excellent parameter ﬁt to the microdialysis concentration–time data. Kreilgaard et al. (2001) then studied the cutaneous bioequivalence of lidocaine, a lipophilic model compound applied topically using a microemulsion formulation, in comparison with a conventional oil-in-water emulsion, in human subjects, using a compartmental pharmacokinetic/pharmacodynamic model to interpret the microdialysis sampling results. Dermal delivery of lidocaine was measured by microdialysis in eight

subjects. Absorption constants and lag times were determined by pharmacokinetic modeling of the microdialysis data. In addition, the anesthetic effect of the treatments was assessed by mechanical stimuli in 12 volunteers. The compartmental pharmacokinetic model described the concentration–time curves extremely well. The authors concluded that the microdialysis technique combined with an appropriate pharmacokinetic model could be used to assess the results of bioequivalence studies of topically applied substances. Mathy et al. (2001) investigated the utility of subcutaneous microdialysis sampling to study the pharmacokinetics of ﬂurbiprofen, a nonsteroidal anti-inﬂammatory drug, and its plasma protein binding in the conscious, freely moving rat. Naproxen was validated for use as a retrodialysis calibrator. Flurbiprofen was injected intraperitoneally and intravenously. Traditional blood sampling and subcutaneous microdialysis sampling were performed simultaneously for both modes of administration. Concentration-dependent protein binding of ﬂurbiprofen was found. The measured unbound fractions were comparable to those reported previously. The authors demonstrate convincingly that subcutaneous microdialysis sampling, if performed with adequate calibration, is an important tool for studying the pharmacokinetics and plasma protein binding of ﬂurbiprofen in vivo in the conscious, freely moving rat. The duration of local anesthesia achieved with biodegradable microcapsules is often significantly prolonged. Kopacz et al. (2003) used microdialysis sampling in human volunteers over a 96-h period to determine the kinetics of bupivacaine and dexamethasone release from microcapsules at a subcutaneous injection site. Bupivacaine exhibited extended release, producing concentrations that increased for 24–34 h following injection of the microcapsule formulation. Furthermore, analgesia was observed to track the tissue bupivacaine concentration determined by microdialysis sampling. Analgesia was apparent at 78% of microcapsuleinjected sites after 96 h, significantly longer than that observed for aqueous bupivacaine injections. Bupivacaine produced higher peak plasma levels after aqueous injection than after microcapsule

612

injection even though the dose associated with the microcapsule injection was threefold larger. Gattringer et al. (2004) used microdialysis to measure concentrations of a ketolide antibiotic, telithromycin, in muscle and adipose tissue of healthy volunteers. Following a single dose, ratios of the AUC from 0 to 24 h to the minimum inhibitory concentration (MIC) suggested that pharmacologically active free concentrations of telithromycin in tissue as well as in the plasma might be effective against infections caused by Streptococcus pyogenes but not against those related to staphylococci or pathogens associated with animal bites. In a novel application of microdialysis sampling, Stagni et al. (2004) investigated the pharmacokinetics of acyclovir (ACV) in skin and plasma following administration in the rabbit via iontophoresis, intravenous bolus dosing, and topical ointment application. Using a crossover design with ﬁve treatment periods, each separated by at least 1 week, rabbits received either an intravenous bolus, ACV iontophoresis for 1 h at three different current densities, or a commercially available ointment for 2 h. Blood samples were collected serially up to 6 h. Concentrations of ACV in the skin were monitored via microdialysis using linear microdialysis probes. In skin dialysates following iontophoresis, Cmax, AUC, and half-life increased with current density; skin exposure to ACV was 40, 22, and 11% of that following intravenous bolus dosing. Systemic exposure to ACV during iontophoresis was extremely marginal since plasma concentrations were always below the limit of quantiﬁcation. Mathy et al. (2005) investigated the distribution of the topical antifungal, ﬂuconazole, into skin following both intravenous and topical administration in the conscious, freely moving rat. After an intravenous injection, microdialysate concentrations of the drug were measured by on-line HPLC in both the jugular vein and the dermis in rats. Further, cutaneous absorption was investigated using dermal microdialysis sampling following topical application of a 0.5% ﬂuconazole gel formulation in 12 rats. A ﬂuorinated analog of ﬂuconazole with equivalent recovery was used as a retrodialysis calibrator. Upon intravenous bolus

dosing, ﬂuconazole quickly penetrated into the dermis. Dermal concentrations of ﬂuconazole assessed by cutaneous microdialysis sampling were similar to the unbound plasma levels assessed by blood microdialysis. After topical application of 0.5 g of the ﬂuconazole gel formulation, free concentrations in dermis were assayed by cutaneous microdialysis for 11 h following dosing. Interestingly, the AUC of ﬂuconazole in dermal dialysate was unchanged at implantation depths up to 350 mM. However, below this, the AUC decreased with increasing depth of the probe. III.A.8. Bone Stolle et al. (2004) were able to examine the time course of concentrations of the aminoglycoside antibiotic, gentamicin, in bone using microdialysis. Pigs (N ¼ 8) received an intravenous bolus of 240 mg of gentamicin. Microdialysates and bone samples were obtained over 6 h. AUCs of the curve of the two microdialysates and bone samples were not significantly different. Excellent reproducibility of the microdialysis measurements was demonstrated, showing that microdialysis is a reproducible method for the measurement of gentamicin levels in experimental bone research. III.B. Kinetics of drug transport, metabolism, and interactions III.B.1. Distribution Although it has been shown that morphine is a substrate for P-glycoprotein, which limits its entry into the central nervous system, the role of transporters in the brain tissue distribution of its metabolite morphine-3-glucuronide (M3G) is unclear. Xie et al. (2000) investigated the possible effect of PBD, a nonspeciﬁc inhibitor of organic anion transport, on the brain distribution of M3G. Two groups of rats (control and treated) received an infusion of M3G over 4 h on each of the 2 consecutive days. PBD was co-administered in the treatment group on day 2. Microdialysis was used to estimate unbound M3G concentrations in brain ECF and also in blood. In vivo recoveries of M3G were determined using retrodialysis by drug,

613

preceding the drug administration. The kinetics of blood-brain barrier transport were modeled using nonlinear mixed effects modeling. In the PBD group, the ratio of the steady-state concentration of free M3G in brain ECF to that in blood was 0.0870.02 in control and 0.1670.05 in PBDtreated rats (p ¼ 0.001), whereas in the control group, the ratio was not significantly different between the 2 days. At least two possible explanations for the PBD effect exist – the inhibitor may reduce the efﬂux clearance of M3G or it may increase its inﬂux clearance. The authors concluded that the inﬂux clearance into the brain was altered by PBD (0.11 and 0.17 mL/min/g, in the absence and presence of PBD, respectively). The blood pharmacokinetics of the metabolite were not inﬂuenced by treatment with PBD. Assuming that the cephalosporin antibiotic cefepime may be a substrate of P-glycoprotein, Chang et al. (2001) hypothesized that cefepime and cyclosporine had the potential to interact. They examined the pharmacokinetics of cefepime in rats when given alone and in combination with cyclosporine using microdialysis. Cefepime, when given intravenously at three dose levels, exhibited linear pharmacokinetics. In the presence of cyclosporine, the MRT of cefepime was increased from 34.9 to 48.6 min (po0.05), and the area under the plasma concentration versus time curve increased from 4,775 to 6,960 min mg/mL (po0.01). Moreover, brain AUC increased from 64.3 to 110.2 min mg/mL. In summary, because the increase in brain AUC for cefepime in cyclosporine-treated animals was greater than that seen in the plasma AUC, it was speculated that cyclosporine alters cefepime pharmacokinetics in both blood and brain. Tsai et al. (2001) also used microdialysis to study the pharmacokinetics of camptothecin in rat blood, brain, and bile using putative inhibitors of P-glycoprotein transport (cyclosporin A (CsA), berberine, quercetin, naringin, and naringenin). Noncompartmental pharmacokinetic analysis of camptothecin data was performed. The results of this rather complicated study suggest that P-glycoprotein may modulate hepatobiliary secretion and/or brain penetration of camptothecin. However, in view of potential multi-site effects of

the modulators, the mechanism of the interactions involved – and the dose–response relationship – need to be further elucidated. Gacyclidine is a chiral noncompetitive NMDA antagonist. To determine whether the enantiomers of this agent exhibited pharmacokinetic differences, Hoizey et al. (2001) studied the pharmacokinetics of this drug in plasma and spinal cord ECF after IV administration of single enantiomers, either (+)- or ()-gacyclidine, in rats. Using microdialysis probes implanted in spinal cord, concentrations of gacyclidine in plasma and spinal cord ECF dialysates were determined using a chiral GC/MS assay. Plasma protein binding was estimated by equilibrium dialysis in vitro. Plasma levels declined in a biphasic manner with no significant difference between the two enantiomers. Clearance and volume of distribution and protein binding were not stereoselective. Both gacyclidine enantiomers were quantiﬁable in spinal cord ECF 10 min after drug administration. These concentrations remained relatively constant over the remaining study even though blood concentrations declined. Surprisingly, spinal cord ECF levels were not correlated with plasma AUCs. The authors concluded that both gacyclidine enantiomers penetrate rapidly and extensively into spinal cord ECF, and their distribution may involve an active transport system. As an extension of their previous study on opioid epidural pharmacokinetics (Bernards et al., 2003a), and recognizing that drug concentrations in the epidural space have not been characterized during epinephrine co-administration, Bernards et al. (2003b) performed a microdialysis study in pigs to determine the effect of epinephrine on the local (epidural and CSF) and plasma pharmacokinetics of opioids given epidurally. Morphine, in addition to alfentanil, fentanyl, or sufentanil, was injected epidurally with and without epinephrine. Opioid concentrations were measured in the epidural space, central venous plasma, and epidural venous plasma. The data were interpreted using a noncompartmental pharmacokinetic model and estimation of residence times. The effects of epinephrine on the pharmacokinetics of the co-administered drugs varied by opioid and by sampling site. In the lumbar epidural space,

614

epinephrine prolonged the MRT of morphine but, surprisingly, shortened the MRT of fentanyl and sufentanil. Epinephrine did not alter the pharmacokinetics of alfentanil, fentanyl, or sufentanil in the lumbar intrathecal CSF, but it increased the AUC of morphine and decreased its elimination half-life. The results of the study demonstrate that epinephrine’s effects on the pharmacokinetics of these opioids in the region of the spinal cord are complicated and not easily predicted. III.B.2. Metabolism Because rutaecarpine, a quinazolinocarboline alkaloid with vasodilator activity, was known to be a selective inhibitor of cytochrome P450 (CYP) 1A in mouse and human liver microsomes, Ueng et al. (2005) examined a potential interaction between rutaecarpine and theophylline, a wellknown substrate of CYP 1A2. Sprague–Dawley rats received 50 mg/kg rutaecarpine orally for 3 days resulting in a four- and threefold increase in hepatic microsomal 7-ethoxyresoruﬁn O-deethylation (EROD) and 7-methoxyresoruﬁn O-demethylation activity, respectively. The authors used blood microdialysis sampling and noncompartmental analysis to determine relevant pharmacokinetic parameters of theophylline in control and treated animals. Rats pretreated with rutaecarpine exhibited a threefold increase in clearance of theophylline and an 70% decrease in the AUC, MRT, and half-life. III.B.3. Excretion Lee et al. (2002) investigated the biliary excretion of the antifungal drug ﬂuconazole, using HPLC coupled with microdialysis. Microdialysis probes were implanted in the jugular vein and bile duct of male Sprague–Dawley rats. Fluconazole was dosed through the femoral vein. Rapid equilibration between blood and bile was observed. The investigators also examined a potential pharmacokinetic drug–drug interaction between ﬂuconazole and CsA by co-administering both drugs in this animal model, but no evidence could be found for ﬂuconazole being secreted by the P-glycoprotein transport system.

A number of pharmacologic actions have been attributed to berberine, a bioactive alkaloid. Tsai and Tsai (2004) examined the pharmacokinetics of berberine in rats, including studies to determine if biliary excretion of berberine could be altered by pretreatment with CsA, SKF-525A, quinidine, and PBD. They employed in vivo microdialysis interfaced with HPLC. Berberine exhibited linear pharmacokinetics in the dosage range from 10 to 20 mg/kg. In addition, berberine was shown to be actively secreted by the hepatobiliary system based on the bile-to-blood AUC ratio. Co-administration of CsA or quinidine with berberine signiﬁcantly decreased berberine secretion into bile, suggesting that the active berberine secretion may be modulated by P-glycoprotein and organic cation transporters (OCT). Although the metabolism of berberine was somewhat reduced by SKF525A, PBD had no effect on the glucuronidation of berberine in this study. Hesperidin is an anti-inﬂammatory bioﬂavonoid derived from citrus fruit. Tsai and Liu (2004) examined the pharmacokinetics of hesperidin using microdialysis sampling of blood and bile of anesthetized rats. A further goal was to investigate the potential interaction between hesperidin and CsA in this animal model. Microdialysis probes were surgically implanted in the jugular vein and bile duct. Hesperidin was dosed intravenously. Dialysates of both blood and bile were directly injected into the liquid chromatographic system. The AUC for hesperidin in bile was much greater than that in blood (AUCbile:AUCblood ¼ 8.972.5). Following CsA treatment, the distribution ratio was reduced to 3.270.6. The authors concluded that hesperidin is secreted via the hepatobiliary system against a concentration gradient and that this excretion may be modulated by P-glycoprotein. Huang et al. (2005) investigated the pharmacokinetics of unbound ranitidine, a competitive, reversible inhibitor of the action of histamine, in rat blood and bile. The authors used microdialysis coupled to a liquid chromatograph with simultaneous dual-site analysis. Control rats received only ranitidine (10 and 30 mg/kg, intravenously); treated rats were administered ranitidine and cyclosporin or quinidine (an organic cation transport inhibitor and P-gp inhibitor). Probes

615

were surgically implanted in the bile duct and jugular vein for simultaneous bile and blood sampling. Ranitidine exhibited maximum concentrations in bile between 20 and 30 min after drug dosing, with a bile-to-blood distribution ratio (AUCbile/AUCblood)>10. Since the AUC of ranitidine in bile was decreased with co-administration of cyclosporin or quinidine, the authors speculated that the biliary excretion of ranitidine may have been regulated in part by P-glycoprotein or OCT.

III.C. Distribution kinetics in pathophysiology, disease, or surgical intervention III.C.1. Trauma Only limited information concerning the bloodbrain distribution of drugs in injured or traumatized brain tissue in humans is available. Bouw et al. (2001) used brain microdialysis to investigate the blood-brain barrier transport of morphine in a patient with a severe brain injury. They simultaneously monitored unbound ECF concentrations of morphine in human brain as well as in subcutaneous adipose tissue. These levels were compared with those in arterial blood. Increased morphine levels in the brain close to the injured site were found in comparison with normal brain tissue. Half-lives of morphine in normal and injured brain tissues were similar, but longer than those in blood and fat, showing that morphine’s residence in brain tissue is greater than that expected from the blood level–time proﬁle.

III.C.2. Malignancy The novel cisplatin derivative, 4-pyridoxate diammine hydroxy platinum (PyPt), exhibits antitumor activity. Tokunaga et al. (2000) examined the pharmacokinetics of PyPt and its distribution into brain tumor in rats, and compared this with results obtained with cisplatin. PyPt (5.0 mg/kg) and cisplatin were given as intracarotid infusions over 30 min. ECF dialysates from probes implanted in tumor and nontumor brain tissues were collected simultaneously. Plasma concentrations of both total and protein-unbound platinum, and the tissue

distribution of total platinum were determined by microdialysis and atomic absorption spectrophotometry. Platinum accumulated in the brain tumor tissue ECF, but little distribution into normal tissue ECF of the brain was observed. The ratio of the AUC for brain tumor ECF platinum to that for plasma protein-unbound platinum for PyPt was 0.85, slightly higher than that for cisplatin (0.69), indicating that the local amount of platinum distributed into the glioma is somewhat enhanced by PyPt compared with cisplatin. Microdialysis sampling in a rat model was employed by Johansen et al. (2002) to simultaneously characterize the pharmacokinetics of cisplatin and carboplatin in blood, tumors, and several peripheral tissues. Following intravenous bolus administration, dialysate samples were collected up to 6 h using probes implanted in the jugular vein, kidney, and either liver or subcutaneously growing breast tumor tissue in Fisher 344 rats maintained under anesthesia. Correction of the data was performed using in vivo recovery. Cisplatin peak renal concentrations were greater than peak plasma and hepatic concentrations. For carboplatin, doses also produced high peak renal concentrations. Although tumor cisplatin and carboplatin AUCs were similar to those in blood, they were quite variable (range relative to plasma AUCs was 52–109%). Extreme increases in hepatic carboplatin exposure with increasing dose suggest a possible mechanism for carboplatin-induced hepatotoxicity where relatively high doses are employed. Although these investigators concluded that microdialysis is a reliable methodology for characterizing the disposition of platinum anticancer agents in various tissues, it should be noted that, for some drugs, extreme heterogeneity in interstitial concentrations may exist within a given organ system, for example, kidney or liver. This may be a complicating factor for the interpretation of microdialysis data where urinary and/or biliary concentrations are extremely high relative to those in blood or ECF. The penetration of the oral prodrug of 5-ﬂuorouracil (FU), capecitabine, and its metabolites into malignant and healthy tissues was the focus of studies performed by Mader et al. (2003). The authors examined the distribution of

616

capecitabine and its metabolites into tumor tissue in vivo in 10 patients with breast cancer-derived skin metastases who received capecitabine for 2 weeks. Microdialysis probes were inserted into a cutaneous metastasis and also into a subcutaneous connective tissue to evaluate the interstitial tissue pharmacokinetics of capecitabine and three of its metabolites – 50 -deoxy-5-ﬂuorocytidine (DFCR), 50 -deoxy-5-ﬂuorouridine (DFUR), and FU – by capillary electrophoresis. FU was present in very low levels in tumor ECF compared with those of capecitabine, DFCR, and DFUR. Capecitabine and its metabolites rapidly entered malignant and healthy tissues. No significant difference in ECF tissue exposure between healthy and malignant tissues was observed. Dukic et al. (2004) investigated the disposition of MTX in plasma, brain, and the ECF of two different brain tumors (C6 and CNS1 glioma) in rats. Plasma and brain ECF dialysates were sampled frequently following intravenous dosing of MTX (50 mg/kg) for 4 h. In vivo recoveries (reverse dialysis) were very low but were reported to be significantly different in the two tumor types (C6: 8.073.8%; CNS1: 4.972.5%), and in the contralateral hemisphere (C6: 6.0 7 4.0%; CNS1: 3.9 7 2.5%) between the two tumors. Area under the MTX concentration–time curve in plasma was 30% higher in CNS1 than that in C6 tumor animals, and this was attributed to a lower systemic clearance. Maximum MTX levels in brain tumor ECF were significantly higher in CNS1 than those in C6 tumor-bearing rats. The extremely low microdialysis recoveries achieved in these studies may have contributed to bias or imprecision in the results. Bergenheim et al. (2005) investigated the ECF kinetics of boron in three intracerebral compartments in vivo using microdialysis – solid tumor, brain adjacent to the tumor (BAT), and normal brain tissue. In addition, they examined delivery of boron into subcutaneous tissue before, during, and subsequent to BNCT in four patients. The patients received a 6-h infusion of BPA (900 mg/kg) that was completed 2–3 h before neutron irradiation. The extracellular concentration of BPA in tumor tissue mirrored that in blood, showing a maximal ratio (tumor:blood) of 1.07. The

concentration–time course of BPA in normal brain was considerably lower than that in the blood and tumor.

III.C.3. Inflammation or infection Huang et al. (2001) developed an experimental freely moving, conscious chinchilla model for the study of antibiotic drug distribution into MEF. Eustachian tube obstruction (ETO) in control animals and direct intrabullar inoculation with type 3 Streptococcus pneumoniae in infected animals were performed. Following surgery to implant microdialysis probes in the jugular vein and middle ear, amoxicillin was given intravenously as a bolus or infusion. Unbound amoxicillin concentrations in blood and MEF were monitored by microdialysis using cefadroxil as a retrodialysis calibrator. MEF-to-blood amoxicillin concentration ratios at steady state during intravenous infusion were similar, being 0.2670.06 and 0.2870.11 for normal and infected ears, respectively. The ratio of the area under the curve of drug concentration in MEF relative to that in blood after bolus administration was less than unity, as was the steady-state concentration ratio following constant-rate intravenous infusion, suggesting that an active transport mechanism may be involved in the efﬂux of amoxicillin from the middle ear of chinchilla. Amoxicillin distribution across the middle ear mucosal membrane in infected ears (validated by culture) was not significantly different from that in normal ears. Tomaselli et al. (2003a) conducted an in vivo microdialysis investigation to measure penetration of antibiotics into ECF of normal or pneumonic human lung tissue. They examined the lung penetration of two b-lactam antibiotics – cefpirome in elective thoracic surgery patients and piperacillin in septic thoracic surgery. Both drugs exhibit low protein binding. Microdialysis probes were inserted into normal or pneumonic lung tissue during surgery, and into healthy skeletal muscle post-operatively to obtain reference values. Serum as well as microdialysis samples were collected for 8 h or longer. The pulmonary ECF concentrations of both antibiotics exceeded the MICs for most commonly observed pathogens for 4–6 h.

617

Tomaselli et al. (2003b), in an extension of the work described above, investigated the penetration of piperacillin and tazobactam (a compound that inhibits the action of bacterial b-lactamases) into pulmonary tissue when administered simultaneously. Both drugs exhibit modest protein binding in plasma. Microdialysis probes were inserted into lung tissue in ﬁve patients during pulmonary surgery, and then into skeletal muscle after surgery for a basis of comparison. A single dose of 4 g of piperacillin and 500 mg of tazobactam was given intravenously. Dialysate and serum samples were collected for 8 h or more. Unbound maximum piperacillin concentrations in the ECF of infected lung tissue were approximately half than those in serum, and AUCs reﬂected this difference. There was a statistically significant difference between AUClung and AUCserum as well as between AUClung and AUCmuscle. The results suggest that unbound concentrations of b-lactam antibiotics in serum and muscle may not be predictive of unbound concentrations in infected pulmonary tissue. Joukhadar et al. (2003) undertook a study to determine how well moxiﬂoxacin, a ﬂuoroquinolone antibiotic, was able to penetrate into both healthy and inﬂamed subcutaneous adipose tissues. Twelve patients with soft tissue infections (STIs) were studied. In vivo microdialysis was used to measure the distribution of moxiﬂoxacin into the ECF of healthy and inﬂamed subcutaneous adipose tissues after a single intravenous dose in diabetic (N ¼ 6) and nondiabetic (N ¼ 6) patients with STIs. Overall, the mean concentration–time proﬁle of free moxiﬂoxacin in plasma was not significantly different from that in tissue. The ratios of the 8-h mean AUC for inﬂamed tissue /AUC for free moxiﬂoxacin in plasma were 0.570.4 for diabetic patients and 1.270.8 for nondiabetic patients. These ratios were not significantly different because of the large interpatient variability. Conditions such as meningitis or traumatic brain injury have the potential to alter the function and integrity of the blood-brain barrier. Tunblad et al. (2004a) investigated the pharmacokinetic pattern of morphine distribution in the intact brain and during experimentally induced meningitis in an anesthetized pig model using

microdialysis. Three microdialysis probes (one occipital and two frontal) were implanted in the brain. A fourth probe was placed in the jugular vein. Morphine was infused intravenously and concentrations were measured for 3 h. After meningitis was established via lipopolysaccharide injection into the CM, the study was repeated. The area under the morphine concentration–time curve ratio (brain to blood) doubled during meningitis, suggesting decreased active efﬂux and increased passive diffusion of morphine across the bloodbrain barrier. The systemic pharmacokinetics of levoﬂoxacin in patients with serious bacterial infections are comparable to those observed in healthy subjects. However, little is known about its distribution to tissue sites where infections may reside. Bellmann et al. (2004) conducted a microdialysis study to investigate the distribution of this drug into both inﬂamed and healthy subcutaneous adipose of patients with STIs. After a single intravenous dose of 500 mg, levoﬂoxacin was analyzed in microdialysates and plasma. They found no difference in the mean concentration–time proﬁles of unbound levoﬂoxacin in plasma from those in inﬂamed or healthy tissues. However, interpatient variability in tissue distribution was extremely large (CV ¼ 82%). Sauermann et al. (2005) studied the penetration of cefpirome, a broad-spectrum cephalosporin, into subcutaneous adipose tissue of infected and healthy subjects. Cefpirome was given as an intravenous injection of 2 g. Tissue cefpirome concentrations in septic patients (N ¼ 11) and healthy controls (N ¼ 7) were determined over a period of 4 h by microdialysis. To estimate inhibition of bacterial growth of speciﬁc strains of Staphylococcus aureus and Pseudomonas aeruginosa at the target site, a pharmacokinetic/pharmacodynamic model was used to simulate concentration–time proﬁles at the site of infection. Tissue exposure of cefpirome was significantly reduced (by 47%) in septic patients compared with that in healthy subjects. Nonetheless, effective inhibition of bacterial growth was observed in all simulations in vitro. The results suggest that cefpirome would be an appropriate agent for the treatment of STIs in septic patients, assuming that dosing intervals of 8 h or less are used.

618

III.C.4. Poisoning A novel study by Bueters et al. (2003) examined the effect of sarin poisoning on the pharmacokinetics and brain distribution of the adenosine A1 receptor partial agonist, 20 -deoxy-N6-cyclopentyladenosine (20 -dCPA), and the effect of 20 -dCPA (20 mg/kgIV) on acetylcholine (ACh) release and protection against sarin poisoning. A multi-compartment model was used to describe the concentration–time proﬁle of 20 -dCPA in brain microdialysate and blood. The volume of distribution of 20 -dCPA in blood was increased in sarinpoisoned rats, but the clearance of 20 -dCPA from blood to brain was unchanged. In addition, bloodto-brain ratios of the area under the curve were identical in control and poisoned rats. Attempting to protect sarin-poisoned rats with 20 -dCPA did not prevent accumulation of ACh in the central nervous system, perhaps because insufﬁcient amounts of 20 -dCPA entered the central nervous system. Although 20 -dCPA did slow the development of symptoms, only 29% of the sarin-treated animals survived for 24 h. The proﬁle of 20 -dCPA in blood was slightly altered by sarin, but its distribution into the brain was unchanged. III.C.5. Surgery Patients with limited blood perfusion of soft tissue because of peripheral arterial occlusive disease may suffer from tissue infections and therefore require antibiotic therapy. These patients may undergo angioplasty in an effort to enhance arterial blood ﬂow. Joukhadar et al. (2001) studied the effect of this procedure on antibiotic concentrations in soft tissues in patients scheduled to undergo percutaneous transluminal angioplasty. Following ciproﬂoxacin doses of 400 mg, plasma concentrations in ischemic and healthy soft tissues were measured before and after angioplasty. Unbound antibiotic concentrations were determined at the site of infection by in vivo microdialysis. Before angioplasty, the median area under the ciproﬂoxacin concentration–time curve measured over 5 h was lower by 37% in ischemic tissue, but following angioplasty, these median areas were not significantly different. Based on tissue concentration data and in vitro antimicrobial data, the authors suggest that the observed improvement in

arterial blood ﬂow would be associated with increased cure rates of STIs in these patients. Kenkel et al. (2004) studied systemic and tissue exposure to lidocaine and its pharmacologically active metabolite, monoethylglycinexylidide, in ﬁve female patients undergoing liposuction and receiving lidocaine by inﬁltration. Concentrations of lidocaine and monoethylglycinexylidide in blood and lipoaspirate were measured during the procedure. Lidocaine and monoethylglycinexylidide levels were measured after surgery using microdialysis in vivo. The combined peak concentration (lidocaine plus monoethylglycinexylidide) occurred 8–28 h after inﬁltration began. Absorbed lidocaine was estimated to represent, on average, 64% of the inﬁltrated dose. Analysis of lipoaspirate indicated that an average of 9.7% of the inﬁltrated dose was removed during liposuction. Combined peak levels of drug and metabolite were deemed to be within safe limits. Microdialysis results demonstrated that tissue lidocaine levels may be subtherapeutic within 4–8 h of the procedure. III.C.6. Anesthesia Dias and Mitra (2003) established a conscious rabbit model to study the effects of anesthesia and microdialysis probe implantation on the ocular pharmacokinetics of GCV. Rabbits were studied in three groups. Group I consisted of rabbits that were anesthetized throughout the experiment, with no recovery period after probe implantation. Group II consisted of animals that had a more than 5-day recovery period after probe implantation and were unanesthetized during the pharmacokinetic study. Group III consisted of rabbits that had more than 5 days of recovery, and were anesthetized during the study. Tritiated GCV was administered (50 mL) intravitreously in all animals, and ocular vitreous levels were monitored for 10 h. Noncompartmental analysis of the data was performed. The anesthetized animals (Groups I and III) exhibited higher areas under the curve than did the unanesthetized group, whereas the vitreous half-life of GCV was shown to be significantly shorter in the groups with a recovery period of more than 5 days (Groups II and III) compared with Group I (no recovery from anesthesia after probe implantation).

619

To investigate intranasal administration as a potential route of enhancing brain delivery of stavudine (d4 T), Yang et al. (2005) employed microdialysis coupled on-line to HPLC–UV to sample brain frontal cortex, caudate putamen, and ventricular CSF. AZdU was used as a retrodialysis calibrator throughout the sampling periods. Sprague–Dawley rats were divided into two groups of seven animals. One group received intranasal administration of 5 mg/kg d4 T, whereas the other group received the same dose intravenously following sham nasal surgery. After intranasal administration, d4 T was rapidly and completely absorbed into the systemic circulation. The brain:plasma AUC ratios in the lateral ventricle, caudate putamen, and frontal cortex in the anesthetized (intraperitoneal 50 mg/kg sodium pentobarbital) and nasal surgery-operated rats were 0.3670.090, 0.4770.089, and 0.4170.087, respectively, whereas these ratios were 0.637 0.077, 0.6270.17, and 0.6070.13, respectively, following IV dosing to sham animals. The halflife of d4 T in the various brain regions was significantly longer than that in plasma. In addition, the systemic clearance of d4 T was significantly reduced in these anesthetized and nasal surgeryoperated animals. The extent of the brain distribution, however, as measured by brain:plasma AUC ratios, was not significantly affected by anesthesia. The authors question the utility of intranasal administration for direct delivery of small molecules into brain tissues, particularly where passive diffusion predominates.

IV. The importance of characterizing recovery – the good, the bad, and the ugly in microdialysis sampling in pharmacokinetic studies The critical need to carefully characterize relative recovery of a drug or its metabolites during microdialysis sampling in pharmacokinetic studies cannot be overstated. Dialysate concentrations reﬂect the concentration at the probed site, but unless the relationship between these two variables is known, this sampling technique cannot provide quantitative information. Even though careful attention is given to study design and validation of

analytical methodology, if recovery is not properly measured, important pharmacokinetic parameters (e.g., volume and clearance) will not be characterized with adequate precision or accuracy. It is clear that recoveries must be determined in vivo, unless the investigator has validated the use of recoveries assessed in vitro under carefully controlled experimental conditions and has provided convincing evidence that these are not significantly different. In addition, if recovery or probe function is measured prior to the pharmacokinetic study, some assurance should be provided that this has remained unchanged over the course of the experiment. For example, retrodialysis by drug may be performed prior to dosing and also after the study has been completed, to assure that recoveries have not drifted during the course of microdialysis sampling. One should be very skeptical about pharmacokinetic parameters derived from studies utilizing microdialysis sampling where investigators fail to determine recovery in the same time-frame of the study, but rather have adopted recovery values from previous experiments, even when the same drug and experimental conditions are involved. Finally, the results of pharmacokinetic studies that rely upon assumed relative recoveries for unrelated molecules (or for the same compound determined under different conditions), or that do not address the need for recovery measurements at all, should not be deemed reliable.

References Anand, B.S., Atluri, H. and Mitra, A.K. (2004) Validation of an ocular microdialysis technique in rabbits with permanently implanted vitreous probes: systemic and intravitreal pharmacokinetics of ﬂuorescein. Int. J. Pharm., 281: 79–88. Bagger, M.A. and Bechgaard, E. (2004) The potential of nasal application for delivery to the central brain – a microdialysis study of ﬂuorescein in rats. Eur. J. Pharm. Sci., 21: 235–242. Bellmann, R., Kuchling, G., Dehghanyar, P., Zeitlinger, M., Minar, E., Mayer, B.X., Muller, M. and Joukhadar, C. (2004) Tissue pharmacokinetics of levoﬂoxacin in human soft tissue infections. Br. J. Clin. Pharmacol., 57: 563–568. Benveniste, H. (1989) Brain microdialysis. J. Neurochem., 52: 1667–1679. Bergenheim, A.T., Capala, J., Roslin, M. and Henriksson, R. (2005) Distribution of BPA and metabolic assessment in glioblastoma patients during BNCT treatment: a microdialysis study. J. Neurooncol., 71: 287–293.

620 Bernards, C.M., Shen, D.D., Sterling, E.S., Adkins, J.E., Risler, L., Phillips, B. and Ummenhofer, W. (2003a) Epidural, cerebrospinal ﬂuid, and plasma pharmacokinetics of epidural opioids (part 1): differences among opioids. Anesthesiology, 99: 455–465. Bernards, C.M., Shen, D.D., Sterling, E.S., Adkins, J.E., Risler, L., Phillips, B. and Ummenhofer, W. (2003b) Epidural, cerebrospinal ﬂuid, and plasma pharmacokinetics of epidural opioids (part 2): effect of epinephrine. Anesthesiology, 99: 466–475. Boschi, G. and Scherrmann, J. (2000) Microdialysis in mice for drug delivery research. Adv. Drug Deliv. Rev., 45: 271–281. Bouw, R., Ederoth, P., Lundberg, J., Ungerstedt, U., Nordstrom, C.H. and Hammarlund-Udenaes, M. (2001) Increased blood-brain barrier permeability of morphine in a patient with severe brain lesions as determined by microdialysis. Acta Anaesthesiol. Scand., 45: 390–392. Brunner, M. and Muller, M. (2002) Microdialysis: an in vivo approach for measuring drug delivery in oncology. Eur. J. Clin. Pharmacol., 58: 227–234. Bueters, T.J., Ijzerman, A.P., van Helden, H.P. and Danhof, M. (2003) Characterization of the pharmacokinetics, brain distribution, and therapeutic efﬁcacy of the adenosine A1 receptor partial agonist 20 -deoxy-N6-cyclopentyladenosine in sarin-poisoned rats. Toxicol. Appl. Pharmacol., 192: 86–94. Bungay, P.M., Dedrick, R.L., Fox, E. and Balis, F.M. (2001) Probe calibration in transient microdialysis in vivo. Pharm. Res., 18: 361–366. Cano-Cebrian, M.J., Zornoza, T., Polache, A. and Granero, L. (2005) Quantitative in vivo microdialysis in pharmacokinetic studies: some reminders. Curr. Drug Metab., 6: 83–90. Chang, Y.L., Chou, M.H., Lin, M.F., Chen, C.F. and Tsai, T.H. (2001) Effect of cyclosporine, a P-glycoprotein inhibitor, on the pharmacokinetics of cefepime in rat blood and brain: a microdialysis study. Life Sci., 69: 191–199. Chang, Y.L., Tsai, P.L., Chou, Y.C., Tien, J.H. and Tsai, T.H. (2005) Simultaneous determination of nicotine and its metabolite, cotinine, in rat blood and brain tissue using microdialysis coupled with liquid chromatography: pharmacokinetic application. J. Chromatogr. A, 1088: 152–157. Chu, J. and Gallo, J.M. (2000) Application of microdialysis to characterize drug disposition in tumors. Adv. Drug Deliv. Rev., 45: 243–253. Clement, R., Malinovsky, J., Le Corre, P., Dollo, G., Chevanne, F. and Le Verge, R. (2000) Spinal biopharmaceutics of bupivacaine and lidocaine by microdialysis after their simultaneous administration in rabbits. Int. J. Pharm., 203: 227–234. Dabrosin, C. (2005) Increased extracellular local levels of estradiol in normal breast in vivo during the luteal phase of the menstrual cycle. J. Endocrinol., 187: 103–108. Dai, H., Chen, Y. and Elmquist, W.F. (2005) Distribution of the novel antifolate pemetrexed to the brain. J. Pharmacol. Exp. Ther., 315: 222–229. de la Pena, A., Liu, P. and Derendorf, H. (2000) Microdialysis in peripheral tissues. Adv. Drug Deliv. Rev., 45: 189–216.

de Lange, E.C., de Boer, A.G. and Breimer, D.D. (2000) Methodological issues in microdialysis sampling for pharmacokinetic studies. Adv. Drug Deliv. Rev., 45: 125–148. Dias, C.S. and Mitra, A.K. (2003) Posterior segment ocular pharmacokinetics using microdialysis in a conscious rabbit model. Invest. Ophthalmol. Vis. Sci., 44: 300–305. Dukic, S.F., Kaltenbach, M.L., Heurtaux, T., Hoizey, G., Lallemand, A. and Vistelle, R. (2004) Inﬂuence of C6 and CNS1 brain tumors on methotrexate pharmacokinetics in plasma and brain tissue. J. Neurooncol., 67: 131–138. Egawa, M., Hoebel, B.G. and Stone, E.A. (1988) Use of microdialysis to measure brain noradrenergic receptor function in vivo. Brain Res., 458: 303–308. Elmquist, W.F. and Sawchuk, R.J. (1997) Application of microdialysis in pharmacokinetic studies. Pharm. Res., 14: 267–288. Freddo, R.J. and Dalla Costa, T. (2002) Determination of norfloxacin free interstitial levels in skeletal muscle by microdialysis. J. Pharm. Sci., 91: 2433–2440. Galvan, A., Smith, Y. and Wichmann, T. (2003) Continuous monitoring of intracerebral glutamate levels in awake monkeys using microdialysis and enzyme ﬂuorometric detection. J. Neurosci. Methods, 126: 175–185. Garrison, K.E., Pasas, S.A., Cooper, J.D. and Davies, M.I. (2002) A review of membrane sampling from biological tissues with applications in pharmacokinetics, metabolism and pharmacodynamics. Eur. J. Pharm. Sci., 17: 1–12. Gattringer, R., Urbauer, E., Traunmuller, F., Zeitlinger, M., Dehghanyar, P., Zeleny, P., Graninger, W., Muller, M. and Joukhadar, C. (2004) Pharmacokinetics of telithromycin in plasma and soft tissues after single-dose administration to healthy volunteers. Antimicrob. Agents Chemother., 48: 4650–4653. Hoizey, G., Kaltenbach, M.L., Dukic, S., Lamiable, D., Millart, H., D’Arbigny, P. and Vistelle, R. (2001) Pharmacokinetics of gacyclidine enantiomers in plasma and spinal cord after single enantiomer administration in rats. Int. J. Pharm., 229: 147–153. Huang, Y., Ji, P., Inano, A., Yang, Z., Giebink, G.S. and Sawchuk, R.J. (2001) Microdialysis studies of the middle ear distribution kinetics of amoxicillin in the awake chinchilla. J. Pharm. Sci., 90: 2088–2098. Huang, S.M., Tsai, T.R., Yeh, P.H. and Tsai, T.H. (2005) Measurement of unbound ranitidine in blood and bile of anesthetized rats using microdialysis coupled to liquid chromatography and its pharmacokinetic application. J. Chromatogr. A, 1073: 297–302. Hunter, B.A., Lee, S., Odland, R.M. and Juhn, S.K. (2003) Estimation of perilymph concentration of agents applied to the round window membrane by microdialysis. Acta Otolaryngol., 123: 453–458. Jacobson, J., Sandberg, M. and Hamberger, A. (1985) Mass transfer in brain dialysis devices: a new method for estimation of extracellular amino acids concentration. J. Neurosci. Methods, 15: 263–268. Johansen, M.J., Thapar, N., Newman, R.A. and Madden, T. (2002) Use of microdialysis to study platinum anticancer

621 agent pharmacokinetics in preclinical models. J. Exp. Ther. Oncol., 2: 163–173. Joukhadar, C., Klein, N., Frossard, M., Minar, E., Stass, H., Lackner, E., Herrmann, M., Riedmuller, E. and Muller, M. (2001) Angioplasty increases target site concentrations of ciproﬂoxacin in patients with peripheral arterial occlusive disease. Clin. Pharmacol. Ther., 70: 532–539. Joukhadar, C., Stass, H., Muller-Zellenberg, U., Lackner, E., Kovar, F., Minar, E. and Muller, M. (2003) Penetration of moxiﬂoxacin into healthy and inﬂamed subcutaneous adipose tissues in humans. Antimicrob. Agents Chemother., 47: 3099–3103. Juan, Y.P. and Tsai, T.H. (2005) Measurement and pharmacokinetics of vincamine in rat blood and brain using microdialysis. J. Chromatogr. A, 1088: 146–151. Kenkel, J.M., Lipschitz, A.H., Shepherd, G., Armstrong, V.W., Streit, F., Oellerich, M., Luby, M., Rohrich, R.J. and Brown, S.A. (2004) Pharmacokinetics and safety of lidocaine and monoethylglycinexylidide in liposuction: a microdialysis study. Plast. Reconstr. Surg., 114: 516–524. Kopacz, D.J., Bernards, C.M., Allen, H.W., Landau, C., Nandy, P., Wu, D. and Lacouture, P.G. (2003) A model to evaluate the pharmacokinetic and pharmacodynamic variables of extended-release products using in vivo tissue microdialysis in humans: bupivacaine-loaded microcapsules. Anesth. Analg., 97: 124–131. Kreilgaard, M. (2001) Dermal pharmacokinetics of microemulsion formulations determined by in vivo microdialysis. Pharm. Res., 18: 367–373. Kreilgaard, M. (2002) Assessment of cutaneous drug delivery using microdialysis. Adv. Drug Deliv. Rev., 54: S99–S121. Kreilgaard, M., Kemme, M.J., Burggraaf, J., Schoemaker, R.C. and Cohen, A.F. (2001) Inﬂuence of a microemulsion vehicle on cutaneous bioequivalence of a lipophilic model drug assessed by microdialysis and pharmacodynamics. Pharm. Res., 18: 593–599. Krupp, J.L. and Bernards, C.M. (2004) Pharmacokinetics of intrathecal oligodeoxynucleotides. Anesthesiology, 100: 315–322. Lee, C.H., Yeh, P.H. and Tsai, T.H. (2002) Hepatobiliary excretion of ﬂuconazole and its interaction with cyclosporin A in rat blood and bile using microdialysis. Int. J. Pharm., 241: 367–373. Liu, P., Fuhrherr, R., Webb, A.I., Obermann, B. and Derendorf, H. (2005) Tissue penetration of cefpodoxime into the skeletal muscle and lung in rats. Eur. J. Pharm. Sci., 25: 439–444. Lonnroth, P. (1991) Microdialysis: a new and promising method in clinical medicine. J. Intern. Med., 230: 363–364. Lonnroth, P., Jansson, P.A. and Smith, U. (1987) A microdialysis method allowing characterization of intercellular water space in humans. Am. J. Physiol., 253: E228–E231. Macha, S. and Mitra, A.K. (2001) Ocular pharmacokinetics of cephalosporins using microdialysis. J. Ocul. Pharmacol. Ther., 17: 485–498. Macha, S. and Mitra, A.K. (2002) Ocular disposition of ganciclovir and its monoester prodrugs following intravitreal

administration using microdialysis. Drug Metab. Dispos., 30: 670–675. Mader, R.M., Schrolnberger, C., Rizovski, B., Brunner, M., Wenzel, C., Locker, G., Eichler, H.G., Mueller, M. and Steger, G.G. (2003) Penetration of capecitabine and its metabolites into malignant and healthy tissues of patients with advanced breast cancer. Br. J. Cancer, 88: 782–787. Marchand, S., Chenel, M., Lamarche, I., Pariat, C. and Couet, W. (2003) Dose ranging pharmacokinetics and brain distribution of norfloxacin using microdialysis in rats. J. Pharm. Sci., 92: 2458–2465. Mathy, F.X., Ntivunwa, D., Verbeeck, R.K. and Preat, V. (2005) Fluconazole distribution in rat dermis following intravenous and topical application: a microdialysis study. J. Pharm. Sci., 94: 770–780. Mathy, F.X., Preat, V. and Verbeeck, R.K. (2001) Validation of subcutaneous microdialysis sampling for pharmacokinetic studies of ﬂurbiprofen in the rat. J. Pharm. Sci., 90: 1897–1906. Muller, M. (2000) Microdialysis in clinical drug delivery studies. Adv. Drug Deliv. Rev., 45: 255–269. Nishiyama, A., Seth, D.M. and Navar, L.G. (2002) Renal interstitial ﬂuid angiotensin I and angiotensin II concentrations during local angiotensin-converting enzyme inhibition. J. Am. Soc. Nephrol., 13: 2207–2212. Olson, R.J. and Justice, J.B. Jr. (1993) Quantitative microdialysis under transient conditions. Anal. Chem., 65: 1017–1022. Persky, A.M., Muller, M., Derendorf, H., Grant, M., Brazeau, G.A. and Hochhaus, G. (2003) Single- and multiple-dose pharmacokinetics of oral creatine. J. Clin. Pharmacol., 43: 29–37. Plock, N. and Kloft, C. (2005) Microdialysis – theoretical background and recent implementation in applied life-sciences. Eur. J. Pharm. Sci., 25: 1–24. Plontke, S., Siedow, N., Hahn, H., Wegener, R., Zenner, H.P. and Salt, A.N. (2004) 1D- and 3D-computer simulation for experimental planning and interpretation of pharmacokinetic studies in the inner ear after local drug delivery. AltexAlternativen Zu Tierexperimenten, 21(Suppl. 3): 77–85. Qian, M., West, W., Wu, J.T., Lu, B. and Christ, D.D. (2003) Development of a dog microdialysis model for determining synovial ﬂuid pharmacokinetics of anti-arthritis compounds exempliﬁed by methotrexate. Pharm. Res., 20: 605–610. Riese, J., Boecker, S., Hohenberger, W., Klein, P. and Haupt, W. (2003) Microdialysis: a new technique to monitor perioperative human peritoneal mediator production. Surg. Infect., 4: 11–15. Robinson, D.L., Brunner, L.J. and Gonzales, R.A. (2002) Effect of gender and estrous cycle on the pharmacokinetics of ethanol in the rat brain. Alcohol. Clin. Exp. Res., 26: 165–172. Rojas, C., Nagaraja, N.V., Webb, A.I. and Derendorf, H. (2003) Microdialysis of triamcinolone acetonide in rat muscle. J. Pharm. Sci., 92: 394–397. Sasongko, L., Ramzan, I., Williams, K.M. and McLachlan, A.J. (2002) Muscle distribution of the neuromuscular blocker gallamine using microdialysis. J. Pharm. Sci., 91: 769–775.

622 Sauermann, R., Delle-Karth, G., Marsik, C., Steiner, I., Zeitlinger, M., Mayer-Helm, B.X., Georgopoulos, A., Muller, M. and Joukhadar, C. (2005) Pharmacokinetics and pharmacodynamics of cefpirome in subcutaneous adipose tissue of septic patients. Antimicrob. Agents Chemother., 49: 650–655. Skilling, S.R., Smullin, D.H., Beitz, A.J. and Larson, A.A. (1988) Extracellular amino acid concentrations in the dorsal spinal cord of freely moving rats following veratridine and nociceptive stimulation. J. Neurochem., 51: 127–132. Solliga˚rd, E., Juel, I.S., Bakkelund, K., Jynge, P., Tvedt, K.E., Johnsen, H., Aadahl, P. and Grønbech, J.E. (2005) Gut luminal microdialysis of glycerol as a marker of intestinal ischemic injury and recovery. Crit. Care Med., 33: 2278–2285. Song, Y., Cartier, L., Cheung, B. and Sawchuk, R. (2004) Pharmacokinetics of Zidovudine and Stavudine in the Cerebrospinal Fluid during Intrathecal Administration in a Novel Rabbit Model. American Association of Pharmaceutical Scientists, Annual Meeting Abstracts, p. 6. Stagni, G., Ali, M.E. and Weng, D. (2004) Pharmacokinetics of acyclovir in rabbit skin after i.v.-bolus, ointment, and iontophoretic administrations. Int. J. Pharm., 274: 201–211. Stahle, L. (1993) Zero and ﬁrst moment area estimation from microdialysis data. Eur. J. Pharmacol., 45: 477–481. Stolle, L.B., Arpi, M., Holmberg-Jorgensen, P., Riegels-Nielsen, P. and Keller, J. (2004) Application of microdialysis to cancellous bone tissue for measurement of gentamicin levels. J. Antimicrob. Chemother., 54: 263–265. Tokunaga, Y., Nakashima, M., Sasaki, H., Tomiyama, N., Nakashima, M.N., Ichikawa, M., Kaminogo, M. and Shibata, S. (2000) Local distribution into brain tumor and pharmacokinetics of 4-pyridoxate diammine hydroxy platinum, a novel cisplatin derivative, after intracarotid administration in rats with 9L malignant glioma: simultaneous brain microdialysis study. Biol. Pharm. Bull., 23: 1491–1496. Tomaselli, F., Dittrich, P., Maier, A., Woltsche, M., Matzi, V., Pinter, J., Nuhsbaumer, S., Pinter, H., Smolle, J. and SmolleJuttner, F.M. (2003a) Penetration of piperacillin and tazobactam into pneumonic human lung tissue measured by in vivo microdialysis. Br. J. Clin. Pharmacol., 55: 620–624. Tomaselli, F., Maier, A. and Smolle-Juttner, F.M. (2003b) Pharmacokinetics of antibiotics in inﬂamed and healthy lung tissue. Wien. Med. Wochenschr., 153: 342–344. Tong, X. and Patsalos, P.N. (2001) A microdialysis study of the novel antiepileptic drug levetiracetam: extracellular pharmacokinetics and effect on taurine in rat brain. Br. J. Pharmacol., 133: 867–874. Tossman, U. and Ungerstedt, U. (1986) Microdialysis in the study of extracellular levels of amino acids in the rat brain. Acta Physiol. Scand., 128: 9–14. Tsai, T.H., Lee, C.H. and Yeh, P.H. (2001) Effect of P-glycoprotein modulators on the pharmacokinetics of camptothecin using microdialysis. Br. J. Pharmacol., 134: 1245–1252.

Tsai, T.H. and Liu, M.C. (2004) Determination of extracellular hesperidin in blood and bile of anaesthetized rats by microdialysis with high-performance liquid chromatography: a pharmacokinetic application. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 806: 161–166. Tsai, P.L. and Tsai, T.H. (2004) Hepatobiliary excretion of berberine. Drug Metab. Dispos., 32: 405–412. Tunblad, K., Ederoth, P., Gardenfors, A., HammarlundUdenaes, M. and Nordstrom, C.H. (2004a) Altered brain exposure of morphine in experimental meningitis studied with microdialysis. Acta Anaesthesiol. Scand., 48: 294–301. Tunblad, K., Hammarlund-Udenaes, M. and Jonsson, E.N. (2004b) An integrated model for the analysis of pharmacokinetic data from microdialysis experiments. Pharm. Res., 21: 1698–1707. Ueng, Y.F., Tsai, T.H., Don, M.J., Chen, R.M. and Chen, T.L. (2005) Alteration of the pharmacokinetics of theophylline by rutaecarpine, an alkaloid of the medicinal herb Evodia rutaecarpa, in rats. J. Pharm. Pharmacol., 57: 227–232. Ungerstedt, U. and Hallstrom, A. (1987) In vivo microdialysis – a new approach to the analysis of neurotransmitters in the brain. Life Sci., 41: 861–864. Verbeeck, R.K. (2000) Blood microdialysis in pharmacokinetic and drug metabolism studies. Adv. Drug Deliv. Rev., 45: 217–228. Wang, X., Ratnaraj, N. and Patsalos, P.N. (2004) The pharmacokinetic inter-relationship of tiagabine in blood, cerebrospinal ﬂuid and brain extracellular ﬂuid (frontal cortex and hippocampus). Seizure, 13: 574–581. Wang, Y. and Sawchuk, R.J. (1995) Zidovudine transport in the rabbit brain during intravenous and intracerebroventricular infusion. J. Pharm. Sci., 84: 871–876. Wang, Y., Wong, S.L. and Sawchuk, R.J. (1991) Comparison of in vitro and in vivo calibration of microdialysis probes using microdialysis. Curr.Sep., 10: 87. Xie, R., Bouw, M.R. and Hammarlund-Udenaes, M. (2000) Modeling of the blood-brain barrier transport of morphine3-glucuronide studied using microdialysis in the rat: involvement of probenecid-sensitive transport. Br. J. Pharmacol., 131: 1784–1792. Yang, Z., Huang, Y., Gan, G. and Sawchuk, R.J. (2005) Microdialysis evaluation of the brain distribution of stavudine following intranasal and intravenous administration to rats. J. Pharm. Sci., 94: 1577–1588. Zhang, J., Qu, F.R., Nakatsuka, A., Nomura, T., Nagai, M. and Nomoto, M. (2003) Pharmacokinetics of L-dopa in plasma and extracellular ﬂuid of striatum in common marmosets. Brain Res., 993: 54–58. Zhu, T., Cheung, B.W., Cartier, L.L., Giebink, G.S. and Sawchuk, R.J. (2003) Simultaneous intravenous and intramiddleear dosing to determine cefditoren inﬂux and efﬂux clearances in middle ear ﬂuid in freely moving chinchillas. J. Pharm. Sci., 92: 1947–1956.

CHAPTER 7.1

Microdialysis in clinical drug delivery studies Martin Brunner and Markus Mu¨ller Department of Clinical Pharmacology, Medical University of Vienna, A-1090 Vienna, Austria

Abstract: Appropriate methods to directly measure drug concentrations at their actual sites of action within tissues and organs have not been available in clinical drug delivery studies for many years. Consequently, pharmacokinetic research was long restricted to drug concentration measurements from biological specimens that are relatively easy to obtain, such as tissue biopsies, urine, saliva, or skin blister ﬂuid, or to indirect modeling of tissue concentrations from plasma concentration curves. Among several techniques, which have become available for in vivo tissue distribution studies in the last years, in vivo MD in particular is a rational and scientiﬁcally sound means to directly measure concentrations of unbound drugs at their site of action in virtually every tissue and organ of the human body, as it provides access to the interstitial space, the site of action for many drugs. During the last 10–15 years, an increasing number of clinical MD studies have substantially expanded the existing knowledge about the concept of tissue distribution in healthy volunteers and patients for a variety of compounds including antibiotics, anticancer drugs, and transdermally applied substances. In combination with pharmacodynamic simulations, data derived from MD studies have the potential to serve as the basis for the prediction of drug effects at the target site and drug administration recommendations. MD will also continue to be an important clinical research tool to address various issues in different clinical settings and might help to deﬁne meaningful surrogate markers for drug efﬁciency along the critical path of drug development later, we are in the fortunate position to know more about distinct mechanisms and pathways that contribute to dose–response variability. Besides polymorphisms in genes encoding for enzymes, drug transporters or receptor proteins (Evans and McLeod, 2003; Wilkinson, 2005), it has been recognized that most drugs, with few notable exceptions like heparin, exert their action in tissues rather than in plasma and do not distribute uniformly in the body but rather attain varying concentrations in different tissues. Assessing tissue chemistry and pharmacology has, thus, for long been viewed as a more rational way to provide clinically meaningful data on dose–response variability than gaining information from blood samples. The in vivo assessment of drug distribution and target site pharmacokinetics (PK), however, has long been treated as a ‘‘forgotten relative’’ by

I. Tissue distribution and drug response variability Drug response in patients receiving comparable doses of the same medication might range from therapeutic failure to toxic reactions. This frequently encountered variability in the dose–response cascade is often unpredictable and thus represents a signiﬁcant challenge for physicians. The notion that drug response is highly variable is not new. Sir William Osler, one of the most inﬂuential American physicians at the turn of the 20th century commented on this problem in 1903 as follows: ‘‘Variability is the law of life, and as no two faces are the same, so no two bodies are alike, and no two individuals react alike, and behave alikey’’ (Osler, 1903). Approximately one century Corresponding author: E-mail: markus.mueller@meduniwien. ac.at

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

625

DOI: 10.1016/S1569-7339(06)16033-6 Copyright 2007 Elsevier B.V. All rights reserved

626

physicians, pharmacists, and clinical pharmacologists alike (Eichler and Mu¨ller, 1998). The main reason for this neglect was largely the lack of appropriate methodology providing in vivo access to the target sites in tissues and organs and consequently, PK research was long restricted to drug concentration measurements from biological specimens that are relatively easy to obtain, such as tissue biopsies, urine, saliva, or skin blister ﬂuid, or to indirect modeling of tissue concentrations from plasma concentration curves, which in most cases only served as surrogates for true target site concentrations. Data obtained by these approaches, however, caused considerable confusion about drug distribution and target site delivery, as their interpretation has been ﬂawed by several misconceptions (Mu¨ller et al., 2004). First, it has been assumed that indirect modeling of tissue drug levels from plasma drug concentration provides useful information in many cases. However, it must be kept in mind that this assumes rapid, unrestricted, and homogeneous diffusion processes in hypothetical spaces – assumptions that do not hold true in reality. Furthermore, the body is a multimillion-compartment model, and studies of capillary physiology and PK have shown that rate constants for analyte transfer from plasma to tissues are heterogeneous and tissue speciﬁc. Second, tissue is not a uniform matrix. Although drug concentrations may be readily measured from total tissue biopsy specimens, total concentration measurements may be misleading for several reasons. Most importantly, it must be considered that the actual target space is not just ‘‘tissue’’ but deﬁned compartments within tissues, such as the interstitial space ﬂuid (ISF) for most anti-infectives. Biopsy samples, however, represent a homogenous matrix of intracellular, extracellular and vascular compartments and the concentrations of antibacterials are determined in the supernatant after centrifugation and correction for the weight of the biopsy sample. If only total tissue drug concentrations are measured by taking biopsies, then the effective site concentrations of drugs that equilibrate exclusively with the extracellular space may be underestimated (Mu¨ller et al., 1996). This situation in turn will lead to an

overestimation of the effective target site concentrations of intracellularly accumulating drugs if the actual effect site is the ISF (Mu¨ller et al., 1996). Thus, the admixture of various compartmental ﬂuids leads to a hybrid tissue drug concentration, which is difﬁcult to interpret. A third misconception is the notion that the entire drug fraction present in various tissue spaces exerts pharmacologic activity. In fact, it has been shown that only the unbound drug fraction has the ability to exert drug effects (Merrikin et al., 1983; Craig and Ebert, 1989). Besides the fact that only this free fraction exerts activity, it is also only the free fraction that has the ability to be distributed to the target site. This information was experimentally shown by various investigators, who found that differences in penetration were directly related to the free drug concentrations in serum (Liu et al., 2005a). Although this concept has been best described for antibacterial agents, it seems reasonable to assume that similar concepts hold true for any other drug formulation. Altogether, these considerations have led to the conclusion that an appropriate deﬁnition of tissue drug concentrations should imply in many cases the meaning of unbound drug concentrations at anatomically distinct sites such as the ISF and that a suitable method for the measurement of ‘‘tissue’’ drug concentrations should allow for the direct measurement of unbound drug concentrations in a clearly deﬁned space within the tissue of a given organ. The last years have seen the introduction of several new techniques and approaches for the assessment of drug distribution and target tissue PK in humans (Langer and Mu¨ller, 2004), including in vivo microdialysis (MD) and imaging techniques, such as magnetic resonance spectroscopy (MRS) and positron emission tomography (PET). Results from studies using these techniques have underlined the importance of the previously neglected drug distribution process to the target site as a crucial determinant for clinical outcome. Furthermore, regulatory guidance documents issued by the Food and Drug Adminstration (FDA) in the US and the Committee for Proprietary Medicinal Products (CPMP) in Europe have emphasized value and importance of human tissue drug

627

concentration data and conceptually support, in particular, the use of clinical MD to obtain this information.1,2 Besides other documents, and in light of the critical path initiative,3 the Centre for Drug Evaluation and Research (CDER) report to the nation 2003 indicates a need for tools that enable the measurement of tissue concentrations by stating that: CDER continues to extend its ‘‘long-standing interest in the application of dose–response principles by viewing drugs and their actions directly at the level of the drug target, rather than indirectly via plasma concentrations’’.4

II. In vivo microdialysis in healthy human subjects and patients Compared with other techniques, MD in healthy human subjects or patients is comparably cheap and theoretically, MD experiments could take place in any clinical research institution. Probe insertion into soft tissues is fairly easy and can be performed by physicians without the need to require extensive additional skills, as the process is only slightly different from standard intramuscular or subcutaneous injections. For the subjects the pain from probe insertion is minimal and many do not report any discomfort during the entire procedure at all. As MD provides selective access to the ISF of tissues and allows for the measurement of unbound target site PKs of most drugs, it can be regarded as a suitable scientiﬁc tool to satisfy regulatory requirements for PK distribution studies. However, much experience and individual probe calibration are required for the correction of the relative dialysate concentration data to absolute tissue concentrations. The roots of MD date back to the early 1960s, when push–pull cannulas, dialysis sacs, and dialytrodes were inserted into animal tissues to directly

1 http://www.fda.gov/cder/present/anti-infective798/ 073198.pdf [last accessed December 3, 2005]. 2 http://www.fda.gov/cder/guidance/2580dft.pdf [last accessed December 3, 2005]. 3 http://www.fda.gov/oc/initiatives/criticalpath/whitepaper.html [last accessed December 3, 2005]. 4 http://www.fda.gov/cder/reports/rtn/2003/rtn2003.PDF [last accessed December 3, 2005].

study tissue biochemistry. In the following years, these sampling devices were steadily improved and ﬁnally resulted in the type of MD probes currently in use in humans. Cornerstones of a constant development toward MD as a widely used clinical technique, at present with more than 1,500 original publications in humans reported in the PubMed database,5 were (1) the ﬁrst publication of MD in humans characterizing interstitial glucose concentrations in healthy volunteers in 1987 (Lo¨nnroth et al., 1987), (2) ﬁrst reports on MD in clinical drug delivery studies in 1991 (Lo¨nnroth et al., 1991), (3) clinical applications for in vivo studies in brain (Benveniste, 1989) or lung tissue (Herkner et al., 2002), and (4) the availability of FDA- and CE-approved MD probes for the use in peripheral human tissues. Today, MD equipment is marketed by several companies and special probes are available for insertion into different human tissues, such as soft tissues, brain, liver, and the peritoneal cavity. MD is feasible in most human organs and is performed in a clinical setting in >50 centers worldwide. The cost of MD probes ranges from h100 to h250 per probe. II.A. Microdialysis in clinical drug delivery studies in humans: theoretical considerations II.A.1. In vivo probe calibration Although different techniques have been developed for in vivo calibration of MD probes, the ‘‘equilibrium method’’ is considered the method of choice (Lo¨nnroth et al., 1987). This technique, however, is time consuming and requires steadystate conditions, which are usually not attainable in clinical drug delivery studies. A simpliﬁed, less time consuming and comparably reproducible technique is called ‘‘retrodialysis’’ or ‘‘reverse dialysis’’. This approach has initially been proposed by Sta˚hle et al. (1991) and has proven practical for the use in human studies. The principle of this method relies on the fact that the diffusion process is quantitatively equal in both directions across the semipermeable membrane. Therefore, the study

5 http://www.ncbi.nlm.nih.gov/entrez [last accessed December 3, 2005].

628

drug can be added to the perfusion medium at a known concentration (Cperfusate) and its disappearance rate across the membrane, which is termed recovery, is determined. The in vivo recovery value can then be calculated by the following formula: Recovery (%) ¼ 100–(100 Cdialysate/Cperfusate). In vivo recovery is dependent on the pharmacological and chemical properties of the analyte, temperature, membrane characteristics, perfusion ﬂow rate, and the diffusion coefﬁcient of the compound in the extracellular ﬂuid. It is independent from the solute concentration gradient between tissue and perfusate. Provided proper in vivo calibration procedures, intraindividual variation for tissue concentration measurements was shown to range between 10% and 20% depending on the analyte (Sta˚hle et al., 1991). II.A.2. Tissue damage Insertion of MD probes has been shown to cause minimal tissue trauma, which could theoretically inﬂuence the results of MD experiments. To allow ‘‘tissue equilibration’’, that is, to provide time for the initial trauma to subside, probe perfusion for more than half an hour has shown to be practical before starting probe calibration (Joukhadar and Mu¨ller, 2005). This is based on the ﬁnding that several markers of tissue trauma (e.g., thromboxane B2, adenosine triphosphate, adenosine, K+, glucose, lactate, and lactate/pyruvate ratio) are elevated after probe insertion and reach baseline or become undetectable within this time range. Furthermore, changes in local skin circulation after insertion of MD probes have been investigated using laser Doppler perfusion imaging and have been shown to return to baseline within 40 min (Anderson et al., 1994). Histological studies have shown that the implantation of small diameter dialysis tubings does not induce major edema, foreign body reactions or bleeding if implantation times are kept below 24 h (Hickner et al., 1995). II.A.3. Limitations As MD probes are usually perfused with aqueous solutions (physiological Ringer’s solution or phosphate buffered saline), the technique is conceptually limited to the study of water-soluble drugs. Several attempts to measure highly lipophilic compounds

have more or less failed (Groth, 1996). However, there are few reports on the successful measurement of more lipophilic compounds, for example, estradiol (Mu¨ller et al., 1995b). To enable the measurement of lipophilic compounds routinely, in vitro experiments have demonstrated the usefulness of lipid emulsion as perfusate instead of aqueous solution, but this approach has not worked in vivo so far (Carneheim and Sta˚hle, 1991). Low sensitivity of conventional analytical methods is often the bottleneck for the successful detection of low drug concentrations in the small dialysate sample volumes, which are in the microliter range depending on the studied drug and the desired temporal resolution. However, the recent introduction and improvement of analytical methods such as high-performance liquid chromatography (HPLC), microbore/capillary LC methods, mass spectrometry, and biosensors are likely to overcome this challenging issue (Petsch et al., 2004; Traunmu¨ller et al., 2005). As opposed to imaging techniques, which allow simultaneous drug distribution studies in several organs and tissues, MD only provides focal information on tissue PK from a limited number of sites, largely due to ethical objections to simultaneously inserting multiple probes into one subject/ patient. Furthermore, the sampling period is commonly limited to approximately 12 h when MD probes with steel shafts are used and subjects are requested to remain in a supine resting position throughout the study period. If ﬂexible probes are used and subjects are allowed to move, sampling over longer periods, that is, days or even weeks, is feasible. Further drawbacks stem form the semi-invasive nature of the technique. Consequently, most human studies have so far been performed in easily accessible tissues such as skeletal muscle, subcutaneous adipose tissue and skin (Joukhadar and Mu¨ller, 2005), tendons (Alfredson and Lorentzon, 2003), superﬁcially located tumors (Brunner and Mu¨ller, 2002), or blood (Elshoff and Laer, 2005). Combined with surgical procedures, however, almost every human tissue is in reach for MD probe implantation as demonstrated by studies in brain (Clausen et al., 2005), lung (Hutschala et al., 2005;

629

Zeitlinger et al., 2005b), bone (Thorsen et al., 1996), heart (Bahlmann et al., 2004), liver (Silva et al., 2005), or the peritoneal cavity (Jansson et al., 2005). For the latter two applications special probes for the use in humans have become available recently.

III. Current clinical applications in drug delivery studies In clinical research, MD is currently employed to address various issues in different clinical settings, such as monitoring of secondary ischemia in neurointenisve care (Ungerstedt and Rostami, 2004) or glucose monitoring for long-term metabolic control in patients with diabetes mellitus (Pickup et al., 2005). Further areas of research comprise studies on the local physiology and metabolism of peripheral tissues (de la Pena et al., 2000) or local drug administration by means of MD with the aim to achieve high target site concentrations without inducing systemic side effects (Ronquist et al., 1992; Hamrin and Henriksson, 2005). This approach also allows the simultaneous measurement of the corresponding tissue response in one experiment. Clinical drug delivery studies focus on the use of MD to measure target site concentrations of antibiotics (Joukhadar et al., 2001a) or anticancer drugs (Brunner and Mu¨ller, 2002) in different tissues and organs and to subsequently relate target site PK to pharmacodynamics (PD) (Delacher et al., 2000). Equally challenging is the characterization of skin penetration of the active drug fraction from transdermal therapeutic systems (Bur et al., 2005). In the following sections an upto-date overview about the main areas of application of MD in clinical drug delivery studies will be given. III.A. Assessment of tissue delivery of anti-infective substances Inadequate tissue delivery of anti-infective substances can lead to therapeutic failure and bacterial resistance. Ideally, a full PK evaluation of antibiotics should include tissue concentration measurements, as in most cases, only the concentration

of free unbound antibiotic in the ISF at the infection site promotes the antibacterial effect. Currently, anti-infective drugs are still mainly selected on an empirical basis. This approach, however, might lead to a skewed understanding of antibiotic target site distribution, as underlined by studies reporting high intertissue and intersubject variability of antibiotic tissue distribution, with antibiotic tissue concentrations considerably different from corresponding plasma levels (Mu¨ller et al., 2004). During the last years, numerous clinical MD studies have contributed to clarifying several questions regarding antibiotic tissue delivery in vivo. It could be demonstrated that complete plasma-totissue equilibration of anti-infectives cannot be taken for granted for many clinical situations (Mu¨ller et al., 2004; Joukhadar and Mu¨ller, 2005). Despite the knowledge about incomplete plasmato-tissue distribution in organs, such as the central nervous system, the prostate gland, eye, placenta and lung, detailed information on plasma-to-tissue equilibration for skeletal muscle and subcutaneous adipose tissue has been largely unavailable. Penetration of antimicrobial agents into soft tissues, however, may be substantially impaired in certain clinical settings as shown by MD studies, either by unspeciﬁc mechanisms and barriers hampering transendothelial drug transfer to the ISF of soft tissues or by the development of barriers during pathological processes, such as tissue ﬁbrosis or sclerosis (Mu¨ller et al., 1996; Hollenstein et al., 2001; Joukhadar et al., 2001b, 2002, 2003b; Tegeder et al., 2002a). Further important determinants inﬂuencing tissue penetration of antimicrobial drugs comprise plasma protein binding, local blood ﬂow, capillary density, capillary permeability, interstitial diffusion coefﬁcients, and transcapillary oncotic and osmotic pressure gradients, factors, whose effects on tissue distribution are usually not directly studied in vivo during development of new antibiotics, because appropriate techniques are not readily available (Curry, 1986; Brunner et al., 2000; Joukhadar et al., 2001b; Clough et al., 2002; Liu et al., 2005a). These issues are well known to regulatory authorities in the US and in Europe. Consequently, they are encouraging pharmaceutical companies to submit PK data

630

of new anti-infective drugs at the site of action during the drug application process rather than concentrating on plasma PK. MD provides exclusive and continuous access to the interstitial space, the target compartment for most bacterial pathogens, in a minimally invasive way and thus appears particularly attractive for clinical distribution studies of anti-infective substances, which has also been acknowledged by FDA advisory committees.1,2 Initially, MD has only been applied in preclinical research. The reﬁnement of chemical analytics and marketing of commercially available probes for the use in humans has increasingly broadened the ﬁeld of applications and has catalyzed the adoption of MD for human PK studies of antiinfectives within the last 15–20 years (Joukhadar et al., 2001a). The ﬁrst report on MD for studying antibiotic tissue penetration has been published in 1995 (Mu¨ller et al., 1995a). In this study, gentamicin concentrations in skeletal muscle and subcutaneous tissue were measured in healthy volunteers after intravenous bolus administration. Based on the results of this study, it was concluded that MD is a readily applicable, relatively noninvasive, reproducible method for PK characterization of anti-infective drugs. Since that time, MD studies have addressed tissue PK and factors potentially inﬂuencing tissue distribution, such as protein binding (Liu et al., 2005a), administration schedules (Hollenstein et al., 2000), or different drug galenics (de PA et al., 2002). Furthermore, it was shown that MD allows to extend the concept of PK–PD markers, for example, time above minimum inhibitory concentration (T>MIC), AUC/ MIC ratio, Cmax/MIC ratio or AUIC, from serum PK to peripheral tissue PK (Delacher et al., 2000; Liu et al., 2005b). A study in healthy volunteers describing soft tissue PK of two orally available cephalosporine antibiotics with similar plasma kinetics and in vitro MIC values for relevant micro-organisms exempliﬁes the potentials of MD to address important issues in human target site drug delivery (Liu et al., 2005a). After administration of comparable cephalosporine doses, it could be shown that tissue penetration was closely related to plasma protein binding. Average penetration into

skeletal muscle was higher for cefopodoxime with a protein binding of 25% as compared with ceﬁxime (protein binding 65%). Consistent with the tissue PK, an in vitro kinetic model showed that the antibacterial potency of cefpodoxime against Streptococcus pneumoniae was higher than that of ceﬁxime (Liu et al., 2005b). As demonstrated for other antibiotics in humans, free antibiotic tissue levels of both drugs corresponded to unbound plasma concentrations (Mu¨ller et al., 1995a; Brunner et al., 1999). Conversely, other MD studies indicated soft tissue accumulation for the ketolide telithromycin (Gattringer et al., 2004) and the ﬂuoroquiolone gemiﬂoxacin (Islinger et al., 2004), a ﬁnding, which has not been described so far. Altogether, healthy volunteer studies employing MD led to a reappraisal of formerly believed concepts about ‘‘tissue penetration’’ of antimicrobial drugs (Mu¨ller et al., 2004). Besides protein binding, clinical conditions such as septicaemia (Joukhadar et al., 2002; Zeitlinger et al., 2003a, b; Sauermann et al., 2005), septic shock (Joukhadar et al., 2001b; Karjagin et al., 2005), intensive care procedures (Brunner et al., 2000), or critical illness (Tegeder et al., 2002b) might inﬂuence antibiotic tissue delivery. For piperacillin (Brunner et al., 2000; Joukhadar et al., 2001b) and levoﬂoxacin (Zeitlinger et al., 2003a), for example, tissue concentrations in septic patients were subinhibitory even though effective concentrations were attained in serum. This may provide an explanation for therapeutic failures and the emergence of drug resistant bacteria, as suboptimal target site concentrations are regarded as key triggers of bacterial resistance. In patients undergoing cardiac surgery, soft tissue concentrations of piperacillin were shown to be markedly altered and decreased (Brunner et al., 2000). However, a pharmacodynamic in vitro simulation indicated that due to impaired renal function, prolonged tissue and plasma half-lives and subsequent prolongation of the relevant surrogate marker T>MIC, bacterial killing was effective despite relatively low target site concentrations (Sauermann et al., 2003). For another b-lactam antibiotic, cefpirome, equilibration of plasma and subcutaneous tissue concentrations was considerably delayed in septic patients compared to

631

healthy controls (Sauermann et al., 2005), a ﬁnding that closely resembled previous investigations with piperacillin (Joukhadar et al., 2001b), levoﬂoxacin (Zeitlinger et al., 2003a), and cefpirome (Joukhadar et al., 2002) in muscle tissue. It thus appears that systemic inﬂammation with subsequent changes in vascular permeability associated with septicemia or postoperative trauma signiﬁcantly affects interstitial concentrations. This ﬁnding may have clinical implications in that current dosing guidelines may result in inadequate target site concentrations that could, conceivably, lead to therapeutic failure in some patients or patient populations. In contrast, rapid tissue equilibration was observed for fosfomycin in septic patients (Joukhadar et al., 2003a), which might be attributed to different pharmacological drug properties, such as low molecular weight, low plasma protein binding, and high hydrophilicity or differences in the pharmacological treatment of sepsis between the patient populations. A recent study also demonstrated favorable tissue penetration of metronidazole in patients with septic shock (Karjagin et al., 2005). Microdialysis in patients was furthermore used to quantify the effect of obesity (Hollenstein et al., 2001) and simulated microgravity (Schuck et al., 2005) on peripheral drug distribution. When obese subjects with a mean weight of 122 kg were compared with age- and sex-matched lean control subjects (Hollenstein et al., 2001) approximately 50% lower tissue AUC/plasma AUC ratios of ciproﬂoxacin were observed in obese subjects. Thus, the process of penetration into the ISF was markedly impaired in obese subjects, most likely because of a reduced capillary permeability surface area in fat tissue. A recent study in healthy volunteers (Schuck et al., 2005) evaluated changes in the disposition and tissue penetration of an antibiotic during simulated microgravity with the goal to characterize antibiotic PK for the use during space ﬂights. Tissue penetration in simulated microgravity was slightly lower than under normal conditions, suggesting that ciproﬂoxacin tissue penetration might be impaired in microgravity. Although the differences were not statistically signiﬁcant, this study pointed on the need for further studies.

Other studies addressed the effect of inﬂammation on antibiotic tissue penetration. The penetration of ciproﬂoxacin (Mu¨ller et al., 1999), levoﬂoxacin (Bellmann et al., 2004), and fosfomycin (Legat et al., 2003) into unaffected and inﬂamed tissue was not signiﬁcantly altered by inﬂammation. In contrast, a signiﬁcantly reduced target site penetration could be observed for ciproﬂoxacin in patients suffering from arterial occlusive disease (Joukhadar et al., 2001c) and for fosfomycin in long-term diabetes mellitus patients (Legat et al., 2003) probably due to an impaired capillary density. Thus, it appears that local blood ﬂow (Joukhadar et al., 2005) and alterations in the capillary surface area are important determinants of ISF concentrations whereas acute inﬂammatory events seem to have little inﬂuence on tissue penetration. These observations were in clear contrast to reports on increased target site availability of antibiotics by macrophage drug uptake and preferential release of antibiotics at the target site (Schentag and Ballow, 1991), a concept that is also used as a marketing strategy by industry. Beyond its application in easily accessible soft tissues, MD has also been adopted for measurements in the human lung (Herkner et al., 2002; Tomaselli et al., 2003, 2004; Zeitlinger et al., 2005b) and the human brain (Mindermann et al., 1998; Brunner et al., 2002). Several recent studies have shown that it is possible to study penetration of different antibiotics into inﬂamed and healthy lung in patients undergoing lung surgery with MD. These studies provided evidence for the feasibility of clinical lung MD for the continuous assessment of target site PK in the ISF of lung tissue and corroborated the use of the studied antibiotics in the treatment of lung infections caused by extracellular bacteria (Tomaselli et al., 2003, 2004). The ﬁrst report on antibiotic PK in the human brain by MD has been published in 1998, when the in vivo penetration of rifampin into various compartments of the human brain in patients undergoing craniotomy for resection of primary brain tumors was determined (Mindermann et al., 1998). Rifampin concentrations in all compartments exceeded the MIC for staphylococci and streptococci. However, ISF concentrations were below the MICs for some mycobacterial strains. In

632

another study, intracerebral penetration of fosfomycin was studied by MD in intensive care patients with intracerebral hemorrhage requiring neurosurgical intensive care including CSF drainage (Brunner et al., 2002). In this study intracerebral Cmax values of fosfomycin were above the MIC values for clinically relevant bacteria of CNS infections and it was concluded that fosfomycin might qualify as a therapeutic option in the treatment of CNS infections, such as meningitis, meningoencephalitis, or brain abscesses. III.A.1. Pharmacokinetic/pharmacodynamic (PK/PD) studies of anti-infective agents The traditional approach of linking antibiotic concentrations to antibiotic effects is to relate a static parameter, the MIC, to the concentration in serum (Derendorf et al., 2000; Liu et al., 2002). This approach is usually applied by using cumulative PK–PD variables, such as AUC/MIC ratios, T>MIC, or Cmax/MIC ratios. However, these approaches do not take into account the complex interactions among an administered drug, patient, and infective agent, since in practice a PD effect in vivo is the result of a dynamic exposure of the infective agent to the unbound drug fraction at the relevant target site rather than a static interaction of two variables. Therefore, several authors proposed PK to be linked to PD in a more dynamic way by using several PK–PD models (Nolting et al., 1996; Craig, 1998; Delacher et al., 2000; Derendorf et al., 2000). All of the above techniques not only may provide information on PK but also may lend themselves to studies of antibiotic PD. This is particularly true for MD, as it monitors free antibiotic concentrations in the ﬂuid that directly surrounds the infective agent; the antimicrobial effect linked to the time–drug concentration proﬁle obtained by MD may be simulated easily in an in vitro setting with bacterial cultures (Zeitlinger et al., 2005a). This dynamic simulation may thus provide a rational approach for describing and predicting PD at a relevant target site. Some recent publications described an MD-based in vivo PK/in vitro PD model (Brunner et al., 2000; Zeitlinger et al., 2005a) that is based on a previously described modiﬁed maximum effect (Emax) model (Nolting et al., 1996; de la

Fig. 1. In vivo concentration versus time proﬁles of piperacillin for plasma (squares) and subcutis (circles) in the in vivo PK/in vitro PD model. MICs (minimal inhibitory concentrations) for pathogens employed in the subsequent PD simulation and the NCCLS breakpoints are indicated by horizontal dashed lines. (Reproduced from Zeitlinger et al., 2005a with permission from Oxford University Press.)

Pena et al., 2004; Liu et al., 2005b) and that may be used to predict drug effects at the target site. This three-step approach is based on (1) the in vivo measurement of ISF drug PK at the target site (Fig. 1) and (2) the subsequent PD simulation of the time–drug concentration proﬁle in an in vitro setting (Fig. 2). In a third step, the data are analyzed with an integrated PK–PD model to link unbound antibiotic concentrations to bacterial killing rates by using Emax. Such experiments enable the simulation of different dosing scenarios without the need for large clinical trials (Liu et al., 2005b). These data could provide strong support for PK–PD modeling procedures, might assist in dose optimization, and might replace current concepts for establishing dosing guidelines for selected tissue infections. This approach is also in accordance with the current reasoning of regulatory authorities, such as the FDA and the EMEA, which not only require measurements of the distribution of antibiotics to unaffected and infected target sites but also require the unbound drug concentration at the site of action to be related to the in vitro susceptibility of the infecting microorganism.2,6 On the basis of the results from clinical MD studies, a re-evaluation of previous concepts of plasma to tissue penetration of antimicrobial 6 http://www.emea.eu.int/pdfs/human/ewp/265599en.pdf [last accessed December 3, 2005].

633

Fig. 2. In vitro PD simulation. Time–kill curves of Staphylococcus aureus (left panels) and Pseudomonas aeruginosa (right panels) with MICs of 4 mg/L (squares, upper panels), 8 mg/L (circles, middle panels), and 16 mg/L (triangles, bottom panels) after exposure of bacteria to the concentration versus time proﬁles of piperacillin derived for plasma (ﬁlled symbols) and subcutaneous adipose tissue (open symbols). Bacterial growth controls for S. aureus and P. aeruginosa are shown as solid diamonds. All data are presented as means7SD (n ¼ 6). The detection limit is indicated for plasma by a dotted line and for subcutis by a dashed line. Plasma piperacillin concentrations were able to effectively inhibit bacterial growth of all bacterial strains irrespective of their MIC values. In contrast, concentration versus time proﬁles of subcutaneous adipose tissue were only effective in killing isolates with MICs of 4 and 8 mg/L, while bacterial growth of S. aureus and P. aeruginosa with MICs of 16 mg/L was not inhibited. Thus, prediction of microbiological outcome based on concentrations of piperacillin in plasma resulted in a marked overestimation of antimicrobial activity at the site of infection. These results underline the importance of viewing drugs and their actions directly at the level of the drug target, rather than indirectly via plasma concentrations. (Reproduced from Zeitlinger et al., 2005a with permission from Oxford University Press.)

634

substances seems justiﬁed. These results furthermore favor the idea of individual drug administration in special patient populations, such as intensive care patients, to reach sufﬁciently high effective tissue concentrations. MD may also be used to address the question whether antibiotics are available at the infection site for a sufﬁciently long time to eradicate bacterial pathogens without promoting the emergence of resistance due to too low target site concentrations. This knowledge will help to appropriately modify dosing regimens in different disease conditions. Integrating target site concentration measurements into drug development of anti-infectives at an early stage potentially provides a means for a more rational and effective drug development process. III.B. Topical drug application Topical drug application aims at delivering therapeutic drug concentrations to an affected region beneath the area of drug application, while keeping systemic drug concentration as low as possible to avoid systemic adverse drug effects. However, it is often unclear whether effective tissue drug concentrations are attained. Techniques such as tape stripping, currently used for the measurement of dermal drug PK and bioequivalence (Shah, 2001), usually fail to yield information on drug distribution into deeper skin layers, as only the uppermost layer of the skin, the stratum corneum, is removed and analyzed. Other methods, on the other hand, such as skin biopsies and indirect radiochemical methods, are not applicable in routine clinical settings because of their invasiveness or radioactive exposure of subjects and personnel. Nonetheless, recommendations for the drug dose and dosage regimen of transdermally applied drugs are still based on these methods. MD has been suggested as a promising alternative approach for the assessment of cutaneous drug delivery and is recognized by regulatory authorities as a potential tool for bioequivalence evaluation of topical dermatological dosage forms, because it allows for continuous measurement of ISF drug concentrations in deﬁned tissue layers (Shah, 2004). In a recently published paper, dermal concentrations of a newly developed diclofenac formulation

were compared with concentrations attained after standard oral treatment (Brunner et al., 2005). By means of MD, it could be demonstrated that topical drug application led to substantial dermal penetration of diclofenac with 250-fold lower plasma concentrations as compared with the oral treatment scheme. Dermal MD studies have furthermore corroborated the notion that the stratum corneum is the main barrier for the penetration of hydrophilic compounds into deeper skin layers and that disruption of this barrier dramatically increases skin penetration (Benfeldt and Serup, 1999). To date, MD studies in the human skin have assessed cutaneous drug delivery for methoxypsoralen (Tegeder et al., 2002a), propranolol (Stagni et al., 2000), isopropanol (Anderson et al., 1996), methyl-nicotinate (Boelsma et al., 2000), lidocaine (lignocaine) (Railcard, 2001), salicylic acid (Benfeldt and Serup, 1999), nicotine (Mu¨ller et al., 1995b; Bur et al., 2005), diclofenac (Mu¨ller et al., 1997b, 1998b; Dehghanyar et al., 2004; Brunner et al., 2005), estradiol (Mu¨ller et al., 1995b), ibuprofen (Tegeder et al., 1999), alcohol (ethanol) (Anderson et al., 1991), aciclovir, and penciclovir (Morgan et al., 2003). Some studies have addressed the penetration of dermatologically active compounds like antihistamines and non-steroidal anti-inﬂammatory drugs (NSAIDs) following systemic administration. In addition, the effect of iontophoresis (Cormier et al., 1999; Fang et al., 1999; Stagni et al., 1999), chemical skin irritation, and tape stripping (Benfeldt and Serup, 1999) on transdermal drug penetration have been studied. Results from these experiments demonstrated that the degree of mechanical barrier impairment, as assessed by measuring transepidermal water loss, positively correlates to the amount of transdermal drug absorption (Morgan et al., 2003). An increase of transcutaneous drug absorption was also demonstrated by the use of iontophoresis (Cormier et al., 1999; Fang et al., 1999; Stagni et al., 1999) and chemical skin irritation (Mu¨ller et al., 1998b; Benfeldt and Serup, 1999). Cutaneous blood ﬂow is another important factor, which inﬂuences dermal drug concentrations after a locally administered drug has successfully passed the stratum corneum. Cutaneous blood ﬂow was shown to rapidly clear the absorbed drug, which

635

may result in low tissue levels despite excellent drug penetration characteristics into superﬁcial tissue layers. Modiﬁcations of cutaneous blood ﬂow by vasoconstrictors or vasodilators were closely linked to changes in deep subepidermal drug tissue concentrations (Boutsiouki et al., 2001; Clough et al., 2002; Morgan et al., 2003). Thus, changes in microcirculatory blood ﬂow might substantially affect transdermal drug absorption. Employing MD in combination with 2D ultrasound has opened the opportunity to relate drug concentrations in deﬁned tissue layers to depth, that is, the distance from the skin surface to the tip of the MD probe. For the more lipophilic compound nicotine, a close correlation between the probe depth and nicotine tissue concentrations was detected after transdermal application (Mu¨ller et al., 1995b). In contrast, for the hydrophilic compound penciclovir, no signiﬁcant relationship was found between the probe depth and penciclovir tissue concentrations (Morgan et al., 2003). These data suggest that once a hydrophilic compound has reached the aqueous environment of the dermis, movement of drug appears unrestricted. A consistent ﬁnding of in vivo transdermal studies was the considerable interindividual variability in the skin penetration of drugs, which currently hampers the conduction of bioequivalence studies in transdermal research. This variability might be explained by factors such as the nature of the drug and vehicle, as well as skin integrity. However, several other parameters which are not present in conventional PK studies, that is, dose control, individual skin properties, hydration status, difﬁculty in standardizing pertubation measures like shaving or rubbing, and local blood ﬂow must also be taken into account. To capture the inﬂuence of these variables during development of novel transdermally applied substances, MD could not only be used for addressing bioequivalence issues, but might also be employed to identify formulations and doses of topically applied drugs that lead to effective local concentrations. III.C. Anticancer drugs Each year oncological research devotes intense efforts and spends enormous amounts of money

for the development of novel anticancer drugs. However, only few drugs are eventually introduced to clinical practice, and drugs that show major effects on mortality are scarce. Two major obstacles that permanently accompany drug development for nonsurgical treatment of malignant solid tumors are (a) acquired drug resistance resulting from genetic and epigenetic mechanisms reducing the effectiveness of available drugs and (b) pathophysiological characteristics of solid tumors that compromise the delivery and effectiveness of both conventional cytotoxic and molecular targeted therapies. Physiological barriers that inevitably develop during tumor formation comprise pathological tumor vessels, alterations in local blood ﬂow, and an abnormal interstitial matrix, characterized by a large interstitial space, high collagen, low proteoglycan, and hyaluronate concentrations, and the absence of a functional lymphatic network. Hyperpermeable tumor vessels contribute to interstitial hypertension, which further impairs the transcapillary ﬂow of ﬂuid and macromolecules. Finally, chaotic and variable blood ﬂow with abnormal shunting and areas of reversed blood ﬂow and oxygen consumption by neoplastic and endothelial cells, in addition to poor oxygen delivery, creates hypoxia and acidity within the tumor. These barriers have been acknowledged to potentially affect tissue drug delivery into the interstitial space of solid tumors and might represent the rate-limiting step in clinical response to chemotherapy (Jain, 1998). As for anti-infectives and transdermal substances, MD provides the opportunity to directly study ISF exposure of antitumor medication including the assessment of local drug metabolism (Brunner and Mu¨ller, 2002). So far, MD has been widely used particularly in preclinical studies either to explore tissue PK of new or approved drug candidates in various tumor models including human xenotransplants, or to determine tumor drug metabolism in the brain (Palsmeier and Lunte, 1994; de Lange et al., 1995; Nakashima et al., 1997). Some preclinical MD experiments have been performed to study the development of multidrug resistance, which is frequently associated with the expression of P-glycoprotein within the tumor cell membrane and acts as a nonspeciﬁc

636

energy dependent efﬂux pump for chemotherapeutic agents (Tsai et al., 2001; Sakata et al., 1994) or in an attempt to clarify the role of angiogenesis inhibitors on tumor neovascularization and intratumoral concentrations of cytotoxic agents (Teicher et al., 1995; Ma et al., 2001). Although frequently used in preclinical tumor research, only a limited number of MD studies have so far been performed in a clinical setting, which is due to both ethical and methodological considerations (Brunner and Mu¨ller, 2002). One important ethical aspect relates to the fact that metastases may be induced by the puncture of solid tumors. Following ﬁne needle biopsy of tumors, the incidence of these events was quantiﬁed in large studies and was shown to be o0.005% (Weiss, 1989; Lundstedt et al., 1991). Thus, a clinical MD study in oncologic patients needs to be limited to ethically appropriate conditions, although there is no evidence from literature that the puncture of tumor lesions inﬂuences the course or prognosis of the underlying disease. The ﬁrst MD study on drug delivery of anticancer drugs was performed in 1996 in patients with metastatic malignant melanoma, where carboplatin PK were measured in melanoma metastases (Blo¨chl-Daum et al., 1996). This study conﬁrmed the practical and ethical feasibility of clinical MD in cancer patients. Since this time MD has been used to evaluate the tumor disposition of commonly used anticancer drugs, including 5ﬂuorouracil (5-FU) (Mu¨ller et al., 1997a), capecitabine (Mader et al., 2003), methotrexate (Mu¨ller et al., 1998a), cisplatin (Tegeder et al., 2003), dacarbazine (Joukhadar et al., 2001d) and its metabolite, epirubicine (Hunz et al., 2001), and melphalan (Thompson et al., 2001) in tumor lesions accessible to MD probe insertion, such as breast cancer (Mu¨ller et al., 1997a, 1998a; Mader et al., 2003), malignant melanoma (Blo¨chl-Daum et al., 1996; Joukhadar et al., 2001d), osteosarcoma (Thompson et al., 2001), basal cell carcinoma (Wennberg et al., 2000), malignant glioma (de Micheli et al., 2000), Merkel cell tumor (Thompson et al., 2001), and oral cancer (Tegeder et al., 2003). A key ﬁnding of all these studies was a high interindividual variability in intratumoral drug concentrations. This provided circumstantial

evidence that insufﬁcient drug penetration into the ISF of solid tumors might explain why drugs that had initially raised enthusiasm concerning their therapeutic potential failed to meet these expectations in clinical studies. Thus, transcapillary drug transfer of cytotoxic agents into the ISF of solid tumors was suggested to be closely linked with clinical response to chemotherapy. The measurement of plasma PK was furthermore shown to be inappropriate to predict cytotoxic concentrations of the drugs in tumor lesions as conﬁrmed by the lack of any correlation between tumor and plasma PK for most cytotoxic agents (Blo¨chl-Daum et al., 1996; Mu¨ller et al., 1997a, 1998a; Mader et al., 2003). The importance of tissue penetration of cytotoxic agents for clinical outcome has been exempliﬁed in a study published in 1997 (Mu¨ller et al., 1997a). In this study, MD probes were inserted into breast cancer lesions and healthy reference tissue in patients scheduled to receive chemotherapy with 5FU. The investigators found very low AUC values of ﬂuorouracil in the tumor ISF of two patients. Interestingly, these two patients were those who did not respond to chemotherapy as assessed by radiography methods. Conceptually, plasma and reference tissue AUC values in these two patients were comparable with therapy responders. A high tumor dose intensity of ﬂuorouracil was associated with favorable tumor response, while this association was not found for AUC values in plasma and healthy subcutaneous adipose tissue. This result was recently conﬁrmed following administration of epirubicine (Hunz et al., 2001) in breast cancer patients. In patients with histological signs of remission, intratumoral epirubicin AUC values were higher than corresponding concentrations in subcutaneous fat, whereas in patients with no response to treatment, the highest epirubicin concentrations were detected in subcutaneous tissue. Thus, for theses two drugs, tissue penetration might be regarded as a rate-limiting step in tumor response to antineoplastic therapy. A recent study in patients with limb malignancies receiving regional chemotherapy by isolated limb infusion of melphalan could not corroborate the correlation between tumor drug load and response (Thompson et al., 2001). Melphalan

637

concentrations in tumor and adipose tissue were not signiﬁcantly different, and there was no signiﬁcant relationship between the severity of toxic reactions in the limb and peak melphalan plasma or MD concentrations. However, there was a signiﬁcant association between tumor response and melphalan concentrations measured over time in subcutaneous tissue. A probable reason for this ﬁnding is the impracticability to exactly position the MD probe into comparable regions in different tumor lesions in vivo. Thus, probes may be located in necrotic tumor regions, which would lead to MD drug concentrations not representative for other, better vascularized regions, a factor that has to be taken into account when studies in solid tumors are performed. Another study in patients with metastatic malignant melanoma lesions indicated favorable tumor penetration characteristics of dacarbazine and its metabolite 5-aminoimidazole-4-carboxamide after single intravenous administration of different dosages (Joukhadar et al., 2001d). Plasma and tumor AUC values were almost identical for dacarbazine and its metabolite for all dosages administered to patients. These ﬁndings were explained by peculiarities of the capillary bed in malignant melanoma, since it was recently proposed that blood-conducting channels in melanoma may not be formed by the cells of endothelial origin but by cancer cells. These abnormal capillaries are likely to be leakier than physiological endothelial capillaries and may thereby rather facilitate equilibration between blood and tumor ISF. In addition to its use in PK studies, the application of MD for therapeutic purposes was proposed by Ronquist et al. (1992), who used MD for site-speciﬁc drug delivery in the treatment of inoperable gliomas. For this purpose, MD probes were directly implanted into tumor tissue and the probes were perfused with deﬁned concentrations of L2,4-diaminobutyric acid for a total of 14–21 days. The procedure was well tolerated and direct drug delivery showed promising antitumor activity in all three investigated patients, whose survival time was three times longer than the mean in patients with malignant glioma receiving only radiotherapy. In conclusion, the issue of tumor drug exposure has until now been mostly neglected in oncological

studies. The use of MD provides the opportunity to study tumor drug exposure directly in vivo. The combination with in vivo PK/in vitro PD simulations, as successfully employed for anti-infectives, might be also employed to link intratumoral target site concentration measurements with in vitro inhibition of tumor cell growth (Mu¨ller et al., 2000). This approach might assist the identiﬁcation of new anticancer compounds with favorable tumor penetration and to select appropriate drug candidates and dosing intervals for anticancer drugs. To date MD has almost exclusively been conﬁned to the study of low molecular weight substances. The recent development of large-pore MD membranes, however, offers the possibility to measure tumor exposure of new high-molecular-weight antitumor medications. In recent years classic cytotoxic anticancer drugs have exerted their relatively unspeciﬁc action mainly by interfering with DNA replication in malignant, but also in healthy tissue, which usually causes a broad range of side effects. New strategies have led to the development of substances rationally directed against tumor- or non-tumor-speciﬁc targets. Owing to their high speciﬁcity for processes directly involved into tumor growth they are expected to cause minimal toxicity. Consequently, paradigms used for the development of classic cytotoxic compounds may not be appropriate for this new class of agents. In contrast to cytotoxic agents, where the endpoint of phase I studies is the maximum tolerated dose, dose-limiting toxicities often do not occur with antiproliferative agents. As some new agents are completely devoid of side effects, it might not even be possible to recommend a dose for phase II studies, and instead the optimal biologically effective dose must be deﬁned based on other endpoints. In this context, techniques for measuring target site PK such as MD and PET are particularly useful approaches (Langer and Mu¨ller, 2004). Imaging techniques may be employed to measure tumor blood ﬂow, angiogenesis, glucose uptake, thymidine metabolism, gene delivery, and PK of radiolabeled anticancer drugs. MD, on the other hand, selectively provides access to the interstitial tumor compartment, the place where receptors for many of the newer anticancer drugs are located, whereas the

638

intercellular biophase for traditional cytotoxic agents is not readily accessible for MD measurements. In both preclinical and clinical phase I studies MD may also address drug distribution during antiangiogenetic therapy and characterization of tumor metabolism. As novel anticancer drugs steadily deviate from the classical cytotoxic proﬁle, the possibility of deﬁning new surrogate parameters for antitumor activity will become increasingly crucial.

IV. In vivo microdialysis and combinatory use with positron emission tomography In the last 20 years imaging techniques have evolved as powerful tools for the noninvasive study of drug distribution in vivo as well as for studying drug effects at their target sites. Imaging techniques that lend themselves to the study of drug distribution in humans comprise MRS and PET (Fischman et al., 1997, 2002; Port and Wolf, 2003; Pien et al., 2005). Although these techniques were initially introduced to clinical medicine for diagnostic purposes and for the study of tissue metabolism and blood ﬂow (Phelps, 2000), they also opened a unique opportunity for PK research by providing a means for noninvasive measurement of drug distribution from the plasma compartment to anatomically deﬁned regions and for the visualization of the entire pattern of drug distribution in given organs (Fischman et al., 1998; Brunner et al., 2004). Studies using imaging techniques such as PET and MRS provided evidence that tissue distribution contributes more to total variability in the dose–effect relationship than the combination of factors determining plasma PK (Fischman et al., 1997; Eichler and Mu¨ller, 1998; Mu¨ller and Eichler, 1999). PET and MRS were also shown to support the deﬁnition of optimal dosing schedules in phase I and II studies and the design of phase III studies (Fischman et al., 1997). In brief, PET is a nuclear imaging technique based on the use of molecules labeled with positron-emitting radioisotopes. The emitted positrons pass through tissue and are ultimately annihilated when combined with an electron, resulting in two 511 keV photons emitted in opposite directions.

Detectors are arranged in a ring around the tissue of interest, and only triggering events that arrive near-simultaneously at diametrically opposite detectors are recorded (coincidence detection). The resulting PET images might yield three-dimensional information on tissue distribution of the positron-emitting molecules. The most commonly employed PET radionuclides are oxygen-15 (15O), nitrogen-13 (13N), carbon-11 (11C), and ﬂuorine-18 (18F). Owing to its comparably long half-life, 18F is the most attractive PET radioisotope for drug distribution studies, since it allows for imaging durations of up to 10 h. So far, PET has been used for studying tissue distribution of radiolabeled antibiotics, antifungals, and inhaled drugs in patients and volunteers (Fischman et al., 1993, 1996, 1998; Berridge et al., 2000; Brunner et al., 2004). Important limitations stem from the fact that only drugs that lend themselves to radiolabeling may be studied. Secondly, the PET signal produced is not necessarily a measure of the intact drug concentration and PET is not able to provide information about speciﬁc tissue compartments, such as the ISF. PET studies are furthermore bound to specialized centers, with on-site access to a cyclotron, radiochemistry, and a PET camera, which substantially contributes to the high costs of PET studies. The combination of MD and PET has the potential to exactly explore and describe the fate and PK of a drug in the body. Exploiting the strengths of both approaches appears to be a straightforward way to predict drug action and therapeutic success and may be used for decision making in drug research and development in the future. However, the combinatory use of both techniques in clinical drug delivery studies has not been used for this purpose in humans so far. A recent study has combined MD and PET to assess intracellular drug PK in vivo (Langer et al., 2005). PET yields a combined signal comprising the intracellular, the extracellular, and the intravascular fraction of a radiolabeled drug and its metabolites in a given volume of tissue, whereas MD describes unbound extracellular drug concentrations. As for several drugs, such as certain anti-infective and anticancer agents, the site of drug action is not the biophase surrounding the cells but rather an intracellular

639

compartment, knowledge of intracellular rather than extracellular or total drug concentrations is relevant in many cases. Fluorine-18-labelled ciproﬂoxacin ([18F]ciproﬂoxacin) (Langer et al., 2003) was used as a model compound to perform simultaneous PET imaging and MD in healthy volunteers to describe intracellular drug PK in human muscle tissue for several hours. A three-compartment PK model was ﬁtted to the tissue concentration–time proﬁles of ciproﬂoxacin measured by PET to estimate the rate constants of ciproﬂoxacin uptake and transport. The predicted extracellular concentration–time proﬁles from the compartmental modeling were in good agreement with the measured MD data and the results were in accordance with previous in vitro data describing cellular ciproﬂoxacin uptake and retention. The setting of this study is unique in the sense that it constitutes one of the rare occasions where the results of the compartmental modeling of PET data were directly validated by an independent measurement technique, that is, HPLC quantiﬁcation of samples collected by MD, in humans. The authors therefore concluded that the employed MD/ PET combination might be useful during research and development of new drugs, for which knowledge of intracellular concentrations is of interest. All other recently published MD/PET combinations in humans so far come from the neuroscience ﬁeld. MD has been introduced as an intracerebral sampling method for clinical neurosurgery in 1990 (Hillered et al., 2005) and has since then been embraced as a safe and effective monitoring technique to measure the neurochemistry of acute brain injury and epilepsy (Cavus et al., 2005; Clausen et al., 2005). Although data from brain MD studies strongly suggest that changes in local markers of brain metabolism might precede the onset of secondary neurological deterioration, cerebral MD is still mainly used as a clinical research tool in neurosurgery and its use to inﬂuence clinical therapeutic decision making has been restricted to only a few institutions worldwide (Hillered et al., 2005). Still, brain MD is one of few methods for neurochemical measurements in the interstitial compartment of the human brain and has become a valuable translational research tool, providing new and important insights into the

neurochemistry of acute human brain injury. Isolated interpretation of biomarkers derived from brain MD experiments, however, should be done cautiously and might require additional validation in particular in clinical studies, in which experimental conditions cannot be easily standardized. Therefore, the simultaneous use of complementary imaging techniques such as PET might be crucial for biomarker interpretation (Hillered et al., 2005). V. Conclusions The lack of appropriate methods to directly measure drug concentrations at their actual sites of action within tissues and organs has long conﬁned clinical PK studies to drug concentration measurements in blood and easy-to-sample body ﬂuids. Among several innovative techniques, which have become available for in vivo tissue distribution studies in the last years, in vivo MD in particular is a rational and scientiﬁcally sound means to directly measure concentrations of unbound drugs at their site of action in virtually every tissue and organ of the human body, as it provides access to the interstitial space, the site of action for many drugs. During the last 10–15 years, an ever-increasing number of clinical MD studies have substantially expanded the existing knowledge about the concept of tissue distribution in healthy volunteers and patients. Combined with pharmacodynamic simulations, MD data will potentially be the basis for the prediction of drug effects at the target site and drug administration recommendations. MD will continue to serve as an important research tool and might help to deﬁne meaningful surrogate markers for drug efﬁciency along the critical path of drug development in particular in combination with other approaches, such as imaging techniques. References Alfredson, H. and Lorentzon, R. (2003) Intratendinous glutamate levels and eccentric training in chronic achilles tendinosis: a prospective study using microdialysis technique. Knee Surg. Sports Traumatol. Arthrosc., 11: 196–199. Anderson, C., Andersson, T. and Molander, M. (1991) Ethanol absorption across human skin measured by in vivo microdialysis technique. Acta. Derm. Venereol., 71: 389–393.

640 Anderson, C., Andersson, T. and Wardell, K. (1994) Changes in skin circulation after insertion of a microdialysis probe visualized by laser Doppler perfusion imaging. J. Invest. Dermatol., 102: 807–811. Anderson, C., Andersson, T., Boman, A. and Molander, M. (1996) Cutaneous microdialysis for the measurement in vivo of the percutaneous absorption of organic solvents. Curr. Probl. Dermatol., 25: 37–46. Bahlmann, L., Misfeld, M., Klaus, S., Leptien, A., Heringlake, M., Schmucker, P., Sievers, H.H., Ungerstedt, U. and Kraatz, E.G. (2004) Myocardial redox state during coronary artery bypass grafting assessed with microdialysis. Intensive Care Med., 30: 889–894. Bellmann, R., Kuchling, G., Dehghanyar, P., Zeitlinger, M., Minar, E., Mayer, B.X., Mu¨ller, M. and Joukhadar, C. (2004) Tissue pharmacokinetics of levoﬂoxacin in human soft tissue infections. Br. J. Clin. Pharmacol., 57: 563–568. Benfeldt, E. and Serup, J. (1999) Effect of barrier perturbation on cutaneous penetration of salicylic acid in hairless rats: in vivo pharmacokinetics using microdialysis and non-invasive quantiﬁcation of barrier function. Arch. Dermatol. Res., 291: 517–526. Benveniste, H. (1989) Brain microdialysis. J. Neurochem., 52: 1667–1679. Berridge, M.S., Lee, Z. and Heald, D.L. (2000) Regional distribution and kinetics of inhaled pharmaceuticals. Curr. Pharm. Des., 6: 1631–1651. Blo¨chl-Daum, B., Mu¨ller, M., Meisinger, V., Eichler, H.G., Fassolt, A. and Pehamberger, H. (1996) Measurement of extracellular ﬂuid carboplatin kinetics in melanoma metastases with microdialysis. Br. J. Cancer., 73: 920–924. Boelsma, E., Anderson, C., Karlsson, A.M. and Ponec, M. (2000) Microdialysis technique as a method to study the percutaneous penetration of methyl nicotinate through excised human skin, reconstructed epidermis, and human skin in vivo. Pharm. Res., 17: 141–147. Boutsiouki, P., Thompson, J.P. and Clough, G.F. (2001) Effects of local blood ﬂow on the percutaneous absorption of the organophosphorus compound malathion: a microdialysis study in man. Arch. Toxicol., 75: 321–328. Brunner, M. and Mu¨ller, M. (2002) Microdialysis: an in vivo approach for measuring drug delivery in oncology. Eur. J. Clin. Pharmacol., 58: 227–234. Brunner, M., Dehghanyar, P., Seigfried, B., Martin, W., Menke, G. and Mu¨ller, M. (2005) Favourable dermal penetration of diclofenac after administration to the skin using a novel spray gel formulation. Br. J. Clin. Pharmacol., 60: 573–577. Brunner, M., Hollenstein, U., Delacher, S., Jager, D., Schmid, R., Lackner, E., Georgopoulos, A., Eichler, H.G. and Mu¨ller, M. (1999) Distribution and antimicrobial activity of ciproﬂoxacin in human soft tissues. Antimicrob. Agents Chemother., 43: 1307–1309. Brunner, M., Langer, O., Dobrozemsky, G., Mu¨ller, U., Zeitlinger, M., Mitterhauser, M., Wadsak, W., Dudczak, R., Kletter, K. and Mu¨ller, M. (2004) [18F]Ciproﬂoxacin, a new positron emission tomography tracer for noninvasive assessment of the tissue distribution and pharmacokinetics of ciproﬂoxacin in humans. Antimicrob. Agents Chemother., 48: 3850–3857.

Brunner, M., Pernerstorfer, T., Mayer, B.X., Eichler, H.G. and Mu¨ller, M. (2000) Surgery and intensive care procedures affect the target site distribution of piperacillin. Crit. Care Med., 28: 1754–1759. Brunner, M., Reinprecht, A., Illievich, U., Spiss, C.K., Dittrich, P., van Houte, M. and Mu¨ller, M. (2002) Penetration of fosfomycin into the parenchyma of human brain: a case study in three patients. Br. J. Clin. Pharmacol., 54: 548–550. Bur, A., Joukhadar, C., Klein, N., Herkner, H., Mitulovic, G., Schmid, R., Agneter, E., Mu¨ller, M. and Brunner, M. (2005) Effect of exercise on transdermal nicotine release in healthy habitual smokers. Int. J. Clin. Pharmacol. Ther., 43: 239–243. Carneheim, C. and Sta˚hle, L. (1991) Microdialysis of lipophilic compounds: a methodological study. Pharmacol. Toxicol., 69: 378–380. Cavus, I., Kasoff, W.S., Cassaday, M.P., Jacob, R., Gueorguieva, R., Sherwin, R.S., Krystal, J.H., Spencer, D.D. and Abi-Saab, W.M. (2005) Extracellular metabolites in the cortex and hippocampus of epileptic patients. Ann. Neurol., 57: 226–235. Clausen, T., Alves, O.L., Reinert, M., Doppenberg, E., Zauner, A. and Bullock, R. (2005) Association between elevated brain tissue glycerol levels and poor outcome following severe traumatic brain injury. J. Neurosurg., 103: 233–238. Clough, G.F., Boutsiouki, P., Church, M.K. and Michel, C.C. (2002) Effects of blood ﬂow on the in vivo recovery of a small diffusible molecule by microdialysis in human skin. J. Pharmacol. Exp. Ther., 302: 681–686. Cormier, M., Chao, S.T., Gupta, S.K. and Haak, R.R. (1999) Effect of transdermal iontophoresis codelivery of hydrocortisone on metoclopramide pharmacokinetics and skin-induced reactions in human subjects. J. Pharm. Sci., 88: 1030–1035. Craig, W.A. (1998) Pharmacokinetic/pharmacodynamic parameters: rationale for antibacterial dosing of mice and men. Clin. Infect. Dis., 26: 1–10. Craig, W.A. and Ebert, S.C. (1989) Protein binding and its signiﬁcance in antibacterial therapy. Infect. Dis. Clin. North Am., 3: 407–414. Curry, F.E. (1986) Determinants of capillary permeability: a review of mechanisms based on single capillary studies in the frog. Circ. Res., 59: 367–380. de la Pena, A., Grabe, A., Rand, K.H., Rehak, E., Gross, J., Thyroff-Friesinger, U., Mu¨ller, M. and Derendorf, H. (2004) PK–PD modelling of the effect of cefaclor on four different bacterial strains. Int. J. Antimicrob. Agents, 23: 218–225. de la Pena, A., Liu, P. and Derendorf, H. (2000) Microdialysis in peripheral tissues. Adv. Drug Deliv. Rev., 45: 189–216. de Lange, E.C., de Vries, J.D., Zurcher, C., Danhof, M., de Boer, A.G. and Breimer, D.D. (1995) The use of intracerebral microdialysis for the determination of pharmacokinetic proﬁles of anticancer drugs in tumor-bearing rat brain. Pharm. Res., 12: 1924–1931. De Micheli, E., Alﬁeri, A., Pinna, G., Bianchi, L., Colivicchi, M.A., Melani, A., Pedata, F., Della Corte, L. and Bricolo, A. (2000) Extracellular levels of taurine in tumoral, peritumoral and normal brain tissue in patients with malignant glioma: an intraoperative microdialysis study. Adv. Exp. Med. Biol., 483: 621–625.

641 de, P.A., Brunner, M., Eichler, H.G., Rehak, E., Gross, J., Thyroff-Friesinger, U., Mu¨ller, M. and Derendorf, H. (2002) Comparative target site pharmacokinetics of immediate- and modiﬁed-release formulations of cefaclor in humans. J. Clin. Pharmacol., 42: 403–411. Dehghanyar, P., Mayer, B.X., Namiranian, K., Mascher, H., Mu¨ller, M. and Brunner, M. (2004) Topical skin penetration of diclofenac after single- and multiple-dose application. Int. J. Clin. Pharmacol. Ther., 42: 353–359. Delacher, S., Derendorf, H., Hollenstein, U., Brunner, M., Joukhadar, C., Hofmann, S., Georgopoulos, A., Eichler, H.G. and Mu¨ller, M. (2000) A combined in vivo pharmacokinetic-in vitro pharmacodynamic approach to simulate target site pharmacodynamics of antibiotics in humans. J. Antimicrob. Chemother., 46: 733–739. Derendorf, H., Lesko, L.J., Chaikin, P., Colburn, W.A., Lee, P., Miller, R., Powell, R., Rhodes, G., Stanski, D. and Venitz, J. (2000) Pharmacokinetic/pharmacodynamic modeling in drug research and development. J. Clin. Pharmacol., 40: 1399–3418. Eichler, H.G. and Mu¨ller, M. (1998) Drug distribution. The forgotten relative in clinical pharmacokinetics. Clin. Pharmacokinet., 34: 95–109. Elshoff, J.P. and Laer, S. (2005) Development of an intravenous microdialysis method for pharmacokinetic investigations in humans. J. Pharmacol. Toxicol. Methods, 52: 251–259. Evans, W.E. and McLeod, H.L. (2003) Pharmacogenomics – drug disposition, drug targets, and side effects. N. Engl. J. Med., 348: 538–549. Fang, J.Y., Sung, K.C., Lin, H.H. and Fang, C.L. (1999) Transdermal iontophoretic delivery of diclofenac sodium from various polymer formulations: in vitro and in vivo studies. Int. J. Pharm., 178: 83–92. Fischman, A.J., Alpert, N.M. and Rubin, R.H. (2002) Pharmacokinetic imaging: a noninvasive method for determining drug distribution and action. Clin. Pharmacokinet., 41: 581–602. Fischman, A.J., Alpert, N.M., Babich, J.W. and Rubin, R.H. (1997) The role of positron emission tomography in pharmacokinetic analysis. Drug. Metab. Rev., 29: 923–956. Fischman, A.J., Alpert, N.M., Livni, E., Ray, S., Sinclair, I., Callahan, R.J., Correia, J.A., Webb, D., Strauss, H.W. and Rubin, R.H. (1993) Pharmacokinetics of 18F-labeled ﬂuconazole in healthy human subjects by positron emission tomography. Antimicrob. Agents Chemother., 37: 1270–1277. Fischman, A.J., Babich, J.W., Bonab, A.A., Alpert, N.M., Vincent, J., Callahan, R.J., Correia, J.A. and Rubin, R.H. (1998) Pharmacokinetics of [18F]trovaﬂoxacin in healthy human subjects studied with positron emission tomography. Antimicrob. Agents Chemother., 42: 2048–2054. Fischman, A.J., Livni, E., Babich, J.W., Alpert, N.M., Bonab, A., Chodosh, S., McGovern, F., Kamitsuka, P., Liu, Y.Y., Cleeland, R., Prosser, B.L., Correia, J.A. and Rubin, R.H. (1996) Pharmacokinetics of [18F]ﬂeroxacin in patients with acute exacerbations of chronic bronchitis and complicated urinary tract infection studied by positron emission tomography. Antimicrob. Agents Chemother., 40: 659–664.

Gattringer, R., Urbauer, E., Traunmu¨ller, F., Zeitlinger, M., Dehghanyar, P., Zeleny, P., Graninger, W., Mu¨ller, M. and Joukhadar, C. (2004) Pharmacokinetics of telithromycin in plasma and soft tissues after single-dose administration to healthy volunteers. Antimicrob. Agents Chemother., 48: 4650–4653. Groth, L. (1996) Cutaneous microdialysis. Methodology and validation. Acta Derm. Venereol. Suppl. (Stockh.)., 197: 1–61. Hamrin, K. and Henriksson, J. (2005) Local effect of vanadate on interstitial glucose and lactate concentrations in human skeletal muscle. Life Sci., 76: 2329–2338. Herkner, H., Mu¨ller, M.R., Kreischitz, N., Mayer, B.X., Frossard, M., Joukhadar, C., Klein, N., Lackner, E. and Mu¨ller, M. (2002) Closed-chest microdialysis to measure antibiotic penetration into human lung tissue. Am. J. Respir. Crit. Care Med., 165: 273–276. Hickner, R.C., Ekelund, U., Mellander, S., Ungerstedt, U. and Henriksson, J. (1995) Muscle blood ﬂow in cats: comparison of microdialysis ethanol technique with direct measurement. J. Appl. Physiol., 79: 638–647. Hillered, L., Vespa, P.M. and Hovda, D.A. (2005) Translational neurochemical research in acute human brain injury: the current status and potential future for cerebral microdialysis. J. Neurotrauma., 22: 3–41. Hollenstein, U.M., Brunner, M., Mayer, B.X., Delacher, S., Erovic, B., Eichler, H.G. and Mu¨ller, M. (2000) Target site concentrations after continuous infusion and bolus injection of cefpirome to healthy volunteers. Clin. Pharmacol. Ther., 67: 229–236. Hollenstein, U.M., Brunner, M., Schmid, R. and Mu¨ller, M. (2001) Soft tissue concentrations of ciproﬂoxacin in obese and lean subjects following weight-adjusted dosing. Int. J. Obes. Relat. Metab. Disord., 25: 354–358. Hunz, M., Jetter, A., Wilde, S., Konig, E., Warm, M., Kurbacher, C. and Fuhr, U. (2001) Plasma and tissue pharmacokinetics of epirubicine in nine patients with primary breast cancer. Eur. J. Clin. Pharmacol., 57: A31. Hutschala, D., Skhirtladze, K., Zuckermann, A., Wisser, W., Jaksch, P., Mayer-Helm, BX., Burgmann, H., Wolner, E., Mu¨ller, M. and Tschernko, E.M. (2005) In vivo measurement of levoﬂoxacin penetration into lung tissue after cardiac surgery. Antimicrob. Agents Chemother., 49: 5107–5111. Islinger, F., Bouw, R., Stahl, M., Lackner, E., Zeleny, P., Brunner, M., Mu¨ller, M., Eichler, H.G. and Joukhadar, C. (2004) Concentrations of gemiﬂoxacin at the target site in healthy volunteers after a single oral dose. Antimicrob. Agents Chemother., 48: 4246–4249. Jain, R.K. (1998) Delivery of molecular and cellular medicine to solid tumors. J. Control. Release, 53: 49–67. Jansson, K., Jansson, M., Andersson, M., Magnuson, A., Ungerstedt, U. and Norgren, L. (2005) Normal values and differences between intraperitoneal and subcutaneous microdialysis in patients after non-complicated gastrointestinal surgery. Scand. J. Clin. Lab. Invest., 65: 273–281. Joukhadar, C. and Mu¨ller, M. (2005) Microdialysis: current applications in clinical pharmacokinetic studies and its potential role in the future. Clin. Pharmacokinet., 44: 895–913.

642 Joukhadar, C., Dehghanyar, P., Traunmu¨ller, F., Sauermann, R., Mayer-Helm, B., Georgopoulos, A. and Mu¨ller, M. (2005) Increase of microcirculatory blood ﬂow enhances penetration of ciproﬂoxacin into soft tissue. Antimicrob. Agents Chemother., 49: 4149–4153. Joukhadar, C., Derendorf, H. and Mu¨ller, M. (2001a) A novel tool for clinical studies of anti-infective agents. Eur. J. Clin. Pharmacol., 57: 211–219. Joukhadar, C., Frossard, M., Mayer, B.X., Brunner, M., Klein, N., Siostrzonek, P., Eichler, H.G. and Mu¨ller, M. (2001b) Impaired target site penetration of beta-lactams may account for therapeutic failure in patients with septic shock. Crit. Care Med., 29: 385–391. Joukhadar, C., Klein, N., Dittrich, P., Zeitlinger, M., Geppert, A., Skhirtladze, K., Frossard, M., Heinz, G. and Mu¨ller, M. (2003a) Target site penetration of fosfomycin in critically ill patients. J. Antimicrob. Chemother., 51: 1247–1252. Joukhadar, C., Klein, N., Frossard, M., Minar, E., Stass, H., Lackner, E., Herrmann, M., Riedmu¨ller, E. and Mu¨ller, M. (2001c) Angioplasty increases target site concentrations of ciproﬂoxacin in patients with peripheral arterial occlusive disease. Clin. Pharmacol. Ther., 70: 532–539. Joukhadar, C., Klein, N., Mader, R.M., Schrolnberger, C., Rizovski, B., Heere-Ress, E., Pehamberger, H., Strauchmann, N., Jansen, B. and Mu¨ller, M. (2001d) Penetration of dacarbazine and its active metabolite 5-aminoimidazole-4carboxamide into cutaneous metastases of human malignant melanoma. Cancer, 92: 2190–2196. Joukhadar, C., Klein, N., Mayer, B.X., Kreischitz, N., DelleKarth, G., Palkovits, P., Heinz, G. and Mu¨ller, M. (2002) Plasma and tissue pharmacokinetics of cefpirome in patients with sepsis. Crit. Care Med., 30: 1478–1482. Joukhadar, C., Stass, H., Mu¨ller-Zellenberg, U., Lackner, E., Kovar, F., Minar, E. and Mu¨ller, M. (2003b) Penetration of moxiﬂoxacin into healthy and inﬂamed subcutaneous adipose tissues in humans. Antimicrob. Agents Chemother., 47: 3099–3103. Karjagin, J., Pahkla, R., Karki, T. and Starkopf, J. (2005) Distribution of metronidazole in muscle tissue of patients with septic shock and its efﬁcacy against Bacteroides fragilis in vitro. J. Antimicrob. Chemother., 55: 341–346. Langer, O. and Mu¨ller, M. (2004) Methods to assess tissuespeciﬁc distribution and metabolism of drugs. Curr. Drug Metab., 5: 463–481. Langer, O., Mitterhauser, M., Brunner, M., Zeitlinger, M., Wadsak, W., Mayer, B.X., Kletter, K. and Mu¨ller, M. (2003) Synthesis of ﬂuorine-18-labeled ciproﬂoxacin for PET studies in humans. Nucl. Med. Biol., 30: 285–291. Langer, O., Karch, R., Mu¨ller, U., Dobrozemsky, G., Abrahim, A., Zeitlinger, M., Lackner, E., Joukhadar, C., Dudczak, R., Kletter, K., Mu¨ller, M. and Brunner, M. (2005) Combined PET and microdialysis for in vivo assessment of intracellular drug pharmacokinetics in humans. J. Nucl. Med., 46: 1835–1841. Legat, F.J., Maier, A., Dittrich, P., Zenahlik, P., Kern, T., Nuhsbaumer, S., Frossard, M., Salmhofer, W., Kerl, H. and Mu¨ller, M. (2003) Penetration of fosfomycin into inﬂammatory lesions in patients with cellulitis or diabetic foot syndrome. Antimicrob. Agents Chemother., 47: 371–374.

Liu, P., Mu¨ller, M. and Derendorf, H. (2002) Rational dosing of antibiotics: the use of plasma concentrations versus tissue concentrations. Int. J. Antimicrob. Agents, 19: 285–290. Liu, P., Mu¨ller, M., Grant, M., Obermann, B. and Derendorf, H. (2005a) Tissue penetration of cefpodoxime and ceﬁxime in healthy subjects. J. Clin. Pharmacol., 45: 564–569. Liu, P., Rand, K.H., Obermann, B. and Derendorf, H. (2005b) Pharmacokinetic-pharmacodynamic modelling of antibacterial activity of cefpodoxime and ceﬁxime in in vitro kinetic models. Int. J. Antimicrob. Agents, 25: 120–129. Lo¨nnroth, P., Carlsten, J., Johnson, L. and Smith, U. (1991) Measurements by microdialysis of free tissue concentrations of propranolol. J. Chromatogr., 568: 419–425. Lo¨nnroth, P., Jansson, P.A. and Smith, U. (1987) A microdialysis method allowing characterization of intercellular water space in humans. Am. J. Physiol., 253: E228–E231. Lundstedt, C., Stridbeck, H., Andersson, R., Tranberg, K.G. and Andren-Sandberg, A. (1991) Tumor seeding occurring after ﬁne-needle biopsy of abdominal malignancies. Acta Radiol., 32: 518–520. Ma, J., Pulfer, S., Li, S., Chu, J., Reed, K. and Gallo, J.M. (2001) Pharmacodynamic-mediated reduction of temozolomide tumor concentrations by the angiogenesis inhibitor TNP-470. Cancer Res., 61: 5491–5498. Mader, R.M., Schrolnberger, C., Rizovski, B., Brunner, M., Wenzel, C., Locker, G., Eichler, H.G., Mu¨ller, M. and Steger, G.G. (2003) Penetration of capecitabine and its metabolites into malignant and healthy tissues of patients with advanced breast cancer. Br. J. Cancer, 88: 782–787. Merrikin, D.J., Briant, J. and Rolinson, G.N. (1983) Effect of protein binding on antibiotic activity in vivo. J. Antimicrob. Chemother., 11: 233–238. Mindermann, T., Zimmerli, W. and Gratzl, O. (1998) Rifampin concentrations in various compartments of the human brain: a novel method for determining drug levels in the cerebral extracellular space. Antimicrob. Agents Chemother., 42: 2626–2629. Morgan, C.J., Renwick, A.G. and Friedmann, P.S. (2003) The role of stratum corneum and dermal microvascular perfusion in penetration and tissue levels of water-soluble drugs investigated by microdialysis. Br. J. Dermatol., 148: 434–443. Mu¨ller, M., Bockenheimer, J., Zellenberg, U., Klein, N., Steger, G.G., Eichler, H.G. and Mader, R.M. (2000) Relationship between in vivo drug exposure of the tumor interstitium and inhibition of tumor cell growth in vitro: a study in breast cancer patients. Breast Cancer Res. Treat, 60: 211–217. Mu¨ller, M., Brunner, M., Hollenstein, U., Joukhadar, C., Schmid, R., Minar, E., Ehringer, H. and Eichler, H.G. (1999) Penetration of ciproﬂoxacin into the interstitial space of inﬂamed foot lesions in non-insulin-dependent diabetes mellitus patients. Antimicrob. Agents Chemother., 43: 2056–2058. Mu¨ller, M., Brunner, M., Schmid, R., Mader, R.M., Bockenheimer, J., Steger, G.G., Steiner, B., Eichler, H.G. and Blo¨chl-Daum, B. (1998a) Interstitial methotrexate kinetics in primary breast cancer lesions. Cancer Res., 58: 2982–2985.

643 Mu¨ller, M., de la Pena, A. and Derendorf, H. (2004) Issues in pharmacokinetics and pharmacodynamics of anti-infective agents: distribution in tissue. Antimicrob. Agents Chemother., 48: 1441–1453. Mu¨ller, M. and Eichler, H.G. (1999) Tissue pharmacokinetics during early clinical drug development. Appl. Clin. Trials, 8: 56–60. Mu¨ller, M., Haag, O., Burgdorff, T., Georgopoulos, A., Weninger, W., Jansen, B., Stanek, G., Pehamberger, H., Agneter, E. and Eichler, H.G. (1996) Characterization of peripheralcompartment kinetics of antibiotics by in vivo microdialysis in humans. Antimicrob. Agents Chemother., 40: 2703–2709. Mu¨ller, M., Mader, R.M., Steiner, B., Steger, G.G., Jansen, B., Gnant, M., Helbich, T., Jakesz, R., Eichler, H.G. and Blo¨chlDaum, B. (1997a) 5-ﬂuorouracil kinetics in the interstitial tumor space: clinical response in breast cancer patients. Cancer Res., 57: 2598–2601. Mu¨ller, M., Mascher, H., Kikuta, C., Schafer, S., Brunner, M., Dorner, G. and Eichler, H.G. (1997b) Diclofenac concentrations in deﬁned tissue layers after topical administration. Clin. Pharmacol. Ther., 62: 293–299. Mu¨ller, M., Rastelli, C., Ferri, P., Jansen, B., Breiteneder, H. and Eichler, H.G. (1998b) Transdermal penetration of diclofenac after multiple epicutaneous administration. J. Rheumatol., 25: 1833–1836. Mu¨ller, M., Schmid, R., Georgopoulos, A., Buxbaum, A., Wasicek, C. and Eichler, H.G. (1995a) Application of microdialysis to clinical pharmacokinetics in humans. Clin. Pharmacol. Ther., 57: 371–380. Mu¨ller, M., Schmid, R., Wagner, O., v Osten, B., Shayganfar, H. and Eichler, H.G. (1995b) In vivo characterization of transdermal drug transport by microdialysis. J. Control. Release, 37: 49–57. Nakashima, M., Shibata, S., Tokunaga, Y., Fujita, H., Anda, T., Arizono, K., Tomiyama, N., Sasaki, H. and Ichikawa, M. (1997) In-vivo microdialysis study of the distribution of cisplatin into brain tumour tissue after intracarotid infusion in rats with 9L malignant glioma. J. Pharm. Pharmacol., 49: 777–780. Nolting, A., Dalla Costa, T., Rand, K.H. and Derendorf, H. (1996) Pharmacokinetic-pharmacodynamic modeling of the antibiotic effect of piperacillin in vitro. Pharm. Res., 13: 91–96. Osler, W. (1903) On the educational value of the medical society. Boston Med. Surg. J., 148: 275–279. Palsmeier, R.K. and Lunte, C.E. (1994) Microdialysis sampling in tumor and muscle: study of the disposition of 3-amino1,2,4-benzotriazine-1,4-di-N-oxide (SR 4233). Life Sci., 55: 815–825. Petsch, M., Mayer-Helm, B.X., Sauermann, R., Joukhadar, C. and Kenndler, E. (2004) Capillary electrophoresis analysis of fosfomycin in biological ﬂuids for clinical pharmacokinetic studies. Electrophoresis, 25: 2292–2308. Phelps, M.E. (2000) PET: the merging of biology and imaging into molecular imaging. J. Nucl. Med., 41: 661–681. Pickup, J.C., Hussain, F., Evans, N.D. and Sachedina, N. (2005) In vivo glucose monitoring: the clinical reality and the promise. Biosens. Bioelectron., 20: 1897–1902.

Pien, H.H., Fischman, A.J., Thrall, J.H. and Sorensen, A.G. (2005) Using imaging biomarkers to accelerate drug development and clinical trials. Drug. Discov. Today, 10: 259–266. Port, R.E. and Wolf, W. (2003) Noninvasive methods to study drug distribution. Invest. New Drugs, 21: 157–168. Railcard, M. (2001) Dermal pharmacokinetics of microemulsion formulations determined by in vivo microdialysis. Pharm. Res., 18: 367–373. Ronquist, G., Hugosson, R., Sjolander, U. and Ungerstedt, U. (1992) Treatment of malignant glioma by a new therapeutic principle. Acta Neurochir. (Wien.), 114: 8–11. Sakata, A., Tamai, I., Kawazu, K., Deguchi, Y., Ohnishi, T., Saheki, A. and Tsuji, A. (1994) In vivo evidence for ATPdependent and P-glycoprotein-mediated transport of cyclosporin A at the blood-brain barrier. Biochem. Pharmacol., 48: 1989–1992. Sauermann, R., Delle-Karth, G., Marsik, C., Steiner, I., Zeitlinger, M., Mayer-Helm, B.X., Georgopoulos, A., Mu¨ller, M. and Joukhadar, C. (2005) Pharmacokinetics and pharmacodynamics of cefpirome in subcutaneous adipose tissue of septic patients. Antimicrob. Agents Chemother., 49: 650–655. Sauermann, R., Zeitlinger, M., Erovic, B.M., Marsik, C., Georgopoulos, A., Mu¨ller, M., Brunner, M. and Joukhadar, C. (2003) Pharmacodynamics of piperacillin in severely ill patients evaluated by using a PK/PD model. Int. J. Antimicrob. Agents, 22: 574–578. Schentag, J.J. and Ballow, C.H. (1991) Tissue-directed pharmacokinetics. Am. J. Med., 91: 5S–11S. Schuck, E.L., Grant, M. and Derendorf, H. (2005) Effect of simulated microgravity on the disposition and tissue penetration of ciproﬂoxacin in healthy volunteers. J. Clin. Pharmacol., 45: 822–831. Shah, V.P. (2001) Progress in methodologies for evaluating bioequivalence of topical formulations. Am. J. Clin. Dermatol., 2: 275–280. Shah, V.P. (2004) Topical drug products – microdialysis: regulatory perspectives. Int. J. Clin. Pharmacol. Ther., 42: 379–381. Silva, M.A., Richards, D.A., Bramhall, S.R., Adams, D.H., Mirza, D.F. and Murphy, N. (2005) A study of the metabolites of ischemia-reperfusion injury and selected amino acids in the liver using microdialysis during transplantation. Transplantation, 79: 828–835. Stagni, G., O’Donnell, D., Liu, Y.J., Kellogg, D.L., Morgan, T. and Shepherd, A.M. (2000) Intradermal microdialysis: kinetics of iontophoretically delivered propranolol in forearm dermis. J. Control. Release, 63: 331–339. Stagni, G., O’Donnell, D., Liu, Y.J., Kellogg, D.L. Jr. and Shepherd, A.M. (1999) Iontophoretic current and intradermal microdialysis recovery in humans. J. Pharmacol. Toxicol. Methods, 41: 49–54. Sta˚hle, L., Arner, P. and Ungerstedt, U. (1991) Drug distribution studies with microdialysis. III: Extracellular concentration of caffeine in adipose tissue in man. Life Sci., 49: 1853–1858.

644 Tegeder, I., Brautigam, L., Podda, M., Meier, S., Kaufmann, R., Geisslinger, G. and Grundmann-Kollmann, M. (2002a) Time course of 8-methoxypsoralen concentrations in skin and plasma after topical (bath and cream) and oral administration of 8-methoxypsoralen. Clin. Pharmacol. Ther., 71: 153–161. Tegeder, I., Brautigam, L., Seegel, M., Al-Dam, A., Turowski, B., Geisslinger, G. and Kovacs, A.F. (2003) Cisplatin tumor concentrations after intra-arterial cisplatin infusion or embolization in patients with oral cancer. Clin. Pharmacol. Ther., 73: 417–426. Tegeder, I., Muth-Selbach, U., Lotsch, J., Rusing, G., Oelkers, R., Brune, K., Meller, S., Kelm, G.R., Sorgel, F. and Geisslinger, G. (1999) Application of microdialysis for the determination of muscle and subcutaneous tissue concentrations after oral and topical ibuprofen administration. Clin. Pharmacol. Ther., 65: 357–368. Tegeder, I., Schmidtko, A., Brautigam, L., Kirschbaum, A., Geisslinger, G. and Lo¨tsch, J. (2002b) Tissue distribution of imipenem in critically ill patients. Clin. Pharmacol. Ther., 71: 325–333. Teicher, B.A., Dupuis, N.P., Robinson, M.F., Emi, Y. and Goff, D.A. (1995) Antiangiogenic treatment (TNP-470/minocycline) increases tissue levels of anticancer drugs in mice bearing Lewis lung carcinoma. Oncol. Res., 7: 237–243. Thompson, J.F., Siebert, G.A., Anissimov, Y.G., Smithers, B.M., Doubrovsky, A., Anderson, C.D. and Roberts, M.S. (2001) Microdialysis and response during regional chemotherapy by isolated limb infusion of melphalan for limb malignancies. Br. J. Cancer, 85: 157–165. Thorsen, K., Kristoffersson, A.O., Lerner, U.H. and Lorentzon, R.P. (1996) In situ microdialysis in bone tissue. Stimulation of prostaglandin E2 release by weight-bearing mechanical loading. J. Clin. Invest., 98: 2446–2449. Tomaselli, F., Dittrich, P., Maier, A., Woltsche, M., Matzi, V., Pinter, J., Nuhsbaumer, S., Pinter, H., Smolle, J. and Smolle-Juttner, F.M. (2003) Penetration of piperacillin and tazobactam into pneumonic human lung tissue measured by in vivo microdialysis. Br. J. Clin. Pharmacol., 55: 620–624.

Tomaselli, F., Maier, A., Matzi, V., Smolle-Juttner, F.M. and Dittrich, P. (2004) Penetration of meropenem into pneumonic human lung tissue as measured by in vivo microdialysis. Antimicrob. Agents Chemother., 48: 2228–2232. Traunmu¨ller, F., Gattringer, R., Zeitlinger, M.A., Graninger, W., Mu¨ller, M. and Joukhadar, C. (2005) Determination of telithromycin in human plasma and microdialysates by highperformance liquid chromatography. J. Chromatogr. B Analyt. Technol. Biomed. Life. Sci., 822: 133–136. Tsai, T.H., Lee, C.H. and Yeh, P.H. (2001) Effect of P-glycoprotein modulators on the pharmacokinetics of camptothecin using microdialysis. Br. J. Pharmacol., 134: 1245–1252. Ungerstedt, U. and Rostami, E. (2004) Microdialysis in neurointensive care. Curr. Pharm. Des., 10: 2145–2152. Weiss, H. (1989) Metastases caused by ﬁne needle puncture? Ultraschall. Med., 10: 147–151. Wennberg, A.M., Larko, O., Lonnroth, P., Larson, G. and Krogstad, A.L. (2000) Delta-aminolevulinic acid in superﬁcial basal cell carcinomas and normal skin-a microdialysis and perfusion study. Clin. Exp. Dermatol., 25: 317–322. Wilkinson, G.R. (2005) Drug metabolism and variability among patients in drug response. N. Engl. J. Med., 352: 2211–2221. Zeitlinger, M.A., Dehghanyar, P., Mayer, B.X., Schenk, B.S., Neckel, U., Heinz, G., Georgopoulos, A., Mu¨ller, M. and Joukhadar, C. (2003a) Relevance of soft-tissue penetration by levoﬂoxacin for target site bacterial killing in patients with sepsis. Antimicrob. Agents Chemother., 47: 3548–3553. Zeitlinger, M.A., Erovic, B.M., Sauermann, R., Georgopoulos, A., Mu¨ller, M. and Joukhadar, C. (2005a) Plasma concentrations might lead to overestimation of target site activity of piperacillin in patients with sepsis. J. Antimicrob. Chemother., 56: 703–708. Zeitlinger, M.A., Marsik, C., Georgopoulos, A., Mu¨ller, M., Heinz, G. and Joukhadar, C. (2003b) Target site bacterial killing of cefpirome and fosfomycin in critically ill patients. Int. J. Antimicrob. Agents, 21: 562–567. Zeitlinger, M., Mu¨ller, M. and Joukhadar, C. (2005b) Lung microdialysis-a powerful tool for the determination of exogenous and endogenous compounds in the lower respiratory tract (mini-review). AAPS J., 7: E600–E608.

CHAPTER 7.2

Transport of glucose to a probe in adipose tissue K.J.C. Wientjes1,, K. Hoogenberg2 and A.J.M. Schoonen1 1

Department of Pharmacy, University of Groningen, Groningen, The Netherlands Department of Internal Medicine, Martini Hospital, Groningen, The Netherlands

2

Abstract: This chapter describes a model of transport of glucose in subcutaneous tissue from capillary to microdialysis probe. It is based on the well-known extraction equation and the histology of the subcutaneous tissue. We present arguments for a simple scheme in which the mean capillary concentration of glucose is the driving force for diffusion of glucose from the capillary to a microdialysis probe. Insertion effects seen as a microdialysis probe is placed in the subcutaneous tissue can be explained by a lower glucose concentration around the probe due to an inﬂammation reaction and by a slow increase on the number of functioning capillaries around the probe due to wound healing.

Theoretically Bungay et al. (1990) laid down an impressive piece of work to understand the transport mechanism of a solute to a microdialysis probe. In this chapter, the general model of Bungay, in which long diffusion paths are assumed, is confronted with histological and physiological data that permit a simpler model for transport of solutes to a microdialysis probe.

I. Introduction Continuous measurement of glucose in vivo is a long-standing goal for researcher’s worldwide. Especially type 1 diabetic patients could beneﬁt from a glucose sensor to keep their glucose values within a reasonable range. At the time of writing a number of sensor designs based on microdialysis are available or being developed (Maran and Poscia, 2002; Kapitza et al., 2003; Schoemaker et al., 2003). A sensor, working reliably for 2 weeks or more, is however, still not available. For a wearable artiﬁcial pancreas such a sensor is needed for a safe operation of the device. Sensors working reasonably well on the laboratory bench did not measure reliably in the living body. Clearly the ignorance about the relevant factors in vivo governing the measurement of glucose should be lifted. To keep things as simple as possible only sampling by a hollow ﬁber was used to investigate the behavior of glucose in the neighborhood of a sampling device.

II. Adipose tissue As most sensors are measuring in subcutaneous adipose tissue, the literature about this tissue was selected to learn about its structure and function. ‘‘Adipose tissue is a special type of connective tissue in which adipose cells predominate. Adipose cells are polyhedral in adipose tissue where they are closely packed. Although blood vessels are not always apparent, adipose tissue is richly vascularized. Considering the amount of cytoplasm in fat cells, that exists as a small rim surrounding the lipid droplet, the ratio of blood volume to cytoplasm

Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

645

DOI: 10.1016/S1569-7339(06)16034-8 Copyright 2007 Elsevier B.V. All rights reserved

646

volume is greater in adipose tissue than in striated muscle’’ (Jungueira et al., 1998). ‘‘Fat cells are surrounded by a meshwork of reticular ﬁbers and collagen ﬁbers with a rich capillary network’’ (Meadows, 1980). ‘‘The fat lobule itself characteristically forms around or within a capillary network in which an artery divides into the smallest capillary vessels quite abruptly, in the shortest possible distance’’ (Ryan and Curri, 1989; Thulesius, 1992). ‘‘The venous drainage ﬂows into a small vein on the surface of the lobule rather than in its interior’’ (Clara, 1956; Ryan and Curri, 1989). ‘‘Capillaries are normally localized in the interstitial space between three fat cells or found around the circumference of an adipocyte. The capillary lumen is not larger in diameter than 4–5 mm, so erythrocytes (7 mm) must deform in order to penetrate the capillaries of the intralobular network. The diffusion distance for blood oxygen is nearly zero. The basal membranes of the capillaries are in direct contact with the fat cell cytoplasmic membrane. This peculiar membrane-membrane contact has been found only in adipose tissue, the central nervous system and the spinal chord. The arterial and venous capillaries in the fat differ structurally from those in the upper and lower dermis by having walls that are only 0.03 mm thick. Usually in other organs the capillary wall is 0.1 mm’’ (Braverman and Yen, 1977; Braverman and Keh-Yen, 1981; Ryan and Curri, 1989). ‘‘The fat tissue of humans is devoid of typical arteriovenous anastomoses, as described in some skin regions’’. ‘‘In a meshwork such as that found in human fat, which has many arterio-arterious and veno-venous connections, the control of the blood ﬂow requires that any low resistance pathway has some

controlling device to prevent blood from ﬂowing through it freely, short circuiting the high resistance pathways through the capillary bed’’ (Braverman and Keh-Yen, 1981). ‘‘These regulating controls consist of a muscular apparatus within the arterial wall that can partly or totally block the lumen, to keep the arterioles from acting as shunts or preferential channels for bulk blood ﬂow. All the evidence concerning the development of adipose tissue suggest that it must have a very sluggish blood supply in which ﬂow must nevertheless be maintained’’ (Curri, 1984; Ryan and Curri, 1989). ‘‘In humans, the thicker the subcutaneous fat layer the smaller is the blood ﬂow per unit weight. In the dog, resting blood ﬂow is signiﬁcantly lower in subcutaneous adipose tissue than in tissue with smaller adipocytes. Indeed, Gersh and Still also noted that capillary density increases as the fat cells become smaller’’ (1945). The cited literature above presents a picture of adipose tissue in which the cells are closely packed. Most of the volume of an adipose cell is reserved for a lipid drop. The living part of the cell is a thin rim of tissue surrounding the lipid drop. Due to the closed packing of the cells, there is only a very thin layer of interstitial ﬂuid in between the cell walls. Such a layer of cells presents an effective barrier for diffusion of hydrophilic solutes as glucose. It is also clear that adipose tissue is richly vascularized. Capillaries are even in direct contact with the membranes of adipose cells. So we conclude that a capillary only supplies nutrients to the adipose cell to which it is attached. If a microdialysis probe is present only the capillaries in the near vicinity of the probe supply the probe on one side and the adipose cell on the other side (Fig. 1). Another important point is the existence of muscular pads within the wall of arterioles to regulate blood ﬂow. This strongly suggests that solute concentrations may be different within the adipose tissue and probably between tissues. So the usual

647 Ccap = concentration in blood Capillary

Cin

Fat cell

P

Cout

concentration profile Fig. 1. Representation of microdialysis model in adipose tissue based on the histological structure of adipose tissue. Adipose tissue is richly vascularized with capillaries. The adipocytes are closely packed together with a minimal interstitial space. In this model, only the capillaries in the near vicinity of the microdialysis probe are supplying solutes to the microdialysis probe.

assumption of a homogenous concentration in the blood compartment of an injected hydrophilic solute (after the distribution phase), may not be justiﬁed. This may explain the different results for blood glucose values at various sites after an oral load or a bolus injection of glucose (Van Der Valk et al., 2002). Finally, the capillary density depends on the size of the fat cells. The capillary density increases as fat cells become smaller.

III. The model A model, based on the structure of adipose tissue, allows diffusion of a solute supplied by the vascular system only from capillaries in a thin layer around the probe. From such a model two predictions can be made to verify the model: (1) The short distance between capillary and probe membrane implies a concentration gradient of glucose between capillary and probe across a thin interstitial layer. (2) The time needed for a molecule of glucose to diffuse from capillary to probe, should be quite short. Crone and Levitt (1984) calculated for a hydrophilic substance as glucose a time of 1 s from capillary to cell for a diffusion distance of 20–50 mm in most tissues. Therefore, this diffusion time should not be exceeded in adipose tissue, for the model to be correct.

ϕ

Seff = determined by number of capillaries Fig. 2. Representation of a microdialysis hollow ﬁber in the subcutaneous adipose tissue. Cin denotes the concentration of a substance in the in-let of the dialysate, Cout denotes the concentration in the out-let, P represents the total permeability of the tissue, the capillary and probe membrane whereas f represents the ﬂow. At a certain ﬂow, there will be an exchange of molecules between the dialysate and capillaries governed by the concentration difference (diffusion).

IV. The extraction equation This equation is a mass balance. In words: Change ¼ Input – Output+Production. As there is no production or destruction of substrate in the system, production ¼ 0, so Change ¼ Input – Output. What if input and output depends on the system. We are only interested in nutrients, so at the arterial end of a capillary, a relatively high concentration of the solute is present. Blood ﬂow transports the solute to the venous end of the capillary. The input is the average intracapillary concentration. Pores in the capillary wall allow diffusion of the solute into the interstitial ﬂuid: the output. In steady state, the input is equal to the output and, therefore, change ¼ 0. For a microdialysis probe, the input is diffusion of a solute from the interstitial ﬂuid across the membrane into the probe. The output is the ﬂow of solute out of the probe. Here also a steady state rapidly develops, so input ¼ output (Fig. 2). Along the length of the probe the concentration of solute in the probe is increasing, causing a decreasing concentration gradient for diffusion ﬂow to the probe. This is a ﬁrst-order process. Derivation of the equation starts by considering a very small segment of the probe. In steady state,

648

the concentration of solute at a certain point is constant, so for that point diffusion of solute across the wall is also constant. Integration over the entire length of the probe yields a logarithmic function, mostly presented as an exponential function: PS E ¼ 1 exp (1) j E¼

C out C in C i C in

(2)

R¼

C out Ci

(3)

P ¼ permeability of the diffusion path; S ¼ surface area of the probe membrane; j ¼ perfusion ﬂow; E ¼ the extraction efﬁciency, deﬁned as the gain of solute in the probe (Cout Cin) compared with the concentration difference between (Ci Cin). Cout ¼ solute concentration leaving the probe. Cin ¼ solute concentration in perfusion ﬂuid entering the probe. Ci ¼ Driving force concentration of solute in interstitial ﬂuid. R (recovery) ¼ E, for Cin ¼ 0. As long as diffusion is the rate-limiting step in transport of solute to a probe, the extraction equation is valid. Reﬂecting on the extraction equation, we may note the following points: 1. To measure a solute by microdialysis, for instance glucose, it is necessary that the recovery or the extraction efﬁciency should be constant. If the recovery is changing, there is no ﬁxed relation between the interstitial glucose concentration, and the dialysate concentration Cout; then no accurate measurement is possible. Therefore, P, S, and f must be constant. The perfusion ﬂow is under our control and should be constant during a measurement. If the permeability P is changing, there is some process operating around the probe that affects the diffusion path. The surface S of the probe is constant of course, but parts of the surface may be inaccessible by fouling or encapsulation or by some other mechanism.

2. In Fig. 3, two experiments are shown (Rebrin et al., 1999; Hullegie et al., 2000), one in dogs (Fig. 3a) and one in type I diabetic patients (Fig. 3B). In both cases, glucose was kept constant (‘‘clamp’’ conditions) at a normal value. Then a bolus of glucose was given, after which the level of glucose was ‘‘clamped’’ at a high value of glucose. In the dogs, a needle type electrode was used and in the patients a microdialysis probe. Also blood glucose was measured. In both cases, we see an increase in glucose concentration in blood as well as an increase measured by the electrode and the microdialysis probe in tissue. The starting point of the increase was the same for the blood and tissue measurements, but the slope of the increase was different for blood and tissue. This difference is usually explained as a lag-time phenomenon. Referring to the extraction equation, at least for microdialysis, this interpretation cannot be true as a difference in slope means a changing recovery during the increase. For a lag time, each point on the tissue lines should be equally displaced to the right parallel to the bloodline. Then the recovery is constant and only the start of the increase is at a later time, for instance due to a long diffusion path. It is important to note that this phenomenon apparently is independent of the measurement method employed, as the used electrodes in dogs showed the same phenomenon. 3. Finally, there is a problem with Ci. which concentration in the interstitial ﬂuid is the real-glucose concentration that is the driving force for diffusion of glucose from tissue to probe? Investigating to what height the recovery may increase, experiments with microdialysis were performed by using longer probes up to 3 cm to increase the probe surface S and also to lower the perfusion ﬂow rate f to a range of 0.5–0.2 mL/min. As could be expected by the extraction equation, these measures clearly showed an increased recovery until 95% of the blood value or more (Bolinder et al., 1992; Rosdahl et al., 1998).

649

(a)

Time (min)

Glucose concentration (mM)

14

(b)

12 10 8 6 4 venous

2 0

0

60

120

180 240 Time (min)

dialysate

300

360

420

Fig. 3. (a) Plasma glucose (K; left axis), sensor signal (solid line; right axis), and model ﬁt (dashed line; right axis) during hyperglycemic clamps in absence of changes in endogenous insulin (A), hyperglycemic clamps with changes in endogenous insulin secretion (B), and hyperglycemic clamps with changes in endogenous insulin (C; 0oto180) but with plasma glucose renormalized to basal with exogenous insulin (180oto240; Rebrin et al., 1999). (b): Average values of 14-microdialysis probes during a stepped hyperglyceamic hyperinsulinaemic clamp (slightly adapted; microdialysis dialysate values corrected for 30-min system lag time and multiplied by a factor of 1.1 to compare with venous blood-glucose values (Hullegie et al., 2000).

650

If we leave the probe in tissue without pumping the perfusion ﬂuid, the solute concentration in tissue and probe are equilibrating. So this concentration, Cequi, is the starting point of the solute concentration gradient, when a perfusion ﬂow is applied. Lo¨nnroth was the ﬁrst who made use of a microdialysis probe in which Cin was varied from 0 to values higher than the blood glucose concentration in volunteers (Lonnroth et al., 1987). He found values close to the blood glucose values for Ci. Bungay et al. (1990) showed a derivation of the extraction equation that theoretically underpinned the results of Lo¨nnroth: C out C in PS ¼ 1 exp (4) j C equi C in Substitute ‘‘A’’ for the right term of Eq. 4, then: C out C in ¼ A C equi A C in

(5)

When Cout – Cin ¼ 0, then Cin ¼ Cequi (see also Fig. 4). Here Ci is replaced by Cequi, as Cequi is the concentration of glucose in the probe that is in equilibrium with its surroundings. Experimentally various concentrations of glucose (Cin) are perfused, for instance: Cin ¼ 0, 2.8, 8.3, and 11.1 mM at a constant blood glucose of about 5.5 mM.

If the line in the graph is a straight line, this is proof for a rate-limiting diffusion process, validating the use of the equation. The point where the line is crossing the horizontal axis gives the Cequi: the point of no-net-ﬂux. Consequently, this method is called the no-net-flux method. Using the no-net-flux method in four volunteers, after insertion of four probes in the abdominal region of each volunteer, the Cequi was measured during 3 weeks. After 12–24 h, the Cequi was equal to the measured capillary blood glucose, obtained by ﬁnger prick (Fig. 5). Therefore, it is the average intracapillary concentration of glucose (Ccap) that drives diffusion of glucose to the probe and not some interstitial glucose concentration. For Cequi, we may now substitute Ccap in the extraction efﬁciency: E¼

C out C in C cap C in

(6)

In fact, a microdialysis probe is measuring Ccap and it is allowed to calibrate the obtained dialysate concentration on the capillary blood glucose to calculate the recovery, if permeability and surface area are constant. V. Features of the no-net-ﬂux method Considering the no-net-flux equation, it is evidently a straight line in a graph: C out C in ¼ A C in þ A C equi

(7)

which follows the general equation of y ¼ ax þ b; where a ¼ the slope, and b ¼ the part of the Y-axis. The no-net-flux method is not only a theoretically sound method to determine the driving force concentration but also a method that discriminates between two types of processes, that play a roll in tissue adjacent to the probe.

Fig. 4. Regression analysis according to the zero-net-ﬂux method. The Y-axis shows the net increase or decrease in dialysate glucose. At y ¼ 0, Cin ¼ Cout, so there is no net inﬂux or efﬂux of glucose in the probe: Cin ¼ Cequi. Changes in recovery (Seff) are reﬂected by changes in the slope of the regression line: Cout/Cequi. Changes in Cequi ( ¼ Ccap around the probe) are reﬂected by a parallel displacement of the regression line.

(1) A change in Cequi produces a parallel shift of the line. (2) If the line is pivoting the Cequi, a change in recovery is occurring. If blood glucose is decreasing, as measured by ﬁnger prick, Cequi should also decrease, as it is

651

Fig. 5. Dialysate glucose concentration, Cout (’), equilibrium glucose concentration Cequi ( ) and blood glucose obtained by ﬁngerprick (K) in a healthy volunteer. Error bars show the SD of the measurements during a day. Glucose concentrations (Cin) used: 0, 2.8, 8.3, and 11.1 mM (Wientjes et al., 1998).

the driving force concentration. In the plot, this will show up as a parallel shift of the line to the left.

VI. Implantation effects Referring to the results shown in Fig. 5, it is clear that the measured values by microdialysis showed signiﬁcant deviations during the ﬁrst days. The equilibrium concentration as well as the recovery was changing at constant blood glucose. Some process locally at the site of the probes caused a lower concentration for Cequi. After 12–24 h, the effect vanished. Besides this effect, we also have seen a decrease of 40% in the slope after insertion of the probes. This effect clearly affected the permeability P and/or the surface S of the probe. This effect lasted for 4–6 days dependent on the volunteer. The ﬁrst effect, a parallel shift of the line to lower values after insertion, may be caused by a sterile inﬂammation effect (Fig. 6). Insertion of any device in tissue causes damage to cells and blood vessels. The body reacts almost immediately

to stop bleeding and often the ﬂair response is seen, due to the release of histamine. It is known that an inﬂammation process, consumes glucose. A lower glucose concentration around the probe can explain the parallel shift of the line to the left. For the measurement of glucose, the effect vanishes within 12–24 h. This is not the case for the increase of recovery during almost a week. This slow effect is probably caused by a process to repair the microstructure of tissue at the site of the probe. For the measurement of glucose, this slow effect can be circumvented by using a probe of about 3 cm and a very low ﬂow rate of 0.3 mL/min or lower (Rosdahl et al., 1998). With these settings, one almost reaches the equilibrium concentration at the end of the probe membrane, resulting in almost 100% recovery. Looking at the extraction equation for an explanation of the slow increasing recovery, we have three parameters: permeability (P), probe surface (S), and the ﬂow rate (f). The last one is kept constant, so the changing recovery can only be caused by a change in permeability or a change in surface. At ﬁrst sight, it is strange to regard the

652

Fig. 6. glucose concentration, Cout (’), equilibrium glucose concentration Cequi ( ) and blood glucose obtained by ﬁngerprick (K) in a healthy volunteer during the ﬁrst day. Note the decrease of Cequi after 4 h, whereas the recovery (Cout/Cequi) remains about half the value of Cequi. This means a parallel displacement of the regression line in Fig. 4 (Wientjes et al., 1998).

probe surface as a variable, but one should bear in mind that the probe surface is covered by adipose cells and capillaries. Adipose cells are not releasing glucose, so the part of the probe surface, adjacent to cell walls, is not available for glucose transport to the probe. Therefore, it is useful to deﬁne an effective surface (Seff) of the probe membrane that actually receives glucose from the capillaries. If it is assumed, as in the model, that only capillaries in the near vicinity of the probe are delivering glucose, it is quite simple to relate the fall in recovery immediately after insertion to the closing up of damaged capillaries around the probe by the inﬂammation process. Less capillaries around the probe means a lower effective surface of the probe membrane. If longer diffusion paths are assumed, capillaries and cells further away from the probe should not be damaged, or far less so, as the needle is only cutting capillaries and cells on its route inside tissue. Then the permeability of the diffusion path may even increase due to the damaged tissue around the probe that may present a lower resistance. It is difﬁcult to assess a changing permeability in tissue around the probe, but it is possible

Fig. 7. Relation between skinfold thickness at the site of probe implantation and the mean recovery of glucose per patient divided into K, dayo4, r ¼ 0.82, n ¼ 8, po0.02; and J, dayZ4, r ¼ 0.69, n ¼ 8, po0.05 (Lutgers et al., 2000).

to obtain some information about the capillary density around the probe. The capillary density is higher in adipose tissue containing small cells (Gersh and Still, 1945). Fig. 7 shows for eight patients with type 1 diabetes, an inverse correlation between the skin fold thickness of the abdominal fat and the recovery of glucose (Lutgers

653

et al., 2000). Until a certain limit, the lipid drop inside the adipose cells is increasing if people are growing more obese. A lean person has a relatively small lipid drop, so more cells with capillaries are present in a unit tissue. These people have a higher recovery and only a small increase in recovery as compared with more obese patients. Although a change in permeability cannot be excluded, the result is far better explained by a changing Seff, due to an increasing number of capillaries in the vicinity of the probe. For lean and more obese patients, the difference in recovery remains, during the entire period of measurement, indicating an effect of capillary density around the probe membrane. Long-term stability: After 4–6 days, the recovery is stable during the time of measurement, in total 3 weeks (Wientjes et al., 1998). Permeability of tissue and probe membrane and also the effective surface is constant. Contrary to rats, humans are not encapsulating the probe quickly (Wisniewski et al., 2002). Clearly, a microdialysis probe is not alarming the immune system. Only simple ions are perfused that are also present in the interstitial ﬂuid, so the biocompatibility looks quite good. A microdialysis probe is a safe sampling device, but due to the changes in recovery during the ﬁrst week, long probes and very low ﬂow rates have to be applied. A microdialysis/sensor assembly measured reliably glucose values in type 1 diabetic patients during 4 days (Kapitza et al., 2003). However, the system is bulky and technical difﬁculties still have to be solved.

VII. The ‘‘delay time’’ phenomenon The function of the microcirculation is to reduce the distance for the exchange of materials to a length at which diffusion is sufﬁciently rapid. For dissolved substances as glucose, diffusion time from capillary to cell is about 1 s for a distance of 20–50 mm (Crone and Levitt, 1984). Experimentally substantial ‘‘delay times’’ are reported for a change in blood glucose to appear in subcutaneous tissue (from 2 min up to 40 min), using microdialysis or electrochemical measurement techniques

(Aussedat et al., 2000; Hagstrom et al., 1990; De Boer et al., 1993). These results are not compatible to the theoretical calculation of 1 s for the diffusion time. If some detection device is placed in subcutaneous tissue in the morning and measurements are done during the day, after with the experiment was ended in the evening, the implantation effects are fully operational: a changing Cequi, due to inﬂammation and a changing recovery due to disrupted capillaries. This may partly explain the wide variation in ‘‘delay time’’ results. The best-controlled experiments are clamp studies; during with a bolus injection of the solute is applied. Two experiments of clamp studies are mentioned earlier: a study in dogs and a study in type 1 diabetic patients (Figs. 3a and b). In both cases, a bolus injection of glucose was delivered. The interesting point of these studies is the onset of the rise in glucose that was equal for blood and tissue values and a difference of 15–20 min between blood glucose and tissue glucose at the end of the increase. It should be noted that for two different species and two different measurement devices the same results were obtained. So it seems that a general phenomenon is involved. In these experiments, venous blood glucose was measured frequently. In dogs, the sensor measured continuously glucose and in the patients glucose was sampled each 5 min from the probe. The only parameter we do not know is the blood glucose concentration around sensor and probe. This value is difﬁcult to measure, but it is a very important parameter, being the driving force concentration for diffusion of glucose to the measuring device. What can be measured easily is the blood glucose from the skin at two sites. Blood glucose by ﬁnger prick and blood glucose from the skin of the abdomen were measured. After 20 min of measurement, a bolus of glucose was injected. The results are clear-cut. In both cases, the increase was fast and for both sites the same, so the recovery is constant. However, there was an important difference between the blood-glucose concentration taken from the skin of the abdomen and blood glucose taken from the ﬁngertip. After 15 min, the difference between the two bloodglucose concentrations vanished. Obviously, there

654

is a mechanism regulating blood-glucose concentrations in various skin tissues after a sudden increase of glucose. The histological evidence for a muscle apparatus in the wall of arterioles in adipose tissue controlling blood ﬂow supports this assertion. So the distribution of blood-glucose concentrations at different sites is regulated, but not the rate of increase. This is in agreement with the extraction equation: no ‘‘delay time’’, but a concentration difference. Returning to the clamp studies; an increasing difference is seen between blood-glucose and the glucose concentration measured by sensor and probe, after a bolus of glucose was delivered. It is a violation of the extraction equation to see this increasing difference between venous blood glucose and the glucose values measured by sensor and probe as a ‘‘delay time’’ that has something to do with the diffusion time of glucose. Permeability and effective surface should not be dependent on a change in blood glucose. When the increasing difference between blood-glucose and the glucose concentration measured by sensor and probe is seen as a concentration difference, it is in agreement with the extraction equation. The time period of 15–20 min after which the concentration difference has vanished, was also seen in the ‘‘two skin sites’’ experiment. However, in the clamp studies blood glucose and sensor/probe measurements are compared; in the ‘‘two skin sites’’ experiment only blood glucoses were compared. In the last experiment, an equal rate of increase was seen, but that is not seen in the clamp studies after a bolus of glucose. Again the extraction equation should be the guide to explain the behavior of sensor and probe. Both devices are measuring the average intracapillary glucose. As the onset of the increase in blood and devices is equal for both, sensor and probe should measure the increase of the capillary blood glucose in adipose tissue. So there is a discrepancy between the reference blood glucose and blood glucose concentration in the capillaries of adipose tissue. After the bolus, the reference blood glucose is increasing slowly, but reaches the new clamp value again after 15–20 min.Considering the sluggish blood ﬂow in the capillaries and the heavy regulation between high-resistance and low-resistance pathways, this

could be the reason for the slow increase of glucose concentration in the capillaries of adipose tissue. In general, the message regarding the ‘‘delay phenomenon’’ is not that the measuring devices are causing ‘‘delay times’’, but that we have underestimated the differences in glucose concentrations in various tissues after a sudden increase of blood glucose.

VIII. Conclusion The results of experiments presented in this chapter have changed our ideas about the relevant parameters, governing transport of glucose in adipose tissue. In adipose tissue, the living cytoplasm of the cells is only a thin rim around a drop of lipid, like an inﬂated balloon. Also the cells are closely packed, making it highly unlikely for hydrophilic solutes to diffuse through the narrow interstitial sheet between the cell walls. This idea is supported by the peculiar membrane–membrane contact between the thin walled capillaries (0.03 mm) and the adipose cells (without interstitial ﬂuid in between), suggesting a direct ﬂux of nutrients from capillary to adipose cells. For instance, the diffusion distance for blood oxygen is nearly zero (Ryan and Curri, 1989). So if a microdialysis probe is present in adipose tissue, only capillaries in the vicinity of the probe may deliver glucose to the probe. It was shown, by using the no-net-flux method that the driving force for diffusion of glucose to the probe is the intracapillary concentration Ccap. Tissue around the probe consists of adipose cells and capillaries. Only capillaries deliver glucose to the probe membrane, so the effective surface Seff of the probe membrane is smaller. Capillary density depends on the size of the lipid drop in the adipose cell. Tissue with smaller cells has a higher capillary density per unit tissue (Gersh and Still, 1945). When a probe is inserted with the aid of a needle, capillaries around the probe are damaged and the Seff becomes smaller. During a period of 4–6 days, wound healing occurs, increasing the number of capillaries around the probe, resulting in an increased Seff and an increased recovery. This view was conﬁrmed by measuring the skin fold thickness in diabetic

655

(50–150 mm). Also, when the onset of a change in blood glucose is equal to the onset of a glucose change in tissue, but both curves have a different slope, then it is a violation of the extraction equation to regard this ‘‘delay time’’ as a diffusion time (Van Der Valk et al., 2002). Obviously, the notion of a ‘‘delay time’’ has another meaning. In a concentration/time plot, there are two possible interpretations: ﬁrst, we may see a difference between two curves along the horizontal time axis, at a certain concentration, as a time lag and second we may see a difference along the vertical concentration axis, as a concentration difference, at a certain time point. Referring to Fig. 8, there is no time lag during and after a steep increase of blood glucose

Blood glucose (mmol/l)

patients. For lean patients, with a higher capillary density, the fall in recovery after insertion of the probe, was far smaller than for more obese patients and also after 4 days the recovery of lean patients remained higher. Therefore, it is concluded, that Seff in the extraction equation is the variable and permeability, P, is the constant. Numerous investigators have reported ‘‘delay times’’ of 2–40 min after a change in blood glucose to appear in a measurement device in tissue. Glucose is a hydrophilic solute that in 1 s travels 20–50 mm by diffusion. If we take 2 min as the lowest ‘‘delay time’’ measured, a glucose molecule has to travel 2,400–6,000 mm to ﬁnd the measuring device. That is an impressive distance, if we compare that with the size of adipose cells 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3

* finger tip abdomen

* * * *

*

*

*

* * *

*

% difference

20 10 0 ****

-10

**

* * * *

*

-20 -30 -25

gradual decline -20

-15

-10

-5

sudden increase 0

5

10

15

20

25

30

35

40

time (min) Fig. 8. Meeasurement of capillary blood glucose in ﬁngertip and the abdominal skin before, during, and after an intravenous bolus injection of glucose in 12 healthy volunteers. During 15 min after the bolus, a signiﬁcant lower glucose concentration in abdominal skin was measured; 10–18% (po0.01), by ‘‘repeated measurement ANOVA’’ (Van Der Valk et al., 2002).

656

measured at two skin sites. There is only a temporary difference of 15–20 min in the concentration between the two sites. If we use some measuring device in tissue in a clamp study (Figs. 3a and b) and we also deliver a bolus of glucose at some time point, we are inclined to judge the difference as a delay time. In any case, it is not a diffusion time, as an increasing diffusion time would mean an increasing diffusion path, or a decreasing permeability. So if the onset is equal, it is not the measuring device that is causing a delay, but it is the intracapillary blood glucose in adipose tissue that causes the delay, for instance due to a sluggish blood ﬂow. Here again there is a concentration difference in blood glucose between two sites, after a sudden increase of glucose, which vanishes after 15 min.These effects are seen only after a rather steep increase of glucose and may, therefore, be regarded as distribution effects.

References Aussedat, B., Dupire-Angel, M., Gifford, R., Klein, J.C., Wilson, G.S. and Reach, G. (2000) Interstitial glucose concentration and glycemia: implications for continuous subcutaneous glucose monitoring. Am. J. Physiol. Endocrinol. Metab., 278(4): E716–E728. Bolinder, J., Ungerstedt, U. and Arner, P. (1992) Microdialysis measurement of the absolute glucose concentration in subcutaneous adipose tissue allowing glucose monitoring in diabetic patients. Diabetologia, 35(12): 1177–1180. Braverman, I.M. and Keh-Yen, A. (1981) Ultrastructure of the human dermal microcirculation III. The vessels in the midand lower dermis and subcutaneous fat. J. Invest. Dermatol., 77(3): 297–304. Braverman, I.M. and Yen, A. (1977) Ultrastructure of the human dermal microcirculation. II. The capillary loops of the dermal papillae. J. Invest. Dermatol., 68(1): 44–52. Bungay, P.M., Morrison, P.F. and Dedrick, R.L. (1990) Steady-state theory for quantitative microdialysis of solutes and water in vivo and in vitro. Life Sci., 46(2): 105–119. Clara, M. (1956) Die Arterio-Venoesen Anastomosen. Springer, Wien, New York. Crone, C. and Levitt, D.G., 1984 Capillary permeability to small solutes. In: Handbook of Physiology, Vol. IV – The Cardiovascular System (Part 1),American Physiology Society, pp. 411–466. Curri, S.B., 1984. Curri SB: Fat tissue microangiopathy; their possible signiﬁcance in the genesis of local obesity. In: Proceedings of the Congress of Angiology. Athens, International Union of Angiology.

De Boer, J., Plijter Groendijk, H. and Korf, J. (1993) Microdialysis probe for transcutaneous monitoring of ethanol and glucose in humans. J. Appl. Physiol., 75(6): 2825–2830. Gersh, I. and Still, M. (1945) Blood vessels in fat tissue. Relation to problems of gas exchange. J. Exp. Med., 81: 219–232. Hagstrom, E., Arner, P., Engfeldt, P., Rossner, S. and Bolinder, J. (1990) In vivo subcutaneous adipose tissue glucose kinetics after glucose ingestion in obesity and fasting. Scand. J. Clin. Lab. Invest., 50(2): 129–136. Hullegie, L.M., Lutgers, H.L., Dullaart, R.P., Sluiter, W.J., Wientjes, K.J., Schoonen, A.J. and Hoogenberg, K. (2000) Effects of glucose and insulin levels on adipose tissue glucose measurement by microdialysis probes retained for three weeks in Type 1 diabetic patients. Neth. J. Med., 57(1): 13–19. Jungueira, L.C., Cauneiro, J. and O’Kelly, R.O., 1998. Adipose tissue. In: Basic Histology, 9th ed. McGraw-Hill/Appleton & Lange, Stamford, CT, pp. 121–126. Kapitza, C., Lodwig, V., Obermaier, K., Wientjes, K.J., Hoogenberg, K., Jungheim, K. and Heinemann, L. (2003) Continuous glucose monitoring: reliable measurements for up to 4 days with the SCGM1 system. Diabetes Technol. Ther., 5(4): 609–614. Lonnroth, P., Jansson, P.A. and Smith, U. (1987) A microdialysis method allowing characterization of intercellular water space in humans. Am. J. Physiol., 253(2 Pt 1): E228–E231. Lutgers, H.L., Hullegie, L.M., Hoogenberg, K., Sluiter, W.J., Dullaart, R.P., Wientjes, K.J. and Schoonen, A.J. (2000) Microdialysis measurement of glucose in subcutaneous adipose tissue up to three weeks in Type 1 diabetic patients. Neth. J. Med., 57(1): 7–12. Maran, A. and Poscia, A. (2002) Continuous subcutaneous glucose monitoring: the GlucoDay system. Diabetes Nutr. Metab., 15(6): 429–433. Meadows, R. (1980) In: Pocket Atlas of Human Histology. Oxford University Press, New York, p. 40. Rebrin, K., Steil, G.M., van Antwerp, W.P. and Mastrototaro, J.J. (1999) Subcutaneous glucose predicts plasma glucose independent of insulin: implications for continuous monitoring. Am. J. Physiol., 277(3 Pt 1): E561–E571. Rosdahl, H., Hamrin, K., Ungerstedt, U. and Henriksson, J. (1998) Metabolite levels in human skeletal muscle and adipose tissue studied with microdialysis at low perfusion ﬂow. Am. J. Physiol., 274(5 Pt 1): E936–E945. Ryan, T.J. and Curri, S.B. (1989) Blood vessels and lymphatics. Clin. Dermatol., 7(4): 25–36. Schoemaker, M., Andreis, E., Roper, J., Kotulla, R., Lodwig, V., Obermaier, K., Stephan, P., Reuschling, W., Rutschmann, M., Schwaninger, R., Wittmann, U., Rinne, H., Kontschieder, H. and Strohmeier, W. (2003) The SCGM1 System: subcutaneous continuous glucose monitoring based on microdialysis technique. Diabetes Technol. Ther., 5(4): 599–608. Thulesius, O. (1992) Capillaroscopy with ﬁbreoptic video microscope. Vasa, 21(1): 87–88. Van Der Valk, P.R., Van Der Schatte Olivier-Steding, I., Wientjes, K.J., Schoonen, A.J. and Hoogenberg, K. (2002)

657 Alternative-site blood glucose measurement at the abdomen. Diabetes Care, 25(11): 2114–2115. Wientjes, K.J., Vonk, P., Vonk-van Klei, Y., Schoonen, A.J. and Kossen, N.W. (1998) Microdialysis of glucose in subcutaneous adipose tissue up to 3 weeks in healthy volunteers. Diabetes Care, 21(9): 1481–1488.

Wisniewski, N., Rajamand, N., Adamsson, U., Lins, P.E., Reichert, W.M., Klitzman, B. and Ungerstedt, U. (2002) Analyte ﬂux through chronically implanted subcutaneous polyamide membranes differs in humans and rats. Am. J. Physiol. Endocrinol. Metab., 282(6): E1316–E1323.

CHAPTER 7.3

Neurochemical monitoring in neurointensive care using intracerebral microdialysis Thomas Lieutaud1, Fre´de´ric Dailler1, Franc- ois Artru1 and Bernard Renaud2,3, 1

Service d’Anesthe´sie-Re´animation, Hoˆpital Neurologique, Groupement Hospitalier Est, Bron Cedex, France 2 Service de Biochimie, Hoˆpital Neurologique, Groupement Hospitalier Est, Bron Cedex, France 3 Plateforme Neurochem, Faculte´ de Pharmacie, Universite´ Claude Bernard et Institut Fe´de´ratif des Neurosciences de Lyon, Lyon Cedex 08, France

Abstract: For more than a decade, bedside cerebral microdialysis (C-MD) has been commercially available for biochemical monitoring of brain-injured patients. This has led to deﬁned normal biochemical range values, which are not yet fully validated for clinical use under pathological conditions. We have carefully reviewed the literature regarding the threshold levels proposed for various biochemical markers detected by microdialysis in stroke, brain trauma, subarachnoid hemorrhage and brain death. We also considered studies comparing C-MD data with PET-scan ﬁndings. In general, the results obtained by C-MD are in accordance with those obtained with PET-scan in brain-injured patients. However, we found very few papers where the absolute biochemical values in microdialysates were signiﬁcantly correlated with the clinical outcome, except for long-lasting very low glucose levels that were strongly related to brain death. For a reliable interpretation of C-MD values, the present review shows the importance of taking into account both the implantation site of the probe and the time resolution of the dialysis system. The changes observed through C-MD monitoring after therapeutic approaches such as modifying CPP, hyperglycemia, insulin therapy, hypo-, and hyperthermia for different brain injuries are also considered in this review. However, the interest to treat patients on the basis of dialysate biochemical data has not yet been demonstrated in appropriate multicentric clinical trials.

injury. The prognosis of SAH depends on the onset of a vasospasm that is likely to occur in the vascular territory of the aneuristical vessel. A severe vasospasm leads to an ischemia close to the one observed during thromboembolic stroke, due to a decrease in the cerebral blood ﬂow below the minimal threshold necessary for providing essential substrates for cell survival. After a ﬁrst diagnosis and, if required, a surgical intervention, the patient is admitted to a NICU. The management of brain injury in intensive care is intended to prevent secondary ischemic brain injury, penumbra in stroke or brain trauma, and vasospasm in SAH.

I. Introduction During the last decades, management in neurointensive care unit (NICU) of patients with brain lesions, especially with traumatic brain injury but also with stroke and subarachnoid hemorrhage (SAH), has much improved. Traumatic brain injury starts when a rapid acceleration or deceleration of the brain within the cranium occurs and induces edema, contusions, epi- or subdural hematoma, and diffuse axonal Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

659

DOI: 10.1016/S1569-7339(06)16035-X Copyright 2007 Elsevier B.V. All rights reserved

660

If occurring, this secondary ischemic brain injury decreases the patient prognosis in terms of mortality and morbidity (Elf et al., 2002). For the management of systemic secondary brain insult after traumatic brain injury, special guidelines are executed in pre-hospital and intensive care units (Maas et al., 1997; The Brain Trauma Foundation, 2000). To support the treatment, different types of monitoring are used: (1) Intracranial pressure (ICP) measurement to calculate cerebral perfusion pressure (CPP) that is the result of mean arterial pressure (MAP) ICP. An ICP value >25 mmHg is accompanied by a high risk of brain herniation. A CPP value o50 mmHg is associated with a risk of brain ischemia (Hlatky et al., 2003). (2) Trans-cranial Doppler to measure blood ﬂow velocity in cerebral arteries. The velocity variations correlate with cerebral blood ﬂow variations in stroke patients (Sugimori et al., 1995). (3) The use of an internal jugular catheter to measure oxygen saturation, lactate, and glucose concentration in venous cerebral blood (Schoon et al., 2002). (4) An oxygen probe in the brain to measure oxygen brain tissue pressure (PbtO2) (Van Santbrink et al., 1996). However, none of these monitoring methods measures directly brain metabolism and especially anaerobic metabolism. To reach this goal, cerebral microdialysis (C-MD) has been developed 15 years ago. This method allows physicians to follow the concentrations of brain substrates or metabolites that enter or exit the cerebral extracellular ﬂuid (ECF) under physiological as well as pathological conditions. The general aim of the present book chapter is to review some of the recent data and developments on the use of clinical microdialysis, especially in a NICU. More speciﬁcally, we tried to answer, at least partially, the following three questions: (i) How do compare results obtained by CMD with the more classical data collected in a NICU?

(ii) What is the potential signiﬁcance and also what are the limitations of C-MD data obtained under various pathophysiological conditions in a NICU? (iii) What is the inﬂuence of cares provided to patients on the biochemical data obtained with C-MD? II. How do compare results of cerebral microdialysis with more clinical data? II.A. Cerebral hemodynamic status and cerebral perfusion pressure as compared with microdialysis data The self-organizing map of Kohonen is a computing method allowing a recognition pattern and adaptative algorithms with a mathematic model. Using this model, Nelson et al. (2004) found a pronounced individualistic pattern that persisted throughout the whole period of analysis. However, it was not possible to establish a correlation between CPP, ICP, and C-MD data in 26 brain trauma patients. The authors did not ﬁnd a common underlying pattern of ICP or CPP that correlated to a MD pattern, due to the great initial variation of the cluster for each patient in comparison with others. They discussed this conclusion in the light of the complex physiopathology of brain trauma including global (hemorrhage, age, and genetics) as well as local (penumbra), external (severe brain trauma), and internal (plasma glucose concentrations and acidosis) factors. II.B. PbtO2 as compared with microdialysis data The correlation between PbtO2 and C-MD data has been studied in severe trauma patients (Hlatky et al., 2004). In patients with stable hemodynamic conditions (CPP>80 mmHg), 7 out of 57 had at least 10 periods of PbtO2 decreasing from 2471.8 to 3.671.4 mmHg. This was followed by a decrease in CPP due to an increase in ICP. C-MD glucose paralleled closely the PbtO2 values and dropped from 1.370.2 to 0.470.1 mM at the time of the lowest PbtO2 values. Pyruvate concentrations did not show any signiﬁcant variation during

661

episodes of low PbtO2. The concentrations of C-MD lactate and glutamate, recognized as markers of ischemia, exhibited a signiﬁcant increase but only for PbtO2 values o10 mmHg. The C-MD lactate/pyruvate (La/Py) ratio, classically interpreted in case of increase as a switch from aerobic to anerobic cellular metabolism, was immediately elevated in these patients and exhibited a twofold increase at the time of lowest PbtO2. All these variations were unrelated to changes observed in plasma concentrations of these substrates or metabolites. Thus, a decrease in PbtO2 was interpreted as a consequence of hypoperfusion, as well as a decrease in C-MD glucose concentration, which is also closely related to the level of perfusion. Other bedside measures were correlated with various biochemical markers: adenosine and xanthine were found to be increased in human brain ECF during episodes of jugular desaturation after brain trauma (Bell et al., 2001). II.C. Interpretation of ratios calculated between microdialysis data Although various ratios have been proposed, the only ratio widely used in C-MD is the La/Py ratio. The La/glucose and Py/glucose ratios may be of interest since they reﬂect the ability of the considered brain area to use the available substrates in the aerobic or anerobic pathways. In some cases, these ratios were modiﬁed sooner than the absolute values. Nevertheless, neither extensive studies nor a recent task force conference has recognized such ratios as clearly identiﬁed diagnosis or prognosis markers of brain injury in a NICU. III. What is the interest and what are the limitations of microdialysis? One of the main limitations of C-MD is its restricted spatial resolution, due to which it is unable to evaluate the functioning of the entire pathophysiological brain. Thus, the choice of the position where the probe has to be inserted, as well as the actual position reached for the probe, may interfere with the interpretation of the C-MD data. Moreover, although arterial pressure, ICP, and

cardiac rate are recorded beat by beat, C-MD is recorded for a much longer time, usually on an hourly basis. However, even if this time interval is shorter than the interval required for the other paraclinical examinations that are usually obtained in a NICU (computed tomography (CT), magnetic resonance imaging (MRI), PET, etc.), C-MD does not reﬂect immediately the changes occurring in brain perfusion or oxygenation, or the resulting changes in brain metabolism.

III.A. Influence of the site of microdialysis probe implantation under different pathological conditions In animal studies, dialysis probes have been implanted in a large number of brain areas, resulting eventually in very different biochemical values. However, in clinical studies, MD catheters are mostly inserted in brain cortical regions, mainly in frontal but less often in temporal or parietal regions in postoperative cases, resulting in a limited variability of the biochemical values in the microdialysates. Severe brain trauma patients revealed pronounced regional variations in blood ﬂow as shown by imaging techniques such as CT, MRI, and positron emission tomography-scan (PETscan). These blood ﬂow variations led to deﬁne different areas, such as the penumbra zone. In the penumbra, a reduced blood ﬂow prevents the normal functioning of brain tissue, although the area is still able to survive as basal blood ﬂow maintains the membrane polarization. However, a more pronounced reduction of substrate supply will induce irreversible damages enlarging the core zone, whereas correction of substrate supply (i.e., by increasing blood ﬂow or substrate concentrations) might induce a recovery of its functions without any signiﬁcant neurological deﬁcit. To evaluate the impact of the location of the catheter implantation site on the interpretation of the C-MD measures, we have focused our review on studies with multiple catheter insertions. In such a situation, one catheter is placed in ‘‘worse’’ position either surgically in the peri-contusional area or hematoma, or after frontal insertion veriﬁed by a CT scan. Another catheter is placed

662

in the controlateral brain hemisphere devoid of ischemia or brain lesion (called the ‘‘better’’ position). In some cases, the probes are placed in the core lesion or in an abdominal subcutaneous area. III.A.1. Brain death In a stroke patient, Berger et al. (1999) recorded brain death after malignant swelling. Two MD catheters were inserted in both hemispheres, and although the ipsilateral catheter immediately recorded data consistent with a central core ischemia, the controlateral one showed results similar to those of the ischemic side when the brain swelling started to compromise the cerebral blood ﬂow. In another study on patients with brain trauma leading to brain death, a decrease in glucose was seen that preceded, by several hours, the increase in ICP or the drop in CPP (Sta˚hl et al., 2001a). However, this glucose decrease was seen later in the ‘‘better’’ position when compared with the ‘‘worse ‘‘position. The cell death markers (glutamate and glycerol, Hillered et al., 1998) increased very signiﬁcantly but more than several hours after the fall in CPP had occurred. Lactate did not appear to be a good indicator for brain death as it was not consistently modiﬁed in comparison to baseline levels. In the ‘‘better’’ position controlateral to the initial traumatic or ischemic lesion, the deterioration of microdialysis results was generally observed when the CPP dropped to zero. Nevertheless, microdialysis probe inserted in the initially insulted brain hemisphere is generally able to display signiﬁcant changes earlier than classical monitoring in stroke or brain trauma patients leading to brain death.

III.A.2. Stroke Insertion of multiple catheters was carried out in stroke patients to evaluate and conﬁrm the pathophysiology of the penumbra or peri-infarcted zone (Berger et al., 2002). In the peri-infarcted lesion, concentrations of cell death indicators (glutamate and glycerol) were signiﬁcantly higher than those in the non-infarcted area, whereas the metabolic rate seemed decreased as shown by low concentrations of lactate as well as pyruvate. The La/Py ratio was higher in the penumbra than that in the non-infarcted brain. The central core catheter monitored concentrations of lactate, glutamate, and glycerol as high as in brain death patients (Table 1). This emphasizes that the great variability in results obtained in stroke patients depends on the location of the C-MD catheter. Thus, results have to be analyzed cautiously by taking into account this variability and also their dynamic pattern over time to evaluate the effects of therapeutic interventions such as hypothermia or modiﬁcation of CPP (see below). III.A.3. Brain trauma In 22 unifocal brain-lesioned patients treated with surgical evacuation, a C-MD probe was inserted in the peri-contusional zone suspected to be the penumbra zone, which was later conﬁrmed by CT scan. A second probe was inserted in normal tissue either contro- or ipsilaterally far away from the main lesion (Engstro¨m et al., 2005). Glucose, lactate, glutamate, glycerol, and the La/Py ratio revealed signiﬁcant differences between C-MD results obtained from penumbra zone in comparison with ipsi- or controlateral normal tissue (Table 2). Interestingly, glucose levels in the normal controlateral tissue were higher than those in other

Table 1. Mean concentrations of various metabolites or substrates in different brain areas during stroke ischemic injury (adapted from Berger et al., 2002)

Non-infarcted Penumbra Core

Lactate (mM)

Pyruvate (mM)

La/Py ratio

Glycerol (mM)

Glutamate (mM)

3.0 1.3 12

95 36 o4

37 69 582

22 82 1,187

3.6 12.6 453

Note that the difference in metabolite or substrate concentrations appears directly related to the distance of the probe from the ischemic core.

663 Table 2. Glucose concentrations (mean 7 SEM) in dialysates after brain trauma injury (data from Engstro¨m et al., 2005)

C-MD glucose (mM)

Penumbra

Ipsilateral

Controlateral

1.270.1

2.270.1

3.170.1

Note the inﬂuence of the site of probe insertion on the glucose brain content: the levels are signiﬁcantly higher in controlateral brain cortex considered as injury-free than those in the ipsilateral cortex; lowest levels are found in the peri-contusional zone (‘‘penumbra’’).

areas as well as in conscious patients (Reinstrup et al., 2000). This ﬁnding raise questions about the interpretation of cerebral ECF glucose concentrations in brain-injured patients without the knowledge of the plasma concentrations. In another study (Sta˚hl et al., 2001b) evaluating the effect of decreasing CPP in the management of brain trauma, patients were implanted with multiple probes in the ‘‘better’’ and ‘‘worse’’ positions. The results show that glucose in the ‘‘worse’’ position was lower than that in the ‘‘better’’ position. In contrast, lactate concentrations were higher in the ‘‘worse’’ position than those in the ‘‘better’’ position; these remained higher during the 72-h observation period, whereas the glucose levels decreased (1.7 vs. 2.2 mM) during the initial 6 h and then stabilized during the following 66 h. No signiﬁcant difference was observed at any time for La/ Py ratio (28 vs. 26). Glycerol concentrations were signiﬁcantly higher in the ‘‘worse’’ position than those in the better position during the ﬁrst 12 h. In conclusion, in the trauma patients, brain metabolism changes early and remains altered for a longer time in the most pathological zone, whereas the normal zone exhibits moderate changes. Thus, interpretation of C-MD has to be performed in a temporal manner, regarding especially the site of insertion, the type of brain injury, and the further development of the pathological condition. The latter depends mainly on cerebral blood ﬂow and cerebral metabolic rate of oxygen or glucose consumption, which can be determined by PET. III.B. Evaluation of brain metabolism: PET findings compared with microdialysis data Considering that C-MD provides direct information about brain metabolism, we compared these data with those obtained by the present ‘‘gold-standard’’ PET monitoring using radio-

markers such as 15O-H2O or cose (18F-FDG).

18

F-ﬂuorodeoxyglu-

III.B.1. Brain death PET and C-MD in a brain death patient have never been simultaneously performed. Meyer (1996) has reported that 18F-FDG–PET in a brain death patient did not show any intracerebral uptake or retention of the tracer. However, it is to be expected that the lack of 18F-FDG uptake, associated with the drop of CBF in brain death, will correspond to a very low or near zero concentration of glucose measured in the ECF by CMD. In poor-prognosis SAH patients, Schulz et al. (2000) have reported very signiﬁcant differences between patients dying from brain death after severe vasospasm and those without neurological deﬁcit. Lower levels of glucose and pyruvate and higher concentrations of glutamate, and at a lesser level of lactate, were observed in patients with developing brain death in comparison with a subgroup of patients without neurological deﬁcit.

III.B.2. SAH Lactate and glutamate in microdialysates have been found signiﬁcantly higher in SAH patients with symptomatic deﬁcit than in patients without deﬁcit (Sarrafzadeh et al., 2003). However, glucose levels were not initially altered except for the day after the probe insertion. In another study, a smaller number of SAH patients has been investigated by both microdialysis and 15O-H2O PETscan (Sarrafzadeh et al., 2004). A correlation was found between the decrease of rCBF from 54 to 33 mL 100 g1 min1 and the presence of symptomatic neurological deﬁcits. Microdialysis results correlated the best between glutamate and rCBF, followed by glycerol and lactate, whereas glucose levels and La/Py ratio were not correlated with

664

rCBF. On the other hand, for a more dramatic decrease in rCBF (below 20 mL 100 g1 min1), La/Py ratio displayed the best sensitivity and speciﬁcity to detect such a pronounced and prolonged ischemia. In contrast, glutamate levels, despite a very good sensitivity, were also altered in periods unrelated to a low rCBF. III.B.3. Stroke In stroke patients suffering from middle cerebral artery occlusion (MCAO) (Dohmen et al., 2003), 11C-ﬂumazenil PET and microdialysis with a probe inserted into the frontal lobe of the infarcted hemisphere were performed. Only PET measurements within 24 h after the onset of stroke can predict a malignant course and help in the selection of patients who might beneﬁt from invasive therapeutic strategies such as early hemicraniectomy. Microdialysis parameters identiﬁed the deterioration of the critical stage later than PET measurements. III.B.4. Brain trauma In 11 patients with traumatic brain injury, both 18-FDG-PET evaluating regional cerebral metabolic consumption of glucose (rCMRglc) and CMD were performed (O’Connell et al., 2005). The increase observed for rCMRglc was closely correlated with an increase in dialysate lactate and pyruvate concentrations as well as in La/Glu and Py/Glu ratios. Dialysate glucose, La/Py, or glutamate were not correlated with rCMRglc. In this study, the results obtained from 18-FDG-PET and microdialysis indicated a rise in metabolism rather than a shift toward non-oxidative metabolism because La/Py ratio remained in normal range. However, oxygen metabolism was not measured through either PbtO2 or CMRO2. III.C. How to analyze microdialysis data? Since brain metabolism is changing over time (either spontaneously or in response to treatments), the analysis of the microdialysis data must take the time resolution into account. Depending on the time scale chosen for the analysis of

microdialysis data, we have reviewed the literature data in the following four sections. III.C.1. Analysis of microdialysis data based on the entire microdialysis period In several studies, the analysis of microdialysis data is based on the entire microdialysis period. In this respect, one can determine the time spent below or over a threshold value for each substrate, metabolite, or ratio reported for the entire microdialysis period. This determines the number of dialysates concerned by the events (i.e., a value below or over the chosen threshold) compared with the total number of dialysates analyzed. Next, the frequency of these biochemical events can be correlated with clinical indicators. Carre´ et al. (2006) have found in brain trauma patients that the most signiﬁcant correlation was obtained between the number of events (La/Py ratio>25) and ICP. We have proposed a similar ‘‘time frequency index (TFI)’’ for the analysis of the dialysate results and have correlated some of these indexes with the NICU outcome (Lieutaud et al., 2006). In 28 patients suffering from severe SAH, we have determined the median TFI and then looked at the clinical outcome at discharge from the NICU. In bad outcome patients (Glascow outcome score (GOS) 1–2), the median TFI with ECF glucose concentrations o0.15 mM was 6.9%, a value that was signiﬁcantly higher (p ¼ 0.011) than in patients with better outcome (0%) (GOS 3–5). At a threshold of 0.6 mM, the difference failed to reach signiﬁcance (Fig. 1). The median TFI for the La/ Py ratio was not signiﬁcantly different at a value as high as 40 as well as 50. One could conclude that the more the patient spent time with ECF for glucose below 0.2, the worse the clinical outcome is. III.C.2. Analysis of microdialysis data based on 24-h periods To compare MD data and PET-scan results, Sarrazfadeh et al. (2002, 2003, 2004) have chosen to average the C-MD values for the 24 h before and after a PET-scan examination. Another possibility is to choose the worst 24 h of the whole microdialysis monitoring period.

665 1.6

GOS A GOS B

1.4

Mediane TFI

1.2 1 0.8 0.6 0.4 0.2 0 MD < 0.15 MD < 0.2

MD < 0.3 MD < 0.4 MD < 0.5 Glucose MD Threshold

MD < 0.6

MD < 0.7

Fig. 1. Comparison of the time spent below different low glucose levels for two groups of patients with bad (GOS A) or better (GOS B) outcome score. Note that the GOS A patients spent signiﬁcantly much more time o0.2 mM than GOS B patients. For higher glucose thresholds, the difference between the two groups is no longer signiﬁcant. These data suggest that the higher the incidence of very low glucose levels, the worse the clinical outcome is likely to be. The group entitled ‘‘GOS A’’ includes patients (n ¼ 13) with poor values of Glasgow outcome score (dead or severely disabled), whereas the group entitled GOS B includes patients (n ¼ 15) with better values of the Glasgow outcome score. The time spent below the considered glucose levels is expressed as the median of the ratios between the number of dialysates below this level over the total dialysate measurements: this ratio was called time frequency index and abbreviated TFI. The bars show the median value of the TFI for different levels of glucose in microdialysates (glucose MD), whereas the upper and lower vertical lines represent the 75 and 25 percentiles, respectively (*po0.05).

Regarding the clinical outcome scale (better vs. bad), our recent observations (unpublished data) showed a signiﬁcant difference between the two groups of patients for glucose in microdialysates. The group of patients with a bad outcome (GOS ¼ 1–2) had signiﬁcantly lower ECF glucose concentrations in comparison with the group of the better outcome. III.C.3. Point-to-point analysis for microdialysis data: taking into account the acute biochemical changes Another method for analyzing the microdialysis data is to compare each result with the immediately preceding result. This was applied to the La/Py ratio based on an increase of La/Py ratio> 20%. More than 80% of acute elevations of the La/Py ratio (i.e., a >20% increase between two

consecutive hourly samples) were explained by changes in one or several pathological conditions such as an increase of the ICP (16% of cases), hyperthermia (13%), and a decrease in PbtO2 (11%). Moreover, almost half of these acute elevations of the La/Py ratio were associated with a signiﬁcant decrease in ECF glucose concentration (Murat et al., 2006). Finally, preliminary data obtained with SAH patients show a signiﬁcant correlation between the number of acute elevations of the La/Py ratio and the worsening of the clinical outcome: the more these events occur, the worse the outcome is. III.C.4. High temporal resolution microdialysis: on-line data analysis An automated ﬂow-injection assay of microdialysates from the cerebral cortex has been developed

666

for clinical research in acute brain injury (Georganopoulou et al., 2000). A rapid-sampling monitoring at each 30-s interval is now available and has been linked to EEG activity in post-traumatic brain injury. The probes were inserted in the perilesion cerebral cortex after emergency craniotomy for traumatic lesions. It appeared that a rapid decrease of glucose was associated with an increase in lactate concentrations. Using an EEG strip cortical recording device adjacent to the microdialysis catheter, modiﬁcations of dialysate concentrations were linked to lowering amplitude of >50% suggesting spreading-depression episodes (Parkin et al., 2005). This elegant rapid-sampling MD device permits very high temporal resolution of dynamic changes in both biochemical and EEG functioning. The aggregate number of EEG events displayed the best correlation with a progressive reduction of glucose concentration. This result suggests that each spreading-depression event is associated with a decrease of 0.11 mM in brain ECF glucose. Changes in lactate concentration were not related to EEG activity. Thus, this new on-line analysis approach will greatly help to understand the pathophysiology of spreading depression in brain trauma patients and might improve signiﬁcantly the outcome of treatments in response to variations in ECF concentrations of substrates, metabolites, or ratios (Parkin et al., 2005). However, this on-line method is not routinely available and requires a pronounced time-investment in obtaining numerous samples for each patient. Another limitation of this new system is its lack of CE label, and consequently, the need to obtain the agreement to monitor each patient.

conditions, in which an active treatment was performed during C-MD.

IV. What is the inﬂuence of cares provided to patients on the biochemical data obtained with microdialysis?

IV.B. Body temperature

Before considering a treatment of patients based on C-MD data, it is necessary to address the reciprocal question regarding the inﬂuence of an active treatment on biochemical parameters of CMD. In the following sections, we have reviewed some data available for various pathological

IV.A. Hyperoxia After brain trauma injury, a low PbtO2 is signiﬁcantly associated with a reduction of rCBF, leading to a decrease of ECF glucose levels that does not correlate with the systemic blood concentrations of glucose or oxygen (Hlatky et al., 2004). This is followed, if clinical conditions worsen, by increases of lactate and glutamate concentrations in microdialysates, suggesting the development of a pronounced ischemia. Correction of PbtO2 values by exposure to normobaric hyperoxia for 24 h has been evaluated in a cohort of matched patients (Tolias et al., 2004). In this study, in which the microdialysis probe was inserted in the intact brain area controlateral to the most traumatized hemisphere, dialysate glucose concentrations were signiﬁcantly higher in the hyperoxia group in comparison with the control group. Lactate and glutamate levels diminished signiﬁcantly and this was associated with a signiﬁcant decrease of ICP from 15 to 12 mmHg without any change in CPP. Hyperoxia induced a signiﬁcant increase of pyruvate concentrations in contrast with changes observed in controls, suggesting an improvement of the aerobic metabolism of glucose in the brain during hyperoxia exposure. However, hyperoxia induced a continuous decrease of glucose concentrations more signiﬁcant in the control group than in hyperoxia group. This suggests a vasoconstrictive response induced by hyperoxia leading to limitations of blood supply despite the higher dissolved oxygen concentrations.

Hyperthermia has been considered for a long time as a brain stress resulting under a hypermetabolic condition, suggesting that hypothermia should help to treat patients with brain injury. For several years, hypothermia has been used in brain trauma without signiﬁcant clinical effects (Clifton et al., 2001) but with more success in stroke (Schwab et al., 1998). We have reviewed C-MD

667

studies that involved active or passive variations of body temperature in brain-injured patients. IV.B.1. Hypothermia Therapeutic hypothermia has been studied in stroke patients with a microdialysis catheter inserted in the frontal lobe at the infarcted site (Berger et al., 2002). Results obtained during hypothermia were compared with those obtained after passive rewarming (if ICP remained in a normal range). After obtaining conﬁrmation by CT scan that the catheter was inserted into noninfarcted, penumbra, or core areas, hypothermia (o34 1C) was applied for 48–118 h. A signiﬁcant reduction of all compounds related to anerobic metabolism was observed, except in the central core lesion where no effect was seen. Rewarming resulted in a >50% increase in the mean lactate, glutamate, and pyruvate concentrations in noninfarcted tissue from 2.6 to 3.6 mM, 1.8 to 3.0 mM, and 50 to 95 mM, respectively, whereas the glycerol content was elevated during rewarming (62 vs. 22 mM). Consequently, La/Py ratio was not signiﬁcantly modiﬁed by hypothermia or rewarming. In the penumbra area, hypothermia induced a signiﬁcant decrease in the glutamate, glycerol concentrations, and La/Py ratio, whereas pyruvate was signiﬁcantly increased. These variations were signiﬁcantly more pronounced than those in the non-infarcted areas. Correlations were obtained between lactate and pyruvate concentrations in non-infarcted zone and between glycerol and glutamate in the penumbra area. For microdialysates collected from the central core lesion, the results were consistent with those obtained in brain death patients (Berger et al., 1999), with a lack of effects of hypothermia on the concentrations of different substrates or metabolites. It can be concluded that studying metabolites in different brain areas of stroke patients by means of microdialysis improves our knowledge about the pathophysiology and clinical effectiveness of hypothermia. IV.B.2. Hyperthermia The incidence and detection of cerebral hyperthermia (CHT) in traumatic brain injury is much more frequent when the temperature is measured

by a brain thermistance instead of a Swan-Ganz catheter, with a mean temperature gradient of 0.16 1C between the pulmonary artery and the brain. This gradient rises to 0.41 1C when the temperature peaks. CHT has a signiﬁcant effect on ICP as an increase of 0.7 1C is associated with an elevation of the ICP of 7 mmHg (Rossi et al., 2001). Stocchetti et al. (2005) have performed a multimodal monitoring in brain trauma patients. When the brain temperature rose from 3870.5 to 39.370.3 1C, arteriojugular oxygen content difference decreased and PbtO2 rose signiﬁcantly, whereas ICP slightly increased while no metabolic change was observed. When the brain temperature returned to baseline, all variables of the multimodal monitoring returned to pre-hyperthermic values with still no change in microdialysis data. Thus, hyperthermia is unlikely to induce by itself changes in brain metabolism. IV.C. Cerebral perfusion pressure (CPP) CPP values that lead to secondary ischemic injury have been widely questioned for decades. Presently, in brain trauma, recommendations tend to limit the highest value of CPP to 70 mmHg (The Brain Trauma Foundation, 2000). However, the lowest suitable CPP has not yet been deﬁned. IV.C.1. Low CPP The Lund’s team has carried out two studies to deﬁne the minimal suitable CPP. The ﬁrst study analyzed dialysates in patients while the CPP was decreased using metoprolol and clonidine. They lowered the CPP (o60 mmHg) in 48 patients that were implanted with one or more probes. No signiﬁcant modiﬁcations of the metabolites or substrates were observed in the ‘‘worse’’ or ‘‘better’’ position of the dialysis probe, and no signiﬁcant rise in glutamate or glycerol concentrations was recorded if the CPP remained in the 50–70 mmHg range (Sta˚hl et al., 2001b). For CPP values o50 mmHg, no signiﬁcant alteration was observed on the better side, whereas lactate and La/Py ratio exhibited signiﬁcant increases on the worse side. Another study revealed that, depending on the CPP levels, lactate concentrations and La/Py

668

ratios of the ‘‘worse’’ probe were signiﬁcantly increased if the CPP decreased o50 mmHg. However, this difference was not observed in the ‘‘better’’ probe, suggesting the lack of low CPP effects on brain metabolism. If the comparison was made between CPP o50 and >70, the dialysates obtained from the ‘‘worse’’ probe were not different, suggesting a possible adverse effect of high CPP values. Finally, glucose values in the ‘‘worse’’ probe were not related to changes in CPP (Nordstro¨m et al., 2003). These results suggest that a high CPP (>70 mmHg) was as toxic as a low CPP (o50 mmHg) in the peri-contusional brain. It was concluded that a switch to anerobic metabolism in the penumbra zone is suspected when CPP decreased o50. Some questions have been raised by these studies. The PbtO2 was not monitored and the hypothesis based on reduction of rCBF consecutive to low CPP levels was not documented by changes in glucose concentrations. In addition, the clinical management was different among patients, as some patients received dihydroergotamine, clonidine, or metoprolol alone, or in combination (Nordstro¨m et al., 2003).

IV.C.2. High CPP Johnston et al. (2005) have studied the effects of a drug-induced increase of the CPP from 65 to 85 mmHg by using either dopamine (DA) or noradrenaline (NA) in a cross-over study. By using PbtO2 recording, microdialysis, and Swan-Ganz monitoring, they observed during NA infusion a signiﬁcant increase in PbtO2, but no changes in the biochemical parameters. Moreover, they noted more variation in the biochemical response with DA than with NA, probably because the cardiovascular system was more activated with DA than with NA. In conclusion, NA signiﬁcantly increases CPP without inducing metabolic changes while PbtO2 was signiﬁcantly increased. In another study, a PET-scan was included. After raising mean CPP from 70 to 90 mmHg using NA, similar results were obtained: PbtO2 increased from 1778 to 2278 mmHg and rCBF from 27.575.1 to 29.776.1 mL 100 mL1 min1. However, these changes were not accompanied by detectable metabolic effects (microdialysis data

and CMRO2) (Johnston et al., 2005). On the other hand, the treatment with DA was not associated with PbtO2 changes. This suggests that increasing CPP may prove to be of no beneﬁt. Another possibility is that the modiﬁcations in CPP were too short to cause signiﬁcant metabolic changes in brain-injured patients. We have preliminary results in SAH patients, indicating that patients with impaired PbtO2 and near zero glucose microdialysis levels might improve by pharmacotherapy that increases their mean CPP level by 20 mmHg. Counteracting the vessel narrowing might be the mechanism involved.

IV.D. Hyperglycemia and active treatment with insulin It is likely that hyperglycemia is related to mortality in patients with trauma (Van den Berghe et al., 2005), stroke (Williams et al., 2002), and SAH (Frontera et al., 2006). In this regard, insulin treatment has been shown to reduce the 12-month mortality rate in surgical intensive care patients (Van den Berghe et al., 2001) from 8.0% (conventional treatment) to 4.6% (intensive treatment). In the latter study, the 66 brain-injured patients were analyzed separately from the total 1,548 patients. A higher mortality rate was noticed (23.3% for conventional treatment and 18.2% for insulin treatment). In the subgroup treated with insulin, there was a signiﬁcant reduction of ICP, incidence of seizure and diabetes insipidus, reduced treatment of vasopressor, and a better outcome in comparison with the control group (Van den Berghe et al., 2005). This corroborates a study of Rovlias and Kotsou (2000) showing that hyperglycemia was more often related to the severity of injury, whereas a signiﬁcant correlation was established between hyperglycemia and bad outcome. Hyperglycemia was also associated with a poor clinical outcome in SAH patients since Frontera et al. (2006), in a prospective study on 281 patients, found that a mean daily plasma glucose >7.6 mM was associated with both a signiﬁcant increase in the length of stay in ICU and a rise of

669

incidence of cardiorespiratory and neurological complications (brain-stem herniation). In a multivariate analysis, increased glucose level was an independent factor for death or severe disability after 3-month treatment. The authors emphasized the use of intensive insulin therapy to achieve adequate levels of plasma glucose in patients with SAH. A correlation was established between the Hunt and Hess grade, APACHE-2 score, and the glucose level. Consequently, the patients with the most pronounced hyperglycemia should be considered as the most severe SAH patients. The different mechanisms by which hyperglycemia leads to detrimental effects are based on experimental studies showing deterioration of brain metabolism, more pronounced brain edema, and worse morphological presentations. However, whether insulin treatment really improves the clinical outcome has not been strictly demonstrated. Moreover, if the effects of insulin treatment on glucose plasma concentrations are well known, this is not the case with respect to cerebral glucose in dialysates as these levels often change independently of the plasma levels. It is evident that CMD might help to reveal the relationship between plasma and cerebral ECF glucose levels under pathological conditions. Under physiological conditions, this relationship is well established: a ratio of 0.4 was calculated between the vascular and the brain ECF compartments. In brain trauma patients, hyperglycemia was recorded by C-MD. In this study, patients were treated according to the Lund’s recommendations (Diaz-Parejo et al., 2003). Monitoring with two microdialysis catheters in the ‘‘worse’’ and ‘‘better’’ positions was carried out in 14 out of 24 patients. Comparisons were made between the microdialysates obtained before, during, and after hyperglycemia. The authors did not observe any signiﬁcant modiﬁcation of brain metabolism under intermediate hyperglycemia conditions (12–15 mM), whereas they found only a signiﬁcant increase in lactate concentrations under higher hyperglycemia conditions (>15 mM). The catheter location did not inﬂuence the results. ‘‘Worse’’ or ‘‘better’’ position displayed similar results in case of focal mass as well as diffuse brain-injured patients. However, the effects of

intensive treatments in brain-injured patients were suspected to be deleterious regarding the correlation between low glucose in dialysates and brain death. Intensive insulin treatment with a target range of plasma glucose 4.50–6.00 mM was compared with historical control group receiving loose insulin treatment as previously recommended (target ¼ 6.00–7.50 mM) (Vespa et al., 2006). Insulin treatment induces a signiﬁcantly larger incidence of low glucose ECF concentrations (o0.2 mM), an increase in La/Py ratio >40 incidence, and a higher frequency of increased glutamate concentrations (>5 mM). This was associated with a better oxygen extraction ratio in the treatment group estimated with PET-scan. The non-randomized design of this study led to a difference in the mean C-MD levels of glutamate and the La/Py ratio, which were higher in the conventional treatment group in comparison with the insulin-treated group and lower for C-MD glucose levels. Moreover, the authors published several years ago that the lower the glucose levels in the ECF, the worse is the outcome (Vespa et al., 2003). By monitoring 30 brain trauma patients with C-MD, the incidence rate of extracellular glucose values o0.2 mM (at a microdialysis perfusion rate of 2 mL min1 corresponding to 0.66 mM at a perfusion rate of 0.3 mL min1) varied between 18 and 30% during the 7 days of monitoring. Hypoglycemia was never related to low ECF glucose. Terminal herniation (6%) and seizures (10%) were the clinical conditions most frequently associated with low ECF glucose. However, in most cases, the origin of low cerebral ECF glucose levels was unexplained (72%). Extremely low cerebral ECF glucose concentrations, lasting more than 4 h, were all related to terminal herniation. In the same study, in the 479 dialysates (31% of total) in which glucose was found o0.2 mM, the La/Py ratio was >40, suggesting anerobic metabolism or ischemia. Increased glutamate (>5 mM) and elevated glycerol values were also associated in 91 and 39% of the dialysates, respectively. La/Py ratio increases were never associated with signiﬁcant modiﬁcation of ICP, PbtO2, or SvjO2. The clinical outcomes measured after 3 and 6 months were not signiﬁcantly related to ECF glucose.

670

In a case report, we have proposed to correct low ECF glucose levels (Lieutaud et al., 2006). As long as we have artiﬁcially maintained the microdialysis glucose levels >0.6 mM (at a microdialysis perfusion rate of 0.3 mL min1) by using glucose infusion to induce high plasma values, the multimodal parameters including PbtO2, ICP, CPP, C-MD, and SvjO2 remained between normal limits. IV.E. Hydrocephalus The biochemical effects of a short-lasting increased ICP in adult idiopathic hydrocephalus syndrome have been recorded in 10 patients during C-MD (Agren-Wilsson et al., 2003). The day after insertion of the shunt associated with the ICP transducer, PbtO2, and microdialysis probes into the right frontal lobe, an increase in ICP was induced by injection of artiﬁcial CSF. A ﬁrst increase to 35 mmHg and then a second to 45 mmHg were maintained during a 10-min period each. The microdialysis results showed that before reducing the ICP with the shunt opening, disrupted energy metabolism was observed as in SAH, as shown by a high La/Py ratio. The hydrodynamic modiﬁcations induced a signiﬁcant increase of both lactate and pyruvate in comparison with the control period of 2 h observed before ICP increase. No other signiﬁcant variations were observed. In contrast, Berger et al. (2005) did not ﬁnd any signiﬁcant modiﬁcation of microdialysis results after injection of 25 g of glycerol to lower ICP in malignant stroke patients. Thus, reduction of ICP has not been consistently proved to modify C-MD results. However, the short duration of the elevation of ICP by infusion of artiﬁcial CSF may have been unable to signiﬁcantly modify the brain metabolism. Similarly, a rapid decrease of ICP by osmotic therapy was equally unable to affect brain metabolism as revealed by C-MD. IV.F. Sedation Benzodiazepines and barbiturates are commonly used in a NICU to decrease ICP. Only thiopental was evaluated with C-MD. When used to manage severely increased ICP in seven patients, it was

associated with 37% reduction of lactate, 59% reduction of glutamate, and 66% reduction in aspartate in the extracellular space of the brain (Goodman et al., 1996). Therefore, it can be concluded that an intravenous bolus injection of barbiturates was able to reduce the altered cellular metabolism in severe brain injury. IV.G. Epilepsia and spreading-depression-like events Spreading-depression-like events occur frequently in areas of cerebral cortex adjacent to contusions in the injured human brain. In 11 patients with intracranial hematomas requiring surgery, a CMD probe was inserted into peri-lesional cerebral cortex and a four-channel electrocorticogram (ECoG) was recorded from a subdural strip adjacent to the catheter (Parkin et al., 2005). Progressive reduction in dialysate glucose was strongly correlated with the aggregate number of ECoG events. The adverse impact of low dialysate glucose on clinical outcome may be caused by recurrence of spontaneous spreading-depression-like events in the peri-lesion cortex. The relationship between treatment with antiepileptic drugs and spreading-depression-like events and the impact on biochemical parameters has not been studied by microdialysis at the present time. IV.H. Hyperventilation Hyperventilation is currently used, in case of emergency, to decrease high ICP. However, prolonged hyperventilation is not recommended due to the adverse vasoconstrictive effects of low PaCO2, which may induce ischemia. Hutchinson et al. (2002) have studied the effects of moderate hyperventilation in brain trauma patients. A decrease of 0.9 kPa induced a signiﬁcant decrease of glucose in dialysates without changes in the other biochemical parameters (lactate, pyruvate, La/Py ratio, and glutamate). IV.I. Anemia Sta˚hl et al. (2001) reported monitoring by C-MD of a traumatic brain death patient belonging to a

671 Table 3. Summary of the biochemical changes obtained with C-MD under several different spontaneous or induced pathophysiological conditions (see text for explanation)

Lactate Pyruvate Glucose Glutamate or glycerol La/Py ratio

Hyperoxia

Hypothermia

Hyperthermia

High CPP

Low CPP (o50 mmHg)

Hyperglycemia (>15 mM)

Intensive insulin treatment

Hyperventilation

k k ¼ ¼

k k NA k

¼ ¼ ¼ ¼

¼ ¼ ¼ ¼

m NA ¼ NA

m ¼ ¼ ¼

m NA k m

¼ ¼ k ¼

¼

¼

¼

¼

m

¼

m

¼

Adapted from Berger et al. (2002), Hutchinson et al. (2002), Diaz-Pajero et al. (2003), Magnoni et al. (2003), Nordstro¨m et al. (2003), and Vespa et al. (2006). NA: data not available.

religious group refusing blood transfusion. During the recording period, hemoglobin concentration ranged between 15 and 18 g L1 (normal values: 125–145 g L1). The microdialysis data revealed signiﬁcantly altered biochemical results consistent with the occurrence of brain death, 10–20 h before the increase of ICP. The blood as well as dialysate glucose concentrations remained at normal levels. Unfortunately, PbtO2 was not recorded as one could suspect that anerobic metabolism was implicated. Further studies will reveal the impact of anemia in brain-injured patients and will give indications for the threshold of the hemoglobin concentration that is relevant for a transfusion. V. Conclusion As summarized in Table 3, C-MD has allowed the characterization of various spontaneous and treatment-induced changes occurring commonly in a NICU. The present data and additional studies should contribute to treat patients in the future, at least partially, on the basis of microdialysis data (Hillered et al., 2005). However, although a ﬁrst consensus paper has been written by the main contributors to the ﬁeld of C-MD in a NICU (Bellander et al., 2004), there are still many questions to address or difﬁculties to clarify for a relevant interpretation and clinical use of C-MD data. First, the dialysis probe has to be implanted in the location supposed to be the most ‘‘at risk’’ for secondary ischemia. For the past few years, the location of the dialysis probe has been made possible by inclusion of gold at their tip, which may signiﬁcantly help to interpret C-MD data. As one

single isolated biochemical value in a microdialysate cannot verify the actual cerebral biochemistry, a temporal analysis by C-MD has to be performed to show the clinical evolution of the patient. Increasing the time resolution of C-MD and the use of additional biochemical markers will be necessary to (i) improve the clinical relevance of this neurochemical monitoring and (ii) increase the efﬁcacy of medical interventions for patients hospitalized in a NICU.

References Agren-Wilsson, A., Roslin, M., Eklund, A., Koskinen, L.O., Bergenheim, A.T. and Malm, J. (2003) Intracerebral microdialysis and CSF hydrodynamics in idiopathic adult hydrocephalus syndrome. J. Neurol. Neurosurg. Psychiatry, 74: 217–221. Bell, M.J., Robertson, C.S., Kochanek, P.M., Goodman, J.C., Gopinath, S.P., Carcillo, J.A., Clark, R.S., Marion, D.W., Mi, Z. and Jackson, E.K. (2001) Interstitial brain adenosine and xanthine increase during jugular venous oxygen desaturations in humans after traumatic brain injury. Crit. Care Med., 29: 399–404. Bellander, B.M., Cantais, E., Enblad, P., Hutchinson, P., Nordstr, C.H., Robertson, C., Sahuquillo, J., Smith, M., Stocchetti, N., Ungerstedt, U., Unterberg, A. and Vidiendal Olsen, N.V. (2004) Consensus meeting on microdialysis in neurointensive care. Intensive Care Med., 30: 2166–2269. Berger, C., Annecke, A., Aschoff, A., Spranger, M. and Schwab, S. (1999) Neurochemical monitoring of fatal middle cerebral artery infarction. Stroke, 30: 460–463. Berger, C., Sakowitz, O.W., Kiening, K.L. and Schwab, S. (2005) Neurochemical monitoring of glycerol therapy in patients with ischemic brain edema. Stroke, 36: e4–e6. Berger, C., Schabitz, W.R., Georgiadis, D., Steiner, T., Aschoff, A. and Schwab, S. (2002) Effects of hypothermia on excitatory amino acids and metabolism in stroke patients: a microdialysis study. Stroke, 33: 519–524.

672 Carre´, E., Cantais, E., Boret, H., Palmier, B. and Risso, J.J. (2006) Improvements of severe head injured patients management: focal (microdialysis and PtiO2) vs global monitoring. In: Di Chiara, G. (Ed.), Monitoring Molecules in Neuroscience. University of Cagliari, Italy, pp. 224–226. Clifton, G.L., Miller, E.R., Choi, S.C., Levin, H.S., McCauley, S., Smith, K.R. Jr., Muizelaar, J.P., Wagner, F.C. Jr., Marion, D.W., Luerssen, T.G., Chesnut, R.M. and Schwartz, M. (2001) Lack of effect of induction of hypothermia after acute brain injury. N. Engl. J. Med., 344: 556–563. Diaz-Parejo, P., Sta˚hl, N., Xu, W., Reinstrup, P., Ungerstedt, U. and Nordstro¨m, C.H. (2003) Cerebral energy metabolism during transient hyperglycemia in patients with severe brain trauma. Intensive Care Med., 29: 544–550. Dohmen, C., Bosche, B., Graf, R., Staub, F., Kracht, L., Sobesky, J., Neveling, M., Brinker, G. and Heiss, W.D. (2003) Prediction of malignant course in MCA infarction by PET and microdialysis. Stroke, 34: 2152–2158. Elf, K., Nilsson, P. and Enblad, P. (2002) Outcome after traumatic brain injury improved by an organized secondary insult program and standardized neurointensive care. Crit. Care Med., 30: 2129–2134. Engstro¨m, M., Polito, A., Reinstrup, P., Romner, B., Ryding, E., Ungerstedt, U. and Nordstro¨m, C.H. (2005) Intracerebral microdialysis in severe brain trauma: the importance of catheter location. J. Neurosurg., 102: 460–469. Frontera, J.A., Fernandez, A., Claassen, J., Schmidt, M., Schumacher, H.C., Wartenberg, K., Temes, R., Parra, A., Ostapkovich, N.D. and Mayer, S.A. (2006) Hyperglycemia after SAH: predictors, associated complications, and impact on outcome. Stroke, 37: 199–203. Georganopoulou, D.G., Carley, R., Jones, D.A. and Boutelle, M.G. (2000) Development and comparison of biosensors for in-vivo applications. Faraday Discuss., 116: 291–303 discussion 335–351. Goodman, J.C., Valadka, A.B., Gopinath, S.P., Cormio, M. and Robertson, C.S. (1996) Lactate and excitatory amino acids measured by microdialysis are decreased by pentobarbital coma in head-injured patients. J. Neurotrauma, 13: 549–556. Hillered, L., Valtysson, J., Enblad, P. and Persson, L. (1998) Interstitial glycerol as a marker for membrane phospholipid degradation in the acutely injured human brain. J. Neurol. Neurosurg. Psychiatry, 64: 486–491. Hillered, L., Vespa, P.M. and Hovda, D.A. (2005) Translational neurochemical research in acute human brain injury: the current status and potential future for cerebral microdialysis. J. Neurotrauma, 22: 3–41. Hlatky, R., Valadka, A.B., Goodman, J.C., Contant, C.F. and Robertson, C.S. (2004) Patterns of energy substrates during ischemia measured in the brain by microdialysis. J. Neurotrauma, 21: 894–906. Hlatky, R., Valadka, A.B. and Robertson, C.S. (2003) Intracranial hypertension and cerebral ischemia after severe traumatic brain injury. Neurosurg. Focus, 14: 1–6. Hutchinson, P.J., Gupta, A.K., Fryer, T.F., Al-Rawi, P.G., Chatﬁeld, D.A., Coles, J.P., O’Connell, M.T., Kett-White, R.,

Minhas, P.S., Aigbirhio, F.I., Clark, J.C., Kirkpatrick, P.J., Menon, D.K. and Pickard, J.D. (2002) Correlation between cerebral blood ﬂow, substrate delivery, and metabolism in head injury: a combined microdialysis and triple oxygen positron emission tomography study. J. Cereb. Blood Flow Metab., 22: 735–745. Johnston, A.J., Steiner, L.A., Coles, J.P., Chatﬁeld, D.A., Fryer, T.D., Smielewski, P., Hutchinson, P.J., O’Connell, M.T., Al-Rawi, P.G., Aigbirihio, F.I., Clark, J.C., Pickard, J.D., Gupta, A.K. and Menon, D.K. (2005) Effect of cerebral perfusion pressure augmentation on regional oxygenation and metabolism after head injury. Crit. Care Med., 33: 189–195 discussion 255–257. Lieutaud, T., Espy, C., Tournoud, M., Bodonian, C., Dailler, F., Convert, J., Grousson, S., Artru, F. and Renaud, B. (2006) Low extracellular glucose levels during cerebral microdialysis after subarachnoid haemorrhage: should we correct them? In: Di Chiara, G. (Ed.), Monitoring Molecules in Neuroscience. University of Cagliari, Italy, pp. 230–232. Magnoni, S., Ghisoni, L., Locatelli, M., Caimi, M., Colombo, A., Valeriani, V. and Stochetti, N. (2003) Lack of improvement in cerebral metabolism after hyperoxia in severe head injury: a microdialysis study. J. Neurosurg., 98: 952–958. Maas, A.I., Dearden, M. and Teasdale, G.M. et al. (1997) EBIC – guidelines for the management of severe head injury in adults. Acta Neurochir., 139: 286–294. Meyer, M.A. (1996) Evaluating brain death with positron emission tomography: case report on dynamic imaging of 18F-ﬂuorodeoxyglucose activity after intravenous bolus injection. J. Neuroimaging, 6: 117–119. Murat, J.B., Limpar, P., Bonvin, A., Convert, J., Grousson, S., Dailler, F., Artru, F. and Renaud, B. (2006) Bedside cerebral microdialysis in subarachnoid haemorrhage: changes in lactate/pyruvate ratio associated with alterations in other biological and physiological parameters. In: Di Chiara, G. (Ed.), Monitoring Molecules in Neuroscience. University of Cagliari, Italy, pp. 227–229. Nelson, D.W., Bellander, B.M., Maccallum, R.M., Axelsson, J., Alm, M., Wallin, M., Weitzberg, E. and Rudehill, A. (2004) Cerebral microdialysis of patients with severe traumatic brain injury exhibits highly individualistic patterns as visualized by cluster analysis with self-organizing maps. Crit. Care Med., 32: 2428–2436. Nordstro¨m, C.H., Reinstrup, P., Xu, W., Gardenfors, A. and Ungerstedt, U. (2003) Assessment of the lower limit for cerebral perfusion pressure in severe head injuries by bedside monitoring of regional energy metabolism. Anesthesiology, 98: 809–814. O’Connell, M.T., Seal, A., Nortje, J., Al-Rawi, P.G., Coles, J.P., Fryer, T.D., Menon, D.K., Pickard, J.D. and Hutchinson, P.J. (2005) Glucose metabolism in traumatic brain injury: a combined microdialysis and [18F]-2-ﬂuoro-2-deoxyD-glucose-positron emission tomography (FDG-PET) study. Acta Neurochir. Suppl. 95: 165–168. Parkin, M., Hopwood, S., Jones, D.A., Hashemi, P., Landolt, H., Fabricius, M., Lauritzen, M., Boutelle, M.G. and Strong, A.J. (2005) Dynamic changes in brain glucose and lactate in

673 pericontusional areas of the human cerebral cortex, monitored with rapid sampling on-line microdialysis: relationship with depolarisation-like events. J. Cereb. Blood Flow Metab., 25: 402–413. Reinstrup, P., Sta˚hl, N., Mellergard, P., Uski, T., Ungerstedt, U. and Nordstro¨m, C.H. (2000) Intracerebral microdialysis in clinical practice: baseline values for chemical markers during wakefulness, anesthesia, and neurosurgery. Neurosurgery, 47: 701–709 discussion 709–710. Rossi, S., Zanier, E.R., Mauri, I., Columbo, A. and Stocchetti, N. (2001) Brain temperature, body core temperature, and intracranial pressure in acute cerebral damage. J. Neurol. Neurosurg. Psychiatry, 71: 448–454. Rovlias, A. and Kotsou, S. (2000) The inﬂuence of hyperglycemia on neurological outcome in patients with severe head injury. Neurosurgery, 46: 335–342 discussion 342–343. Sarrafzadeh, A.S., Haux, D., Lu¨demann, L., Amthauer, H., Plotkin, M., Kuchler, I. and Unterberg, A.W. (2004) Cerebral ischemia in aneurysmal subarachnoid hemorrhage: a correlative microdialysis-PET study. Stroke, 35: 638–643. Sarrafzadeh, A., Haux, D., Sakowitz, O., Benndorf, G., Herzog, H., Kuechler, I. and Unterberg, A. (2003) Acute focal neurological deﬁcits in aneurysmal subarachnoid hemorrhage: relation of clinical course, CT ﬁndings, and metabolite abnormalities monitored with bedside microdialysis. Stroke, 34: 1382–1388. Schoon, P., Mori, L.B. and Orlandi, G. et al. (2002) Incidence of intracranial hypertension related to jugular bulb oxygen saturation disturbances in severe traumatic brain injury patients. Acta Neurochir., 81: 285–287. Schulz, M.K., Wang, L.P., Tange, M. and Bjerre, P. (2000) Cerebral microdialysis monitoring: determination of normal and ischemic cerebral metabolisms in patients with aneurysmal subarachnoid hemorrhage. J. Neurosurg., 93: 808–814. Schwab, S., Schwarz, S., Spranger, M., Keller, E., Bertram, M. and Hacke, W. (1998) Moderate hypothermia in the treatment of patients with severe middle cerebral artery infarction. Stroke, 29: 2461–2466. Sta˚hl, N., Mellergard, P., Hallstro¨m, A., Ungerstedt, U. and Nordstro¨m, C.H. (2001a) Intracerebral microdialysis and bedside biochemical analysis in patients with fatal traumatic brain lesions. Acta Anaesthesiol. Scand., 45: 977–985. Sta˚hl, N., Ungerstedt, U. and Nordstro¨m, C.H. (2001b) Brain energy metabolism during controlled reduction of cerebral perfusion pressure in severe head injuries. Intensive Care Med., 27: 1215–1223.

Stocchetti, N., Protti, A., Lattuada, M., Magnoni, S., Longhi, L., Ghisoni, L., Egidi, M. and Zanier, E.R. (2005) Impact of pyrexia on neurochemistry and cerebral oxygenation after acute brain injury. J. Neurol. Neurosurg. Psychiatry, 76: 1135–1139. Sugimori, H., Ibayashi, S., Fujii, K., Sadoshima, S., Kuwabara, Y. and Fujishima, M. (1995) Can transcranial Doppler really detect reduced cerebral perfusion states? Stroke, 26: 2053–2060. The Brain Trauma Foundation. (2000) The American Association of Neurological Surgeons. The Joint Section on Neurotrauma and Critical Care. Guidelines for cerebral perfusion pressure. J. Neurotrauma, 17: 507–511 Review. Tolias, C.M., Reinert, M., Seiler, R., Gilman, C., Scharf, A. and Bullock, M.R. (2004) Normobaric hyperoxia-induced improvement in cerebral metabolism and reduction in intracranial pressure in patients with severe head injury: a prospective historical cohort-matched study. J. Neurosurg., 101: 435–444. Van den Berghe, G., Schoonheydt, K., Becx, P., Bruyninckx, F. and Wouters, P.J. (2005) Insulin therapy protects the central and peripheral nervous system of intensive care patients. Neurology, 26: 1348–1353. Van den Berghe, G., Wouters, P., Weekers, F., Verwaest, C., Bruyninckx, F., Schetz, M., Vlasselaers, D., Ferdinande, P., Lauwers, P. and Bouillon, R. (2001) Intensive insulin therapy in the critically ill patients. N. Engl. J. Med., 345: 1359–1367. Van Santbrink, H., Maas, A.I. and Avezaat, C.J. (1996) Continuous monitoring of partial pressure of brain tissue oxygen in patients with severe head injury. Neurosurgery, 38: 21–31. Vespa, P., Boonyaputthikul, R., McArthur, D.L., Miller, C., Etchepare, M., Bergsneider, M., Glenn, T., Martin, N. and Hovda, D. (2006) Intensive insulin therapy reduces microdialysis glucose values without altering glucose utilization or improving the lactate/pyruvate ratio after traumatic brain injury. Crit. Care Med., 34: 850–856. Vespa, P., McArthur, D., O’Phelan, K., Glenn, T., Etchepare, M., Kelly, D., Bergsneider, M., Martin, N.A. and Hovda, D.A. (2003) Persistently low extracellular glucose correlates with poor outcome 6 months after human traumatic brain injury despite a lack of increased lactate: a microdialysis study. J. Cereb. Blood Flow Metab., 23: 865–877. Williams, L.S., Rotich, J., Qi, R., Fineberg, N., Espay, A., Bruno, A., Fineberg, S.E. and Tierney, W.R. (2002) Effects of admission hyperglycemia on mortality and costs in acute ischemic stroke. Neurology, 59: 67–71.

CHAPTER 7.4

Microdialysis in the human brain: clinical applications Urban Ungerstedt and Elham Rostami Department of Physiology and Pharmacology, Karolinska Institute, Stockholm, Sweden

Abstract: Microdialysis is a technique for sampling the chemistry of the interstitial ﬂuid of tissues and organs in animal and man. It is minimally invasive and simple to perform in a clinical setting. Although microdialysis samples essentially have all small molecular substances present in the interstitial ﬂuid and the use of microdialysis in neurointensive care has focused on markers of ischemia and cell damage. The lactate/pyruvate ratio is a well-known marker of changes in the redox state of cells caused by, for example, ischemia. Glycerol is an integral component of cell membranes. Loss of energy due to ischemia eventually leads to an inﬂux of calcium and a decomposition of cell membranes, which liberates glycerol into the interstitial ﬂuid. Thus, the lactate/pyruvate ratio and glycerol have become the most important markers of ischemia and cell membrane damage. While the primary insult at the site of the accident is beyond our control, secondary insults during intensive care should be avoided by all means. Therefore, the single most important ﬁnding from microdialysis studies is the profound difference in the vulnerability of the tissue adjacent to a lesion as compared with normal brain tissue allowing early detection of secondary insults after traumatic brain injury as well as the onset of vasospasm after subarachnoid hemorrhage. I. Introduction

II. Methodology

This article focuses on the use of microdialysis in the human brain. The emphasis is on the use and interpretation of bedside microdialysis in clinical practice and we apologize for leaving out several excellent articles on clinical ﬁndings not yet applicable in routine intensive care. We have attempted to highlight those facts where there is unanimous agreement on ﬁndings as well as interpretations. In this way we hope to present knowledge that may be useful for preventing and relieving secondary insults, predicting outcome and guiding therapy during neurointensive care.

Microdialysis is a technique for sampling the chemistry of the interstitial ﬂuid of tissues and organs. It started as an animal research technique but has gradually been applied to man as well as animal (Ungerstedt, 1991). It is minimally invasive and simple to perform in a clinical setting. It has become a standard technique in physiological and pharmacological investigations on animals with over 9,000 published papers. During the last 10 years it has developed into a clinically useful technique with close to a 1,000 papers published on investigations of brain as well as peripheral tissues. The idea of inserting a dialysis membrane into the tissue where a continuous ﬂow of physiological ﬂuid inside equilibrates with the interstitial ﬂuid outside was conceived of more than 30 years ago by Delgado et al. (1972) and Ungerstedt and Pycock (1974) using slightly different technical

Corresponding author: E-mail: [email protected]

B.H.C. Westerink and T.I.F.H. Cremers (Eds.) Handbook of Microdialysis, Vol. 16 ISBN 0-444-52276-X

675

DOI: 10.1016/S1569-7339(06)16036-1 Copyright 2007 Elsevier B.V. All rights reserved

676

approaches. Today the Ungerstedt approach of using a ‘‘hollow ﬁber’’ has become the standard approach. In its simplicity it forms a ‘‘biosensor’’ where samples of the tissue chemistry are transported out of the body for analysis in contrast to the traditional biosensor where the analysis takes place inside the body. The availability of modern analytical techniques has made microdialysis a ‘‘universal’’ biosensor capable of monitoring essentially every small- and medium-sized molecular compound in the interstitial ﬂuid of endogenous as well as exogenous origin. The original microdialysis ‘‘probes’’ for animal use have developed into ﬂexible and sterile microdialysis ‘‘catheters’’ (CMA Microdialysis, Stockholm), which are approved for the use in human brain and peripheral tissues (Fig. 1). The dialysis membrane at the distal end of a microdialysis catheter functions like a blood capillary. Chemical substances from the interstitial ﬂuid diffuse across the membrane into the perfusion ﬂuid inside the catheter. The recovery of a particular substance is deﬁned as the concentration in the dialysate expressed as percent of the concentration in the interstitial ﬂuid. The concept of recovery, however, is not always well deﬁned, especially in clinical investigations

Fig. 1. CMA 70 brain microdialysis catheter. The luer ﬁtting is connected to a syringe. The inﬂow and outﬂow tubes are surrounded by a sliding cuff, which is used to suture the catheter to the skin of the scalp. The two tubes join in the cylindrical ‘‘liquid cross’’, which connects to the shaft and the dialysis membrane. The dialysis membrane has a gold tip making it visible on CT. The needle of the vial holder penetrates the vial membrane when the vial is pushed into the holder. The sample is collected in the neck of the vial just under the membrane.

where it may be deﬁned as the dialysate concentration expressed as percent of the concentration of the particular substance in peripheral blood. This deﬁnition is based on the belief that the concentration of the substance of interest is the same in blood as in the interstitial ﬂuid, which may not always be the case. A low perfusion ﬂow and a long dialysis membrane give a high recovery. If the membrane is long enough and the ﬂow slow enough, the concentration in the dialysate will approach the concentration in the interstitial ﬂuid, that is, recovery will be close to 100%. In case of the human brain the common perfusion ﬂow is 0.3 mL/min and the length of the membrane is usually 10 mm (allowing exact positioning in relation to, e.g., a lesion). Under these conditions the recovery has been estimated to be 70% (Hutchinson et al., 2000a, b). By using a longer membrane, for example, 30 mm, and the same perfusion ﬂow it is possible to reach 100% recovery in the human brain. This is a considerable difference compared with experiments in the animal brain where the membrane length might be 2 mm and the ﬂow 2 mL/min yielding a concentration in the dialysate that is about two potencies lower than in the dialysate from the human brain. It is important to realize that the concentration in the dialysate not only depends on the ﬂow and the length of the membrane but also on the supply of substances from blood capillaries as well as uptake and release from cells. This has been rarely discussed in experiments on animals but is an important factor to consider when interpreting data from the human brain. For example, the supply of glucose to the microdialysis catheter may decrease due to a decrease in the capillary blood ﬂow or due to an increase in the cell uptake of glucose. The high recovery of substances that can be achieved in the human brain makes it possible to analyze most neurotransmitters and energy metabolites but also cytokines (Hillman et al., 2005) and small proteins. The priority of most investigators of the human brain has so far been to arrive at a clinically useful application of microdialysis. Therefore, the majority of studies have been done on patients with severe brain trauma or

677

hemorrhage where an analysis of brain energy state is of direct clinical importance and where such knowledge may improve the outcome of the patient. In this article we have therefore concentrated on the use of microdialysis for detection and monitoring of ischemic states in the human brain. II.A. Biochemical markers of ischemia and cell damage The interstitial ﬂuid is the ‘‘cross road’’ of all substances passing between cells and blood capillaries. By monitoring this compartment in the brain it is possible to get crucial information about the biochemistry of neurons and glia and how seriously brain cells are affected by, for example, ischemia, hyperemia, trauma, hemorrhage, vasospasm as well as various physiological, pharmacological, and surgical interventions during intensive care. Although microdialysis samples essentially have all small molecular substances present in the interstitial ﬂuid and the use of microdialysis in neurointensive care has focused on markers of ischemia and cell damage. The reason is that they are of obvious importance for the survival of the tissue, well understood from a biochemical point of view and easy to interpret in the clinical setting of intensive care. Microdialysis tells us how cells react to an increase or decrease in the supply of oxygen and glucose. However, while normal brain tissue may not suffer when exposed to a moderate decrease in oxygen and glucose, vulnerable cells in the pericontusional penumbra may simply not survive. In this way, severe secondary damage to brain tissue may pass unnoticed if microdialysis is not performed in the most vulnerable tissue of the brain (see below). II.A.1. Lactate/pyruvate ratio The lactate/pyruvate ratio is a well-known marker of changes in the redox state of cells caused by, for example, ischemia (Siesjo¨, 1978). Pyruvate is formed from glucose in the anaerobic part of the glycolysis generating two molecules of ATP. It enters the citric acid cycle provided that oxygen is available. The citric acid cycle is the dominant

producer of energy yielding 32 molecules of ATP. If the tissue is exposed to ischemia (a decrease in blood ﬂow causing an inadequate supply of oxygen and glucose) the production of ATP from the citric acid cycle decreases (Fig. 2). The cells attempt to compensate for the decrease in ATP production by increasing the turn over of glucose in the anaerobic part of the glycolysis. During this process it is necessary to regenerate NAD+ from NADH by converting pyruvate to lactate, which causes an increase in lactate and the lactate/pyruvate ratio. The decrease in glucose delivery from blood capillaries causes a fall in the glucose concentration in the interstitial ﬂuid. This leads to a decreased production of pyruvate due to lack of glucose. In the dialysate this is seen as a fall in pyruvate and a further increase in the lactate/ pyruvate ratio, that is, a worsening of the ischemia. In a recent study Vespa et al. (2005) compared the lactate/pyruvate ratio with positron emission tomography (PET) for metabolism of glucose and oxygen and concluded that an increase in lactate/pyruvate ratio is a sign of metabolic crisis that is not necessarily synonymous with ischemic cell damage. This highlights the usefulness of glycerol as a marker of cell membrane

Fig. 2. Schematic diagram of the glycolysis. Lack of oxygen will block the production of ATP from the citric acid cycle. To generate energy from the anaerobic glycolysis NADH needs to be converted to NAD+, which favors the conversion of pyruvate to lactate. The result will be an increase in lactate. Ischemia eventually leads to a decrease in the available glucose, which decreases the production of pyruvate. The sum effect of these events will be an increase in the lactate/pyruvate ratio.

678

decomposition and cell damage (see below) in combination with the lactate/pyruvate ratio. The use of a ratio between two analytes has the advantage of abolishing the inﬂuence of changes in catheter recovery, as such a change will inﬂuence lactate and pyruvate to a similar degree. Therefore, the lactate/pyruvate ratio may be used to compare the redox state of different tissues in one individual as well as in different individuals. The ratio is essentially the same in all tissues, that is, 20. We consider a ratio above 25 as a sign of tissue ischemia. Lactate alone is a less good marker of the redox state of cells as an increase in lactate may be due to hypoxia, ischemia, as well as hyper-metabolism (Persson et al., 1996).

II.A.2. Glycerol Glycerol is an integral component of cell membranes. Loss of energy due to ischemia leads to an inﬂux of calcium into cells, activation of phospholipases, and eventually to a decomposition of cell membranes, which liberates glycerol into the interstitial ﬂuid (Hillered et al., 1998). Considering the fast changes in glycerol concentration in vulnerable peri-contusional brain tissue, which are often related to changes in cerebral perfusion pressure (CPP), it seems likely that cells may react by ‘‘leaking’’ more or less glycerol due to the severity of the ischemia. The normal glycerol concentration in the dialysate from the brain when using a 10 mm dialysis membrane and a perfusion ﬂow of 0.3 mL/min is 50–100 mM (Reinstrup et al., 2000). In subcutaneous adipose tissue, in contrast, glycerol originates from the splitting of fat (triglycerides) into free fatty acids and glycerol. This process is controlled by the local sympathetic noradrenalin nerve terminals. Glycerol in subcutaneous tissue is therefore an indirect marker of sympathetic tone in the dermatome where the catheter is inserted (Hagstrom-Toft et al., 1993). During intensive care a subcutaneous catheter may be inserted in the peri-umbilical region monitoring glycerol as an indicator of sympathetic ‘‘stress’’ and glucose as an indicator of the systemic blood glucose levels (Sta˚hl et al., 2001a, b).

The normal glycerol concentration in the dialysate from subcutaneous tissue of a sedated patient when using a microdialysis catheter with a 30 mm dialysis membrane and a perfusion ﬂow of 0.3 mL/ min is 200 mM. II.A.3. Glutamate During ischemia there is an increased release of glutamate, which may open neuronal calcium channels initiating a pathological inﬂux of calcium provoking cell damage. In this way an increasing level of glutamate in the dialysate from the human brain is an indirect marker of cell damage. However, it is difﬁcult to interpret changes in brain glutamate due to the fact that glutamate release from neurons is mixed with a metabolic pool of glutamate. The normal glutamate concentration in the dialysate from the brain of a sedated patient when using a 10 mm dialysis membrane and a perfusion ﬂow of 0.3 mL/min is 10 mM and somewhat higher in a non-sedated patient (Reinstrup et al., 2000). II.A.4. Glucose As the primary source of energy to the brain, glucose is an important marker of changes in brain metabolism. Glucose levels in the dialysate from human brain may, however, change for several reasons:

Ischemia, that is, a decrease in capillary blood ﬂow. Less glucose is delivered to the microdialysis catheter and the concentration in the dialysate decreases. Hyperemia, that is, an increase in capillary blood ﬂow. More glucose is delivered to the microdialysis catheter and the concentration in the dialysate increases. Hyperglycemia, that is, increased blood glucose concentration due to metabolic stress, insulin resistance, or intravenous infusion of glucose. More glucose is delivered to the microdialysis catheter and the concentration in the dialysate increases. Hyper metabolism or hypo metabolism, that is, increased or decreased uptake of glucose into cells, for example, a shift from aerobic to

679

anaerobic metabolism. This will affect the amount of glucose in the tissue available to the microdialysis catheter causing a decreased or increased concentration in the dialysate. The fact that the glucose concentration in the dialysate may change for several reasons makes it difﬁcult to interpret glucose levels unless glucose is compared with other markers. A subcutaneous microdialysis catheter implanted in the peri-umbilical region of the abdomen is therefore of great value as it will show changes in the systemic levels of blood glucose. If a change in the brain is parallel to a change in subcutaneous glucose the change is in all probability due to a variation in systemic blood glucose. If the change in brain glucose is not parallel to a change in systemic blood glucose it is probably due to changes in brain capillary perfusion, for example, hyperemia or hyper- or hypo-metabolism causing a simultaneous increase or decrease in lactate as well as pyruvate. In contrast if a decrease in brain dialysate glucose coincides with an increase in lactate and a decrease in pyruvate it is in all probability due to ischemia. In this way it is possible to interpret the changes in glucose by observing changes in lactate and pyruvate. II.A.5. Other markers There are several papers in the literature describing changes in other biochemical markers than the above during neurointensive care. They include urea, NO-related compounds such as nitrate, nitrite, arginine and citrulline, pH, xanthenes, etc. As our understanding of the role of these compounds for evaluating tissue pathology in a clinical setting is still in its infancy they will not be discussed in this review. II.B. Implanting and positioning of microdialysis catheters Microdialysis monitors the local chemistry of an area of the brain roughly corresponding to the length of the catheter membrane and a diameter of a few millimeters. The interpretation of microdialysis

data therefore depends on the position of the catheter in relation to the existing pathology. The ﬁrst clinical microdialysis catheters appearing on the market were not visible on CT. Several clinical studies in the literature are therefore difﬁcult to interpret as we do not know if the microdialysis catheters ended up in normal tissue, penumbra tissue surrounding a contusion, or in damaged or dead tissue. Catheters available today are visible due to their gold tip. It is of great importance for the interpretation of bedside microdialysis data that the position of a catheter in the brain is determined from CT (Sta˚hl et al., 2003). Only then is it possible to use microdialysis data effectively to provide an early warning for secondary insults and to evaluate the result of various clinical interventions aimed at improving the condition of brain tissue during neurointensive care. It is important to adopt a consistent strategy of where to place catheters, for example, in the penumbra surrounding a mass lesion, in the region most likely to be affected by vasospasm after subarachnoid hemorrhage, and/or in ‘‘normal’’ brain tissue. In the ICU it is often convenient to place catheters through cranial bolts avoiding the need to bring the patient into the operating room. However, it is difﬁcult to position the catheter in a select region of the brain when using bolts as there is no provision for changing the depth or angle of the catheter. Therefore, catheters are often tunnelated and placed through burr holes in the ICU making it possible to better aim for a predeﬁned region of the brain. When the patient is subjected to a craniotomy it is easy to place the catheter under visual inspection into the peri-contusional penumbra of a lesion or in the territory of the parent vessel of an aneurysm. The catheter is tunnelated under the scalp and a small incision is made through the dura, subarachnoidea, and pia. The dialysis membrane is positioned in the penumbra, usually 1 cm from the border of the lesion or in the region most likely to be affected by vasospasm after hemorrhage. In our own experience we have seen no consistent difference in the chemistry if the catheter is placed in white or gray matter. However, it is to be expected that the levels of neurotransmitters may vary depending on the position.

680

The cuff, which slides over the inﬂow and outﬂow tubing, is sutured tightly to the skin. In this way the catheter will not be pulled out if, for example, the attached microdialysis pump drops out of the patient’s bed. Regardless of how the catheter is introduced into the brain it is important to locate the gold tip of the catheter on the ﬁrst CT performed after implantation. The location of the catheter will determine how relevant the biochemical data are for the interpretation of brain pathology and the early warning of a secondary insult.

II.C. Selecting perfusion flow The ﬂow of the standard microdialysis pump available on the market is 0.3 mL/min, which gives 70% recovery when using a 10 mm dialysis membrane (Hutchinson et al., 2000a, b, 2002). Catheters with 20 or 30 mm membrane will give next to 100% recovery in the brain. There may be situations when it is necessary to use a higher perfusion ﬂow, for example, when sampling is frequent and the 0.3 mL/min ﬂow does not give enough sample volume to permit analysis. This may be the case during intra-operative microdialysis when samples are changed every minute to, for example, monitor ischemia during temporary clipping. A ﬂow of 1 mL/min will give a recovery of 30% using a 10 mm membrane (Hutchinson et al., 2002). Another reason to use a high perfusion ﬂow is to reduce the time delay occurring when the dialysate ﬂows from the brain to the microvial. The delay is in the range of 20 min when the perfusion ﬂow is 0.3 mL/min but can be reduced to close to 1 min if the ﬂow is increased to 5 mL/min. In this way it is possible to react to changes in brain chemistry during surgical procedures, for example, ischemia occurring as a result of brain retraction (Xu et al., 2002). However, during neurointensive care chemical changes usually take place over several hours and a 20-min delay is of minimal consequence. During normal intensive care it is highly advisable to use the standard ﬂow and standard catheter as this will make it possible to compare data over time between patients in the same or different clinical departments.

II.D. Multimodal monitoring To make effective use of microdialysis data it is essential to relate them to other data collected bedside. This may be done by software, which allows for integrating data from the microdialysis analyzer, the ICU monitor displaying intra cranial pressure (ICP) and CPP, a tissue oxygen analyzer, the ventilator, the infusion pumps, etc. This ‘‘multimodal monitoring’’ allows for the display of all data as trend curves on one computer screen. It creates the framework for individualizing therapy on the basis of clinical status, brain tissue chemistry, and the effect of therapeutical interventions.

II.E. Interpreting microdialysis data During intensive care brain chemistry often changes profoundly in the patient. At our present state of knowledge it is impossible to interpret every change, however, major pathological states manifest themselves as dramatic increases or decreases of the chemical markers. The ﬁrst hours of microdialysis data give an indication of how severely brain tissue is affected in the peri-contusional penumbra. This information gives a reference value for determining if tissue physiology is improving or deteriorating. The implantation of a microdialysis catheter inﬂicts a certain amount of trauma to the brain tissue. This is well known from animal studies and it usually takes an hour or more before baseline values are reached after an implantation. In the human brain this is particularly evident for glutamate and sometimes for glycerol. However, in a clinical setting the time between implantation of the catheter and the actual use of microdialysis data is often longer than an hour due to all procedures taking place around the patient. The range from normal to pathological levels of different analytes are well known from normal brain tissue in patients with posterior fossa tumors (Reinstrup et al., 2000) and from damaged as well as ‘‘normal’’ brain tissue in traumatic brain injury (TBI) and subarachnoid hemorrhage (SAH) patients. Normal levels differ strongly from

681

pathological levels in, for example, severe brain trauma (see below). III. Clinical studies The following is an account of studies where microdialysis has been applied to neurosurgery and neurointensive care. The purpose is to present a selection of papers that describe biochemical ﬁndings of particular relevance for the early detection of secondary damage and the evaluation of therapeutic interventions during surgery and neurointensive care. Microdialysis of the human brain was ﬁrst performed in 1987 at the Karolinska institute in a Parkinson patient subject to thalamic lesion for alleviating tremor (Meyerson et al., 1990).The catheter was introduced stereotaxically and samples were collected every 10 min and analyzed for a large number of neurotransmitters and metabolites. We found that baseline levels of the various analytes in the dialysate were much higher than in animals due to the possibility of using a much larger dialysis membrane. Even more important, baseline levels were reached much faster probably due to the small implantation trauma in relation to the size of the human brain. We performed the ﬁrst study on brain ischemia in Uppsala monitoring the brain chemistry in tissue, which was resected during tumor surgery (Hillered et al., 1990). This led to a study of microdialysis during neurointensive care of TBI and SAH describing changes in especially lactate, pyruvate, and glutamate (Persson and Hillered, 1992). The experience from Karolinska and Uppsala prompted us, in cooperation with neurosurgery in Lund, to develop ﬂexible catheters more suitable for implantation in human brain and a microdialysis analyzer (CMA Microdialysis) designed for bedside use. III.A. Subarachnoid hemorrhage (SAH) Microdialysis has been used extensively for monitoring ischemia in SAH patients. Hamberger et al. (1995) and Runnerstam et al. (1997) found that the level of consciousness in the post-operative phase

was inversely related to the total amino acid concentration in the neuronal environment adjacent to a SAH. Persson et al. (1996) found that severe ischemia, for example, temporary clipping, severe hypoxemia, or infarct development within or close to the probe area resulted in a clearly detectable increase in lactate/pyruvate ratio. In many instances this was followed by an increase in glutamate when the lactate/pyruvate ratio reached values of 25 or above. Lactate/pyruvate ratio appeared to be a more reliable marker compared with lactate alone and there was a statistically signiﬁcant correlation between lactate/pyruvate ratio and clinical outcome during day 0–4, which did not exist for lactate. In a study combining microdialysis and PET Enblad et al. (1996) concluded that the energyrelated metabolites (lactate, lactate/pyruvate ratio, and hypoxanthine) may be used as extracellular markers of ischemia. Sa¨veland et al. (1996) found that increased levels of glutamate correlate well with clinical course and neurological symptoms. However, rise of glutamate in one region was not necessarily parallel to the rise in the other regions. Nilsson et al. (1999) described the detailed biochemistry of vasospasm and concluded that lactate and glutamate may be the most sensitive and early markers for incipient ischemia followed by the lactate/pyruvate ratio and glycerol during manifest ischemia and cell degeneration. They found that metabolic changes preceded the increase in blood ﬂow velocity as recorded by trans cranial doppler (TCD). In a series of studies Unterberg et al. (2001) placed microdialysis catheters 25–35 mm into the parenchyma of the vascular territory most likely to be affected by vasospasm. Sakowitz et al. (2001) concluded that microdialysis ‘‘can be carried out routinely in the ICU-setting to detect and monitor patterns of metabolic impairment. Compared with TCD it has a remarkable speciﬁcity making it a well-suited method to monitor delayed ischemic neurological deﬁcits following aneurysmal haemorrhage’’. Sarrafzadeh et al. (2002) found that lactate and glutamate are early markers of clinical vasospasm followed by lactate/pyruvate ratio and glycerol

682

during manifest vasospasm in patients with SAH. She states that bedside cerebral microdialysis is a safe technique for the indication of acute and delayed ischemic neurological deﬁcits in SAH patients when inserted into the region of interest and suggests that ‘‘early detection of metabolic changes might also allow optimization of standard intensive care treatments, such as triple-H therapy.’’ Skjøth-Rasmussen et al. (2004) found that the ischemic pattern after SAH preceded the occurrence of delayed ischemic neurological deﬁcits by a mean interval of 11 h. III.B. Traumatic brain injury Persson and Hillered (1992) made the ﬁrst microdialysis studies of the human brain after traumatic brain injury. They found that microdialysis can be used for long-term studies of energy-related metabolites and amino acids, for example, glutamate, and that the ﬂuctuation of these substances corresponded to various clinical events ‘‘presumably involving hypoxia/ischemia’’. They used the lactate/pyruvate ratio as a marker for energy disturbance in the brain. This ratio is known to reﬂect the redox potential of the tissue and thereby the severity of ischemia. They presented several arguments for the reliability of the lactate/pyruvate ratio in comparison to the use of the absolute concentration of other substances in the dialysate: (1) Owing to the structural similarity of lactate and pyruvate any change in the in vivo diffusion conditions during a pathological state could be expected to affect both metabolites similarly. (2) Being a ratio it is independent of probe characteristics. (3) On the basis of a review of 13 papers in the literature describing the normal brain lactate/pyruvate ratio in different species they concluded that the basal level of the lactate/ pyruvate ratio is below 20. This ﬁts with the basal lactate/pyruvate ratio of 23 that we found in normal brains of patients operated for posterior fossa tumors (Reinstrup et al., 2000).

Bullock, Zauner, and co-workers made the important observation that when placing the microdialysis catheter next to a cerebral contusion sustained cerebral blood ﬂow reductions, caused massive release of excitatory amino acids while in patients without secondary ischemic complications or focal contusions post traumatic glutamate release appears to be only transient (Zauner et al., 1996). They conclude that sustained high ICP and poor outcome were signiﬁcantly correlated to high levels of glutamate (>20 mM) (Bullock et al., 1998). In 1995 CE-labeled microdialysis catheters intended for human use and an instrument for bedside analysis of glucose, lactate, pyruvate, glycerol, and glutamate became available in the market (CMA Microdialysis). This enabled us to start routine monitoring of all patients with severe head injuries in Lund and a few years later in Stockholm. In our ﬁrst report on normal brain we established baseline values for the energy-related metabolites (see beow) (Reinstrup et al., 2000). In view of previous ﬁndings we routinely placed one catheter in peri-contusional, penumbra tissue and a second catheter in normal tissue, usually through a second burr hole in front of the intraventricular ICP catheter. We found that microdialysis could be performed on a routine basis by the regular staff in an ICU and that the data could be used for detecting global as well as local complications (Sta˚hl et al., 2001a, b). Our most important observations were:

The metabolites measured give information that is of direct clinical importance regarding substrate availability (glucose), redox state of the tissue (lactate/pyruvate ratio), degradation of glycerophospholipids in cell membranes (glycerol), and extracellular concentration of excitatory amino acids (glutamate). There was a great difference in the energy metabolism of the peri-contusional tissue as compared with normal tissue in the same patients. The biochemical consequences of severe anemic hypoxia were observed several hours before the deterioration was detected by conventional methods (ICP–CPP).

683

We were able to compare the mean levels of the various markers in the peri-contusional area of patient with fatal traumatic lesion (Sta˚hl et al., 2001a, b) with values obtained during wakefulness in normal human brain (Reinstrup et al., 2000): Glucose: 0.1 (fatal); 1.7 (normal), Lactate: 8.9 (fatal); 2.9 (normal), Pyruvate: 31 (fatal); 166 (normal), Lactate/pyruvate ratio: 458 (fatal); 23 (normal), Glycerol: 573 (fatal); 82 (normal), and Glutamate: 381 (fatal); 16 (normal). In a study of 27 patients, treated according to the Lund concept, we documented the brain chemistry in patients with favorable outcome (Sta˚hl et al., 2001a, b) in contrast to the previous study of fatal outcome. Our intention was that such data may serve as reference data for bedside prediction of outcome in the individual patient. The introduction of microdialysis catheters with a gold tip visible on CT marked a quantum leap in the use of microdialysis in routine monitoring during neurointensive care. It became possible to visualize the position of the catheters in relation to the contusion or hemorrhage and thereby determine the relevance of the microdialysis data. In our ﬁrst study where the catheter position was veriﬁed we received further proof of the great difference in sensitivity to secondary insults between normal brain and the tissue of the peri-contusional penumbra (Sta˚hl et al., 2003).

III.C. Intra-operative microdialysis Mendelowitsch, Landholt, and coworkers introduced microdialysis as a technique of pre-operative monitoring during neurosurgery (Bachli et al., 1996). In patients undergoing extra-intracranial bypass operations they placed a catheter 1–1.5 cm from the bypass site immediately after craniotomy. pH was measured on-line by a ﬂow through sensor, while glucose and lactate were determined colorimetrically and ﬂuorometrically, respectively. During the ﬁrst four bypass operations there were no marked changes in any of the measured parameters while in the ﬁfth patients the last temporary clipping caused a reduction of glucose to unmeasurable values, an increase in lactate and a slight reduction in pH during 20 min. In a

subsequent study of 10 patients they added the analysis of glutamate and found three patients with signiﬁcant intra-operative glutamate increase two of whom awoke with hemiparesis (Mendelowitsch et al., 1998). They also monitored the brain during aneurysm surgery inserting the catheter into the cortex frontally in the region where retractor compression was expected. On-line pH was reduced and lactate increased during retraction and glucose was decreased in most but not all patients (Bachli et al., 1996). They concluded that microdialysis is a sensitive method for detecting intra-operative changes in cerebral metabolism and emphasized that glutamate is an excellent marker of neuronal damage. The tissue damage caused by brain retraction was also evaluated by Xu et al. (2002) in a study of patients operated on sub-frontally for pituitary adenoma. The catheters were placed in the cerebral cortex beneath the brain retractor. It was found that glucose was within normal range while the lactate/pyruvate ratio, glutamate, and glycerol were considerably above the normal range. This was interpreted as an incomplete ischemia, which, however, seemed to cause tissue damage as indicated by the high glutamate and glycerol levels. Hutchinson et al. (2000a, b) demonstrated how combined microdialysis and tissue oxygen monitoring was able to assist in decision-making in a patient undergoing aneurysm surgery where the level of oxygen, glucose, and pyruvate decreased while lactate and glutamate increased during intraoperative ischemia as well as post-operative hydrocephalus.

III.D. Middle cerebral artery infarction Severe hemispheric stroke carry a high mortality because of the formation of fatal brain edema. Berger et al. (1999) reported a case of fatal middle cerebral artery (MCA) infarction where the neurochemical alterations contralateral to the infarction preceded clinical signs of herniation for several hours. Schneweis et al. (2001) used ICP and microdialysis in the ipsilateral frontal lobe in an attempt

684

to identify MCA patients at risk and decide on invasive therapies such as decompressive hemicranectomy or hypothermia. They found that chemical changes varied in accordance with clinical course, size of infarction, and brain edema. Stable ICP and chemistry was found in patients without progressive space-occupying infarcts while increase in ICP, glutamate, and lactate/pyruvate ratio was followed by massive edema and large infarcts. Berger et al. (2002) assessed the effect of therapeutic moderate hypothermia with microdialysis and were able to characterize three different brain regions with different reaction to hypothermia: (1) Non-infarcted tissue with stable chemistry and a moderate lowering of glutamate, lactate, and pyruvate during hypothermia. (2) Peri-infarct tissue where hypothermia caused a pronounced lowering of glutamate, glycerol, lactate, and pyruvate. (3) Irreversively damaged tissue with excessive increases of glutamate, glycerol, and lactate and lowering of pyruvate. They conclude that microdialysis is a safe and feasible method for neurochemical monitoring indicating normal brain tissue, salvageable tissue, and irreversibly damaged tissue and the effect of hypothermia on these different compartments. Thus, future treatment strategies for life-threatening stroke should be guided by close neurochemical monitoring. III.E. Drug distribution Microdialysis monitors the concentration of endogenous as well as exogenous substances in the interstitial ﬂuid of the brain. This offers a unique possibility to monitor the distribution and the free fraction of drugs in the brain at the same time as changes in endogenous substances reﬂect the effect of the drug on brain neurotransmission and metabolism. In a neurointensive care setting it is of obvious interest to determine the penetration of anti-infective agents over the blood-brain barrier into the brain interstitial ﬂuid where the drug becomes

available to brain cells. Mindermann et al. (1998) published the ﬁrst study on the penetration of antibiotics into the human brain. They determined the absolute concentration of free Rifampin in the interstitial ﬂuid by equilibrium dialysis (nonet-ﬂux method) in patients receiving a standard dose of Rifampin before resection of primary brain tumors. Rifampin concentration was highest within tumors followed by perifocal regions and normal brain tissue. Another important aspect of drug penetration into the brain during neurointensive care is the possibility that drugs may penetrate more effectively into areas where the blood-brain barrier is defective due to, for example, trauma and stroke. Bouw et al. (2001) found that there was an increase in morphine levels near a trauma site, that is, the peri-contusional penumbra, when compared with uninjured brain tissue. This is in all probably due to damage of the barrier in the penumbra region allowing drugs to penetrate more freely. It means that the vulnerable penumbra cells, which are most prone to suffer secondary damage, are exposed to unknown doses of drugs. It may on one hand offer an opportunity to target drug treatment to the vulnerable cells but at the same time we may unknowingly impair their situation making them more susceptible to secondary damage.

IV. Conclusions Today microdialysis is a well-established technique for monitoring brain tissue during neurointensive care. A large number of studies have clariﬁed the chemistry of ischemia during SAH, TBI, and stroke. These studies have identiﬁed markers that make it possible to assess the extent of tissue pathology, its change over time, and its relationship to other variables monitored bedside such as ICP and CPP. The discovery that the peri-contusional penumbra tissue has an increased sensitivity to secondary insults is well-documented (Sta˚hl et al., 2001a, b; Marion et al., 2002; Hlatky et al., 2004) and has given the clinician a unique opportunity to identify changes in chemistry that may develop into secondary damage before it has manifested itself in

685

clinical signs. Most important, it gives the clinician the possibility to determine the effect of a therapeutic intervention aimed at correcting brain chemistry and avoiding secondary damage provided that the catheters are located in the vulnerable pericontusional tissue (Engstro¨m et al., 2005). The implications of microdialysis data obtained in clinical studies have been discussed in a recent, Consensus paper on microdialysis procedures in patients with SAH and TBI (Bellander et al., 2004). References Bachli, H., Langemann, H., Mendelowitsch, A., Alessandri, B., Landolt, H. and Gratzl, O. (1996) Microdialytic monitoring during cerebrovascular surgery. Neurol. Res., 18: 370–376. Bellander, B.-M., Cantais, E., Enblad, P., Hutchinson, P., Nordstro¨m, C.-H., Robertson, C., Sahuquillo, J., Smith, M., Stocchetti, N., Ungerstedt, U., Unterberg, A. and Vidiendal Olsen, N. (2004) Consensus meeting on microdialysis in neurointensive care. Intensive Care Med., 30: 2166–2169. Berger, C., Annecke, A., Aschoff, A., Spranger, M. and Schwab, S. (1999) Neurochemical monitoring of fatal middle cerebral artery infarction. Stroke, 30: 460–463. Berger, C., Schabitz, W.R., Georgiadis, D., Steiner, T., Aschoff, A. and Schwab, S. (2002) Effects of hypothermia on excitatory amino acids and metabolism in stroke patients: a microdialysis study. Stroke, 33: 519–524. Bouw, R., Ederoth, P., Lundberg, J., Ungerstedt, U., Nordstro¨m, C.H. and Hammarlund-Udenaes, M. (2001) Increased blood-brain barrier permeability of morphine in a patient with severe brain lesions as determined by microdialysis. Acta Anaesthesiol. Scand., 45: 390–392. Bullock, R., Zauner, A., Woodward, J.J., Myseros, J., Choi, S.C., Ward, J.D., Marmarou, A. and Young, H.F. (1998) Factors affecting excitatory amino acid release following severe human head injury. J. Neurosurg., 89: 507–518. Delgado, J.M., DeFeudis, F.V., Roth, R.H., Ryugo, D.K. and Mitruka, B.M. (1972) Dialytrode for long term intracerebral perfusion in awake monkeys. Arch. Int. Pharmacodyn. Ther., 198: 9–21. Enblad, P., Valtysson, J., Andersson, J., Lilja, A., Valind, S., Antoni, G., La˚ngstro¨m, B., Hillered, L. and Persson, L. (1996) Simultaneous intracerebral microdialysis and positron emission tomography in the detection of ischemia in patients with subarachnoid hemorrhage. J. Cereb. Blood Flow Metab., 16: 637–644. Engstro¨m, M., Polito, A., Reinstrup, P., Romner, B., Ryding, E., Ungerstedt, U. and Nordstro¨m, C.-H. (2005) Intracerebral microdialysis in severe brain trauma: the importance of catheter location. J. Neurosurg., 10: 460–469. Hagstrom-Toft, E., Arner, P., Wahrenberg, H., Wennlund, A., Ungerstedt, U. and Bolinder, J. (1993) Adrenergic regulation

of human adipose tissue metabolism in situ during mental stress. J. Clin. Endocrinol. Metab., 76: 392–398. Hamberger, A., Runnerstam, M., Nystro¨m, B., Starmark, J.E. and von Essen, C. (1995) The neuronal environment after subarachnoid haemorrhage – correlation of amino acid and nucleoside levels with post-operative recovery. Neurol. Res., 17: 97–105. Hillered, L., Persson, L., Ponten, U. and Ungerstedt, U. (1990) Neurometabolic monitoring of the ischaemic human brain using microdialysis. Acta Neurochir. (Wien), 102: 91–97. Hillered, L., Valtysson, J., Enblad, P. and Persson, L. (1998) Interstitial glycerol as a marker for membrane phospholipid degradation in the acutely injured human brain. J. Neurol. Neurosurg. Psychiatry, 64: 486–491. Hillman, J., A˚neman, O., Anderson, Ch., Sjo¨gren, F., Sa¨berg, C. and Mellerga˚rd, P. (2005) A microdialysis technique for routine measurement of macromolecules in the injured human brain. Neurosurgery, 56: 1264–1270. Hlatky, B., Valadka, J., Goodman, C. and Robertson, C.S. (2004) Evolution of brain tissue injury after evacuation of acute traumatic subdural hematomas. Neurosurgery, 55: 1318–1324. Hutchinson, P.J., Al-Rawi, P.G., O’Connell, M.T., Gupta, A.K., Pickard, J.D. and Kirkpatrick, P.J. (2000a) Biochemical changes related to hypoxia during cerebral aneurysm surgery: combined microdialysis and tissue oxygen monitoring: case report. Neurosurgery, 46: 201–205. Hutchinson, P.J., O’Connell, M.T., Al-Rawi, P.G., Maskell, L.B., Kett-White, R., Gupta, A.K., Richards, H.K., Hutchinson, D.B., Kirkpatrick, P.J. and Pickard, J.D. (2000b) Clinical cerebral microdialysis: a methodological study. J. Neurosurg., 93: 37–43. Hutchinson, P.J., O’Connell, M.T., al-Rawi, P.G., Kett-White, R., Gupta, A.K., Kirkpatrick, P.J. and Pickard, J.D. (2002) Clinical cerebral microdialysis – determining the true extracellular concentration. Acta Neurochir. Suppl., 81: 359–362. Marion, D., Puccio, A., Wisniewski, S.R., Kochanek, P., Dixon, C.E., Bullian, L. and Carlier, P. (2002) Effect of hyperventilation on extracellular concentrations of glutamate, lactate, pyruvate, and local cerebral blood ﬂow in patients with severe traumatic brain injury. Crit. Care Med., 30: 2619–2625. Mendelowitsch, A., Sekhar, L.N., Wright, D.C., Nadel, A., Miyashita, H., Richardson, R., Kent, M. and Shuaib, A. (1998) An increase in extracellular glutamate is a sensitive method of detecting ischaemic neuronal damage during cranial base and cerebrovascular surgery. An in vivo microdialysis study. Acta Neurochir. (Wien), 140: 349–355. Meyerson, B., Linderoth, B., Karlsson, H. and Ungerstedt, U. (1990) Microdialysis in the human brain: extracellular measurements in the thalamus of parkinsonian patients. Life Sci., 46: 301–308. Mindermann, T., Zimmerli, W. and Gratzl, O. (1998) Rifampin concentrations in various compartments of the human brain: a novel method for determining drug levels in the cerebral extracellular space. Antimicrob. Agents Chemother., 42: 2626–2629.

686 Nilsson, O.G., Brandt, L., Ungerstedt, U. and Saveland, H. (1999) Bedside detection of brain ischemia using intracerebral microdialysis: subarachnoid hemorrhage and delayed ischemic deterioration. Neurosurgery, 45: 1176–1184. Persson, L. and Hillered, L. (1992) Chemical monitoring of neurosurgical intensive care patients using intracerebral microdialysis. J. Neurosurg., 76: 72–80. Persson, L., Valtysson, J., Enblad, P., Warme, P.E., Cesarini, K., Lewen, A. and Hillered, L. (1996) Neurochemical monitoring using intracerebral microdialysis in patients with subarachnoid hemorrhage. J. Neurosurg., 84: 606–616. Reinstrup, P., Sta˚h, l.N., Mellergard, P., Uski, T., Ungerstedt, U. and Nordstro¨m, C.H. (2000) Intracerebral microdialysis in clinical practice: baseline values for chemical markers during wakefulness, anesthesia, and neurosurgery. Neurosurgery, 47: 701–709. Runnerstam, M., von Essen, C., Nystro¨m, B., Rosengren, L. and Hamberger, A. (1997) Extracellular glial ﬁbrillary acidic protein and amino acids in brain regions of patients with subarachnoid hemorrhage – correlation with level of consciousness and site of bleeding. Neurol. Res., 19: 361–368. Sakowitz, O.W., Sarrafzadeh, A.S., Benndorf, G., Lanksch, W.R. and Unterberg, A.W. (2001) On-line microdialysis following aneurismal subarachnoid hemorrhage. Acta Neurochir. Suppl., 77: 141–144. Sarrafzadeh, A.S., Sakowitz, O.W., Kiening, K.L., Benndorf, G., Lanksch, W.R. and Unterberg, A.W. (2002) Bedside microdialysis: a tool to monitor cerebral metabolism in subarachnoid hemorrhage patients? Crit. Care Med., 30: 1062–1070. Sa¨veland, H., Nilsson, O.G., Boris-Mo¨ller, F., Wieloch, T. and Brandt, L. (1996) Intracerebral microdialysis of glutamate and aspartate in two vascular territories after aneurysmal subarachnoid hemorrhage. Neurosurgery, 38: 12–19. Schneweis, S., Grond, M., Staub, F., Brinker, G., Neveling, M., Dohmen, C., Graf, R., Heiss, W.-D. and Shuaib, A. (2001) Predictive value of neurochemical monitoring in large middle cerebral artery infarction. Stroke, 32: 1863–1867. Siesjo¨, B.K. (1978) Brain Energy Metabolism. John Wiley and Sons, New York.

Skjøth-Rasmussen, J., Schulz, M., Risom Kristensen, S. and Bjerre, P. (2004) Delayed neurological deﬁcits detected by an ischemic pattern in the extracellular cerebral metabolites in patients with aneurysmal subarachnoid hemorrhage. J. Neurosurg., 100: 8–15. Sta˚hl, N., Ungerstedt, U. and Nordstro¨m, C.H. (2001b) Brain energy metabolism during controlled reduction of cerebral perfusion pressure in severe head injuries. Intensive Care Med., 27: 1215–1223. Sta˚hl, N., Mellergard, P., Hallstro¨m, A., Ungerstedt, U. and Nordstro¨m, C.H. (2001a) Intracerebral microdialysis and bedside biochemical analysis in patients with fatal traumatic brain lesions. Acta Anaesthesiol. Scand., 45: 977–985. Sta˚hl, N., Schale´n, W., Ungerstedt, U. and Nordstro¨m, C.H. (2003) Bedside biochemical monitoring of the penumbra zone surrounding an evacuated acute subdural haematoma. Acta Neurol. Scand., 108: 211–215. Ungerstedt, U. (1991) Microdialysis – principles and applications for studies in animals and man. J. Intern. Med., 230: 365–373. Ungerstedt, U. and Pycock, C. (1974) Functional correlates of dopamine neurotransmission. Bull. Schweiz Akad. Med. Wiss, 30: 44–55. Unterberg, A.W., Sakowitz, O.W., Sarrafzadeh, A.S., Benndorf, G. and Lanksch, W.R. (2001) Role of bedside microdialysis in the diagnosis of cerebral vasospasm following aneurysmal subarachnoid hemorrhage. J. Neurosurg., 94: 740–749. Vespa, P., Bergsneider, M., Hattori, N., Wu, H.-M., Huang, S.-C., Martin, N., Glenn, T., McArthur, D. and Hovda, D. (2005) Metabolic crisis without brain ischemia is common after traumatic brain injury: a combined microdialysis and positron emission tomography study. J. Cereb. Blood Flow Metab., 25: 763–774. Xu, W., Mellerga˚rd, P., Ungerstedt, U. and Nordstro¨m, C.H. (2002) Local changes in cerebral energy metabolism due to brain retraction during routine neurosurgical procedures. Acta Neurochir. (Wien), 144: 679–683. Zauner, A., Bullock, R., Kuta, A.J., Woodward, J. and Young, H.F. (1996) Glutamate release and cerebral blood ﬂow after severe human head injury. Acta Neurochir. Suppl. (Wien), 67: 40–44.

Subject Index addiction 419, 428, 490 adenosine 9, 21, 409–410, 437, 440–441, 447 adenosine A1 receptor 447 adenosine A2A receptor 444, 447 ADHD 188, 192, 428 adipocytes 646–647 adipose tissue 645–647, 652, 654, 656 adipsia 354, 356, 363 afﬁnity chromatography 254–255 AIT-082 560–561 akinesia 405 albumin 150, 576 alfentanil 613–614 alovudine 553, 556 ALS 33, 410–411 alpha-chloralose 73 Alzheimer’s disease 205, 207, 435, 445–446 amino acids 115, 123 aminopyrine 145 amisulpride 487 amitryptyline 528 amnesia 73, 445 amoxicillin 591, 595, 616 AMPA 37, 187, 447–448 amperometric 400, 402 amperometric detection 235, 238, 240 amphetamine 9–10, 148, 186, 244, 259, 404–407, 421–423, 552–553 amprenavir 579 amygdala 302, 304, 308–309, 323, 326, 329, 332–333, 338, 341, 353, 424, 427, 429–430, 437, 445, 533 amyloid precursor protein 410–411 amyotrophic lateral sclerosis 207 anaesthesia 301 analgesics 473–474 anesthesia 171, 602, 608–611, 615, 618–619 anesthetics 401 anesthetized 71–73, 80–85 aneurysm 679, 683 angiotensin II 362 anhedonic 357 anterior cingulate cortex 473, 476 anti-epileptic drugs 436 antibiotics 521, 523, 625, 629–632, 634, 638, 684

a-CHCA 258 a2-adrenoceptors 528, 534 b-amyloid protein 205 1,2-diphenylethylenediamine 238–239 2,3-DHBA 575 2,5-DHBA 575 2,5-dihydroxybenzoic acid 258 3-O-methyldopa 235 3-nitropropionic acid 205, 447 4-aminopyridine 23, 84, 205 5-ﬂuorouracil 615, 636 5-HT7 10–11, 405, 409–410, 475–477 5-HT transporter 529–530, 532–534 5-HT1A-KO 409 5-HT1B 409 5-HT1B-KO 408–409 5-HT2C knockout 409 5-HT1A autoreceptors 488–489, 491 5-HT1A receptor agonist 440, 444, 447 5-HT1A receptors 530–531, 533, 535 5-HT1B autoreceptors 531 5-HT2C receptors 533 5-hydroxyindoleacetic acid 301, 303, 308, 312 6-hydroxydopamine 161, 186, 207 6-mercaptourine 561 6-OHDA 284 6-OHDA-lesioned 113 18F-FDG 663 abdominal fat 652 abuse 490 acetaminophen 150, 270, 550, 552 acetonitrile 254 acetyl-L-carnitine 560 acetylcholine 5, 7, 11, 112, 118, 183, 187, 189–190, 233–234, 241–242, 255, 259, 268–269, 352, 355, 357–359, 361, 401, 407, 422, 444–447 acetylcholinesterase 184, 485, 494–496 acetylcholinesterase inhibitors 184 ACh 401, 407–408, 410 acrylonitrile 576 acutely implanted 59, 67 acyclovir 612

687

688 antibody beads 283 anticancer drugs 625, 629, 635–638 antidepressant drugs 527–529, 532, 536–537 antiepileptic drugs 554, 556–557, 559 antinociception 76, 80 antioxidant 577 antipsychotics 408 antipyrine 145 antisense oligonucleotides 206–207 anxiety 304, 306, 427–428 aphagia 354, 356, 363 apolipoprotein 411 apomorphine 81 apoptosis 441–442 appetitive conditioning 326, 328–330 appetitive stimulus 42 arachidonic acid 563 area postrema 76 arecoline 357 arginine 441, 488 aripiprazole 492–493 arousal 72, 76, 78–79, 317–319, 321–323, 328, 331, 341–342, 377–378, 386, 389, 391 artefacts 386 artery occlusion 41, 209 aspartate 9, 21, 24, 78, 80, 84, 239–240, 268, 270–273, 275, 421, 427–428, 474–475, 477 astrocytes 20–23, 35–42, 456, 458, 460–461, 467, 546, 550 astrocytic 25, 384 astrocytic release 22 astrocytic swelling 21 atenolol 150, 550–552 atomoxetine 188 ATP 441, 461–462 atrium 577 attention 72, 377, 379, 386, 389–393, 440, 442, 444–447 auricle vein 578 autoreceptors 402, 404, 406, 409–410 aversion 351–353, 355–357 aversive stimulation 391 baclofen 25, 27 baicalein 579, 583 barbiturates 364 basal concentration 47 basal extracellular levels 401, 403, 408 basal ganglia 183–184, 187, 189, 441, 443–445 BBB damage 551 BDNF 444 behavior stimulations 24 behaviour 301–304, 306–310, 312

benzazapine 516 benzodiazepines 364, 489 benztropine 595 benzylamine 238–239 berberine 561, 582, 613–614 Bessel function 132, 139, 162 beta-endorphin 289 beta-hydroxybutyrate 563 bicuculline- 437 bile 573, 575, 581–583 bile sampling 177–178 biliary excretion 614–615 biocompatibility 118, 120 biosensors 111, 114–115, 121 bipolar depression 527 blood 573, 575–584 blood sampler 175 blood–CSF barrier 519 blood-brain barrier 116, 118, 149, 155–156, 201, 589, 591, 597 bone 219–220, 224, 602, 612, 629 boron 616 bradikinesia 357 bradykinin 244 brain cooling 441 brain death 659, 662–663, 667, 669, 671 brain injury 639 brain metabolism 73 brain trauma 596 bupivacaine 596, 608, 611 bursting mode 324–325 buspirone 306, 489 Ca2+ dependency 292 Ca2+ signals 21 Ca2+-free Ringer 57–58 Ca2+-sensitivity 401 caffeic acid 577 caffeine 145–146, 156, 578, 591 calcium 218–219, 224, 384 calcium-free Ringer 19, 25, 27 calibration 93–97, 99, 103, 105, 131, 144–148, 151, 158–160, 163 camptothecin 561, 579, 581–582, 613 capecitabine 636 capillary electrophoresis 18, 233, 254–255, 258 capillary LC 235–236, 244 capsaicin 475, 477 carbachol 83, 357–358, 361–362 carbamazepine 219, 224, 557, 594 carbon-ﬁber microelectrode 53, 60, 64, 66 carboplatin 615, 636

689 carboxyamidation 289, 291 carboxypeptidase 287–288, 292 catechol-O-methyltransferase 184 catheters 676, 679–683, 685 cats 474 CCK 474, 477 cefadroxil 616 cefepime 579, 613 ceﬁxime 630 cefopodoxime 630 cefoselis 564 cefpirome 596, 616–617, 630–631 cefpodoxime 580, 609–610 ceftriaxone 244 cellulose 223 cephalexin 610 cephalosporine 630 cephalothin 610 cephazolin 610 cerebral artery occlusion 440 cetirizine 244, 591–592 cGMP 485, 488 chemoattractants 120 chemotherapy 635–636 chinchillas 591, 602, 610 chloral hydrate 73–75, 77, 79–83, 401 chloramphenicol 583 chloride 218–219, 224 chlorpromazine 495 cholecystokinin 362, 428, 474 cholinergic 445–447 cholinergic neurons 378, 384, 386–387, 390–393 cholinergic transmission 494 cholinesterase inhibitors 384 chromogranin 288–289 chronic pain 473–474, 478 chronoamperometry 57 chymotrypsin 260 ciproﬂoxacin 560, 594–596, 606, 618, 631, 639 circadian rhythms 112, 322 circadianic variability 551 cisplatin 615, 636 cisterna magna 606 citalopram 410, 437, 487, 528, 530, 534 citrulline 488 citrulline/arginine 235 clindamycin 448 clinic 455, 466–467 clomipramine 528, 530–531, 533 clonidine 83 clorgyline 407 clozapine 423, 486–488, 492–495, 498–499

cocaine 42, 81–82, 148, 161, 253, 259, 404–407, 428, 552–553, 593, 595 cocaine addiction 42–43, 522 cocaine-seeking 42 codeine 554, 591 cognitive 377, 379, 384, 386–387, 393 cognitive decline 445–446 cognitive processes 377, 380, 385, 392 COMT 406–407 conditioning 383, 391 conscious 71, 82 consciousness 393 convection 134, 137, 143, 149 cortical cup technique 6, 378 corticosterone 301–302, 309–312, 486, 500 cortisol 536 cotinine 607–608 CPP 659–660, 662–663, 666–668, 670–671, 678, 680, 684 creatine 609 CRF 306–307, 310–312, 489–490, 500, 527, 536 CRF1 receptor antagonists 536 CRH 441 CSF 112, 115 cue conditions 392 cuprophan 150–151, 223 cutaneous microdialysis 596 cyclic voltammetry 47, 53, 60, 141, 156, 161 cyclodextrin 124 cyclosporin A 561 cyclosporine 520, 579, 582–583, 592, 607, 613 cystine 22, 262–263 cystine–glutamate antiporter 39–41 cystine–glutamate exchange 21, 38–40, 42 cytokines 283, 309, 676 CZE 269–271 D2-KO 406 D2-ligand binding 324, 341 D3-KO 406 D2 receptors 64–65 dacarbazine 636–637 damaged tissue 50, 66 DAT 400, 402–405, 407 DAT-KO 400, 402–405, 409–410 DC potential 204–205, 209–210 delta-9-tetrahydrocannabinol 448 dentate gyrus 385, 459 deprenyl 407 depression 183, 302, 304, 306, 310, 419, 421, 424, 429–430, 437 dermal penetration 634 dermis 579

690 desipramine 161, 534 dexamethasone 611 dextrocetirizine 562–563 diabetes 563 diabetic patients 579 dialysis electrode 121 dialytrode 8 diazepam 186, 427–428, 591 diclofenac 634 diffusion coefﬁcient 50 dihydrexidine 186 discrete sample processing 179 dogs 602, 611 donepezil 447 L-DOPA 48, 235, 445, 606 DOPAC 10, 57, 113, 235, 237, 595, 597 dopamine 7–12, 19, 22, 24–25, 27, 35, 37, 41, 48, 55, 65, 72, 78, 81, 93, 95, 98–106, 112, 141, 146, 160, 162, 183–184, 187, 189–190, 235, 268–270, 272–273, 351–352, 354, 356, 359, 363, 365, 421–430, 437, 440–446, 475–477, 515–518, 522, 590, 597 dopamine systems 534 dopamine transporter 95, 400 dopamine-beta-hydroxylase 410 drug addiction 33 drug-seeking 429 dual amperometric cells 237 dual-probe microdialysis 112–113 dual-quadrupole-TOF 257 duloxetine 528, 535 dynamic-no-net-ﬂux 148, 552 dynorphin A 284 dynorphins 243, 253–254 dysthymia 527 ear 602, 610, 616 ecstasy 259 edema 48, 50, 148, 152–153, 466, 683–684 EEG 204–205, 209–210, 304–305, 312 efﬂux 328, 335 efﬂux transporters 548 electric stimulation 60, 62 electrical stimulation 23, 100–101, 104, 475 electro-osmotic pumping 222 electrochemical detection 235, 241 electrophysiology 203 electroshock 436, 439 electrospray ionization 235, 251, 256–257 eletriptan 519–520 ELISA 124, 251, 280 encephalomyelitis 448 endocytosis 547

endopeptidase 280, 287–288 endothelin-1 203, 205, 210, 440–442, 464 endotoxin 309–311 enfurane 81 enkephalins 243, 254, 259, 261, 355 enterohepatic circulation 573, 581, 583–584 enterostatin 361 epidural 608–609, 613 epilepsies 204, 341, 435–439, 523, 639 epinephrine 613–614 epirubicine 636 epitopes 251 ESI 251, 256–258, 260, 262 estradiol 628, 634 ethanol 133, 138, 149, 153, 155–156, 186, 270, 605–606, 634 exercise 458 exocytosis 35, 39, 402 exploration 303 extended amygdala 354 extracellular glutamate 65–66 extracellular space 49 extraction equation 645, 647–648, 650–651, 654–655 extraction fraction 590 extrasynaptic 50, 65 extrasynaptic glutamate receptors 37–38 extrasynaptic transmission 56 eye 602, 610 FDA 179, 513, 518, 521–524, 626–627, 630, 632 FDG 577–578 fear conditioning 325–328 feeding 351–352, 354–357, 359–364, 379, 382, 387, 392 felbamate 557 femoral vein 576–577, 583 fentanyl 613–614 ﬁbrosis 48 Fick’s ﬁrst law 102, 380, 574 ﬂeroﬂoxacin 560 ﬂesinoxan 487 ﬂuconazole 244, 579, 582, 612, 614 ﬂuorescamine 271 ﬂuorescein 271, 553, 561, 606 ﬂuorescence detection 234, 239–240, 244 ﬂuoroquinolones 560 ﬂuoxetine 306, 409, 428, 486, 489, 528–529, 531–534 ﬂurbiprofen 576, 579, 611 ﬂuvoxamine 528, 531 fMRI 114 food intake 302 food rewards 323, 333 foot shocks 323–328, 338

691 forced swimming 306–307, 311 formalin 474–475, 477 fosfomycin 595, 631–632 Freely moving 71–72, 75, 77, 80–85 frontal cortex 76–83, 184, 187, 189–190, 380, 382, 386–388, 391, 486, 492, 528–532, 535–536, 605–606, 619 functional MRI 208 G-coupled receptors 39 GABA 7, 9, 11, 17–19, 21–22, 25–28, 36–38, 75–78, 80, 187, 239–241, 270–272, 275, 292, 357–359, 361, 401, 421, 423–426, 429, 437–440, 443–444, 446–448, 474–477, 531, 594, 597 GABA receptors 391 GABA transporter 439 GABAB receptor 426 gabapentin 556, 591 gacyclidine 613 galanin 81, 291, 360, 446 galanin deﬁcient 410 galanin-overexpressing 410 gallamine 609 gamma-butyrolactone 404 ganciclovir 610 GAP-junction hemichannels 21–22 gastric dosing 173 GBR-12909 437 GDNF 408, 444–445 gemiﬂoxacin 596, 630 gene expression 399, 411 generalised anxiety 302 genistein 582 gentamicin 610, 612, 630 gerbils 339, 579 GF120918 579, 594–595 gliosis 48, 117 gliotransmission 20–21, 25 globus pallidus 261, 386 glucocorticoids 548 glucoprivation 361 glucose 72, 78, 117, 119–121, 125, 141, 145–147, 151, 217–219, 221, 224–227, 263, 384, 392, 455–467, 550–551, 563, 577–580, 627–629, 637, 659–666, 668–671, 676–679, 682–683 glucose sensor 456, 459–461, 463, 523, 579, 645 D-glutamate 23 glutamate 9, 11, 17–25, 27–28, 33–43, 59, 65–66, 75–76, 78, 80, 83–84, 118, 121, 226, 233, 240–241, 268–275, 357–358, 361, 401, 405, 408, 421–429, 436–444, 446–448, 455–456, 458–460, 462, 464–467, 474–477,

485–486, 489–490, 497–500, 531, 594, 597, 661–664, 666–667, 669–671, 678, 680–684 glutamate microsensors 24, 66 glutamate oxidase 19, 34 glutamate receptors 391 glutamate release 460–461, 465 glutamate signaling 33–35, 37–43 glutamate transporters 20–21 glutamatergic 438, 446–447 glutamine synthetase 36 glutathione-deﬁcient mice 409 glycerol 145, 151, 226, 500, 563, 662–663, 667, 669–671, 675, 677–678, 680–684 glycine 80, 84, 239, 485–486, 497–500 glycolysis 455–456, 461–462, 467 glycoprotein 576 GRK6-KO 406 grooming 303, 308, 458 guide cannulas 301–302 habituation 377, 381, 387–389 haemorrhage 500 haloperidol 82, 148, 487–488, 492, 494, 496 halothane 73–75, 77–84 hamsters 304 heart 629 hemichannels 39–40, 42 hemorrhage 48, 677, 679, 683 heparin 223 hepatobiliary secretion 613 hesperidin 614 high potassium 23 hippocampal 423–424, 426–427, 430 hippocampus 22, 24, 26–27, 76–78, 81, 83, 145, 183–188, 190–191, 254, 323, 337, 353, 377, 380–389, 391–393, 436–440, 445–447, 528, 531–532, 534–536, 594–595, 605–607 histamine 40, 78, 80–81, 83, 360, 401, 405, 410, 448, 477, 486, 489, 500, 614 horses 217, 223 HPA axis 302, 306, 309–312 HPLC 18, 23, 26 human 601, 604, 608, 611, 614–616 Huntington 409–410 Huntington’s disease 205, 435, 447 HVA 10, 113, 235, 237 hydrocephalus 670 hydrogen clearance 208 hydrophilic interaction chromatography 234 hydroxyl radical 410, 440–441, 448 hyperactive 192 hyperdopaminergic 192

692 hyperdopaminergica 402, 404 hyperemia 677–679 hyperglycemia 563 hyperlocomotion 188 hyperoxia 666, 671 hyperphagia 360 hyperthermia 659, 665–667, 671 hypnosis 73, 76 hypoglutamatergic 408 hypothalamus 302, 304, 308, 310, 323, 332, 381, 386, 595 hypothermia 662, 666–667, 671 hypoxia 436, 455, 461–463, 465–467, 523 ibuprofen 634 idazoxan 360 iloperidone 495–496 imipenem 580 imipramine 486, 528 immobility 73, 76, 386 immobilization 458, 460 immune stress 309–311 immunosuppression 120 implantation effects 651, 653 implantation trauma 132, 141, 147, 153, 156, 163 infarct 439–440, 442, 445 infarction 683–684 inﬂammation 219, 223, 645, 651–653 inﬂammatory 48, 60, 118, 124, 473–474 inﬂammatory challenge 309 injured brain 578 insulin 360–361, 659, 668–669, 671 interleukin-2 309 interpeduncular nucleus 78 interstitial space ﬂuid 626 intracranial pressure 467 intracranial self-stimulation 351–352, 363 intrathecal 474–476, 602, 608–609, 614 ion mobility spectrometry 262 iontophoresis 612 ischemia 33, 42, 204–206, 208–210, 225–226, 455, 464, 467, 523, 579, 629, 659–662, 664, 666, 669–671, 675, 677–684 isoﬂurane 73–75, 77, 79, 81 isoprenaline 461 isopropanol 634 jugular vein 261, 559, 561–562, 577, 579 K+-stimulated 261 kainate 37 kappa-opioid 161–162

ketamine 73–74, 76–80, 273 kindling 437 knockout 400–402, 407–411 kynurenate 59 kynurenic acid 363, 439 D-lactate 455, 466 lactate 36–37, 121, 125, 145, 151, 218–219, 224–227, 455–456, 458–467, 563, 579–580, 584, 628, 660–664, 666–667, 669–671 lactate/pyruvate 661 lactate/pyruvate ratio 466, 628, 675, 677–678, 681–684 lamotrigine 554, 557 laser Doppler ﬂowmetry 206, 208 lateral hypothalamus 353 laterodorsal tegmentum 184, 189 LC/MS 251–253, 258, 260–261, 263 LC/MS/MS 260 learning 76, 80, 317, 320, 323, 325–326, 330–331, 333–335, 337–342, 391–392 lecozotan 447 leptin 361 Leu-enkephalin 261, 280–283, 287, 289 lever pressing 332–334, 336, 339–340 levetiracetam 439, 554–555, 557, 605 levocetirizine 562–563 levodopa 578 levoﬂoxacin 595, 617, 630–631 lidocaine 205, 608, 611, 618, 634 lipoproteins 576 liquid swivels 171 lithium 272, 405 liver 573, 575–576, 578, 580–581, 583–584, 627, 629 locomotion 76, 83 locus coeruleus 23, 317, 323, 427–428, 437, 530, 534, 536 low ﬂow rates 603 low perfusion rate 401, 404–407, 410 lung 602, 609–610, 616–617, 627–629, 631 luteinizing hormone releasing hormone 283 LY335979 553

M3G 558 M6G 558–559, 592–595, 597 magnesium 219, 224, 579 magnetic resonance spectroscopy 626 major depression 527–529, 533 MALDI 256, 258 malonate 206 maprotiline 534 m-chloro-phenylpiperazine 306 MDMA 207 medial forebrain bundle 100

693 medial preoptic area 363 medial septum 184–185 melanin-concentrating hormone 289, 490 melanoma 636–637 melperone 496 melphalan 636–637 memantine 446 membranes 116, 120, 122–124, 380, 384 memory 76, 80, 356, 377, 382, 385, 387–389, 391–393, 445–447, 494 meningitis 448, 549, 564, 617 mesolimbic 421–424, 426–430 mesolimbic system 518 met-enkephalin 243, 261, 280–281, 287, 289, 476–477 metabolism cage 172 metabotropic glutamate receptors 408, 448, 460 methamphetamine 81, 428, 536 methotrexate 576, 611, 636 methoxypsoralen 634 methyl-nicotinate 634 methylphenidate 404, 595 metoprolol 595 metrifonate 446–447 metronidazole 580, 582, 595, 631 m-glutamate receptor 23 mianserin 528 microbore LC 236, 238–241, 244–245 microelectromechanical 111, 115 microﬁltration 217–218, 220, 225, 227 microﬂuidic 111, 115, 121, 123 microglia 117, 120 microinjection cannula 203 microneedles 122 microPET 117–119 microwave irradiation 288 milnacipran 528, 535 mirtazapine 488–489, 528 MK-801 405, 408 monkeys 336–337, 339, 385 monoamine oxidase inhibitors 528 monoaminoxidase 184 monolithic columns 235 morphine 244–245, 355, 364, 474, 477, 516, 553, 558–560, 563–564, 578–579, 591–596, 607, 612–615, 617 morphine induced 407 morphine-3-glucuronide 578, 591, 612 morphine-6-glucuronide 591, 597 morphine-induced 407 motor activity 379, 381, 387–389 mouse 385, 399–400, 402, 408, 410–411 moxiﬂoxacin 617

MPP+ 205, 561 MPTP 405, 441, 444 MRI 201, 208–210 MS/MS 233, 242–246, 251–253, 256, 258–261, 263 multiple sclerosis 448 muscarinic 40, 385, 391, 393, 407–408, 410 muscarinic receptors 184–185, 189–192 muscle 220, 223–224, 226, 602, 609–610, 612, 616–617 mutant mouse 301, 309 myocardial 225 N-acetylaspartate 41, 209 nalorphine 411, 554, 560 naloxone 364–365 nano LC 236, 243 naproxen 611 naringenin 561 naringin 561 nefazodone 488, 528 neocortex 183–184, 187 neostigmine 161, 241, 357, 383–385 neuro-inﬂammation 549 neurogranin 261, 285, 287 neurokinin 487, 490 neurokininA 187 neuropeptide Y 500 neuropeptides 112, 123–125, 243, 251–253, 255–256, 258–263, 279–280, 282–291, 435, 444 neuropeptidomics 252, 263 neurotensin 243, 259, 281, 289, 476–477 neurotoxins 9 neurotransmitter overﬂow 42 nicotine 186, 189–192, 408, 515–518, 607–608, 634–635 nicotinic 75–76 nicotinic receptors 183–184, 190–191 nitric oxide 363, 410, 595, 597 nitrous oxide 75–76, 78 NK1 receptor antagonist 535–536 NMDA 37–38, 40–41, 75, 81–84, 187, 190, 268, 272, 405, 408, 410, 421–423, 426 NMDA antagonists 390, 408, 446 NMDA receptors 497–498 NO 437, 439–443, 488–491, 499–501 no net ﬂux 401–402, 404, 406–407, 409, 411 no-net-ﬂux technique 186 nociception 76, 253, 473, 477 nomifensine 53, 57–59, 81, 186 non-vesicular 65 nor-binaltorphimine 161–162 noradrenaline 7, 11, 75, 78–79, 82, 235, 303, 309, 421–422, 427–428, 430, 437, 527–529, 534–535 norepinephrine 37, 161, 360, 590

694 norepinephrine transporter 404 norﬂoxacin 436, 560, 562, 580, 591–593, 606, 609 nortryptiline 534 Novel environment 387–389, 391 novelty 72, 81, 377, 381, 386–390, 445 NPPB 41 NPY 291, 360, 362 NSAIDs 634 NSD-1015 404 nucleus accumbens 23–24, 35, 39–42, 56, 61, 77–80, 82, 148, 162, 184, 187, 189–190, 317, 321, 351–352, 357, 377, 381–382, 386–387, 391, 400, 422–424, 426–430, 437, 473, 475–476, 517–518, 522, 605 nucleus basalis 383, 385 nucleus tractus solitarius 76 obese 653, 655 obesity 359–360, 364, 631 olanzapine 487, 494–495, 578 oleic acid 124 oligodendrocytes 448 oligonucleotides 608 omeprazole 244 on-line analysis 666 on-line coupling 255–256, 261 operant behaviour 317, 319, 322, 331–333, 336, 338–339 opioid peptide 437 opioid receptors 407, 477 opioids 613–614 oral–buccal movements 303 orexin A 410 osmotic minipump 123 ouabain 206 over-expression 399 oxcarbazepine 244–245, 259, 439, 557–558, 594 oxidative stress 402 oxotremorine 82–83 oxygen 72 P-glycoprotein 411, 439, 487, 520, 548, 578–579, 581–584, 590, 592, 594–595, 607, 612–615, 635 panic disorder 302 panic-inducing 306 parafascicular 438–439 paraquat 561 paraventricular nucleus 302, 308, 310, 360, 427 Parkinson’s disease 183, 205, 207, 253–254, 284, 435, 441–442, 444–445 Parkinsonian 341 paroxetine 161, 306, 487–488, 528, 531 patients 608, 610, 616–618 Pavlovian 318–319, 328–329, 332, 338

Pavlovian conditioning 319, 322, 328, 330 PC-12 cells 583 PDC 273 pedunculopontine tegmentum 184, 190 peﬂoxacin 560, 575, 578, 582 pemetrexed 591 penciclovir 579, 634–635 penicillin 261 pentobarbital 73–75, 79–83 penumbra 440–442, 464, 467, 500, 659–663, 667–668, 677, 679–680, 682–684 peptide E 258 pergolide 424–425 periaqueductal gray 473, 475, 477 pericytes 546 perirhinal cortex 385 peritoneal cavity 627, 629 permeability 647–648, 650–656 PET 114–115, 208–210, 549, 661, 663–664 pharmacodynamics 169, 172, 515 pharmacokinetics 169, 172, 179–180, 515, 518, 521, 524, 601–602, 605–618, 625 phasic 324–325, 330–331, 335–336, 341–342 phencyclidine 186, 421, 497 phenobarbital 557 phenol 575, 581 phenylpropanolamine 360 phosphomycin 270 photic stimulation 24 physostigmine 241 phytoestrogen 582 pigs 217, 223, 225 pilocarpine-induced 436–438, 444 pindolol 531–534 pinocytosis 547 piperacillin 596, 616–617, 630–633 PK–PD models 632 PK/PD 175, 180, 589–595, 597 Polyacrylonitrile 218–219, 222–223 polyamide 223 polysulfone 218–219, 223 positron emission tomography 626, 638 posttranslational modiﬁcations 253 PPI 421–422, 424 predator stress 307–308, 311 prefrontal cortex 35, 39, 41–42, 255, 302–303, 307–309, 317, 319, 381, 384, 389–392, 422–423, 425–429, 437, 446 pregabalin 556–557 preoptic area 78–79, 303–304, 310 preproenkephalin-deﬁcient 408 preprohormones 253

695 preprotachykinins 289 prilocaine 611 probe calibration 627–628 probenecid 149, 553, 556–559, 561, 590, 594–595, 607 procholecystokinin 291 proenkephalin 280, 285, 287–289, 292–293 proenkephalin A 261 prohormone convertases 280, 287–289, 291 prohormones 280 proneurotensin 289 proopiomelanocortin 288–289 prooxytocin 289 propofol 76, 79–80 propranolol 459, 461, 463, 634 prostaglandin 244 prostaglandin PGE2 474 protachykinins 288 provasopressin 289 pseudoconditioning 326 ptosis 405 punishments 318 purine 437 purine metabolites 48 purinergic P2X7 receptor 21–22 push–pull 332 push–pull cannula 378–379 push–pull perfusion 6 pyramidal neurons 462 pyrithiamine 448 pyrogen 310 pyruvate 36, 226, 580, 584, 660, 662–664, 666–667, 670–671, 677–679, 681–684

recovery 47–52, 55–56, 60, 64, 67, 117–118, 120, 123–124, 380, 384, 456–457, 460–464, 466, 590–591, 595, 676, 678, 680 reinforcement 351–352, 354–355, 359, 362–364 relative recovery 93, 96–97, 132, 140, 142, 145 REM 302, 304–305, 312 remifentanil 259 reserpine 186 response time 114 restraint 421, 426–427 reticular activating system 317 reticular formation 76, 78, 81 retrodialysis 57–58, 96, 145–146, 156, 159, 603, 606–609, 611–612, 616, 619, 627 reverse dialysis 627 reverse microdialysis 201–203, 205, 207–209 reward 317–319, 323–326, 328–337, 339–342, 351–359, 363–364, 387, 389–392 rhodamine 579 RIA 251, 280 rifampin 631, 684 rigidity 405 risperidone 486–487, 494–495 rivastigmine 447 RNAi 411 rocuronium 580 ropivacaine 244, 608 rostral ventral medulla 473, 477 rotenone 206 RS 102221 533 rutaecarpine 614

quadrupole ion trap 281 quantitative microdialysis 48–53, 56, 67, 93, 95, 97–98, 102, 131, 139, 142, 145, 147, 152, 156, 161, 163, 552, 561, 590 quercetin 561 quinidine 155–156, 582–583, 614–615 quinine 324, 327 quinolinic acid 205, 560 quinpirole 404

S18986 562 salicylic acid 575, 634 saliva 217–218 salvianolic acid 244 sampling time 267–269 sarin 618 SB 242084 533 scar 384 schizophrenia 33–34, 42, 183, 192, 408, 419, 421–426, 430 scopolamine 357, 410, 447, 580 secretogranin 288–289 seizures 205, 436–439 selected reaction monitoring 281 self-administration 352, 354–355, 364 sensory stimulation 72, 82 septal 385 D-serine 21, 497–498 serine 259, 262

rabbits 385, 576, 578, 602, 610, 612, 618 raclopride 358, 426 radioimmunoassay 243 ranitidine 244, 583, 614–615 raphe nuclei 302–304, 529–530, 533 Raturns system 171 rearing 275 reboxetine 528, 534 recognition 385, 387

696 serotonin 55, 66, 72, 78–79, 83, 161, 235, 301–304, 306–310, 312, 360, 363, 421–423, 426–430, 437, 439, 444, 447, 590, 594, 597 serotonin transporter 409 SERT 409–410 sertraline 488, 528 sevoﬂurane 79 sex 352, 362 sexual behavior 362–363 sheep 217, 219, 223–224 sinapinic acid 258 sinomenine 582 skeletal muscle 133 Skinner box 320–323, 329, 333 sleep 301–302, 304–305, 309, 312 sleep deprivation 304 slow perfusion method 50, 148, 159 smoking cessation 515–516, 518, 522 Soman-evoked 437 somatostatin 437 sparﬂoxacin 553, 560, 579 SPECT 114–115 spill-over 20 spinal 602, 608, 613–614 spinal cord 473, 476 spreading depression 204, 208–210, 666 SSRIs 527–529, 531–536 stavudine 244, 608, 619 stearate–graphite past electrode 59 stratum corneum 579 stress 301–302, 304, 306–307, 309–312, 318, 321, 324, 341–342, 355, 419, 421–424, 426–427, 430, 458, 460 stria terminalis 354, 363 striatopallidal pathway 444 striatum 47–48, 53, 56–59, 61, 64–66, 77–82, 84, 100, 104, 184, 187, 189, 253–254, 437, 440–445, 447, 551, 559, 561, 563, 592–593, 606–608 stroke 287, 464, 466, 500, 659–660, 662, 664, 666–668, 670 subarachnoid haemorrhage 466, 672, 675, 679–681 subcutaneous 218–222, 224–226, 611, 615–617 Substance P 9, 37, 40, 80–82, 253, 259, 289, 292, 535 substantia innominata 354 substantia nigra 184, 187, 189, 254, 323, 442–444 sufentanil 613–614 sugar addiction 364 sulpiride 359, 361, 487, 496 suprachiasmatic nucleus 304 synaptic release 384 synovial ﬂuid 602, 611 syringe pumps 171, 174

tachykinins 9 tacrine 190 tactile stimulation 387 tail pinch 458–460, 463 tamoxifen 41 taurine 78, 80, 83–84, 274–275, 292, 440 tazobactam 617 TBOA 23 telemetry 118, 120 telitromycin 612, 630, 244 temperature effect 552 temporal cortex 385 temporal resolution 234–236, 244 tertatolol 531 tetrodotoxin 19, 292, 402 thalamus 386, 473–476 theofylline 244 theophylline 578, 614 thermoregulation 79–80 thioperamide 83 thioridazine 495 threonine 259 tiagabine 439, 592, 606 tight junctions 546–547 tirapazamine 244 tissue reactions 117–118, 120, 222 tissue trauma 48–50, 52–54, 56, 65–67, 550, 628 TMA+-selective electrodes 208 TOF 243 tonic 324–325, 331, 336, 342 tonic release 61–64 topiramate 597 topotecan lacton 244 tortuosity 49, 52, 132, 150, 153 transdermally 625, 634–635 transendothelial 629 transgenes 435, 445 transgenic models 207 transporter 604–605, 607–608, 612, 614 trapping column 236 trauma 112, 114–119, 123, 139, 142, 147–148, 152, 676–677, 680–681, 684 traumatic 204, 209, 500 traumatic brain 455, 466, 521, 549, 564, 597, 659–660, 664, 667 traumatic brain death 670 traumatic brain injury 675, 680, 682 traumatized 615 traumatized tissue 48, 52–55, 59–60 trazodone 528 tremor 405, 444 triamcinolone 609

697 triﬂuoperazine 521 triﬂuoroacetic acid 254 trimebutine 270 triple-probe 594 trypsin 260 TTX 34, 65–66, 459 TTX-dependent glutamate 25 TTX-sensitive 319 tuberomamillary 76 tumors 602, 615–616, 628, 631, 635–637, 680, 682, 684 tumour necrosis factor alpha 309 turning behavior 172 turnover rate 320 tyrosine hydroxylase 324, 399

verapamil 439, 520, 557, 579, 592, 594 vigabatrine 594 vigilance 301, 303–305 vincamine 607 vitreous humor 596 VMAT2 402, 405–406 voltammetry 7, 9–10, 12, 34–37, 57–64, 66, 99–101, 103–106, 112, 114, 147, 161, 203, 320, 325–326, 332, 336, 400, 402, 404–405, 410, 456 volume transmission 112, 393 volume-sensitive organic anion channels (VSOAC) 21, 39, 41 VTA 422, 425–426

UH-301 531 ultra-slow microdialysis 580 ultraﬁltration 217, 220–221, 225, 573–574, 576–577 ultraﬁltration probes 178 ultraslow microdialysis 217 unperturbed 47, 50, 56–57, 67 uptake inhibitors 55–56, 59 urea 226, 580 urethane 73–75, 78–79, 81, 84, 609 urocortins 310

wakefulness 304–305 WAY-100635 531, 533 whisker stimulation 24 withdrawal 355, 364–365 wound healing 120, 645, 654

valproate 557, 576–577 vanoxerine 81 vasospasm 675, 677, 679, 681–682 venlafaxine 528, 535 ventral pallidum 353 ventral tegmental area 79, 82–83, 184, 187, 189–190, 317, 323, 351, 353, 359

xenobiotics 577 xenotransplants 635 YM992 562 zero perfusion ﬂow 98 zero-net-ﬂux 47, 49–50, 54, 146, 603 zidovudine 146, 149, 155–156, 556, 608 ziprasidone 492–493, 495–496 zonisamide 522 zotepine 493