Genomics, Proteomics, and the Nervous System (Advances in Neurobiology, Volume 2)

Advances in Neurobiology

Series Editor Abel Lajtha

For further volumes: www.springer.com/series/8787

James D. Clelland Editor

Genomics, Proteomics, and the Nervous System

Editor James D. Clelland, Ph.D. Nathan S. Kline Institute for Psychiatric Research 140 Old Orangeburg Road Orangeburg, NY 10962 and New York University School of Medicine Langone Medical Center 550 First Avenue New York, NY 10016 USA [email protected]

ISSN 2190-5215 e-ISSN 2190-5223 ISBN 978-1-4419-7196-8 e-ISBN 978-1-4419-7197-5 DOI 10.1007/978-1-4419-7197-5 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2010938371 © Springer Science+Business Media, LLC 2011 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)

Preface

Mammalian central and peripheral nervous systems are highly complex at the structural, genetic and molecular levels, composed of multiple cell types and tissue structures. Thousands of genes, regulated at the genomic level via sequence variation or epigenetic regulation, are expressed at the RNA level and translated into proteins required to develop and maintain these cells and tissues, and along with small regulatory RNA molecules, lipids, and small molecule neurotransmitters, these gene products constitute the physical substrate for learning, memory, emotion, sensory perception, and consciousness itself. The potential for malfunction of this large number of complex biological systems is great, leading to the many behavioral and cognitive deficits observed in human psychiatric and neurological disorders, such as schizophrenia, autism and Alzheimer’s disease. This Volume of Advances of Neurobiology discusses research designed to increase our understanding of the nervous system and its structures and activities, through the utilization of genomic and proteomic technologies, addressing facets including development and epigenetic regulation, functions in learning and memory, and changes associated with neurological and psychiatric disorders. Specifically, the development of high-throughput genomic and proteomic analysis technologies, including microarray and high-throughput DNA sequencing technology, as well as integrated protein separation and mass spectrometry analysis systems, have created the opportunity for researchers to collect datasets that include measurements for all or most of the RNA species or the complement of proteins, within a particular biological sample. These high dimensional datasets are being generated for different nervous system cells and tissues, such as laser-capture microdissected neurons, or samples of postmortem pre-frontal cortical tissue. Different approaches have then been utilized to extract pertinent information, and these range from comparisons of postmortem cells and/or tissues using samples collected from subjects with and without disease states, for example patients with Alzheimer’s disease compared to control subjects, in order to discover differences between the samples that reflect aspects of the disease pathology, and that can then be investigated further to determine their role(s) in disease development. In addition, and in disorders such as autism, genome-wide expression analysis can provide data that allows for more focused investigations to test hypotheses regarding disease etiology, such as immune system dysfunction. Other experimental approaches include the utilization v

vi

Preface

of tissue or animal models to determine the effects of external stimuli on for example, investigation of prenatal viral infection as a model of schizophrenia. Although studies that utilize genomic and proteomic approaches differ widely, they can provide both data to support pre-existing hypotheses, plus they can implicate previously unconsidered biological networks and pathways in mammalian development, in nervous system functioning, and in the etiologies of diseases. A further important aspect in the development of genomic and proteomic approaches to nervous system research is the use of computational interrogation methods, that can be used to extract relevant information from the high dimensional data-sets. These techniques include cluster analysis and group classification algorithms, and the development of software tools allows the bench researcher to perform useful analyses of the multivariate datasets produced in genomics and proteomics experiments. The future ability to collect, store and analyze large-scale datasets will be central to the growing area of personalized medicine, whereby treatment choices and monitoring of individual’s responses to medications will be performed through the utilization of genomic and proteomic methods. Firstly, I would like to thank Dr. Catherine Clelland for her invaluable help in editing this volume. We would like to thank each of the authors who have contributed their outstanding work. We would also like to thank Dr. Abel Lajtha and Ms. Kristine Immediato for their contributions during the completion of this volume. Finally, we would like to dedicate this volume to our wonderful daughter Ayrleigh Clelland. New York, NY 10016

James D. Clelland

Contents

Part I  Development 1  The Genomics of Turner Syndrome and Sex-Biased Neuropsychiatric Disorders...................................................................... Phoebe M.Y. Lynn, Evangelia Stergiakouli, and William Davies 2  Mental Retardation and Human Chromosome 21 Gene Overdosage: From Functional Genomics and Molecular Mechanisms Towards Prevention and Treatment of the Neuropathogenesis of Down Syndrome.............. Mohammed Rachidi and Carmela Lopes 3  Epigenetic Programming of Stress Responses and Trans-Generational Inheritance Through Natural Variations in Maternal Care: A Role for DNA Methylation in Experience-Dependent (Re)programming of Defensive Responses.............................................................................. Ian C.G. Weaver

3

21

87

4  Prenatal Viral Infection in Mouse: An Animal Model of Schizophrenia......................................................................................... 113 S. Hossein Fatemi and Timothy D. Folsom 5  Proteomic Actions of Growth Hormone in the Nervous System........... 137 Steve Harvey and Marie-Laure Baudet Part II  Learning and Memory 6  Gene Expression and Signal Transduction Cascades Mediating Estrogen Effects on Memory.................................................. 161 Kristina K. Aenlle and Thomas C. Foster

vii

viii

Contents

  7  Diagnostic Genome Profiling in Mental Retardation........................... 177 David A. Koolen, Joris A. Veltman, and Bert B.A. de Vries   8  Genomic Imprinting and Sexual Experience-Dependent Learning in the Mouse............................................................................. 195 William T. Swaney and Eric B. Keverne   9  Proteomic Analysis of the Postsynaptic Density.................................... 227 Ayse Dosemeci Part III  Behavior 10  Functional Genomic Dissection of Speech and Language Disorders................................................................................. 253 Sonja C. Vernes and Simon E. Fisher 11  Studying Human Circadian Behaviour Using Peripheral Cells........................................................................................ 279 Lucia Pagani, Anne Eckert, and Steven A. Brown 12  Genome-Wide Expression Profiles of Amygdala and Hippocampus in Mice After Fear Conditioning............................ 303 Zheng Zhao and Yinghe Hu Part IV  Psychiatric Disorders 13  Genetic Studies of Schizophrenia........................................................... 333 Brien Riley 14  Proteomics of the Anterior Cingulate Cortex in Schizophrenia....................................................................................... 381 Danielle Clark, Irina Dedova, and Izuru Matsumoto 15  Proteome Effects of Antidepressant Medications................................. 399 Lucia Carboni, Chiara Piubelli, and Enrico Domenici Part V  Neurological Disorders 16  MicroRNAs in Neurodegenerative Disorders........................................ 445 Catherine L. Clelland and James D. Clelland 17  Specific and Surrogate Cerebrospinal Fluid Markers in Creutzfeldt–Jakob Disease.................................................................. 455 Gianluigi Zanusso, Michele Fiorini, Pier Giorgio Righetti, and Salvatore Monaco

Contents

ix

18  Genome-Wide Expression Studies in Autism-Spectrum Disorders: Moving from Neurodevelopment to Neuroimmunology....................... 469 Roberto Sacco, Antonio M. Persico, Krassimira A. Garbett, and Károly Mirnics 19  Protein Expression Profile of Alzheimer’s Disease Mouse Model Generated by Difference Gel Electrophoresis (DIGE) Approach..................................................................................... 489 Daria Sizova 20  Proteomic Analysis of CNS Injury and Recovery................................. 511 Günther K.H. Zupanc and Marianne M. Zupanc 21  MALDI Imaging of Formalin-Fixed Paraffin-Embedded Tissues: Application to Model Animals of Parkinson Disease for Biomarker Hunting.............................................................. 537 Isabelle Fournier, Julien Franck, Céline Meriaux, and Michel Salzet 22  Comparative Proteomic Analysis as a Method to Investigate Inflammatory and Neuropathic Pain..................................................... 557 Ellen Niederberger Index.................................................................................................................. 583

Contributors

Kristina K. Aenlle Department of Neuroscience, McKnight Brain Institute, University of Florida, 100244, Gainesville, FL 32610-0244, USA Marie-Laure Baudet Department of Physiology, Development and Neuroscience, University of Cambridge, Cambridge CB2 3D4, UK Steven A. Brown Chronobiology and Sleep Research Group, Institut für Pharmakologie und Toxikologie, Medizinische Fakultät, Universität Zürich, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland Lucia Carboni Neurosciences Centre of Excellence for Drug Discovery, GlaxoSmithKline, 37135, Verona, Italy Danielle Clark Discipline of Pathology, The University of Sydney, NSW 2006, Australia James D. Clelland Movement Disorders and Molecular Psychiatry, Nathan S. Kline Institute for Psychiatric Research, 140 Old Orangeburg Road, Orangeburg, NY 10962, USA Department of Psychiatry, New York University School of Medicine, Langone Medical Center, 550 First Avenue, New York, NY 10016, USA Catherine L. Clelland Department of Pathology and Cell Biology and Taub Institute for Research on Alzheimer’s Disease and the Aging Brain, Columbia University Medical Center, 630 West 168th Street, New York, NY 10032, USA William Davies Behavioural Genetics Group, Department of Psychological Medicine, School of Medicine and School of Psychology, MRC Centre for Neuropsychiatric Genetics and Genomics, University of Cardiff, Cardiff, UK xi

xii

Contributors

Irina Dedova Schizophrenia Research Institute, 384 Victoria Street, Darlinghurst, NSW 2010, Australia; School of Biomedical and Health Sciences, University of Western Sydney, NSW 1797, Australia Enrico Domenici Neurosciences Centre of Excellence for Drug Discovery, GlaxoSmithKline, 37135 Verona, Italy Ayse Dosemeci Laboratory of Neurobiology, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD 20892, USA Anne Eckert Psychiatric University Clinic, Basel, Switzerland S. Hossein Fatemi Department of Psychiatry, Division of Neuroscience Research, University of Minnesota, Medical School, MMC 392, 420 Delaware Street S.E, Minneapolis, MN 55455, USA; Departments of Pharmacology and Neuroscience, University of Minnesota Medical School, Minneapolis, MN 55455, USA; Department of Neuroscience, University of Minnesota Medical School, Minneapolis, MN 55455, USA Michele Fiorini Department of Neurological and Visual Sciences, Section of Clinical Neurology, University of Verona, P.le L.A. Scuro, 10, 37134 Verona, Italy Simon E. Fisher Wellcome Trust Centre for Human Genetics, University of Oxford, Roosevelt Drive, Oxford OX3 7BN, UK; Language and Genetics Department, Max Planck Institute for Psycholinguistics, 6525 XD, Nijmegen, The Netherlands Timothy D. Folsom Department of Psychiatry, Division of Neuroscience Research, University of Minnesota, Medical School, MMC 392, 420 Delaware Street S.E, Minneapolis, MN 55455, USA Thomas C. Foster Department of Neuroscience, McKnight Brain Institute, University of Florida, 100244, Gainesville, FL 32610-0244, USA Isabelle Fournier Laboratoire de Neuroimmunologie et Neurochimie Evolutives, FRE CNRS 3249, MALDI Imaging Team, Université Lille Nord de France, Université de Lille1, Bâtiment SN3, 1er étage, 59655 Villeneuve d’Ascq Cedex, France

Contributors

xiii

Julien Franck Laboratoire de Neuroimmunologie et Neurochimie Evolutives, FRE CNRS 3249, MALDI Imaging Team, Université Lille Nord de France, Université de Lille1, Bâtiment SN3, 1er étage, 59655 Villeneuve d’Ascq Cedex, France Krassimira A. Garbett Department of Psychiatry, Vanderbilt University, Nashville, TN 37203, USA Steve Harvey Department of Physiology, University of Alberta, 7-41 Medical Sciences Building, Edmonton, Alberta T6G 2H7, Canada Yinghe Hu Key Laboratory of Brain Functional Genomics, MOE & STCSM, Shanghai Institute of Brain Functional Genomics, East China Normal University, 3663 Zhongshan Road (N), Shanghai, 200062, China Eric B. Keverne Sub-Department of Animal Behaviour, University of Cambridge, High Street, Madingley, Cambridge CB23 8AA, UK David A. Koolen Department of Human Genetics, Nijmegen Centre for Molecular Life Sciences, Radboud University Nijmegen Medical Centre, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands Carmela Lopes University of French Polynesia, Tahiti, French Polynesia Phoebe M.Y. Lynn Behavioural Genetics Group, Department of Psychological Medicine, School of Medicine and School of Psychology, University of Cardiff, Cardiff, UK Izuru Matsumoto Discipline of Pathology, The University of Sydney, NSW 2006, Australia Schizophrenia Research Institute, 384 Victoria Street, Darlinghurst, NSW 2010, Australia Céline Meriaux Laboratoire de Neuroimmunologie et Neurochimie Evolutives, FRE CNRS 3249, MALDI Imaging Team, Université Lille Nord de France, Université de Lille1, Bâtiment SN3, 1er étage, 59655 Villeneuve d’Ascq Cedex, France Károly Mirnics Department of Psychiatry, Vanderbilt University, 8130A MRB III, 465 21st Avenue South, Nashville, TN 37203, USA Salvatore Monaco Department of Neurological and Visual Sciences, Section of Clinical Neurology, University of Verona, P.le L.A. Scuro, 10, 37134 Verona, Italy

xiv

Contributors

Ellen Niederberger Pharmazentrum frankfurt/ZAFES, Institut für Klinische Pharmakologie, Klinikum der Johann Wolfgang Goethe-Universität Frankfurt, Theodor Stern Kai 7, 60590 Frankfurt am Main, Germany Lucia Pagani Psychiatric University Clinic, Basel, Switzerland Antonio M. Persico Laboratory of Molecular Psychiatry and Neurogenetics, University “Campus Bio-Medico”, Via Alvaro del Portillo 21, 00128 Rome, Italy Chiara Piubelli Neurosciences Centre of Excellence for Drug Discovery, GlaxoSmithKline, 37135 Verona, Italy Mohammed Rachidi EA 3508, Laboratory of Genetic Dysregulation Models: Trisomy 21 and Hyperhomocysteinemia, University of Paris 7-Denis Diderot, Tour 54, E2-54-53, Case 7104, 2 Place Jussieu, Paris 75251, France Pier Giorgio Righetti Department of Chemistry, Materials and Chemical Engineering, Polytechnic of Milano, Milano 20131, Italy Brien Riley Departments of Psychiatry and Human & Molecular Genetics, and Virginia Institute for Psychiatric and Behavioral Genetics, Virginia Commonwealth University, Richmond, VA, USA Roberto Sacco Laboratory of Molecular Psychiatry and Neurogenetics, University “Campus Bio-Medico”, Rome, Italy Department of Experimental Neurosciences, I.R.C.C.S. “Fondazione Santa Lucia”, Rome, Italy Michel Salzet Laboratoire de Neuroimmunologie et Neurochimie Evolutives, FRE CNRS 3249, MALDI Imaging Team, Université Lille Nord de France, Université de Lille1, Bâtiment SN3, 1er étage, 59655 Villeneuve d’Ascq Cedex, France Daria Sizova Department of Genetics, University of Pennsylvania School of Medicine, Clinical Research Building, Room 755, 415 Curie Boulevard, Philadelphia, PA 19104-6149, USA Evangelia Stergiakouli Department of Psychological Medicine, School of Medicine, MRC Centre for Neuropsychiatric Genetics and Genomics, University of Cardiff, Cardiff, UK

Contributors

xv

William T. Swaney Behavioural Biology and Helmholtz Institute, Utrecht University, 3508 TB, Utrecht, The Netherlands Joris A. Veltman Department of Human Genetics, Nijmegen Centre for Molecular Life Sciences, Radboud University Nijmegen Medical Centre, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands Sonja C. Vernes Wellcome Trust Centre for Human Genetics, University of Oxford, Roosevelt Drive, Oxford OX3 7BN, UK Bert B.A. de Vries Department of Human Genetics, Nijmegen Centre for Molecular Life Sciences, Radboud University Nijmegen Medical Centre, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands Ian C.G. Weaver The Krembil Family Epigenetics Laboratory, Centre for Addiction and Mental Health (CAMH), 250 College Street, Toronto, ON, Canada M5T 1R8 Gianluigi Zanusso Department of Neurological and Visual Sciences, Section of Clinical Neurology, University of Verona, P.le L.A. Scuro, 10, 37134 Verona, Italy Zheng Zhao Key Laboratory of Brain Functional Genomics, MOE & STCSM, Shanghai Institute of Brain Functional Genomics, East China Normal University, 3663 Zhongshan Road (N), Shanghai 200062, China Günther K.H. Zupanc Department of Biology, Northeastern University, 134 Mugar Life Sciences, 360 Huntington Avenue, Boston, MA 02115, USA Marianne M. Zupanc School of Engineering and Science, Jacobs University Bremen, Campus Ring 1, 28759 Bremen, Germany

Part I

Development

The Genomics of Turner Syndrome and Sex-Biased Neuropsychiatric Disorders Phoebe M.Y. Lynn, Evangelia Stergiakouli, and William Davies

Abstract  Turner Syndrome (TS) is a developmental disorder caused by the absence of all, or part, of an X chromosome. Subjects with TS exhibit a welldefined neurocognitive profile characterised by deficits in aspects of memory, attention and social functioning. In this chapter, we focus on recent work analysing the genomic underpinnings of these cognitive endophenotypes. Through studying TS, we are likely to gain insights into the neural processes impacted upon by X-linked genes. As males and females differ with respect to their complements of X-linked genes, work on TS may provide clues as to the genetic basis of sexspecific vulnerability to certain neuropsychiatric disorders. Keywords  ADHD • Amygdala • Attention • Autism • Behaviour • Brain • Emotion recognition • Genomic imprinting • X chromosome • X-inactivation • X-monosomy

1 Turner Syndrome Turner Syndrome (TS) is a chromosomal disorder affecting approximately 1 in 1,800–2,500 live female births (Rovet, 2004). Interestingly, approximately 99% of affected foetuses do not reach term (Sybert & McCauley, 2004), but those that do survive to birth show a comparatively mild phenotype. TS is caused by complete or partial X-monosomy (Nijhuis-van der Sanden, Eling, & Otten, 2003). Approximately 50% of TS individuals exhibit complete loss of one X chromosome (karyotype 45,X), whilst the remainder comprise of females with cryptic mosaicism (in which a proportion of cells, including those in the brain, can have additional sex-linked sequences besides the single X chromosome) or structural sex chromosome ­abnormalities (Sagi et  al., 2006). The single intact X chromosome is inherited W. Davies (*) Behavioural Genetics Group, Department of Psychological Medicine, School of Medicine and School of Psychology, MRC Centre for Neuropsychiatric Genetics and Genomics, University of Cardiff, Cardiff, UK e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_1, © Springer Science+Business Media, LLC 2011

3

4

P.M.Y. Lynn et al.

maternally in around 70% of TS individuals (45,Xm), and the remainder possess an X chromosome of paternal origin (45,Xp) (Uematsu et al., 2002). Subjects affected by TS display characteristic physiological, neuropsychological and neuroanatomical features. Physiological deficits linked to the syndrome include short stature, ovarian dysgenesis, webbed neck, oedema, cardiovascular defects, renal malformations and sensorineural hearing loss (Hamelin, Anglin, Quigley, & Deal, 2006). Neuropsychological abnormalities associated with TS include deficits in attentional function (subjects show poor performance on attentionally demanding tasks (Romans, Stefanatos, Roeltgenm, Kushner, & Ross, 1998; Ross et  al., 2002; Temple, Carney, & Mullarkey, 1996) and tend to be readily distracted in real life situations (Rovet & Ireland, 1994), in visuospatial skills (Murphy et al., 1997; Ross, Roeltgen, & Zinn, 2006), in some forms of memory (Bishop et  al., 2000; Buchanan, Pavlovic, & Rovet, 1998; Haberecht et al., 2001), in recognising facial emotions (notably fear and anger; Lawrence, Kuntsi, Coleman, Campbell, & Skuse, 2003; Weiss et al., 2007) and in arithmetic skills (Mazzocco, 1998). TS subjects may also demonstrate subtle motor function abnormalities (Nijhuis-van der Sanden et  al., 2003; Rovet, 2004) and difficulties in aspects of social cognition (Lagrou et  al., 2006; McCauley, Feuillan, Kushner, & Ross, 2001; Schmidt et  al., 2006; Skuse et  al., 1997). The lack of X-linked material in TS also seems to confer enhanced vulnerability to disorders of social cognition, behavioural flexibility and attention such as autism, ADHD and schizophrenia (Donnelly et al., 2000; Prior, Chue, & Tibbo, 2000; Russell et al., 2006). Neuroanatomical abnormalities have been found in parietal lobe, amygdala, orbitofrontal cortex and superior temporal gyrus (Good et al., 2003; Kesler et al., 2003; Lawrence et al., 2003), which have been consistently linked to some of the neuropsychological characteristics associated with the TS phenotype. Visuospatial impairments have been linked to abnormalities in the parietal lobe; reduced activation in the parietal-occipital regions during a visuospatially demanding task has been shown in TS subjects (Haberecht et al., 2001). Significantly larger grey matter has been found in the amygdala and orbitofrontal cortex in TS individuals when compared to normal controls, and it has been suggested that the aberrant connections between these two structures contribute to the deficits in affective processing seen in TS subjects (Good et al., 2003; Kesler et al., 2004). There is significantly greater right superior temporal gyrus volume in TS individuals compared to controls (along with a parent-of-origin effect (POE); see below); the superior temporal gyrus is important in language processes, and therefore this difference between TS and control subjects might explain the preserved or superior verbal skills observed in TS females (Kesler et al., 2003; Skuse et al., 1997). Lastly, aberrant frontal cortical function has been shown in TS subjects during a visuospatial working memory task; anomalies in the frontal region could feasibly contribute to the impairments in attention, impulsivity and social functions in TS (Haberecht et al., 2001). Whilst TS arises from the lack of an entire X chromosome (or part thereof), its presentation can also depend upon the parental origin of the remaining intact X chromosome, such that the appearance of 45,Xp and 45,Xm subjects differs. POE on physiological parameters (notably neck webbing, cardiovascular defects and

The Genomics of Turner Syndrome and Sex-Biased Neuropsychiatric Disorders

5

sensorineural hearing loss) have been described in TS (Chu et al., 1994; Hamelin et al., 2006), though their existence is somewhat controversial (Bondy et al., 2007; Mathur et al., 1991). POE have also been noted with respect to the neuropsychology and neuroanatomy of TS. Using image and story recall tasks, Bishop et  al. (2000) have shown that visuospatial working memory is impaired in 45,Xp subjects relative to 45,Xm (and 46,XX) subjects, while the 45,Xm group exhibited poorer verbal material retention than the other two groups. The parental origin of the X chromosome also appears to influence social cognitive functioning, notably the inhibition of a prepotent response (Skuse et al., 1997). Specifically, 45,Xm subjects perform more poorly than 45,Xp (and 46,XX) subjects on a task taxing behavioural inhibition and demonstrate enhanced vulnerability to autism, the archetypal disorder of social cognition (Baron-Cohen, Knickmeyer, & Belmonte, 2005). Using magnetic resonance imaging (MRI) to examine neuroanatomy in  vivo, X-linked POE have been found in the superior temporal gyrus (Kesler et al., 2003), caudate nuclei, thalamus, and the temporal lobe (Cutter et  al., 2006); in 45,Xm subjects, bilateral superior temporal gyrus volume is significantly greater than that in 45,Xp and 46,XX subjects, whereas the bilateral caudate nuclei and thalamus grey matter and the temporal lobe white matter have been shown to be larger in 45,Xp than 45,Xm individuals. Furthermore, there is evidence for an X-linked POE in the hippocampus whereby 45,Xm have a larger right hippocampal volume than 45,Xp subjects, which might explain the differential performance on visuospatial memory performance between these two groups (Cutter et  al., 2006; Moscovitch, Nadel, Winocur, Gilboa, & Rosenbaum, 2006). The existence of POE on TS neurocognitive measures and neuroanatomy, as for the POE on physiology, is disputed, and it is likely that POE, should they exist, will be relatively subtle (Skuse, 2005). However, that is not to say that these effects will be unimportant!

1.1 Genetic Mechanisms Underlying TS Endophenotypes The characteristic features associated with TS outlined above presumably arise due to haploinsufficiency for the products of one (or more) X-linked genes that normally escape X-inactivation (the epigenetic process by which, in the somatic cells of 46,XX females, one of the two X chromosomes is randomly inactivated). Between 15 and 20% of human X-linked genes are thought to escape X-inactivation and hence are expressed from both X chromosomes in 46,XX females (Carrel & Willard, 2005); these genes will only be expressed from the single X chromosome in TS subjects and thus represent candidates for TS endophenotypes. However, it is important to be aware that data on escape from X-inactivation have generally been obtained from cell culture studies and confirmed in a limited number of tissues. The exact degree of escape from X-inactivation for a gene (and therefore its relevance to the TS phenotype) is likely to vary according to the precise tissue examined, and according to the time point at which it is assayed (Bittel et al., 2008; Carrel, Hunt, & Willard, 1996; Carrel & Willard, 2005).

6

P.M.Y. Lynn et al.

The X-linked POE could arise due to the differential expansion of repeat sequences in parental germlines, such that 45,Xp and 45,Xm subjects possess different numbers of repeats in certain key genes (e.g. FMR1) , which may in turn influence expression (Weinhäusel & Haas, 2001). Alternatively, the POE could arise due to the action of one or more so-called X-linked imprinted genes (Davies, Isles, Burgoyne, & Wilkinson, 2006). Imprinted genes, in contrast to most mammalian genes, are inherited in duplicate, but as a consequence of differential epigenetic marking, are only expressed from one allele. For approximately half of all imprinted genes, it is the paternally inherited allele that is preferentially expressed, whilst for the remainder the maternally inherited allele is preferentially expressed. Recent evidence (mainly from mouse models) has implicated these genes as key players in neurodevelopmental and behavioural processes (Davies, Isles, Humby, & Wilkinson, 2006; Davies et  al., 2005; Wilkinson, Davies, & Isles, 2007). As 45,Xp subjects inherit a single X chromosome from their father, their brain and behavioural functioning can only be influenced by paternally expressed X-linked genes. In contrast, the brain and behaviour of 45,Xm subjects can only be modulated by the action of maternally expressed X-linked genes. Skuse et al. (1997) have postulated the existence of a paternally expressed gene that acts to enhance social cognitive function, to explain the superior performance of 45,Xp subjects relative to 45,Xm subjects in this domain.

1.2 Problems with Investigating the Genomics of TS Recently, there has been substantial interest in identifying and beginning to characterise candidate genes underlying TS endophenotypes. Through finding such genes, we are likely to gain important insights into the molecular and neurobiological basis of cognitive constructs such as visuospatial memory, attention and face recognition within the normal population. However, the genomics of TS have been difficult to investigate for a variety of reasons. Firstly, as the disorder does not markedly reduce life expectancy there is a lack of post-mortem tissues available for analysis. Moreover, gene expression in tissues that are available (both post-mortem and from living subjects) may be influenced by the presence of cryptic mosaicism, i.e. the presence of additional sex-linked gene sequences (both X and Y) in a subset of cells within the population (Henn & Zang, 1997), or by skewed X-inactivation (Zinn et  al., 2007). Hence, this may obscure the precise relationship between the action of a gene and its phenotypic manifestation. Finally, to alleviate the short stature and ovarian dysfunction endophenotypes, TS subjects are commonly treated with growth hormone and oestrogen supplements, respectively. Both treatments could potentially influence brain and behavioural functioning (Falleti, Maruff, Burman, & Harris, 2006; Hamelin et al., 2006; Ross et al., 2003) and are likely to elicit significant effects on gene expression. Again, these treatments may make the downstream effects of a specific gene difficult to ascertain.

The Genomics of Turner Syndrome and Sex-Biased Neuropsychiatric Disorders

7

The problems associated with investigating the genomics of TS may be overcome to some extent through investigations in mouse models. The 39,XO mouse, which like TS subjects possesses just one X chromosome, appears to recapitulate some of the key behavioural features of the TS profile (Lynn & Davies, 2007). This model is amenable to large-scale neurobiological and behavioural analysis, is unconfounded by mosaicism, and the mice can have a well-controlled behavioural and treatment history. Limitations of the model include difficulties associated with assaying complex neuropsychological phenotypes, the fact that some phenotypes sensitive to X-monosomy in man may not be in mouse (either due to more extensive X-inactivation in mice, or to orthologues of X-linked human genes being autosomal in mice) and the fact that some murine X-linked genes do not have human orthologues. Nevertheless, despite these caveats, the 39,XO mouse model has provided some intriguing clues as to possible TS candidate genes (see below).

1.3 Identification of Candidate Genes for TS Endophenotypes 1.3.1 Human Studies To date, the main strategy for identifying X-linked regions containing TS candidate genes has been deletion mapping. In this approach, subjects possessing an intact X  chromosome and a second X chromosome with a deletion (generally of the ­terminal region of the short arm of the X chromosome) are compared to 46,XX ­controls. In theory, a TS endophenotype will be present in subjects lacking the underlying gene(s) on their deleted chromosome, but not in subjects whose deletion does not encompass the gene(s). In Skuse et al. (1997), 45,XXp− subjects (possessing an intact maternally inherited X chromosome and a paternally inherited X chromosome with a large terminal deletion of the short arm) resembled 46,XX females in terms of their social cognitive profile. This implied that the paternally expressed gene that they postulated may be influencing social cognitive function must either reside close to the centromere on the short arm of the X or on the long arm. As the deleted chromosome was preferentially inactivated in 46,XXp− ­subjects, this result further implied that the gene escaped X-inactivation. Using a combination of deletion mapping and structural neuroimaging, Good et al. (2003) attempted to localise candidate genes underlying the impaired fear and anger recognition seen in TS females and to determine their effects on neuroanatomy. In this study, subjects with deletion of a critical region of Xp11.3 (of paternal or maternal origin) were shown to have impaired fear recognition and significantly increased grey matter volumes bilaterally in the amygdala and orbitofrontal cortex relative to 46,XX controls. This small region, less than 5 Mb in size, contains 21 confirmed genes, of which there is evidence that 5 may escape X-inactivation in man. The authors thus hypothesised that haploinsufficiency for one or more of these genes could cause aberrant development (and therefore connectivity) of the orbitofrontal cortex-amygdala axis and subsequent impairments in emotion recognition­.

8

P.M.Y. Lynn et al.

Initially, the prime suspects for the effects described above were the MAOA and MAOB genes encoding the monoamine oxidases A and B, respectively. These enzymes are highly expressed in the amygdala, modulate a wide variety of fundamental neurochemical processes, and can influence neurodevelopment (Vitalis et al., 2002). Moreover, MAOB seems to escape X-inactivation (as indexed by reduced activity in TS platelets relative to those of 46,XX subjects (Lawrence et al., 2007) though MAOA appears to be normally X-inactivated (Benjamin, Van Bakel, & Craig, 2000; Nordquist & Oreland, 2006). Despite initial understandable interest in the MAOA and MAOB genes, SNP mapping of the critical region suggested no association between markers close to, and within, these genes and impaired fear recognition in a TS sample. Instead, convincing and reproducible association was found between a second gene within the critical 5 Mb region, EFHC2, and fear recognition (Weiss et al., 2007). EFHC2 is a novel brain-expressed transcript spanning 195  kb, and its predicted protein structure contains an EF-hand domain which is generally found in calcium-binding proteins. The protein product of EFHC2 has 41.6% amino acid identity to EFHC1 (Gu, Sander, Heils, Lenzen, & Steinlein, 2005), a protein involved in neural pruning (Suzuki et  al., 2004). Assuming that the function of EFHC2 is similar to that of EFHC1, one may speculate that its haploinsufficiency could underlie the increased amygdala size observed in TS females. Recently, a second study has failed to find any evidence for association between EFHC2 and fear recognition in a new TS cohort (Zinn, Kushner, & Ross, 2008). This inability to replicate the first study could be due to population stratification considerations. However, given the potential importance of the initial findings to TS, and to behavioural genetics in general, it is clear that further research on the potential function of EFHC2 variants is needed. A second region on the distal short arm of the X chromosome has been implicated by deletion mapping in the overall neurocognitive profile of TS (Ross, Roeltgen, Kushner, Wei, & Zinn, 2000; Zinn et al., 2007). This region of Xp22.3, which encompasses around 8.3  Mb, contains 31 annotated genes, including the short stature gene SHOX. SHOX, which encodes a homeobox protein, has previously been implicated in the TS physical phenotype (Ross et al., 2001), but a lack of correlation between the TS neurocognitive profile and height appears to indicate that its haploinsufficiency is unlikely to play a prominent role in the TS brain and behavioural phenotype. Alternative candidate genes for the characteristic TS cognitive profile include NLGN4X and STS; both of these genes are known to escape X-inactivation in man (Carrel & Willard, 2005; Yen et al., 1988). NLGN4X encodes a member of the synaptic membrane-bound neuroligin family (Bolliger, Frei, Winterhalter, & Gloor, 2001), and has been shown to be mutated in cases of autism and mental retardation (Jamain et al., 2003; Laumonnier et al., 2004). STS encodes steroid sulfatase, an enzymic modulator of neurosteroid activity (Yen et al., 1988). Male subjects with microdeletions encompassing the STS gene can exhibit mental retardation and ADHD (Boycott et  al., 2003; Lonardo et  al., 2007). The NLGN4X and STS gene products will be haploinsufficient in TS subjects, rather than completely absent as in the aforementioned cases, therefore the resultant phenotypic consequences will presumably be more subtle. On the basis of the

The Genomics of Turner Syndrome and Sex-Biased Neuropsychiatric Disorders

9

available human data, other genes within the Xp22.3 interval may be considered as contributors towards the TS neurocognitive profile. However, the absence of a syntenic region in rodents has precluded comprehensive functional analysis of these genes. Besides potentially being involved in the TS neurocognitive profile, the short arm region Xp11.2-Xp22.3 has also been implicated in some physical TS phenotypes, namely: short stature, ovarian failure, high arched palate and possibly autoimmune thyroid disease (Ross et al., 2006; Zinn et al., 1998). Haploinsufficiency of the protein encoded by the pseudoautosomal gene SHOX appears to be partially responsible for the short stature phenotype in TS; it has been suggested that SHOX haploinsufficiency can account for about two-thirds of the height deficit displayed in TS females (Ross et al., 2001), while the rest of the variance could be explained by other putative height loci within Xp11.2-22.1 proximal to PAR1 and/or genetic background (Zinn et al., 1998). Ovarian failure in TS might arise, in part at least, from haploinsufficiency for the candidate gene ZFX at Xp21.2. Mutation of the mouse orthologue of this gene leads to growth retardation and a reduced number of germ cells (Luoh et al., 1997; Simpson & Rajkovic, 1999). Despite several years passing since the first indications that X-linked imprinted genes could contribute to the physical and cognitive aspects of the TS phenotype, still no candidate genes have been identified. Theoretically, X-linked imprinted genes could be identified by microarray comparison of gene expression in 45,Xp and 45,Xm tissues. The messenger RNA from paternally expressed X-linked genes should be present in the former tissue type but absent (or almost absent) from the latter, whilst the reverse would be true for messenger RNA from maternally expressed X-linked genes. The major problem with this approach is the lack of availability of suitable tissues (notably brain), as referred to previously. It is possible that comparing gene expression in surrogate tissues such as blood or skin may provide clues as to imprinted genes influencing brain function. Preliminary investigations in our laboratory comparing gene expression in 45,Xp and 45,Xm blood have so far failed to identify candidate X-linked imprinted genes. Other strategies for identifying novel X-linked imprinted genes may include ascertainment on the basis of differential epigenetic marking, e.g. methylation (Smith & Kelsey, 2001) or on the basis of bioinformatic signatures characteristic of imprinted genes. However, the fact that the sequence content of the X chromosome differs markedly from that of the autosomes (Davies, Isles, & Burgoyne, et  al., 2006) means that proven bioinformatic methods for detecting autosomal imprinting in humans (Luedi et al., 2007) may be of limited utility when investigating this chromosome. 1.3.2 Mouse Studies In mice, only seven genes are known to escape X-inactivation (Brown & Greally, 2003). Thus, identifying candidate genes (and molecular pathways) underlying X-monosomy effects in mice is a much more straightforward procedure than in man.

10

P.M.Y. Lynn et al.

As six of these genes also escape X-inactivation in man, their human ­orthologues may be examined for effects on analogous functions. Several physical phenotypes have now been shown to be sensitive to X-monosomy in mice, including pre- and postnatal growth (Burgoyne, Evans, & Holland, 1983; Burgoyne, Ojarikre, & Turner 2002), oocyte loss (Burgoyne & Baker, 1985) and inner ear development (Hultcrantz, Stenberg, Fransson, & Canlon, 2000). A number of brain and behavioural measures have also recently been shown to be sensitive to X-monosomy in the mouse. 39,XO mice, like TS subjects, display heightened anxiety under aversive conditions (Isles, Davies, Burrmann, Burgoyne, & Wilkinson, 2004). In the same study, the authors demonstrated reduced expression of the GABAA receptor subunit genes Gabra1 and Gabrg2, and increased expression of the GABAA receptor subunit gene Gabra3 in 39,XO brain relative to 40,XX brain. However, the GABAergic perturbations evident in 39,XO mice did not seem to contribute to the fear reactivity phenotype. Instead, the authors proposed the X-escapee Utx, a gene possessing a human orthologue at Xp11.2 (Greenfield et  al., 1998), as a candidate for the observed X-monosomy effects on behaviour. Again like TS subjects, 39,XO mice display impaired response accuracy on a behavioural task taxing attention; this deficit was not apparent in 40,XY*X mice (effectively 39,XO mice with an additional small chromosome, Y*X) implying that a gene on Y*X could affect attention (Davies, Humby, Isles, Burgoyne, & Wilkinson, 2007). Of the nine or so genes on Y*X, two were known to escape X-inactivation in mice: the pseudoautosomal gene Sts and Mid1, and could therefore be considered candidates for the X-monosomy effect. On the basis of its expression pattern and previous knowledge of its protein product’s function, Sts was considered the better candidate for the attentional phenotype. Interestingly, Sts is the mouse orthologue of STS, a gene previously mentioned as a positional candidate for the neurocognitive profile of TS. It is therefore tempting to speculate that having just one copy of STS (and therefore being haploinsufficient for the steroid sulfatase enzyme) in TS specifically contributes to the deficits in attentional functioning. Moreover, mouse work has hinted that the GABAergic abnormalities seen in the 39,XO mice may be dependent upon Sts levels (Isles et  al., 2004). Current work in our laboratory aims to determine the extent to which steroid sulfatase may contribute to attentional phenotypes via GABAergic mechanisms in man and mouse. In order to discover candidate genes underpinning the enlarged amygdala phenotype (and ensuing neurobehavioural abnormalities) associated with TS, Raefski, Carone, Kaur, Krueger, and O’Neill (2007) compared gene expression in the amygdalae of late-stage 39,XO and 40,XX embryos. This study revealed significant expression differences in 161 genes between the two groups, of which several were known to be involved in Wnt signalling cascades. One such X-linked gene, Gpc3, was downregulated approximately twofold in 39,XO brain. Gpc3 encodes a glypican protein, the downregulation of which was hypothesised to lead to activation of the canonical Wnt pathway and subsequent increased cell proliferation and amygdala size. The human orthologue of Gpc3 is located at Xq26, and is apparently subject to X-inactivation (Huber et  al., 1999), hence it appears unlikely to be the causative gene for the X-monosomy effect on amygdala size in TS. It may, however, be a downstream effector. Whatever the precise contributions of individual genes,

The Genomics of Turner Syndrome and Sex-Biased Neuropsychiatric Disorders

11

it seems that investigating the expression of genes involved in Wnt signalling may shed light on some of the neuroanatomical findings in TS. By using two separate crosses, 39,XO mice may be generated whose single X is of paternal or maternal origin (Burgoyne & Evans, 2000; Evans & Phillips, 1975). Davies et  al. (2005) compared these two groups of mice on a reversal learning task designed to assay behavioural flexibility, i.e. the psychological construct that had been shown by Skuse et  al. (1997) to be dependent upon parental origin of the X chromosome. The mouse work accurately recapitulated the human data, in that the mice inheriting their single X chromosome maternally (39,XmO) displayed impaired behavioural flexibility relative to 39,XpO (and 40,XX) subjects, much like 45,Xm TS subjects did relative to 45,Xp subjects. Follow-up studies comparing gene expression in 39,XpO and 39,XmO brains to find possible X-linked imprinted genes underlying this behavioural finding identified Xlr3b as a novel maternally expressed imprinted gene (Davies et al., 2005; Raefski & O’Neill, 2005). The exact role, if any, which this gene and its product may play in murine brain development, brain function and cognition remain to be determined. The closest human orthologue to Xlr3b is FAM9B, located within the critical region for the TS neurocognitive profile on Xp22.3. This gene is known to be highly expressed in testis (Martinez-Garay et  al., 2002), but does not appear to be expressed in embryonic brain or in adult sensorimotor cortex (unpublished results). Hence, despite its promising location, it seems a weak a priori candidate for the TS cognitive phenotype.

2 The Genomics of Sex-Biased Neuropsychiatric Disorders Even taking into account possible ascertainment biases, most of the common neuropsychiatric disorders exhibit a sex bias; this may be evident in terms of incidence, aetiology, underlying neural substrates or response to therapeutics (Cahill, 2006). Presumably, there is something about the brains of males that predisposes them to developing autism spectrum disorders and ADHD (between 4 and 10 times more prevalent in males), and something about the brains of females that predisposes them to developing disorders such as anorexia nervosa and unipolar depression (between 2 and 10 times more prevalent in females) (Holden, 2005). In the case of autism, this idea has been formalised in the “extreme male brain theory”, which posits that the core behavioural and neurological features of autism (notably the impaired empathising and enhanced systematising) represent extreme versions of typical male traits (Baron-Cohen et al., 2005). Neural differences between males and females must ultimately arise from the differential expression of sex-linked genes between the sexes. There are two dissociable routes through which sex-linked gene expression may affect sex-specific neurobiology: the first, and best studied, route is through influencing the development of the gonads (ovaries in females, testes in males). Steroid hormones secreted by the gonads (notably oestrogen and testosterone in different ratios according to

12

P.M.Y. Lynn et al.

the gonadal type) may then act to shape brain function, either by affecting ­neurodevelopment (organisational effects) or by affecting ongoing brain processes (activational effects). The second route is via direct effects in brain tissue. Interestingly, there seems to be an over-representation of genes influencing cognitive functioning on the sex chromosomes (the X chromosome in particular), exemplified by the fact that X-linked mutations are commonly found to cause mental retardation (Zechner et al., 2001). Hence, this second route of sexual differentiation of the brain may turn out to be at least as important, if not more so, than the first. There are a number of genetic mechanisms that may bring about sex-specific gene expression, and these have been reviewed elsewhere (Davies & Wilkinson, 2006). Genes that escape X-inactivation will generally be more highly expressed in female than in male brain, though the fold difference in expression will depend upon the efficiency with which the gene escapes the silencing process. Moreover, some genes that escape X-inactivation have Y-linked homologues; these may, or may not, encode functionally similar products and may, or may not, be expressed at equivalent levels. Hence, male-limited expression of such homologues is a second way in which sex-specific gene expression may be achieved. Vawter et al. (2004) showed by microarray analysis that five Y-linked homologues of X-inactivation escapees (UTY, USP9Y, SMCY, DBY and RPS4Y) were expressed between 2 and 128 times more highly in male than in female post-mortem brain. The same study revealed that levels of the X-inactivation mediating RNA XIST were approximately 14 times higher in female than male brain. X-linked imprinting may also give rise to sexually dimorphic gene expression as a consequence of the fact that females inherit two X chromosomes (one from their father and a second from their mother), whereas males inherit a single X chromosome, invariably from their mother (Davies, Isles, & Burgoyne, et  al., 2006). Therefore, paternally expressed X-linked genes can only be expressed in female brain, whereas maternally expressed X-linked genes can be expressed more highly in male brain (provided they are subject to X-inactivation). A final mechanism that could potentially underpin sex-specific brain development is the expression of Y-linked genes with no X-linked counterparts; the paradigmatic example of such a gene is SRY. This gene encodes a transcription factor which acts to initiate a complex signalling cascade culminating in the development of the testes and testosterone secretion. Thus, it may influence sexual differentiation of the brain indirectly. SRY is also expressed throughout the brain (Mayer, Lahr, Swaab, Pilgrim, & Reisert, 1998), and data in rodents suggests that its protein might regulate tyrosine hydroxylase transcription to effect downstream changes in dopaminergic function (Dewing et al., 2006; Milsted et al., 2004). As dopaminergic dysregulation has been implicated in a number of neuropsychiatric disorders which are more prevalent in males (ranging from ADHD and schizophrenia to compulsive gambling and alcoholism; Blum et al., 2000; Holden, 2005; Staller & Faraone, 2007; Toda & Abi-Dargham, 2007) a role for this gene may be suspected. Ongoing work in our laboratory is aiming to characterise the role of SRY (and other Y-linked brain expressed genes) in rodent and human brain function. In man, Y chromosome variation is captured by using biallelic markers that can be combined to form stable

The Genomics of Turner Syndrome and Sex-Biased Neuropsychiatric Disorders

13

l­ineages of Y chromosome, the so-called haplogroups (Jobling & Tyler-Smith, 2003). Since SRY is a small gene with only one exon, it has limited variation. Thus, markers in the rest of the non-recombining region of Y chromosome are also used. Theoretically, these Y chromosome variants can increase susceptibility to neurodevelopmental disorders through their interactions with brain-expressed genes on other chromosomes (such as TH encoding tyrosine hydroxlase) and/or with environmental factors that are known to increase risk to neurodevelopmental disorders such as smoking exposure during pregnancy (Thapar et al., 2003) and birth complications (Thapar et al., 2005). Finally, Y chromosome variants could have an effect on cognitive function in the normal population and thus have a modifying effect on risk for neurodevelopmental disorders such as ADHD and schizophrenia. We hope that studying Y-linked genes will provide insights into whether Y chromosome variants play an important role not only in male susceptibility to neurodevelopmental disorders but also in variability of cognitive function within the normal population.

3 Insights into Sex-Biased Neuropsychiatric Disorders from Turner Syndrome Hemizygosity for X-linked genes (in males), and X-linked imprinting can contribute towards sexually dimorphic gene expression profiles. The effects of X-monosomy (effectively equivalent to hemizygosity) and X-linked imprinting on brain development, brain function and behaviour can both be readily ascertained in the “experiment of nature” that is Turner Syndrome (taking into account the caveats listed previously). Importantly, in TS, these effects are unconfounded by the presence of Y-linked genes. From the TS data, it should therefore be possible to generate hypotheses about the genetic and neural substrates underlying sex differences in behaviour and in vulnerability to mental disorders. In general, on endophenotypes shown to be sensitive to X-monosomy (for example, emotion recognition, attention and amygdala size), we might expect hemizygous males to resemble TS subjects more than they resemble 46,XX females (assuming a lack of mitigating factors such as hormonal influences). The fact that males exhibit greater attentional difficulties relative to females, and are substantially more vulnerable to ADHD (Holden, 2005), suggests that hemizygosity for one or more X-linked genes may be responsible for this sex bias. Similarly, the fact that males exhibit larger amygdalae than females (Good et al., 2003) may be explained by their hemizygosity for one or more X-linked genes. The larger male amygdala may result from reduced neural pruning, and this may explain why males exhibit impaired discrimination of facial emotions (Thayer & Johnsen, 2000), and an enhanced vulnerability to autistic spectrum disorders in which emotion recognition processes go awry (Boraston, Blakemore, Chilvers, & Skuse, 2007; Humphreys, Minshew, Leonard, & Behrmann, 2007). As discussed previously, studies in TS (and in a TS mouse model) have proposed STS and EFHC2 as candidate genes for the attentional and emotion recognition phenotypes respectively. As STS escapes X-inactivation in females, and as its Y

14

P.M.Y. Lynn et al.

homologue is a non-expressed pseudogene (Yen et  al., 1988), we may expect its expression to be higher in female than in male brain. Indeed, this appears to be the case (unpublished results). Hence, we may hypothesise that the effective hemizygosity of STS in males predisposes them to developing ADHD (Davies et  al., 2007). Currently, little is known about the EFHC2 gene. In mice, the gene appears to escape X-inactivation (unpublished results), and as it resides in a genomic region rich in X-escapees, the same is likely to be true in man. Moreover, the gene does not seem to have a functionally equivalent Y-linked homologue. Therefore, it is possible that male hemizygosity for EFHC2 may lead to reduced neural pruning in the amygdala, thereby predisposing this sex to autism. A third gene, NLGN4X has been proposed as an attractive candidate for aspects of the TS neurocognitive profile (Zinn et al., 2007) and for predisposition to autism (Jamain et al., 2003; Laumonnier et al., 2004). This gene has a Y homologue (NLGN4Y) located ay Yq11 (Ylisaukko-oja et al., 2005). Determining the exact contributions of STS, EFHC2, NLGN4X and NLGN4Y to sexually dimorphic brain development (and therefore to sex-specific vulnerability to neuropsychiatric disorders) represents an exciting avenue for further investigation. Due to the nature of X-linked imprinting, we may expect traits underpinned by X-linked paternally expressed genes to be evident in females but not in males, whilst the expression of traits influenced by X-linked maternally expressed genes may vary between the sexes according to whether the causal genes are X-inactivated or not (Davies, Isles, & Burgoyne, et  al., 2006). Skuse et  al. (1997) have postulated that female-limited expression of the paternally expressed X-linked gene conferring social competence referred to previously may explain why females score better than males in a questionnaire assessing social cognitive impairment. The existence of an X-linked maternally expressed gene that enhances visuospatial memory skills (Bishop et al., 2000) may explain male superiority in this domain (Loring-Meier & Halpern, 1999), assuming that it is X-inactivated, and thus would be expressed at lower levels in female than male brain. The TS data further suggest that sex differences in the structure/function of the superior temporal gyrus (Im et al., 2006), thalamus (Li, Huang, Constable, & Sinha, 2006) and caudate nucleus (Munro et al., 2006) may arise due to the downstream neurodevelopmental effects of one or more X-linked imprinted genes. Clearly, the identification of X-linked imprinted genes in man, together with subsequent analysis of how these genes may contribute towards sexually dimorphic neurobiology, represents an important goal. In mice, the maternally expressed Xlr genes appear to be subject to X-inactivation, and thus are expressed more highly in male tissues than in female tissues (Davies et al., 2005; Raefski & O’Neill, 2005). Likewise, their closest human orthologue, the FAM9B gene, is more highly expressed in male than female gonads (Martinez-Garay et al., 2002).

4 Conclusion Understanding sex differences in neurobiology, in vulnerability to mental disorders, and in response to therapy represents one of the biggest challenges in psychiatric genetics today. A number of mechanisms can give rise to sexually dimorphic gene

The Genomics of Turner Syndrome and Sex-Biased Neuropsychiatric Disorders

15

expression and subsequent differences in brain development and/or function. Two of these mechanisms (X-linked gene dosage and X-linked imprinting) can be studied in the amenable “experiment of nature” provided by Turner Syndrome. Comprehensive phenotyping/genomic assays in TS (and in model systems such as the 39,XO mouse) are likely to provide important insights into brain processes affected by these mechanisms, and to identify underlying genes. Such combined analyses will increase our knowledge about the basis of sexually dimorphic neurobiological processes, and will enable more effective sex-specific therapies to be developed for sex-biased neuropsychiatric disorders. Acknowledgements  P.M.Y.L. is supported by the Biotechnology and Biological Sciences Research Council (BBSRC, UK). E.S. is supported by the Wellcome Trust and the Medical Research Council (MRC, UK). W.D. is an RCUK Fellow in Translational Research in Experimental Medicine.

References Baron-Cohen, S., Knickmeyer, R. C., & Belmonte, M. K. (2005). Sex differences in the brain: implications for explaining autism. Science, 310, 819–823. Benjamin, D., Van Bakel, I., & Craig, I. W. (2000). A novel expression based approach for assessing the inactivation status of human X-linked genes. European Journal of Human Genetics, 8, 103–108. Bishop, D. V. M., Canning, E., Elgar, K., Morris, E., Jacobs, P. A., & Skuse, D. H. (2000). Distinctive patterns of memory function in subgroups of females with Turner syndrome: evidence for imprinted loci on the X-chromosome affecting neurodevelopment. Neuropsychologia, 38, 712–721. Bittel, D. C., Theodoro, M., Kibiryeva, N., Fischer, W., Talebizadeh, Z., & Butler, M. G. (2008). Comparison of X chromosome inactivation patterns in multiple tissues from human females. Journal of Medical Genetics, 45, 309–313. Blum, K., Braverman, E. R., Holder, J. M., Lubar, J. F., Monastra, V. J., Miller, D., et al. (2000). Reward deficiency syndrome: a biogenetic model for the diagnosis and treatment of impulsive, addictive, and compulsive behaviors. Journal of Psychoactive Drugs, 32, 1–112. Bolliger, M. F., Frei, K., Winterhalter, K. H., & Gloor, S. M. (2001). Identification of a novel neuroligin in humans which binds to PSD-95 and has a widespread expression. The Biochemical Journal, 356, 581–588. Bondy, C., Matura, L. A., Wooten, N., Troendle, J., Zinn, A., & Bakalov, V. K. (2007). The physical phenotype of girls and women with Turner syndrome is not X-imprinted. Human Genetics, 121, 469–474. Boraston, Z., Blakemore, S. J., Chilvers, R., & Skuse, D. (2007). Impaired sadness recognition is linked to social interaction deficit in autism. Neuropsychologia, 45, 1501–1510. Boycott, K. M., Parslow, M. I., Ross, J. L., Miller, I. P., Bech-Hansen, N. T., & MacLeod, P. M. (2003). A familial contiguous gene deletion syndrome at Xp22.3 characterized by severe learning disabilities and ADHD. American Journal of Medical Genetics. Part A, 122, 139–147. Brown, C. J., & Greally, J. M. (2003). A stain upon the silence: genes escaping X inactivation. Trends in Genetics, 19, 432–438. Buchanan, L., Pavlovic, J., & Rovet, J. (1998). A re-examination of the visuospatial deficit in Turner Syndrome: contributions of working memory. Developmental Neuropsychology, 14, 341–367. Burgoyne, P. S., & Baker, T. G. (1985). Perinatal oocyte loss in XO mice and its implications for the aetiology of gonadal dysgenesis in XO women. Journal of Reproduction and Fertility, 75, 633–645. Burgoyne, P. S., & Evans, E. P. (2000). A high frequency of XO offspring from X(Paf)Y* male mice: evidence that the Paf mutation involves an inversion spanning the X PAR boundary. Cytogenetics and Cell Genetics, 91, 57–61.

16

P.M.Y. Lynn et al.

Burgoyne, P. S., Evans, E. P., & Holland, K. (1983). XO monosomy is associated with reduced birthweight and lowered weight gain in the mouse. Journal of Reproduction and Fertility, 68, 381–385. Burgoyne, P. S., Ojarikre, O. A., & Turner, J. M. (2002). Evidence that postnatal growth retardation in XO mice is due to haploinsufficiency for a non-PAR X gene. Cytogenetic and Genome Research, 99, 252–256. Cahill, L. (2006). Why sex matters for neuroscience. Nature Reviews. Neuroscience, 7, 477–484. Carrel, L., Hunt, P., & Willard, H. F. (1996). Tissue and lineage-specific variation in inactive X ­chromosome expression of the murine Smcx gene. Human Molecular Genetics, 5, 1361–1366. Carrel, L., & Willard, H. F. (2005). X-inactivation profile reveals extensive variability in X-linked gene expression in females. Nature, 434, 400–404. Chu, C. E., Donaldson, M. D., Kelnar, C. J., Smail, P. J., Greene, S. A., Paterson, W. F., et  al. (1994). Possible role of imprinting in the Turner phenotype. Journal of Medical Genetics, 31, 840–842. Cutter, W. J., Daly, E. M., Robertson, D. M. W., Chitnis, X. A., van Amelsvoort, T. A. M. J., Simmons, A., et al. (2006). Influence of X chromosome and hormones on human brain development: A magnetic resonance imaging and proton magnetic resonance spectroscopy study of Turner syndrome. Biological Psychiatry, 59, 273–283. Davies, W., Humby, T., Isles, A. R., Burgoyne, P. S., & Wilkinson, L. S. (2007). X-monosomy effects on visuospatial attention in mice: A candidate gene and implications for Turner syndrome and atrtention deficit hyperactivity disorder. Biological Psychiatry, 61, 1351–1360. Davies, W., Isles, A. R., Burgoyne, P. S., & Wilkinson, L. S. (2006). X-linked imprinting: Effects on brain and behaviour. Bioessays, 28, 35–44. Davies, W., Isles, A. R., Humby, T., & Wilkinson, L. S. (2006). What are imprinted genes doing in the brain? In J. F. Wilkins (Ed.), Genomic imprinting (pp. 62–70). USA: Landes Bioscience. Davies, W., Isles, A. R., Smith, R., Karunadasa, D., Burrmann, D., Humby, T., et al. (2005). Xlr3b is a new imprinted candidate for X-linked parent-of-origin effects on cognitive function in mice. Nature Genetics, 37, 625–629. Davies, W., & Wilkinson, L. S. (2006). It is not all hormones: alternative explanations for sexual differentiation of the brain. Brain Research, 1126, 36–45. Dewing, P., Chiang, C. W. K., Sinchak, K., Sim, H., Fernagut, P.-O., Kelly, S., et al. (2006). Direct regulation of adult brain function by the male-specific factor SRY. Current Biology, 16, 415–420. Donnelly, S. L., Wolpert, C. M., Menold, M. M., Bass, M. P., Gilbert, J. R., Cuccaro, M. L., et al. (2000). Female with autistic disorder and monosomy X (Turner syndrome): parent-of-origin effect of the X chromosome. American Journal of Medical Genetics, 96, 312–316. Evans, E. P., & Phillips, R. J. (1975). Inversion heterozygosity and the origin of XO daughters of Bpa/+female mice. Nature, 256, 40–41. Falleti, M. G., Maruff, P., Burman, P., & Harris, A. (2006). The effects of growth hormone (GH) deficiency and GH replacement on cognitive performance in adults: a meta-analysis of the current literature. Psychoneuroendocrinology, 31, 681–691. Good, C. D., Lawrence, K., Thomas, N. S., Price, C. J., Ashburner, J., Friston, K. J., et al. (2003). Dosage-sensitive X-linked locus influences the development of amygdala and orbitofrontal cortex, and fear recognition in humans. Brain, 126, 2431–2446. Greenfield, A., Carrel, L., Pennisi, D., Philippe, C., Quaderi, N., Siggers, P., et al. (1998). The UTX gene escapes X inactivation in mice and humans. Human Molecular Genetics, 7, 737–742. Gu, W., Sander, T., Heils, A., Lenzen, K. P., & Steinlein, O. K. (2005). A new EF-hand containing gene EFHC2 on Xp11.4: tentative evidence for association with juvenile myoclonic epilepsy. Epilepsy Research, 66, 91–98. Haberecht, M. F., Menon, V., Warsofsky, I. S., White, C. D., Dyer-Friedman, J., Glover, G. H., et al. (2001). Functional neuroanatomy of visuo-spatial working memory in Turner syndrome. Human Brain Mapping, 14, 96–107. Hamelin, C. E., Anglin, G., Quigley, C. A., & Deal, C. L. (2006). Genomic imprinting in Turner syndrome: Effects on response to growth hormones and on risk of sensorineural hearing loss. The Journal of Clinical Endocrinology and Metabolism, 91, 3002–3010.

The Genomics of Turner Syndrome and Sex-Biased Neuropsychiatric Disorders

17

Henn, W., & Zang, K. D. (1997). Mosaicism in Turner’s syndrome. Nature, 390, 569. Holden, C. (2005). Sex and the suffering brain. Science, 308, 1574. Huber, R., Hansen, R. S., Strazzullo, M., Pengue, G., Mazzarella, R., D’Urso, M., et al. (1999). DNA methylation in transcriptional repression of two differentially expressed X-linked genes, GPC3 and SYBL1. Proceedings of the National Academy of Sciences of the United States of America, 96, 616–621. Hultcrantz, M., Stenberg, A. E., Fransson, A., & Canlon, B. (2000). Characterization of hearing in an X, 0 ‘Turner mouse’. Hearing Research, 143, 182–188. Humphreys, K., Minshew, N., Leonard, G. L., & Behrmann, M. (2007). A fine-grained analysis of facial expression processing in high-functioning adults with autism. Neuropsychologia, 45, 685–695. Im, K., Lee, J. M., Lee, J., Shin, Y. W., Kim, I. Y., Kwon, J. S., et al. (2006). Gender difference analysis of cortical thickness in healthy young adults with surface-based methods. Neuroimage, 31, 31–38. Isles, A. R., Davies, W., Burrmann, D., Burgoyne, P. S., & Wilkinson, L. S. (2004). Effects on fear reactivity in XO mice are due to haploinsufficiency of a non-PAR X gene: implications for emotional function in Turner’s syndrome. Human Molecular Genetics, 13, 1849–1855. Jamain, S., Quach, H., Betancur, C., Råstam, M., Colineaux, C., Gillberg, I. C., et  al. (2003). Mutations of the X-linked genes encoding neuroligins NLGN3 and NLGN4 are associated with autism. Nature Genetics, 34, 27–29. Jobling, M., & Tyler-Smith, C. (2003). The human Y chromosome: an evolutionary marker comes to age. Nature Genetics, 4, 598–612. Kesler, S. R., Blasey, C. M., Brown, W. E., Yankowitz, J., Zeng, C. M., Bender, B. G., et al. (2003). Effects of X-monosomy and X-linked imprinting on superior temporal gyrus morphology in Turner syndrome. Biological Psychiatry, 54, 636–646. Kesler, S. R., Garrett, A., Bender, B. G., Yankowitz, J., Zeng, S. M., & Reiss, A. L. (2004). Amygdala and hippocampal volumes in Turner syndrome: a high-resolution MRI study of X-monosomy. Neuropsychologia, 42, 1971–1978. Lagrou, K., Froidecoeur, C., Verlinde, F., Craen, M., De Schepper, J., François, I., et al. (2006). Psychosocial functioning, self-perception and body image and their auxologic correlates in growth hormone and oestrogen-treated young adult women with Turner syndrome. Hormone Research, 66, 277–284. Laumonnier, F., Bonnet-Brilhault, F., Gomot, M., Blanc, R., David, A., Moizard, M. P., et  al. (2004). X-linked mental retardation and autism are associated with a mutation in the NLGN4 gene, a member of the neuroligin family. American Journal of Human Genetics, 74, 552–557. Lawrence, K., Jones, A., Oreland, L., Spektor, D., Mandy, W., Campbell, R., et al. (2007). The development of mental state attributions in women with X-monosomy, and the role of monoamine oxidase B in the sociocognitive phenotype. Cognition, 102, 84–100. Lawrence, K., Kuntsi, J., Coleman, M., Campbell, R., & Skuse, D. H. (2003). Face and emotion recognition deficits in Turner syndrome: A possible role for X-linked genes in amygdala development. Neuropsychology, 17, 39–49. Li, C. S., Huang, C., Constable, R. T., & Sinha, R. (2006). Gender differences in the neural correlates of response inhibition during a stop signal task. Neuroimage, 32, 1918–1929. Lonardo, F., Parenti, G., Luquetti, D. V., Annunziata, I., Della Monica, M., Perone, L., et  al. (2007). Contiguous gene syndrome due to an interstitial deletion in Xp22.3 in a boy with ichthyosis, chondrodysplasia punctata, mental retardation and ADHD. European Journal of Medical Genetics, 50, 301–308. Loring-Meier, S., & Halpern, D. F. (1999). Sex differences in visuospatial working memory: components of cognitive processing. Psychonomic Bulletin & Review, 6, 464–471. Luedi, P. P., Dietrich, F. S., Weidman, J. R., Bosko, J. M., Jirtle, R. L., & Hartemink, A. J. (2007). Computational and experimental identification of novel human imprinted genes. Genome Research, 17(12), 1723–1730. Luoh, S. W., Bain, P. A., Polakiewicz, R. D., Goodheart, M. L., Gardner, H., Jaenisch, R., et al. (1997). Zfx mutation results in small animal size and reduced germ cell number in male and female mice. Development, 124, 2275–2284.

18

P.M.Y. Lynn et al.

Lynn, P. M., & Davies, W. (2007). The 39, XO mouse as a model for the neurobiology of Turner syndrome and sex-biased neuropsychiatric disorders. Behavioural Brain Research, 79, 173–182. Martinez-Garay, I., Jablonka, S., Sutajova, M., Steuernagel, P., Gal, A., & Kutsche, K. (2002). A new gene family (FAM9) of low-copy repeats in Xp22.3 expressed exclusively in testis: implications for recombinations in this region. Genomics, 80, 259–267. Mathur, A., Stekol, L., Schatz, D., MacLaren, N. K., Scott, M. L., & Lippe, B. (1991). The parental origin of the single X chromosome in Turner syndrome: lack of correlation with parental age or clinical phenotype. American Journal of Human Genetics, 48, 682–686. Mayer, A., Lahr, G., Swaab, D. F., Pilgrim, C., & Reisert, I. (1998). The Y-chromosomal genes SRY and ZFY are transcribed in adult human brain. Neurogenetics, 1, 281–288. Mazzocco, M. M. M. (1998). A process approach to describing mathematics difficulties in girls with Turner syndrome. Pediatrics, 102, 492–496. McCauley, E., Feuillan, P., Kushner, H., & Ross, J. L. (2001). Psychosocial development in adolescents with Turner syndrome. Journal of Developmental and Behavioral Pediatrics, 22, 360–365. Milsted, A., Serova, L., Sabban, E. L., Dunphy, G., Turner, M. E., & Ely, D. L. (2004). Regulation of tyrosine hydroxylase gene transcription by Sry. Neuroscience Letters, 369, 203–207. Moscovitch, M., Nadel, L., Winocur, G., Gilboa, A., & Rosenbaum, R. (2006). The cognitive neuroscience of remote episodic, semantic and spatial memory. Current Opinion in Neurobiology, 16, 179–190. Munro, C. A., McCaul, M. E., Wong, D. F., Oswald, L. M., Zhou, Y., Brasic, J., et al. (2006). Sex differences in striatal dopamine release in healthy adults. Biological Psychiatry, 59, 966–974. Murphy, D. G. M., Mentis, M. J., Pietrini, P., Grady, C., Daly, E. M., Haxby, J. V., et al. (1997). A PET study of Turner’s Syndrome: Effects of sex steroids and the X chromosome on brain. Biological Psychiatry, 41, 285–298. Nijhuis-van der Sanden, M. W. G., Eling, P. A. T. M., & Otten, B. J. (2003). A review of neuropsychological and motor studies in Turner syndrome. Neurosci Biobehav Rev, 27, 329–338. Nordquist, N., & Oreland, L. (2006). Monoallelic expression of MAOA in skin fibroblasts. Biochemical and Biophysical Research Communications, 348, 763–767. Prior, T. I., Chue, P. S., & Tibbo, P. (2000). Investigation of Turner syndrome in schizophrenia. American Journal of Medical Genetics, 96, 373–378. Raefski, A. S., Carone, B. R., Kaur, A., Krueger, W., & O’Neill, M. J. (2007). Wnt pathway anomalies in developing amygdalae of Turner syndrome-like mice. Journal of Molecular Neuroscience, 32, 111–119. Raefski, A. S., & O’Neill, M. J. (2005). Identification of a cluster of X-linked imprinted genes in mice. Nature Genetics, 37, 620–624. Romans, S. M., Stefanatos, G., Roeltgen, D., Kushner, H., & Ross, J. (1998). Transition to young adulthood in Ullrich-Turner syndrome: Neurodevelopmental changes. American Journal of Medical Genetics, 79, 140–147. Ross, J., Roeltgen, D., Kushner, H., Wei, F., & Zinn, A. (2000). The Turner syndrome-associated neurocognitive phenotype maps to distal Xp. American Journal of Human Genetics, 67, 672–681. Ross, J. L., Roeltgen, D., Stefanatos, G. A., Feuillan, P., Kushner, H., Bondy, C., et al. (2003). Androgen-responsive aspects of cognition in girls with Turner syndrome. The Journal of Clinical Endocrinology and Metabolism, 88, 292–296. Ross, J., Roeltgen, D., & Zinn, A. (2006). Cognition and the sex chromosomes: Studies in Turner syndrome. Hormone Research, 65, 47–56. Ross, J. L., Scott, C. J., Marttila, P., Kowal, K., Nass, A., Papenhausen, P., et al. (2001). Phenotypes associated with SHOX deficiency. The Journal of Clinical Endocrinology and Metabolism, 86, 5674–5680. Ross, J., Stefanatos, G. A., Kushner, H., Zinn, A., Bondy, C., & Roeltgen, D. (2002). Persistent cognitive deficits in adult women with Turner syndrome. Neurology, 58, 218–225. Rovet, J. (2004). Turner syndrome: Genetic and hormonal factors contributing to a specific learning disability profile. Learning Disabilities Research & Practice, 19, 133–145. Rovet, J., & Ireland, L. (1994). Behavioral phenotype in children with Turner syndrome. Journal of Pediatric Psychology, 19, 779–790.

The Genomics of Turner Syndrome and Sex-Biased Neuropsychiatric Disorders

19

Russell, H. F., Wallis, D., Mazzocco, M. M. M., Moshang, T., Zackai, E., Zinn, A. R., et al. (2006). Increased prevalence of ADHD in Turner syndrome with no evidence of imprinting effects. Journal of Pediatric Psychology, 31, 945–955. Sagi, L., Zuckerman-Levin, N., Gawlik, A., Ghizzoni, L., Buyukgebiz, A., Rakover, Y., et  al. (2006). Clinical significance of the parental origin of the X chromosome in Turner syndrome. The Journal of Clinical Endocrinology and Metabolism, 92, 846–852. Schmidt, P. J., Cardoso, G. M. P., Ross, J., Haq, N., Rubinow, D. R., & Bondy, C. (2006). Shyness, social anxiety, and impaired self-esteem in Turner syndrome and premature ovarian failure. JAMA, 295, 1374–1376. Simpson, J. L., & Rajkovic, A. (1999). Ovarian differentiation and gonadal failure. American Journal of Medical Genetics, 89, 186–200. Skuse, D. H. (2005). X-linked genes and mental functioning. Human Molecular Genetics, 14, R27–R32. Skuse, D. H., James, R. S., Bishop, D. V. M., Coppin, B., Dalton, P., Aamodt-Lepper, G., et al. (1997). Evidence from Turner’s syndrome of an imprinted X-linked locus affecting cognitive function. Nature, 387, 705–708. Smith, R. J., & Kelsey, G. (2001). Identification of imprinted loci by methylation: Use of methylation-sensitive representational difference analysis (Me-RDA). Methods in Molecular Biology, 181, 113–132. Staller, J. A., & Faraone, S. V. (2007). Targeting the dopamine system in the treatment of attention-deficit/hyperactivity disorder. Expert Review of Neurotherapeutics, 7, 351–362. Suzuki, T., Delgado-Escueta, A. V., Aguan, K., Alonso, M. E., Shi, J., Hara, Y., et  al. (2004). Mutations in EFHC1 cause juvenile myoclonic epilepsy. Nature Genetics, 36(8), 842–849. Sybert, V. P., & McCauley, E. (2004). Turner’s syndrome. The New England Journal of Medicine, 351, 1227–1238. Temple, C. M., Carney, R. A., & Mullarkey, S. (1996). Frontal lobe function and executive skills in children with Turner’s syndrome. Developmental Neuropsychology, 12, 343–363. Thapar, A., Fowler, T., Rice, F., Scourfield, J., van den Bree, M., Thomas, H., et  al. (2003). Maternal smoking during pregnancy and attention deficit hyperactivity disorder symptoms in offspring. The American Journal of Psychiatry, 160, 1985–1989. Thapar, A., Langley, K., Fowler, T. A., Rice, F., Turic, D., Whittinger, N., et  al. (2005). Catechol- O-methyltransferase gene variant and birth weight predict early onset antisocial behavior in children with Attention Deficit Hyperactivity Disorder. Archives of General Psychiatry, 62, 1275–1278. Thayer, J. F., & Johnsen, B. H. (2000). Sex differences in judgement of facial affect: a multivariate analysis of recognition errors. Scandinavian Journal of Psychology, 41, 243–246. Toda, M., & Abi-Dargham, A. (2007). Dopamine hypothesis of schizophrenia: making sense of it all. Current Psychiatry Reports, 9, 329–336. Uematsu, A., Yorifuji, T., Muroi, J., Kawai, M., Mamada, M., Kaji, M., et al. (2002). Parental origin of normal X chromosomes in Turner syndrome patients with various karyotypes: implications for the mechanism leading to generation of a 45, X karyotype. American Journal of Medical Genetics, 111, 134–139. Vawter, M. P., Evans, S., Choudary, P., Tomita, H., Meador-Woodruff, J., Molnar, M., et al. (2004). Gender-specific gene expression in post-mortem human brain: localization to sex chromosomes. Neuropsychopharmacology, 29, 373–384. Vitalis, T., Fouquet, C., Alvarez, C., Seif, I., Price, D., Gaspar, P., et al. (2002). Developmental expression of monoamine oxidases A and B in the central and peripheral nervous systems of the mouse. The Journal of Comparative Neurology, 442, 331–347. Weinhäusel, A., & Haas, O. A. (2001). Evaluation of the fragile X (FRAXA) syndrome with methylation-sensitive PCR. Human Genetics, 108, 450–458. Weiss, L. A., Purcell, S., Waggoner, S., Lawrence, K., Spektor, D., Daly, M. J., et  al. (2007). Identification of EFHC2 as a quantitative trait locus for fear recognition in Turner syndrome. Human Molecular Genetics, 16, 107–113. Wilkinson, L. S., Davies, W., & Isles, A. R. (2007). Genomic imprinting effects on brain development and function. Nature Reviews. Neuroscience, 8, 832–843.

20

P.M.Y. Lynn et al.

Yen, P. H., Marsh, B., Allen, E., Tsai, S. P., Ellison, J., Connolly, L., et al. (1988). The human X-linked steroid sulfatase gene and a Y-encoded pseudogene: Evidence for an inversion of the Y chromosome during primate evolution. Cell, 55, 1123–1135. Ylisaukko-oja, T., Rehnström, K., Auranen, M., Vanhala, R., Alen, R., Kempas, E., et al. (2005). Analysis of four neuroligin genes as candidates for autism. European Journal of Human Genetics, 13, 1285–1292. Zechner, U., Wilda, M., Kehrer-Sawatzki, H., Vogel, W., Fundele, R., & Hameister, H. (2001). A high density of X-linked genes for general cognitive ability: a run-away process shaping human evolution? Trends in Genetics, 17, 697–701. Zinn, A., Kushner, H., & Ross, J. L. (2008). EFHC2 SNP rs7055196 is not associated with fear recognition in 45, X Turner syndrome. American Journal of Medical Genetics. Part B, 147, 507–509. Zinn, A., Roeltgen, D., Stefanatos, G., Ramos, R., Elder, F., Kushner, H., et al. (2007). A Turner syndrome neurocognitive phenotype maps to Xp22.3. Behavioral and Brain Functions, 3, 24. Zinn, A., Tonk, V. S., Chen, Z., Flejter, W. L., Gardner, H. A., Guerra, R., et al. (1998). Evidence for a Turner syndrome locus or loci at Xp11.2-p22.1. American Journal of Human Genetics, 63, 1757–1766.

Mental Retardation and Human Chromosome 21 Gene Overdosage: From Functional Genomics and Molecular Mechanisms Towards Prevention and Treatment of the Neuropathogenesis of Down Syndrome Mohammed Rachidi and Carmela Lopes Abstract  Down syndrome (DS), caused by a genomic imbalance of human chromosome 21 (HSA21), is mainly observed as trisomy 21 and is the major genetic cause of mental retardation (MR). MR and associated neurological and behavioural alterations result from dysregulation in critical HSA21 genes and associated molecular pathways. Gene expression, transcriptome, proteome and functional genomics studies, in human, trisomic and transgenic mouse models have shown similar genotype/phenotype correlation and parallel outcomes suggesting that the same evolutionarily conserved genetic programmes are perturbed by gene-dosage effects. The expression variations caused by this gene-dosage imbalance may firstly induce brain functional variations at cellular level, as primary phenotypes, and finally induce neuromorphological alterations and cognitive deficits as secondary phenotypes. The identification of trisomic genes overexpressed in the brain and their function, their developmental regulated expression and their downstream effects, their interaction with other proteins, and their involvement in regulatory and metabolic pathways is giving a clearer view of the origin of the MR in DS. This led to the identification of potential targets in the altered molecular pathways involved in MR pathogenesis, such as calcineurin, NFATs and MAPK pathways, that may be potentially corrected, in the perspective of new therapeutic approaches. Treatment of DS mouse models with NMDA receptor or GABAA antagonists allowed post-drug rescue of cognitive deficits. Besides these new pharmacotherapies, the regulation of gene expression by microRNAs or small interfering RNAs provide exciting possibilities for exogenous correction of the aberrant gene expression in DS and provide potential directions for clinical therapeutics of MR. Herein, we highlight the genetic networks and molecular mechanisms implicated in the pathogenesis of MR in DS and, thereafter, we outline some of the therapeutic strategies for the treatment of this as yet incurable cognitive disorder with a considerable impact on public health. M. Rachidi (*) EA 3508, Laboratory of Genetic Dysregulation Models: Trisomy 21 and Hyperhomocysteinemia, University of Paris 7-Denis Diderot, Tour 54, E2-54-53, Case 7104, 2 Place Jussieu, Paris 75251, France e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_2, © Springer Science+Business Media, LLC 2011

21

22

M. Rachidi and C. Lopes

Keywords  Trisomy 21 • Mental retardation • Learning and memory • Down syndrome critical region • Genotype-phenotype correlation • Mouse models • Gene-dosage imbalance • Transcriptome • Proteome • MicroRNAs • Gene expression variation • Molecular mechanism model • NFATs/calcineurin pathways • NMDA receptor antagonist • GABAA antagonists • Pharmacotherapy Abbreviations AChEI AD APP ATP BAC BFCN CA1 CA3 CaMKII CBR1 ChAT CIT-K CREB DS DSCAM DSCR DSCR1 DYN1 DYRK1A EGF EPSCs ERG ES ETS2 GABAA GIRK2 HSA21 IQ ITSN1 KCNJ6 LPS LTD LTP MAPK MCIP1 miRNA

Acetylcholinesterase inhibitor Alzheimer’s disease Amyloid precursor protein Adenosine triphosphate Bacterial artificial chromosome Basal forebrain cholinergic neurons Cornu ammonis 1 Cornu ammonis 3 Calcium/calmodulin-dependent protein kinase Carbonyl reductase 1 Choline acetyl transferase Citron kinase c-AMP response element-binding protein Down syndrome Down syndrome cell adhesion molecule Down syndrome critical region Down syndrome critical region gene 1 Dynamin 1 Dual-specificity tyrosine-(Y)-phosphorylation kinase 1A Epidermal growth factor Excitatory postsynaptic currents Ets related gene Embryonic stem cells v-ets erythroblastosis virus E26 oncogene homolog 2 Gamma-aminobutyric acid type A receptor G-protein coupled inward rectifying potassium channel subunit 2 Human chromosome 21 Intelligence quotient Intersectin gene 1 Potassium inwardly rectifying channel J6 Lipopolysaccharide Long-term depression Long-term potentiation Mitogen activated protein kinase Myocyte-enriched calcineurin-interacting protein 1 MicroRNA

Mental Retardation and Human Chromosome 21 Gene Overdosage

MMU16 MR NFATc NGF NMDA NMDA-R PP1 PTZ qRT-PCR RCAN1 S100B SAGE SHH SIM2 SNP SOD1 SYNJ1 TBS TPRD TTC3 YAC

23

Mouse chromosome 16 Mental retardation Nuclear factor of activated T cells Nerve growth factor N-methyl-d-aspartate N-methyl-d-aspartate receptor Protein phosphatase 1 Pentylenetetrazol Quantitative reverse transcriptase polymerase chain reaction Regulator of calcineurin 1 protein S100 calcium-binding protein beta Serial analysis of gene expression Sonic hedgehog Single minded 2 Single nucleotide polymorphism Superoxide dismutase 1 Synaptojanin gene 1 Theta-burst stimulation Tetratricopeptide repeat domain Down syndrome Tetratricopeptide repeat domain 3 Yeast artificial chromosome

1 Mental Retardation in Down Syndrome: An Invalidating Neuropathological Aspect with Hard Impact on Public Health Trisomy of human chromosome 21 (HSA21) is the most frequent genetic cause of mental retardation (MR) and other phenotypic abnormalities, including heart defects, cranio-facial abnormalities, cognitive impairment and Alzheimer’s disease (AD), collectively known as Down syndrome (DS) and affecting 1 in 700 live births (Roizen & Patterson, 2003). While the clinical phenotypes of each DS individual are variable in trait number and intensity, the MR remains the invariable hallmark disorder of DS and the most invalidating pathological aspect contributing to about 30% of all moderate-to-severe cases of MR (Lejeune, 1990; Pulsifer, 1996; Stoll, Alembik, Dott, & Roth, 1990). Early infants show delayed cognitive development, leading to mild–moderate MR and decrease of the intelligence quotient (IQ) from early in the first year to late childhood (Brown, Greer, Aylward, & Hunt, 1990; Hodapp, Ewans, & Gray, 1999). DS patients have difficulties in both learning and memory. Moreover, the learning can be complicated by avoidance strategies when faced with cognitive challenges (Wishart, 1995). Although all domains of development follow the usual sequence, a deficiency in language production relative to other areas of development often causes ­substantial  impairment (Chapman, Seung, Schwartz, & ­Kay-Raining  Bird,  1998).

24

M. Rachidi and C. Lopes

The ­impairment of prefrontal cortex and cerebellar function (Nadel, 2003), speech and articulation are also particularly affected. The lower performances of DS in linguistic tasks may be partially explained in terms of impairment of the frontocerebellar structures involved in articulation and verbal working memory (Fabbro, Alberti, Gagliardi, & Borgatti, 2002). In adult life, the IQ of DS patients persists at low levels (30–70) and also undergoes a decline in cognitive performance (Chapman & Hesketh, 2000; Vicari, 2004, 2006) that has been interpreted as the consequence of accelerated ageing in DS (Devenny et  al., 1996; Lott & Head, 2005). In addition, an early onset of an Alzheimer disease-like neurohistopathology is systematically observed by the fourth decade (Dalton & Crapper-McLachlan, 1986). DS children have more behavioural and psychiatric problems than in other children, but fewer than in other individuals with MR. Adult DS patients can have a similar prevalence of psychiatric problems to other people with intellectual disability. A raised frequency of psychiatric problems is also related to the increased prevalence of depression in people with DS. However, they seem protected from some psychiatric disorders such as personality disorder, schizophrenia and anxiety (Collacot, Cooper, Branford, & McGrother, 1998). On the other hand, DS children show continuous but gradual improvement in mental age throughout childhood; IQs generally decline from early in the first year to late childhood (Hodapp & Zigler, 1990). Improvements in cognitive abilities and in quality of life of individuals with DS have resulted from improvements in medical care, identification and treatment of psychiatric disorders (such as depression, autism, and disruptive behaviour disorders) and early implementation of special educational programmes and interventions with typical educational settings (Connolly, Morgan, Russell, & Fulliton, 1993).

2 Mental Retardation in Down Syndrome: A Consequence of Developmental and Functional Brain Alterations Individuals wih DS have a functionally abnormal brain with developmental alterations in morphogenesis and histogenesis. The brain of DS subjects is characterised by several postmortem macroscopic features that are related to pre- and post-natal abnormalities leading to retardation of brain growth (Schmidt-Sidor, Wisniewski, Shepard, & Sersen, 1990). Infants and children with DS have delayed brain maturation, retardation of growth and delayed and disorganised second phase of cortical development and lamination emergence, cortical dysgenesis, delayed myelination, fewer neurons and lower neuronal density, abnormal synaptic connection (Wisniewski, 1990; Wisniewski & Schmidt-Sidor, 1989), shortened basilar dendrites, decreased number of spines with altered morphology, and defective cortical layering in several cortical areas (Becker, Armstrong, & Chan, 1986; Golden & Hyman, 1994; Marin-Padilla, 1976; Schmidt-Sidor et al., 1990; Takashima, Becker, Armstrong, & Chan, 1981; Takashima, Iida, Mito, & Arima, 1994).

Mental Retardation and Human Chromosome 21 Gene Overdosage

25

Overall, brain volume is reduced in DS subjects (Becker, Mito, Takashima, & Onodera, 1991; Coyle, Oster-Granite, & Gearhart, 1986; Wisniewski, 1990), including cerebellar and cerebral grey and white matter (Kesslak, Nagata, Lott, & Nalcioglu, 1994; Pearlson et al., 1998; Pinter, Eliez, Schmitt, Capone, & Reiss, 2001; Raz et al., 1995; Schapiro, Haxby, & Grady, 1992; Schapiro, Luxenberg, Kaye, Haxby, & Friedland, 1989; Weis, Weber, Neuhold, & Rett, 1991). In particular, the reduced cerebellum shows a decreased volume of lobules VI to VIII (Ayraham, Sugarman, Rotshenker, & Groner, 1991; Jernigan, Bellugi, Sowell, Doherty, & Hesselink, 1993; Raz et al., 1995). Hippocampus volume is disproportionally reduced (Aylward et al., 1999; Kesslak et  al., 1994; Krasuski, Alexander, Horwitz, Rapoport, & Schapiro, 2002; Pearlson et al., 1998; Pinter, Eliez et al., 2001; Raz et al., 1995), in particular at the level of the corpus callosum (Lai & Williams, 1989; Wang, Doherty, Hesselink, & Bellugi, 1992). The anterior cortex, including frontal and anterior temporal lobes, also appears reduced after adjustment for total cerebral grey matter volume (Jernigan et al., 1993; Lai & Williams, 1989; Teipel et al., 2004), whereas amygdala volume reductions do not exceed the overall reduction of brain size (Aylward et al., 1999; Pinter, Brown et al., 2001; Pinter, Eliez et al., 2001). On the other hand, an increased volume is found in other brain areas, such as ventricles (Ikeda & Arai, 2002; Kesslak et  al., 1994; Pearlson et  al., 1998; Schimmel, Hammerman, Bromiker, & Berger, 2006), parahippocampal gyrus after adjustment for overall brain volume (Kesslak et al., 1994; Raz et al., 1995; Teipel & Hampel, 2006; Teipel et al., 2003), temporal, parietal and posterior cortex, lenticular nucleus and thalamus and hypothalamus (Jernigan et al., 1993; Pinter, Eliez et al., 2001), while the occipital lobe and superior temporal gyrus do not show volume changes after adjustment for overall brain volume (Frangou et al., 1997; Pinter, Eliez et al., 2001). In addition, DS brains are characterised by several neurological defects in cortex lamination (Golden & Hyman, 1994) and in cerebellar foliation (Raz et  al., 1995). Morphological and functional defects have also been found at the cellular level determined by alteration in neurogenesis, neuronal differentiation, myelination, dendritogenesis and synaptogenesis (Becker et  al., 1986, 1991; Coyle et  al., 1986; Dierssen & Ramakers, 2006; Huttenlocher, 1974; Marin-Padilla, 1972, 1976; Purpura, 1974; Takashima, Ieshima, Nakamura, & Becker, 1989; Takashima et  al., 1994; Vuksic, Petanjek, Rasin, & Kostovic, 2002; Wisniewski, 1990; Wisniewski & Schmidt-Sidor, 1989). Biochemical alterations also occur in foetal DS brain, which could serve as substrates for the morphological changes (Engidawork & Lubec, 2003), involving a decrease of the choline acetyltransferase and histidine decarboxylase activities and also a decrease of serotonin, histamine, and glutamate levels (Risser, Lubec, Cairns, & Herrera-Marschitz, 1997; Schneider et al., 1997; Wisniewski & Bobinski, 1991). Taken as a whole, these alterations observed in the brain of DS, in particular those in the key regions involved in learning and memory processes, could be the origin of MR (Black, Nadel, & O’Keefe, 1977; Funahashi, Takeda, & Watanabe, 2004; Milner, Squire, & Kandel, 1998; Nadel & Willner, 1980). In addition, although young children with DS appear to be born with a normal septohippocampal cholinergic system (Kish et  al., 1989), an ageing-dependent neurodegeneration of the basal forebrain cholinergic neurons (BFCN) was observed (Casanova, Walker, Whitehouse,

26

M. Rachidi and C. Lopes

& Price, 1985; Yates et al., 1983). Because BFCN provide the major cholinergic input to the hippocampus and neocortex, the degeneration of these neurons may have functional consequences at the level of cholinergic receptors. These dysfunctions could produce additional learning and memory deficits in older individuals with DS (Yates et al., 1983) and could be an outgrowth of AD in these patients. In addition, an early onset of an Alzheimer disease-like neurohistopathology is systematically observed by the fourth decade (Dalton & Crapper-McLachlan, 1986). The short- and long-term memory deficits observed in DS patients (Brown et al., 2003; Clark & Wilson, 2003; Hodapp et  al., 1999; Hulme & Mackenzie, 1992) provide behavioural evidence of hippocampal dysfunction by adolescence (Carlesimo, Marotta, & Vicari, 1997). The spatial learning, also depending on the hippocampus, is particularly affected and there is also evidence for impairment of prefrontal cortex and cerebellar function (Nadel, 2003). In addition to the known effects of the hippocampal formation in spatial memory, the altered cortical layer and cerebellum also may participate to cognitive and behavioural phenotypes in DS (Funahashi et al., 2004). Overall, the MR, the major neurological disorder of DS, is mainly a consequence of functional and developmental brain alterations in neurogenesis, neuronal differentiation, myelination, dendritogenesis and synaptogenesis.

3 Mental Retardation in Down Syndrome: A Consequence of Chromosome 21 Gene Overdosage In the most cases, DS results from the trisomy of the HSA21 in all cells of the afflicted individuals (LeJeune, Gautier, & Turpin, 1959) and the generally accepted molecular origin of DS is the chromosomal imbalance associated to the HSA21 triplication and thus the overdosage of HSA21 genes that could be responsible for the phenotype seen in DS patients (Antonarakis, 1998). In some rare cases, no more than 1% of living trisomic patients, DS results from a partial trisomy 21 showing variable phenotypes depending of the extra copy of the triplicated region. Clinical, cytogenetic and molecular analysis of such patients allowed narrowing a region of HSA21, called Down syndrome critical region (DSCR), localised on the distal part of the long arm, around the marker D21S55, and flanked by D21S17 and Ets related gene (ERG) (Delabar et al., 1993; Korenberg et al., 1994; Rahmani et al., 1989). The chromosomal imbalance due to the extra copy of DSCR is associated with the expression of many features of the disease and contributes significantly, but not exclusively, to MR (Delabar et al., 1993; Korenberg et al., 1994; Rahmani et al., 1989).

3.1 Chromosomal Imbalance Effects on Mental Retardation The DSCR, containing the genes located between the carbonyl reductase 1 (CBR1) and the transcriptional regulator Ets-related gene (ERG) loci (Fig.  1) has been

27

Mental Retardation and Human Chromosome 21 Gene Overdosage ES#21

D2lS5

Tc1

21P Dp(16)1Yu

LIPl CXADR

q21.1

D2lSL922 MRPL39

q21.2

Ms1Ts65

Ts65Dn

Ms1Rhr/ Ts65Dn Ts2Cje

APP q21.3 Ts1Cje

SODl SYNJl ITSNl DSCRl SIM2 DYRKl A

ETS2

q22.11 IFNARl RUNXl CBRl CLDNl4 TTC3 GIRK2 MX1 ZNF295

q22.12

230E8

q22.2

Ts1Rhr

152F7

q22.13 141G6

285E6

q22.3

Sl00B

Fig. 1  Down syndrome mouse models. The boundary localisations of the HSA21 or its mouse syntenic triplicated regions in the DS mouse models are indicated on the right of the HSA21. Black lines indicate partial trisomic 16 and transchromosomal mice. Grey lines indicate segmental transgenic mice carrying human YACs, containing the DS critical region (DSCR). Monogenic transgenic mice are indicated on the left of the HSA21

designated as the DS critical region that, when duplicated, is associated with multiple neurological features of DS, including MR (Delabar et al., 1993; Korenberg et al., 1994; Toyoda et al., 2002). Consequently, the major phenotypes of DS, particularly MR, may have their origin in the over-dosage of genes located in DSCR. To explain the pathogenesis of DS from the genetic over-dosage, two genetic hypotheses have been considered. 3.1.1 Dosage-Sensitive Gene Hypothesis This genetic hypothesis holds that the phenotype is a direct result of the cumulative effects of the dosage imbalance of the individual genes located on the triplicated HSA21 or critical region DSCR (Epstein, 1986, 1990; Korenberg et  al., 1990). According to this “dosage-sensitive gene” hypothesis, a subset of genes on the triplicated HSA21 is directly responsible for particular pathological traits associated with trisomy 21. Consequently, the DSCR was defined as a minimal interval of the HSA21 that carries the dosage-sensitive genes necessary and sufficient for typical features of DS individuals (Delabar et al., 1993; Korenberg et al., 1994; Rahmani

28

M. Rachidi and C. Lopes

et al., 1989). Several phenotypes of DS have also been found in transgenic mice engineered to overexpress HSA21 genes or their mouse orthologs. The observations of cardiac pathology, craniofacial dysmorphology, malformation of cerebellum and overall the deficit of cognitive functions both in DS individuals and in mouse models were in agreement with the “dosage-sensitive gene” hypothesis. 3.1.2 Amplified Developmental Instability Hypothesis This genetic hypothesis, in contrast to the preceding hypothesis, states that dosage imbalance of the hundreds of genes on HSA21 determines a non-specific disturbance of genomic regulation and expression. This global disruption of the correct balance of gene expression in development pathways alters the normal developmental homeostasis and determines most manifestations of DS (Pritchard & Kola, 1999; Shapiro, 1983; Shapiro & Whither-Azmitia, 1997). In agreement to this hypothesis, the variability of the DS phenotype in the different individuals has also been explained by intervention of stochastic factors during development (Kurnit, Aldridge, Matsuoka, & Matthysse, 1985), which can also be involved in normal development (Kurnit, Layton, & Matthysse, 1987). Moreover, several features observed in DS (for example, AD, cardiac malformations and metabolic diseases) can be observed in other trisomies and in the general population at lower frequency. In addition, the significant increase of the individual variability in DS, as compared to euploid individuals, also supports this hypothesis. Nevertheless, the “amplified developmental instability” hypothesis and the “dosagesensitive genes” hypothesis are not mutually exclusive. It is possible that single genes, or a specific subset of genes, may be involved in specific DS phenotypes, while some other DS phenotypes may be due to a more general disturbance in gene dosage imbalance as a result of the extra chromosomal material (Antonarakis, 2001).

3.2 Gene Dosage Imbalance in Down Syndrome Determine Dysregulation of HSA21 Gene Expression 3.2.1 Primary and Secondary Gene Effects The genetic origin of DS is the overdosage of HSA21 genes that could be responsible for the genesis of the neurological and cognitive defects seen in DS patients (Antonarakis, 1998). The 1.5-fold increase of HAS21 gene dosage may determine a primary effect on gene transcription consisting in a 1.5-fold increase of expression level of these genes (Kurnit, 1979; Mao et al., 2005; Mao, Zielke, Zielke, & Pevsner, 2003). Genes on the trisomic HSA21 encoding transcription factors and other proteins, that can directly or indirectly influence gene expression, may produce a secondary genome-wide transcriptional downstream misregulation, which may consist of

Mental Retardation and Human Chromosome 21 Gene Overdosage

29

gene over-expression different from 1.5-fold, and also in down-regulation of genes on both HSA21 and the other chromosomes. This secondary effect of the chromosomal imbalance could be complex and highly variable in the different cells and during the lifespan (Dauphinot et  al., 2005; Epstein, 1986, 1988; Lyle, Gehrig, Neergaard-Henrichsen, Deutsch, & Antonarakis, 2004; Prandini et al., 2007; Saran, Pletcher, Natale, Cheng, & Reeves, 2003; Sultan et al., 2007). An important application of the secondary gene effects on other chromosomes is widely used in antenatal screening programmes for trisomy 21 by detection of abnormal levels of foetal proteins in maternal serum. In particular, the level of the alpha fetoprotein is reduced (Newby et  al., 1997) while human chorionic gonadotrophin is increased (Aitken et al., 1993) in trisomy 21, although the encoding genes are located on the chromosomes 4 and 19, respectively. The molecular effects of the 1.5 gene overdosage may be even more complex at the protein level, as additional regulatory points are introduced, such as posttranscriptional, translational and post-translational regulations and post-translational modifications. Modification in the levels of proteins involved in multicomplex protein formation, in protein–protein intections and in metabolic and regulatory networks can determine alterations in these interactions because of the loss of the correct ratio among the proteins and, finally, can alter the function or stability of the proteins. In the brain, it can be considered that the mere presence of the chromosomal imbalance determines misexpression and interaction of crucial genes/proteins involved in neuromorphogenesis and neurogenic cascades. The developmental errors caused by trisomy 21 during neural patterning and signal transduction pathways may lead to defective neuronal circuitry and could be the biological mechanism responsible for the pathogenesis of MR in DS. 3.2.2 Transcriptional Variation as a Consequence of Trisomy 21 DS may be considered as a multifactorial disease with an unusual aetiology characterised by overdosage of HSA21 genes determining gene expression variation that can be responsible for the complex DS phenotype. Thus, expression studies in normal and trisomic tissues contribute to understanding the role of the HSA21 genes and the contribution of their dosage alterations in DS pathogenesis, allowing the selection of HSA21 genes potentially involved in a given DS phenotype. In particular, a gene expressed in developing and/or adult brain may be selected as a candidate gene for MR, particularly when its transcription is restricted to key regions for cognitive functions, such as the hippocampal formation, the cortex and the cerebellum. DNA sequencing of HSA21 (Hattori et al., 2000; International Human Genome Sequencing Consortium, 2004) and gene annotation improve identification of the gene products. These sequencing data and the development of molecular analysis tools allow large-scale application of gene expression analysis. Transcriptome studies are performed by quantitative RT-PCR (qRT-PCR), microarrays and serial

30

M. Rachidi and C. Lopes

analysis of gene expression (SAGE), to study gene expression variation in trisomic tissues compared to the euploid ones using human tissues and cell lines (Aït ­Yahya-Graison et al., 2007; Chou et al., 2008; Deutsch et al., 2005; FitzPatrick et al., 2002; Giannone et al., 2004; Li et al., 2006; Malago et al., 2005; Mao et al., 2003, 2005; Prandini et  al., 2007), or mouse trisomic model tissues (Amano et  al.,  2004; Chrast, Scott, Madani et  al., 2000; Dauphinot et  al., 2005; Kahlem et al., 2004; Lyle et al., 2004; Potier et al., 2006; Saran et al., 2003; Sultan et al., 2007; Wang et  al., 2004). Transcriptome studies of the brain, the cerebellum or neuronal cell lines are particularly abundant, reflecting the major interest in the understanding of the molecular mechanisms involved in MR pathogenesis in DS (Amano et  al., 2004; Chrast, Scott, Madani et  al., 2000; Dauphinot et  al., 2005; Kahlem et al., 2004; Lyle et al., 2004; Mao et al., 2003, 2005; Potier et al., 2006; Saran et al., 2003; Sultan et al., 2007; Wang et al., 2004). Theoretically, the supernumerary copy of HSA21 is expected to result in a 50% increase in the level of transcripts of all genes mapping to HSA21. Most of these works confirm that transcript levels are elevated about 1.5-fold for the majority of trisomic genes in human trisomic tissues and across a broad range of tissues of trisomic mouse models (Aït Yahya-Graison et al., 2007; Amano et al., 2004; Dauphinot et al., 2005; Epstein, 1986; FitzPatrick et al., 2002; Kahlem et al., 2004; Lyle et al., 2004; Mao et al., 2003, 2005; Potier et al., 2006; Prandini et al., 2007; Sultan et al., 2007; Wang et al., 2004). These results indicate that the triplicated genes are overexpressed in a dosage-dependent manner, supporting the hypothesis that a global HSA21 dosage imbalance causes the heterogeneous phenotypes of DS (Shapiro, 1983, 1997). It cannot be excluded that the overexpression of a limited number of genes on HSA21 is responsible for the DS phenotypic features (Korenberg et al., 1990). In addition, in several studies, it was also found that there is not always a direct correlation between genomic imbalance and not all genes are overexpressed ~1.5fold compared to euploid (Aït Yahya-Graison et al., 2007; Dauphinot et al., 2005; Kahlem et al., 2004; Lyle et al., 2004; Potier et al., 2006; Saran et al., 2003), and a decreased expression is also found for some genes, such as GRIK1 (Saran et al., 2003), Ets2 (Engidawork & Lubec, 2003; Greber-Platzer, Schatzmann-Turhani, Cairns, Balcz, & Lubec, 1999; Wang et  al., 2004), superoxide dismutase 1 (SOD1) (Engidawork & Lubec, 2003; Wang et  al., 2004), DSCR3 (Engidawork & Lubec, 2003), HMGN1 (Engidawork & Lubec, 2003), and CCT8 (Engidawork & Lubec, 2003). For these dysregulated genes, the authors suggested that the initial overexpression of genes from the aneuploid chromosome was amplified by subtle compensatory mechanisms to the gene-dosage effect that may, in turn, result in the extensive variability of the phenotype that characterises DS (Dauphinot et al., 2005; Kahlem et al., 2004; Lyle et al., 2004). The euploid genes, or the euploid region of chromosome 16, in the case of the DS mouse models, were generally found differentially expressed (over or under) (Bahn et  al., 2002; Chrast, Scott, Madani et  al., 2000; Chrast, Scott, Papasavvas et al., 2000; Dauphinot et al., 2005; FitzPatrick et al., 2002; Mao et al., 2003, 2005; Potier et  al., 2006; Saran et  al., 2003). These data support a model of a subtle primary upregulation of genes on the trisomic chromosome resulting in a more generalised secondary transcriptional misregulation (FitzPatrick et  al., 2002).

Mental Retardation and Human Chromosome 21 Gene Overdosage

31

Examination of difference of gene expression in two independent experiments suggests that the global perturbation includes a significant stochastic component. Thus, dosage imbalance of 124 genes in Ts65Dn mice alters the expression of thousands of genes to create a variable trisomic transcriptome (Saran et al., 2003). At the present time, few studies have been performed concerning the HSA21 genes that do not have mouse homologs on the mouse chromosome 16 (MMU16). When gene expression was examined in Ts43H mice, a segmental Ts17 mouse model for DS, 20 brain-specific genes at dosage imbalance gave an average of 1.2fold increased expression of euploid, with expression of only two genes reaching 1.5-fold expression (Vacik et al., 2005). In addition, 12 genes on the nontrisomic portion of chromosome 17 had expression levels that were 90% of euploid level. Brains from Ts2Cje mice exhibited a 1.5-fold expression level of specific trisomic genes comparable to Ts65Dn and different from euploid. Further data and analyses in both humans and mice are needed to reach biologically significant conclusions (Antonarakis & Epstein, 2006; Reeves, 2006). Recently, new aspects of gene expression have acquired more importance in DS studies. The level of the expression variation for a given gene can change in the different tissues, including brain (Dauphinot et al., 2005; Kahlem et al., 2004; Lyle et al., 2004; Mao et al., 2005; Rachidi, Lopes, Charron et al., 2005; Rachidi, Lopes, Delezoide, & Delabar, 2006; Rachidi et al., 2000; Saran et al., 2003; Sultan et al., 2007) and during developmental stages (Dauphinot et al., 2005; Potier et al., 2006; Rachidi, Lopes, Costantine, & Delabar, 2005; Rachidi et al., 2006; Rachidi et al., 2000; Sultan et al., 2007), indicated that there were tissue- and cell-specific changes of gene expression in trisomy 21 during foetal development. Inter-individual gene expression variations can explain at least some phenotypic individual differences, including susceptibility to common disorders. Since the 1970s, quantitative differences in gene expression have been proposed to explain variation in natural populations, participating in evolution and contributing to phenotypic diversity (King & Wilson, 1975). Recent studies indicate that variation in gene expression levels within and among populations is abundant, with significant inter-individual variation (Brem, Yvert, Clinton, & Kruglyak, 2002; Cheung et al., 2003; Oleksiak, Churchill, & Crawford, 2002; Schadt et  al., 2003). Most of the differentially expressed genes had significant heritability (Brem et al., 2002; Monks et al., 2004; Morley et al., 2004; Spielman et al., 2007; Storey et al., 2007; Yvert et al., 2003). This inter-individual gene expression variation has also been observed for HSA21 genes in DS (Aït Yahya-Graison et  al., 2007; Deutsch et  al., 2005; FitzPatrick et al., 2002; Prandini et al., 2007; Stranger et al., 2005; Sultan et al., 2007), and significant eQTLs have been identified (Deutsch et al., 2005). In particular, a cis-eQTL was identified for CCT8 corresponding to a single nucleotide polymorphism (SNP) located within the cis-regulatory region of CCT8 (Deutsch et al., 2005). These results are in agreement with the hypothesis that a molecular mechanism for the variability of phenotypic manifestations of trisomy 21 is a threshold effect of expression of HSA21 genes that show variable levels of ­expression in the population (Antonarakis, Lyle, Dermitzakis, Reymond, & Deutsch, 2004).

32

M. Rachidi and C. Lopes

On the basis of the observed extensive variation in gene expression observed among normal individuals, it has been predicted that for many HSA21 genes there is a considerable overlap in total expression levels between normal and trisomy 21 individuals due to allelic variation (Aït Yahya-Graison et al., 2007; Deutsch et al., 2005; Prandini et al., 2007; Sultan et al., 2007). In this way, two considerations can be proposed. First, the expression variations can explain the differences in the penetrance and variability of the DS phenotypes. It has been proposed that overexpressed genes, showing low levels of expression variation, would be predicted to lead to the more penetrant phenotypes. In contrast, genes with high variation in expression would contribute to incompletely penetrant/variable DS-related phenotypes (Aït Yahya-Graison et al., 2007; Deutsch et al., 2005; Prandini et al., 2007; Sultan et al., 2007). Second, the existence of expression variations suggest caution in the analyses of the gene expression changes, particularly for low variation ratios (less than twofold), because of overlapping of gene expression variation in DS and normal individuals in this variation interval (Aït Yahya-Graison et  al., 2007; Deutsch et al., 2005; Prandini et al., 2007; Sultan et al., 2007). In this way, a recent work analysed inter-individual gene expression variations between HSA21 genes in trisomic and normal cell lines. When pooled RNAs were used, a global gene dosage-dependent expression of chromosome 21 genes was observed. In contrast, when inter-individual gene expression variations were analysed, most of the HSA21 genes results compensated for the gene-dosage effect (Aït Yahya-Graison et  al., 2007). The authors suggested that overexpressed genes are likely to be involved in DS phenotypes, in contrast to the compensated genes. Moreover, a more recent work analysed the differences in euploid gene expression variation between trisomy 21 and euploid tissues, on the hypothesis that these differences may contribute to the phenotypic variations in DS (Chou et al., 2008). The authors found a group of euploid genes showing greater expression variance in human trisomy 21 tissues than in euploid tissues, and that the number of euploid genes with elevated variance was significantly higher in DS tissues than in the euploid tissues (Chou et al., 2008). Recently, new transcripts, the microRNAs (miRNAs), have been identified that play a role in gene expression regulation. MiRNAs are small, non-protein coding RNAs that link specific mRNA targets and lead to translational repression or mRNA cleavage (Bartel, 2004; Bushati & Cohen, 2007; Wang, Stricker, Gou, & Liu, 2007). Moreover, miRNAs have been shown to play a fundamental role in diverse biological and pathological processes, including cell proliferation, differentiation, apoptosis, carcinogenesis, and cardiovascular disease (Bushati & Cohen, 2007; Wang et al., 2007). It has been demonstrated that each miRNA can potentially regulate a large number of protein-coding genes, and many miRNAs can act in combination to regulate the same target genes (Bushati & Cohen, 2007; Wang et al., 2007). Thus, miRNA target genes are not restricted to a particular functional category or biological pathway, but rather are involved in a wide variety of biological processes. Very recently, bioinformatic analyses have demonstrated that HSA21 harbours five miRNA genes: miR-99a, let-7c, miR-125b-2, miR-155, and miR-802 (Kuhn

Mental Retardation and Human Chromosome 21 Gene Overdosage

33

et  al., 2008). HSA21 miRNA expression analyses demonstrate that they are ­overexpressed in foetal brain and heart specimens from individuals with DS when compared with controls (Kuhn et al., 2008; Sethupathy et al., 2007). Moreover, some miRNAs, located on chromosomes other than 21, have been found overexpressed or underexpressed in hippocampus specimens from individuals with DS when compared to controls (Kuhn et  al., 2008). The overexpression of the five HSA21 miRNAs in DS individuals may result in the aberrant expression of a large number of proteins in a variety of tissues. Thus, their inhibition or knock-down should normalise the expression level of all miRNA/mRNA targets back to non-trisomic 21 levels. Very interestingly, these potentialities suggest that HSA21 miRNAs may provide novel therapeutic targets in the treatment of individuals with DS. The application of the global genomic approach to in situ expression analysis allowed the establishment of expression atlas of the HSA21 genes for large gene screening and identification of candidate genes for DS phenotypes (Gitton et  al., 2002; Reymond et  al., 2002). Nevertheless, single gene approaches remain indispensable to determine precise gene expression map in different embryonic, foetal and adult ages in human and mouse. In addition, identification of the brain cell types expressing a given gene supplies fundamental information that helps gene function understanding (Lopes, Chettouh, Delabar, & Rachidi, 2003; Lopes, Rachidi, Gassanova, Sinet, & Delabar, 1999; Rachidi, Lopes, Charron et al., 2005; Rachidi et al., 2000, 2006). These spatio-temporal investigations have been particularly improved with a novel powerful microscopy technology, allowing in situ quantification of mRNA variations in different neuronal cell types in a given key structure of the brain (Rachidi, Lopes, Charron et al., 2005; Rachidi et al., 2000, 2006), and a novel quantitative method (quantitative assessment gene expression, QAGE) for assessment of in situ gene expression (Rachidi et al., unpublished data). 3.2.3 Proteomic Variation as a Consequence of Trisomy 21 It is known that quantity of proteins does always not correspond to the quantity of the corresponding mRNAs, because of several post-translational mechanisms that determine the final protein level in the cells in the given condition. Since the proteins are the final and functional products of the genes, to know how protein levels change in DS cells is a fundamental knowledge for understanding the real genotype/phenotype correlation and, finally, the DS pathogenesis. Initially, western blots have been used to measure the expression level of individual proteins. Several studies, analysing individual or small number of proteins, identified several changes in protein levels. Between the proteins encoded on HSA21, collagen VI A1 chain, COL6A1 (Engidawork, Balic et  al., 2001) was found decreased in DS tissues compared to the normal tissues, while HMG14 (Epstein, 2001), S100B (Griffin et al., 1998), carbonyl reductase (Balcz, Kirchner, Cairns, Fountoulakis, & Lubec, 2001), and synaptojanin (Arai, Ijuin, Takenawa, Becker, & Takashima, 2002) were found increased in DS tissues compared to the normal tissues. In addition, some proteins encoded on chromosomes other than

34

M. Rachidi and C. Lopes

HSA21 also show level changes. In particular, EF1A1 and EF2 (Freidl, Gulesserian, Lubec, Fountoulakis, & Lubec, 2001), Adenosine triphosphate (ATP)-sensitive potassium channels (Kim & Lubec, 2001), synaptosomal associated protein 25 subunits, drebrin, nucleoside diphosphate kinase B, Rab GDP-dissociation inhibitor beta subunit, histidine triad nucleotide-binding protein (Weitzdoerfer et al., 2001), and stathmin (Cheon, Fountoulakis, Dierssen, Ferreres, & Lubec, 2001) were found decreased in DS tissues compared to the normal tissues, while alcohol dehydrogenase (Balcz et  al., 2001) and nicotinic acetylcholine receptor beta 2 subunits (Engidawork, Gulesserian, Balic, Cairns, & Lubec, 2001) were found increased in DS tissues compared to the normal tissues. In a serial study, Lubec et al. analysed expression levels of 31 proteins encoded on HSA21 (Cheon, Bajo, Kim et al., 2003; Cheon, Kim, Ovod et al., 2003; Cheon, Kim, Yaspo et  al., 2003; Cheon, Shim, Kim, Hara, & Lubec, 2003; FerrandoMiguel, Cheon, & Lubec, 2004) and only three proteins showed different expression levels in DS compared to controls: Hematopoietic adapter containing Src homology 3 (SH3) domain and sterile a motifs (HACS1) was decreased in DS, compared to controls (Cheon, Bajo, Kim et al., 2003), Synaptojanin-1 was increased in DS, compared to controls (Cheon, Kim, Ovod et al., 2003), and DSCR5 (PIG-P), a component of glycosylphosphatidylinositol-N-acetylglucosaminyltransferase (GPI-GnT) was overexpressed about twofold in DS, compared to controls (Ferrando-Miguel et al., 2004). Genome-wide proteomic approaches are performed using 2D-gel electrophoresis that, more recently, was associated to mass spectrometry, quantifying the protein spots. One of the first works using a global approach, having a limited resolution power, identified 11 proteins, among the 49 proteins analysed, that are deregulated in cerebral cortex of foetal DS, none of which encoded on HSA21 (Engidawork, Gulesserian, Fountoulakis, & Lubec, 2003). Using the same approach, Kadota et al. (2004) have used an in vitro neuronal differentiation system of mouse Embryonic stem (ES) cells containing a single HSA21 (TT2F/hChr21) (Shinohara et al., 2001), using TT2F parental ES cells as a control. The authors have detected only 18 proteins with significantly altered levels, including SOD1 and CCT8, which are encoded on HSA21 (Kadota et al., 2004). Among the other 16 proteins, they found matrix and structural proteins, heat shock/ stress proteins, protein or translational regulators, nuclear transcriptional factors, and enzymes for energy and macromolecular metabolism (Kadota et  al., 2004). Among these 16 proteins encoded on other human chromosomes, the authors identified 7 that were overexpressed: protein subunits Atp6v1a1 and Atp6v1b2 of the vacuolar ATPase proton pump, which mediate acidification of intracellular organelles for energy production and convention, actin- (T-plastin and Vil2), filament(Krt2–8) and phospholipid- (Anxa4) related cytoskeleton proteins. In constrast, nine proteins were underexpressed significantly in TT2F/hChr21 cells compared with TT2F cells: AI850305, Eef1D and UchL1, involved in protein catabolism or translation regulation, heat shock proteins Hsp84-1, Hsp70 and Hsp86-1, microtubule- (Mapre2) and calmodulin- (Cnn3) related architectural proteins were underexpressed. Moreover, splicing regulatory elements, HnrnpF and HnrnpC, displayed

Mental Retardation and Human Chromosome 21 Gene Overdosage

35

contradictory expression patterns of overexpression and ­underexpression, ­respectively (Kadota et al., 2004). Moreover, in a comparison between mRNA and protein level change in TT2F/hChr21 cells compared with TT2F cells, different features were identified. The expression of Anxa4, Atp6v1a1, Atp6v1b2, Krt2–8, Vil2 (overexpressed), and of HnrnpC, Mapre2, UchL1, AI850305 (underexpressed) showed consistent mRNA transcription and protein translation. In contrast, Cnn3, Eef1D, Hsp70, Hsp84, Hsp86, HnrnpF and T-plastin showed disagreement (Kadota et al., 2004), suggesting the existence of post-transcriptional regulation or translational modification. Recently, Shin, Gulesserian, Verger, Delabar, and Lubec (2006) performed a proteomic approach using a non-mosaic polytransgenic mouse model for DS generated by inserting yeast artificial chromosomes (YACs), containing a fragment of the human critical region DSCR, into the murine genome (Smith et al., 1995). These mice carry 141G6 YAC and are polytransgenic for HSA21 genes DSCR3, 5, 6, 9, and tetratricopeptide repeat domain 3 (TTC3). The authors identified 45 proteins showing altered expression levels, among the 422 polypeptides, which were the products of 239 different genes, in mouse transgenic hippocampus compared to control, although none of DSCR3, 5, 6, 9, and TTC3 proteins was detectable using the low resolution Coomassie staining (Shin et al., 2006). These aberrant protein expressions may lead to impairment of cognitive functions. In particular, calcium/calmodulin-dependent protein kinase (CaMKII) protein was decreased in the 141G6 mouse hippocampus (Shin et al., 2006), and it is known that alteration of the CaMKII-pathway leads to a downstream alteration of the c-AMP response element-binding protein (CREB) pathway associated with impairment of fear memory (Bourtchuladze et  al., 1994). 141G6 mice showed a lower performance in fear conditioning against sound as acoustic conditional stimulus (Chabert et al., 2004) that could be explained by aberrant protein levels of CaMKII (Shin et al., 2006). In contrast, 141G6 mice showed no cognitive defect by using Morris water maze and the multiple T-maze paradigms (Chabert et al., 2004) that could be explained by threshold levels necessary to alter these functions that was not surpassed, or that these tests are not sensitive enough to detect minor cognitive alterations (Shin et al., 2006). In recent years, it is emerging that protein alterations exist as polymorphisms among wild-type mice of different inbred strains. These polymorphic variations complicate the interpretation of the variation of protein level changes and their correlation to a given disease. Recently, Mao et  al. (2007) conceived a simplified approach to analyse the effect of gene-dosage imbalance on proteome in a controlled environment by using mouse ES cells. They investigated four cell lines contained one single overexpressed gene (App, Snca, Dyrk1a, & Dopey2) and two cell lines with a duplication or a deletion, respectively, of a HSA21 segment containing 14 genes. The authors identified globally 255 proteins showing expression variation in the six cell lines. Four features appear in this study: (1) about the same numbers (70–110) of proteins showed expression alterations in each line, with dosage imbalance in only one or 14 genes; (2) dosage alteration of a single gene led to quantitative changes in a large number of proteins; (3) many proteins showed changed ­expression

36

M. Rachidi and C. Lopes

levels in several cell lines (38 proteins have alteration in at least three lines); and (4) 114 proteins were altered only in one cell line (Mao et al., 2007). On the basis of these observations, the authors proposed that the protein level changes may also be explained in part by a global response of the cellular proteome to the gene dosage defect, restoring the balance in the cellular proteome, on the hypothesis that quantitative changes of the proteome by gene dosage effects can be compensated by a rearrangement restoring a new balance. In this way, the cellular proteins were defined as balancer proteins, with altered quantity in several lines, and cell line-specific proteins, when altered only in one cell line. Balancer proteins would function as buffers in the proteome homeostasis without a direct functional correlation with the transgene(s) and among them, in contrast to the cell line-specific proteins, likely including proteins participating to common functional networks of the transgene(s) (Mao et al., 2007). Interestingly, the balancers have more non-synonymous SNPs in coding regions than cell line-specific proteins (Mao et al., 2007), indicating that balancers may have more tolerance towards quantitative changes, whereas cell line-specific proteins need more precise correlation between expression level and function.

4 Modelling Neuronal Alterations and Mental Retardation in Mouse Models of Down Syndrome The alterations observed in brain of DS patients are likely to take place during embryogenesis and cannot be easily investigated at early stages of human development. Developmental studies in humans are extremely difficult and in vitro molecular biology and cell culture systems do not replicate the complex developmental processes perturbed by trisomy. These investigations became possible with the generation of the mouse models of DS, because of the ability to manipulate their genome genetically and the accessibility to all their tissues at different embryonic, foetal and adult stages. Interestingly, a high degree of conservation of the genomes and molecular mechanisms exists between mouse and human, and human genes on chromosome 21 are syntenic to mouse genes on chromosome 16 (~26.5 Mb), chromosome 10 (~2.3 Mb) and chromosome 17 (~1.1 Mb) (Hattori et al., 2000; Mural et al., 2002; Toyoda et al., 2002). This genetic evidence between the two mammalian species supports the essential use of the mouse in animal models to study the disruption of the developmental process caused by trisomy. More interestingly, the elevated gene expression due to trisomy is very comparable between mice and human and shows similar complexity and a comparable genetic effect with the same outcome on the features of the mouse analogous to DS phenotypes. This make the mouse models powerful tools for dissecting the phenotypic consequences of dosage imbalance that affect single genes or chromosome segments, and they have greatly enhanced our understanding of the cellular and biochemical mechanisms of gene dosage effects involved in DS.

37

Mental Retardation and Human Chromosome 21 Gene Overdosage

4.1 Trisomic Mouse Models and Candidate Chromosomal Regions for Mental Retardation in Down Syndrome These trisomic mouse models with segmental or complete trisomy for MMU16, containing the orthologous region of the most part of HSA21, imitate the genetic complexity seen in trisomy 21 and have clinical phenotypes that correspond well to that observed in DS patients (Table 1).

Table 1  Trisomic and transgenic (Tg) mouse models for Down syndrome, their human syntenic chromosomal regions and genes and their neurological alterations Models Human syntenic region and genes Neurological alterations Ts16 (21q21.1–21q22.3) Trisomic for Decreased brain size; LIP1–ZNF295 region: 158 genes cellular hypoplasia; abnormal neuronal migration. Alterations in cerebellum, Ts65Dn (21q21.2–21q22.3) Trisomic for hippocampus, and cortex; MRPL39–ZNF295 region: synapses, neurotransmitters; 136 genes BFCN, learning and memory deficits Reduced cerebellar volume and Ts1Cje (21q22.11–21q22.3) Trisomic for some brain defects similar SOD1–ZNF295 region: to Ts65Dn; learning deficits 83 genes Spatial learning impairement Ms1Ts65 (21q21.2–21q21.3) Trisomic for even less severe than MRPL39–SOD1 region: Ts65Dn and Ts1Cje 53 genes Altered brain volume and Ts1Rhr (21q22.13–21q22.3) Trisomic for CBR1–MX1 region: 33 genes shape Decreased spines density Ts2Cje (21q21.3–21q22.3) Trisomic from APP- to ZNF295: 132 genes of dendrites; enlarged dendritic spines Not identified Dp(16)1Yu (21q21.1–21q22.3) Trisomic for LIP1–ZNF295 region: 158 genes. Altered cerebellar neuronal Tc1 (HSA21 with two gaps: number, synaptic plasticity, Cxadr-D21S1922; Ifnar1-Runx1) learning and memory Trisomic for 92% of HSA21 genes. Impairment in behaviour and ES#21 HSA21 chimaera, Trisomic for HSA21 genes learning Tg SOD1 SOD1 (21q22.11), superoxide dismutase, Decreased serotonin level; neuronal degeneration in key enzyme in the metabolism of brain; learning defects oxygen-derived free radicals Dystrophic neuritis associated Tg APP APP (21q21.3), b-amyloid precursor with involved congophilic protein in senile plaque formation plaques; learning defects in DS and AD Learning and memory defects Tg Synj1 SYNJ1 (21q22.11), synaptojanin 1 polyphosphoinositide phosphatase in synapses (continued)

38

M. Rachidi and C. Lopes

Table 1  (continued) Models

Human syntenic region and genes

Neurological alterations

Tg Ets2

ETS2 (21q22.2), erythroblastosis virus E26 transformation-specific transcription factor

Tg S100b

S100b (21q22.3), calcium-binding protein beta neurotrophic factor released by astrocytes

Tg Dyrk1 TgYac152F7

DYRK1A (21q22.13),Cbr1-Cldn14 (21q22.12-q22.13) containing Dyrk1A dual-specificity tyrosine-(Y)phosphorylation regulated kinase 1A SIM2 (21q22.13), single minded, transcription factor/helix-loop-helix, master regulator in CNS cell fate DSCR1 (21q22-12), Down syndrome critical region 1

Neuonal cell apoptosis; brachycephaly; neurocranial and cervical skeletal defects, Abnormal dendritic development; astrocytosis; learning and memory deficits Abnormal brain structure; increased brain weight and neuronal size; learning deficits Altered behaviour and learning deficits

Tg Sim2

Tg DSCR1

TgYac230E8

TTC3-DYRK1A (21q22.13) containing DOPEY2

Neurological phenotype; impaired working memory in null mice Increased cortical neuronal density; learning deficits

These mouse models represent powerful tools allowing a genetic dissection of the complex DS phenotype, identifying different candidate chromosomal regions, syntenic with HSA21, and candidate genes involved in mouse brain alterations, and permitting a study of the early developmental phenotypes and the molecular and cellular pathogenesis of the brain abnormalities and MR in DS. 4.1.1 Ts16 Mice: Trisomic for Most Part of HSA21 with Three Copies of Complete MMU16 Ts16 mice, the first trisomic model for DS (Epstein, 1986; Lacey-Casem & OsterGranite, 1994), carry three full copies of MMU16, which contain a region orthologous to the larger part of the HSA21 (Cox, Smith, Epstein, & Epstein, 1984; Gearhart, Davisson, & Oster-Granite, 1986). The lethality in utero observed in Ts16 mice, due to the presence of three copies of genomic regions syntenic to other human chromosomes, limited the studies to cell lines and foetal stages. Interestingly, Ts16 foetuses have a number of phenotypes similar to those seen in DS patients, including brain alterations (Behar & Colton, 2003; Cox et al., 1984; Epstein et al., 1985; Gearhart et al., 1986; Lacey-Casem & Oster-Granite, 1994). 4.1.2 Segmental Ts65Dn Mice: Trisomic for Most HSA21 Genes Conserved in Distal End of MMU16 The Ts65Dn is the first segmental trisomy model created and is the most frequently used and the greatest characterised mouse model (Davisson, Schmidt, & Akeson,

Mental Retardation and Human Chromosome 21 Gene Overdosage

39

1990). Ts65Dn mice are trisomic for most of the HSA21 orthologous genes conserved in the distal end of MMU16, extending from the gene for mitochondrial ribosomal protein L39 (Mrpl39) to the Znf295 gene, at the distal telomere (Fig. 1). Moreover, Ts65Dn mice are also trisomic for a c. 6 Mb region of chromosome 17 not syntenic to HSA21 (Akeson et al., 2001; Li et al., 2007). Ts65Dn mice display many features that are reminiscent of those seen in people with DS, particularly the neurological phenotypes, including learning and ­behavioural abnormalities (Sago et  al., 1998, 2000). This mouse model shows delayed brain development, decreased cerebellar volume and granular cell density, decreased ­dentate gyrus, and abnormal synaptic plasticity (Baxter, Moran, Richtsmeier, Troncoso, & Reeves, 2000; Davisson et  al., 1990; Kleschevnikov, Belichenko, Villar, Epstein, & Malenka, 2004). Ts65Dn mice also show age-related ­atrophy, ­neurodegeneration of BFCN, neurotransmitters alterations and extensive astrocyte hypertrophy, which resembles the neuropathology of AD in DS patients (Casanova et al., 1985; Cooper et  al., 2001; Dierssen, Vallna, Baamonde, GarciaCalatayud, & Lumbreras, 1997; Yates et al., 1983). The abnormal learning and behavioural abnormalities, analogous to DS MR, have been demonstrated using different behavioural tests such as T-maze, Y-maze and radial maze. Ts65Dn mice also show important learning defects in the Morris water maze that is the most commonly used test for spatial learning in almost all DS mouse models, and in which the cognitive performances of the different segmental trisomy 16 mouse models can be compared. In the hidden platform test, Ts65Dn mice must learn the spatial relationships between objects in the room and the position of the platform to escape from the water. Ts65Dn mice showed increased search time compared with control and impaired performance that is not improved over successive trials indicating poor learning. In the probe test, in which the platform has been removed, the mice had learned the location of the platform and should search where the platform had been located. In this test, which assesses spatial selectivity, Ts65Dn mice showed a greater preference for the trained quadrant than control mice, providing evidence for learning. However, Ts65Dn mice spent significantly less time in the trained quadrant and crossed the trained site significantly less frequently than did controls. In the reverse platform test, the mice are required to learn a novel position for the hidden platform that has been moved to the quadrant opposite to its original location. Ts65Dn mice showed no decrease in latency, spent more time in the initial trained quadrant and showed increased time to reach the novel position of the platform. In the reverse probe dwell test, Ts65Dn mice continue to show a preference for the initial trained site. In the reverse probe crossing test, Ts65Dn mice failed to show a preference for the trained site. In these different Morris water maze tests, Ts65Dn mice show significant learning and memory deficits with a severe impairment in spatial learning and reversal, but not in visual discrimination learning and reversal (Holtzman et al., 1996; Reeves et al., 1995; Sago et al., 2000). Interestingly, the long-term potentiation (LTP) is reduced in the cornu ammonis (CA1) and dentate gyrus areas of the hippocampus in the Ts65Dn (Kleschevnikov et al., 2004; Siarey et al., 1999; Siarey, Stoll, Rapoport, & Galdzicki, 1997) and the excitatory and inhibitory inputs to pyramidal neurons in cornu ammonis (CA3) of

40

M. Rachidi and C. Lopes

the hippocampus are reduced (Hanson, Blank, Valenzuela, Garner, & Madison, 2007). This is particularly interesting since the hippocampal LTP, a form of synaptic plasticity evoked by a train of electrical stimuli, is considered a physiological model of learning and memory. 4.1.3 Segmental Ts1Cje Mice: Trisomic for Three-Quarters of Genes of Ts65Dn, Including DSCR The Ts(16C-tel)1Cje, or Ts1Cje, present a smaller extra segment of MMU16 than that of the Ts65Dn mice. This mouse model, generated by a fortuitous translocation during the targeting of Sod1 by homologous recombination, carries the translocation from the proximal break-point in Sod1, that is not functional, to Znf295 (Sago et al., 1998) (Fig. 1). The Ts1Cje mouse is trisomic for about three-quarters of the genes that are present in the Ts65Dn mouse. Ts1Cje mice have similar phenotypes to Ts65Dn, often with lower intensity, and fewer similarities to DS than do Ts65Dn mice, but they are important to study the particular effects of trisomy for a subset of genes triplicated in Ts65Dn and not in Ts1Cje. In particular, the neurological phenotypes in Ts1Cje are similar to those observed in Ts65Dn, such as the reduced volume and granule cell of the cerebellum (Olson, Roper et al., 2004) and the reduced LTP in the CA1 and dentate gyrus areas of the hippocampus (Siarey, Villar, Epstein, & Galdzicki, 2005). Ts1Cje mice also show behavioural abnormalities in the Morris water maze tests. These mice displayed moderate to severe impairment in the hidden platform and probe parts of the test. In the reverse platform test, they showed a decrease in latency over the trials of the test, but the rates of decrease were significantly less than the controls and they were not significantly better than Ts65Dn. There was no preference of the trained quadrant in the reverse probe dwell test and Ts1Cje did significantly better than Ts65Dn in reverse crossing and dwell tests (Sago et  al., 1998, 2000). Comparison of the behavioural performances of the Ts1Cje and Ts65Dn in the Morris water maze showed that, except in the reverse probe tests, the learning deficits of Ts1Cje mice are similar to those of Ts65Dn. These findings indicate that an important gene or genes involved in these deficits lie in the overlapping region in these mice, from Sod1 to Mx1, and containing the critical region DSCR. 4.1.4 Segmental Ms1Ts65 Mice: Trisomic for Non-DSCR Genes of Ts65Dn and Missing from Ts1Cje Ts65Dn mice, produced by reciprocal translocation T(17;16)65Dn, and Ts1Cje mice, carrying the reciprocal translocation T(12;16)1Cje, have been mated to produce offspring called Ms1Cje/Ts65Dn, or Ms1Ts65, that are trisomic for the genes present in Ts65Dn and missing from Ts1Cje, corresponding to the segment from MRPL39 to SOD1 (Fig. 1) (Sago et al., 2000).

Mental Retardation and Human Chromosome 21 Gene Overdosage

41

These segmental trisomic mice have fewer similarities to DS than do Ts65Dn and Ts1Cje mice. In Ms1Ts65 mice, a neurological alteration has been identified at the cerebellar level, in which the granule cell density is moderately reduced similarly to Ts1Cje compared to Ts65Dn mice, which show significant reduction (Baxter et al., 2000; Olson, Roper et al., 2004). The spatial learning and memory performances of Ms1Ts65 mice, tested by the Morris water maze (Sago et al., 2000), showed reduced latencies in the hidden platform test. In the probe test, the performance of Ms1Ts65 mice was similar to the controls. In the reverse platform test, mice such as Ts1Cje and Ms1Ts65showed a decrease in latency, but the rates of decrease were significantly less than the controls. Although the difference in latency between Ms1Ts65 and Ts1Cje was not statistically significant, Ms1Ts65 was significantly better than Ts65Dn whereas Ts1Cje was not. In the reverse probe dwell test, Ms1Ts65 was also significantly better than Ts65Dn in reverse crossing and dwell tests (Sago et  al., 2000). Compared with controls, Ms1Ts65 mice show significant deficits in the latencies of the hidden and reverse hidden platform tests, but not in the probe tests. These results indicate that Ms1Ts65 has little impairment in learning the task in the Morris water maze compared with controls, while their deficits are significantly less severe than those of Ts65Dn. Therefore, whereas triplication of the region from Sod1 to Mx1 plays a major role in the abnormalities of Ts65Dn in the Morris water maze, triplication of the region from App to Sod1 also contributes to the poor performance. 4.1.5 Segmental Ts1Rhr and Ms1Rhr Mice: Trisomic and Monosomic for DSCR A duplication, Dp(16Cbr1-Mx1)1Rhr, or Ts1Rhr, and a deletion, Ms1Rhr (Fig. 1), of the MMU16 segment between Cbr1 and Mx1 genes have been created using Cre-loxP and ES cell technologies (Olson, Richtsmeier, Leszl, & Reeves, 2004). These mice provide trisomy and monosomy, respectively, for a smaller segment of MMU16 that is orthologous to the critical region DSCR of HSA21 responsible for many of the features of the DS, including craniofacial abnormalities and MR. Ts1Rhr mice have less severe craniofacial dysmorphology than either Ts1Cje or Ts65Dn (Olson, Richtsmeier et  al., 2004). Both Ts1Rhr and Ms1Rhr mice show changes in volume and shape of both cerebrum and cerebellum, but different from each other and from Ts65Dn mice (Aldridge, Reeves, Olson, & Richtsmeier, 2007). In contrast, the performances of these two mouse models in the Morris water maze were similar to euploid mice (Olson et al., 2007). 4.1.6 Segmental Ts2Cje Mice: Trisomic from APP to the Telomere Ts{Rb[12.17(16)]}2Cje mice, or Ts2Cje (Fig. 1), carry a chromosomal rearrangement of the Ts65Dn genome whereby the marker chromosome has been translocated to chromosome 12 forming a Robertsonian chromosome. This stable rearrangement

42

M. Rachidi and C. Lopes

confers fertility in males and increases the frequency of transmitted segmental trisomy through the female germline. Like Ts65Dn mice, Ts2Cje mice are about 20% smaller in size postnatally compared with euploid control littermates, and this smaller size persists throughout life (Villar et al., 2005). This trisomic model retains a dosage imbalance of HSA21 homologous genes from App to the telomere and expression levels similar to Ts65Dn within the triplicated region. Similarly to Ts65Dn mice, significant decreases in the density on the dendritic spine of dentate granule cell neurons and enlarged dendritic spines are observed in the Ts2Cje mice (Villar et al., 2005). Ts2Cje mice exhibit neurologica1 features comparable with those of Ts65Dn mice, thereby validating the utility of this segmental trisomy model for the study of the molecular, genetic and developmental mechanisms underlying DS. 4.1.7 Segmental Dp(16)1Yu Mice: Trisomic from LIP1 to ZNF295 To generate a more complete trisomic mouse model of DS, a duplication has been established recently spanning the entire HSA21 syntenic region on MMU16 in mice using Cre/loxP-mediated long-range chromosome engineering. This new DS mouse model carries a chromosomal duplication, Dp(16)1Yu (Fig.  1), spanning 22.9 Mb of the complete HSA21 syntenic region 21q11q22.3 of the MMU16, delimited by the mouse orthologs of LIP1 and ZNF295 genes (Li et al., 2007). The analysis of several genes located within Dp(16)1Yu in the brain and heart tissues showed that the segmental trisomy altered the transcript levels of the genes in the brain and heart of the Dp(16)1Yu/+ model, reflecting the dosage imbalance for the duplicated region. This result supports the conclusion that the duplicated genes are expressed with the exception for transcriptionally inactive genes. About 37% of Dp(16)1Yu/+ embryos exhibit structural heart defects, and about 26 and 22% of Dp(16)1Yu/+ embryos exhibit annular pancreas and malrotation of the intestine, respectively. These phenotypes are also observed in patients with DS at higher frequencies than normal individuals. The cardiovascular and gastrointestinal phenotypes of the mouse model were similar to those of patients with DS. This new mouse model is particularly interesting because of the largest duplication of the HSA21 syntenic region and its stability, and it represents a powerful tool to further understand the molecular and cellular mechanisms of DS. 4.1.8 Transchromosomal ES(#21) Mice: Trisomic for a Large Part of HSA21 To maximally mimic the DS phenotypes, transchromosomal mouse models, carrying a HSA21 or a large part of it, have been generated (Fig. 1). These mouse models, containing an additional entire or partial HSA21, have been developed using a microcell-mediated chromosome transfer approach (MMCT) (Shinohara et  al., 2001). The initial transchromosomic mouse model was obtained by transferring a

Mental Retardation and Human Chromosome 21 Gene Overdosage

43

HSA21 into mouse ES cells. ES cell lines retaining HSA21 as an independent ­chromosome were used to produce chimeric mice with a substantial contribution from HSA21-containing cells. Chimeric mice derived from these cells, named ES(#21), in which a high percentage of cells contained a HSA21, demonstrated specific parallels to developmental anomalies seen in DS and a wide range of behavioural abnormalities indicating abnormal brain development and function. Interestingly, these mice present similar phenotypes to those observed in DS, such as thymus and cardiac defects, impairment in learning or emotional behaviour found in open-field, contextual conditioning and forced swim (Shinohara et al., 2001). The high correlation between retention of HSA21 in the brain and behavioural and cognitive alterations found in these transchromosomal mice make them good models to study the complex and critical aspects of DS phenotype, because they provide the complete set of genes that are in dosage imbalance in human with trisomy 21. Further, these genes are introduced in mice into the context of their native cis-acting regulatory elements and chromatin structures; this maximises temporal and tissue-specific gene expression and function under physiologically appropriate conditions. 4.1.9 Transchromosomal TC1 Mice: Trisomic for Almost HSA21 (92% of HSA21 Genes) Another transchromosomal mouse model, Tc1, has been generated containing an almost complete HSA21 with only two deletions. This mouse model represents the most complete animal model for DS currently available and carrying 92% of human genes (Fig. 1) (O’Doherty et al., 2005). Tc1 mice showed alterations in cerebellar neuronal number, in heart development, and in mandible size. In addition, they have impaired short-term recognition memory and display reduced LTP in the dentate gyrus of the hippocampus, as well as showing a deficit in a novel-object recognition task (O’Doherty et  al., 2005). Thus, Tc1 display many aspects of human DS but also recapitulate several of the DS features present in other mouse models (Reeves, 2006). Tc1 mice also have impaired spatial working memory but preserved long-term spatial reference memory in the Morris water maze (Morice et al., 2008). These mice showed a loss of the HSA21 from about 50% of the cells in adult mice, determining a high degree of mosaicism. The effect of this mosaicism may be different in the individuals and contributes to the variability of the phenotype. Therefore, unlike other segmental DS mouse models, Tc1 mice are also trisomic for orthologous genes on mouse chromosomes 10 and 17, consisting of a condition more similar to the trisomy 21 in human (O’Doherty et al., 2005). 4.1.10 Segmental Transgenic Mouse In Vivo Library of Human DSCR The identification of the critical region DSCR and its association to MR suggests that this major and invariable DS trait arises from triplication of one or few genes

44

M. Rachidi and C. Lopes

located in the DSCR. Interestingly, several genes located in the DSCR are involved in brain development and function and the overexpression of these critical genes determine cognitive alterations. In vivo libraries of large insert transgenic mice offer an approach to study the contribution of a genomic region to complex quantitative traits. These mice are frequently transgenic for many genes and, thus, it is possible to investigate the cumulative effects of these genes upon one biological phenotype at a time, allowing multiplex analysis of the relationship between genotype and phenotype. Phenotypic and functional analysis of the in vivo library members could be used to define candidate genes for further analysis in human populations enabling association rather than linkage studies (Risch and Merikangas, 1996) to be employed in the identification of genes contributing to complex traits such as the MR in DS. In this way, transgenic mice containing large fragments of the DSCR have been constructed (Smith et  al., 1997; Smith, Zhu, Zhang, Cheng, & Rubin, 1995). The human genome fragments are 4 YACs, 230E8, 152F7 141G6 and 285E6, spanning 2 Mb of the DSCR (Dufresne-Zacharia et al., 1994; Smith et al., 1995). This panel of YAC transgenic mice propagating targeted megabase regions of the genome constitutes an in vivo library allowing genotype/phenotype comparison studies (Fig. 1). The transgenic lines, carrying the YAC 152F7, containing six genes including TTC3 and dual-specificity tyrosine-(Y)-phosphorylation kinase 1A (DYRK1A) genes, show an increase of brain size and neuronal sizes (Branchi et al., 2004; Rachidi et al., 2007), and exhibit severe spatial learning and memory defects (Smith et al., 1997). The transgenic lines, carrying the YAC 230E8, containing seven genes of the DSCR region, including DOPEY2 gene, present increased cortical cell density (Rachidi, Lopes, Costantine et al., 2005; Smith et al., 1997), and increased length of the anterior lobules of the cerebellar vermis (Rachidi et al., 2007), and exhibit spatial learning and memory defects (Smith et al., 1997). The 141G6 mice showed a lower performance in fearconditioning against sound as acoustic conditional stimulus (Chabert et  al., 2004), although any evident neuroanatomical and cognitive defects have not yet been demonstrated in the 141G6 mice (Smith et al., 1997). Finally, the performances of transgenic lines carrying YACs 285E6 are not significantly different from the controls, and no detectable neurological defects have been found (Smith et al., 1997). It is of greatest interest to dissect the role of these critical genes of DSCR, by separate analysis and study of their different combinations, to better understand the function of each gene or cooperation of gene groups, in neurological alterations and in learning and memory processes.

5 Genetic Dissection of the Role of the Down Syndrome Critical Region in Mental Retardation In human, although the concept of the involvement of the critical region DSCR in the principal phenotypes of DS is largely accepted, its delimitation is not completely defined (Delabar et al., 1993; Korenberg et al., 1994; Rahmani et al., 1989), and its existence was rarely controversial (Shapiro & Whither-Azmitia, 1997).

Mental Retardation and Human Chromosome 21 Gene Overdosage

45

The genotype/phenotype comparison approach has been applied to the mouse, and different mouse models have been generated allowing the evaluation of the role of the DSCR in DS pathogenesis, particularly in the MR. In a first approach, a transgenic mouse in vivo library has been developed, as described above, by inserting human YACs bearing different fragments of the human DSCR into the murine genome (Smith et al., 1995). The neuroanatomical alterations and defects in learning and memory observed in particular in two transgenic lines (152F7 and 230E8 YAC transgenic mice) indicate that these HSA21 fragments of the DSCR are critical for brain alterations and learning and memory defects, and that the correct dosage of critical genes of these DSCR fragments are crucial for brain function and cognitive impairment (Rachidi et  al., 2007; Rachidi, Lopes, Costantine et  al., 2005; Smith et al., 1997), in agreement with an important role of the critical region DSCR in neuronal and cognitive alterations observed in DS patients. In an other approach, three other mouse models have been generated: Ts1Rhr mice trisomic for DSCR; Ms1Rhr mice with deleted DSCR; and Ms1Rhr/Ts65Dn mice obtained by breeding Ms1Rhr with Ts65Dn mice, trisomic for genes triplicated in Ts65Dn but not in the DSCR (Olson, Richtsmeier et al., 2004; Olson et al., 2007). Initially, the authors tested the association of craniofacial phenotypes to DSCR and found that three copies of DSCR alone are not sufficient to generate these phenotypes. Moreover, reducing trisomy of the DSCR to disomy in the Ts65Dn mice did not eliminate this phenotype, indicating that the DSCR is also not necessary to generate the cranio-facial phenotypes in mice (Olson, Richtsmeier et  al., 2004). Recently, these studies have been extended to test the role of the DSCR in hippocampal function, learning and memory (Olson et al., 2007). Unlike Ts65Dn and Ts1Cje mice, no LTP impairment is detected in the CA1 hippocampal area of Ts1Rhr mice, consistent with the normal spatial learning in the Morris water maze showed by these mice. Thus, trisomy for DSCR is not sufficient to produce deficits in this hippocampal-based task (Olson et al., 2007). Ms1Rhr/Ts65Dn mice, with disomic DSCR, show identical performances to euploid in the Morris water maze. This indicates that the restoration of disomy of DSCR in trisomic mice rescues the spatial learning and memory performance, demonstrating that trisomy of DSCR is necessary for this cognitive phenotype (Olson et al., 2007). Thus, in contrast to the craniofacial phenotype, the combination of the behavioural results of Ts65Dn, Ts1Rhr and Ms1Rhr/Ts65Dn mice show that DSCR is necessary although not sufficient to determine the hippocampal dysfunction seen in Ts65Dn mice (Olson et al., 2007).

6 Transgenic Mouse Models of Down Syndrome Contrary to segmental trisomic mice imitating the genetic complexity seen in trisomy 21 with eventual interactions between different genes present at three copies, the transgenic mouse models overexpress one or a few genes and allow a direct genotype/ phenotype correlation. The other interesting kinds of mouse models of DS are the transgenic monogenic mouse models that have been generated to study the effect of cell-specific and

46

M. Rachidi and C. Lopes

­stage-specific overexpression of a unique gene. These mice (Table 1; Fig. 1) include models for overexpression of the Cu-Zn superoxide dismutase1 (SOD1 gene), the neurotrophic factor (S100B gene), the beta amyloid peptide (APP gene), the transcription factor (ETS2 gene), the Drosophila minibrain homolog (DYRK1A gene), and the transcription factor single minded (SIM2 gene), the regulator of calcineurin (RCAN1 or DSCR1 gene), the C21orf5 or (DOPEY2 gene), the tetratricopeptide repeat domain 3 (or tetratricopeptide repeat domain Down syndrome [TPRD] gene), the potassium inwardly rectifying channel (KCJN6 gene), the inersectin (ITSN1 gene), and the synaptoganin (SYNJ1 gene). These transgenic mice showed overexpression of some genes in the key brain regions that play crucial roles in cognitive functions and that were found altered in the brain of DS patients. Moreover, for most of these genes, mouse models overexpressing them have an impaired behaviours and cognitive defects. These transgenic mouse models allow the dissecting of the phenotypic consequences of imbalances that affect single genes and have greatly enhanced our understanding of the cellular and biochemical mechanisms of gene dosage effects involved in the developmental brain alterations and in the MR in DS.

7 Candidate Genes and Genotype/Phenotype Correlation for Mental Retardation in Down Syndrome The final goal of genetic dissection is the identification of the gene(s) responsible of each phenotypic trait in DS. To date, several HSA21 genes have been identified as candidates for neurological alterations and MR in DS (Table 2), on the basis of different criteria. All these candidate genes show a strong expression in the key brain regions that play crucial roles in cognitive functions and that were found altered in the brain of DS patients. They are overexpressed in the brain of DS patients and/or in DS mouse models. Moreover, for most of these genes, mouse models overexpressing them have impaired behaviours and cognitive defects, similar to those observed in DS patients. Consequently, studies of these candidate genes and of the effects of their overexpression may help the understanding of the developmental brain alterations and the MR in DS.

7.1 Cu-Zn Superoxide Dismutase (SOD1) Gene SOD1 gene encodes the Cu-Zn superoxide dismutase, a key enzyme in the metabolism of oxygen-derived free radicals. SOD1 product levels, both mRNA and protein, are increased in human and mouse trisomic tissues (Epstein et al., 1987; Kadota et al., 2004; Lyle et al., 2004; Mao et al., 2005; Saran et al., 2003). Mice lacking SOD1 develop subtle motor symptoms by approximately 6 months of age, in which motor unit numbers are reduced early but decline slowly with age, suggesting that axonal sprouting are functionally impaired in the absence of SOD1 (Shefner et al., 1999).

Table 2  HSA21 genes over-expressed in Down syndrome brain and candidates for mental retardation Genes Brain regions References APP (amyloide beta A4) Cortex, midbrain cerebellum Epstein (2001), Saran et al. (2003), Lyle et al. (2004), and Kahlem et al. (2004) SOD1 (superoxide dismutase 1) Cortex, midbrain cerebellum Saran et al. (2003), Lyle et al. (2004), Kadota et al. (2004), and Mao et al. (2005) SYNJ1 (synaptojanin 1) Cortex, midbrain cerebellum Arai et al. (2002) and Lyle et al. (2004) ITSN1 (intersectin 1) Cortex, cerebellum Pucharcos et al. (1999), Amano et al. (2004), and Lyle et al. (2004) DSCR1 (calcipressin 1) Cortex, midbrain cerebellum Fuentes et al. (2000), Amano et al. (2004), Lyle et al. (2004), and Dauphinot et al. (2005) DOPEY2 (DOPEY2/C21orf5) Cortex, cerebrum Lopes et al. (2003), Lyle et al. (2004), and Rachidi, Lopes, Costantine et al. (2005) SIM2 (single-minded 2) Midbrain Vialard et al. (2000) and Lyle et al. (2004) TTC3 (tetratricopeptide repeat domain 3) Cerebrum, cerebellum Saran et al. (2003), Amano et al. (2004), and Lyle et al. (2004) DYRK1A (dual-specificity tyrosine-(Y)Cortex, midbrain cerebellum Saran et al. (2003) and Lyle et al. (2004) phosphorylation kinase) Cortex, midbrain Saran et al. (2003), Lyle et al. (2004), and Kahlem et al. KCNJ6 (potassium inwardly-rectifying channel J6) (2004) Ets2 (v-ets erythroblastosis virus E26) Cortex, cerebrum, cerebellum Saran et al. (2003), Lyle et al. (2004), and Dauphinot et al. (2005) Cortex, midbrain cerebellum Saran et al. (2003), Lyle et al. (2004), and Kahlem et al. DSCAM (Down syndrome cell adhesion molecule) (2004) Cortex Griffin et al. (1998) and Epstein (2001) S100b (S100 calcium-binding protein b)

Mental Retardation and Human Chromosome 21 Gene Overdosage 47

48

M. Rachidi and C. Lopes

The transgenic mice containing human SOD1 had 1.6–6-fold increased enzyme activity as compared to control, associated with decreased plasma serotonin levels and serotonin accumulation rate in transgenic mouse platelets (Epstein et al., 1987), a phenomenon similar to that reported in DS. Human SOD1 transgenic mice show impairment in the ability to adjust their posture in response to a moving surface and show mild deficits in sensori-motor responsiveness (Lalonde, Dumont, Paly, London, & Strazielle, 2004; Lalonde, Le Pecheur, Strazielle, & London, 2005). The  overexpression of Sod1 in transgenic mice leads to an impairment in LTP (Gahtan, Auerbach, Groner, & Segal, 1998) and defects in distal motor neuron terminals, indicating that this gene can selectively affect motor neurons (Avraham, Sugarman, Rotshenker, & Groner, 1991; Gurney et al., 1994). Moreover, these transgenic mice showed decreased cell number in several brain areas and decreased LTP in the pyramidal neuron CA1 (Harris-Cerruti et al., 2004; Zang et al., 2004). Premature ageing, one of the characteristics of DS that contributes to decreased cognitive performance in DS adults, may involve oxidative stress and impairment of proteasome activity. Transgenic mice overexpressing the human SOD1 gene show a reduction in proteasome activities in the cortex and an associated increase in the content of oxidised SOD1 protein (Le Pecheur et al., 2005). These results suggest a role of this gene in development of axons and motor neurons.

7.2 Amyloid Precursor Protein (APP) Gene Amyloide precursor protein (APP) gene encodes the beta-amyloid precursor protein, a protein involved in senile plaque formation in DS and AD (Kang et al., 1987). APP is widely expressed in axons, dendrites, and synapses in both central and peripheral nervous systems. In DS and Ts65Dn, APP is expressed at more than the expected 1.5-fold (Epstein, 2001; Hunter et al., 2003; Kahlem et al., 2004; Lyle et  al., 2004), suggesting that other genes on HSA21 directly or indirectly can ­further up-regulate the APP gene. The transgenic mice TgAPP exhibited overexpression of APP in the neocortex and hippocampus region mimicking features of DS. These amyloid precurseur protein transgenic models with AD-like pathology showed dystrophic neuritis associated with congopholic plaques (Sturchler-Pierrat et  al., 1997) and also showed learning defects (Lamb et al., 1993). APP transgenic mice have been tested in the Morris water maze tests and they show impairment in the probe test, measuring the reference memory, and impaired performance in the reverse probe test, measuring the spatial working memory (Janus, 2004). APP-null mice show impairment in the formation of LTP in the CA1 hippocampal region, and paired-pulse depression of GABA-mediated inhibitory post-synaptic currents is also attenuated, indicating that the impaired synaptic plasticity in APP deficient mice is associated with abnormal neuronal morphology and synaptic function within the hippocampus (Seabrook et al., 1999). Hippocampal neurons lacking APP show significantly enhanced amplitudes of evoked AMPA- and N-methyl-d-aspartate

Mental Retardation and Human Chromosome 21 Gene Overdosage

49

(NMDA)-receptor-mediated excitatory postsynaptic currents (EPSCs), and increased size of the readily releasable synaptic vesicle pool, indicating that lack of APP increases the number of functional synapses (Priller et  al., 2006). These findings suggest a role of APP in the neurophysiology of AD and DS.

7.3 v-ets Erythroblastosis Virus E26 (ETS2) Gene The HSA21 protooncogene ETS2 encodes a transcription factor ETS2 (Watson et al., 1985), and alteration of its expression has been implicated in the pathophysiological features of DS. ETS2 is expressed in neurons and is crucial for the normal formation of the neuromuscular junction (de Kerchove et al., 2002), and ETS2 is overexpressed in trisomic tissues (Dauphinot et al., 2005; Lyle et al., 2004; Saran et al., 2003). Null mice homozygous for mutation of ETS2 are embryonic lethal and show trophoblast alteration (Yamamoto et  al., 1998). Transgenic mice overexpressing ETS2 develop neurocranial and cervical skeletal abnormalities (Sumarsono et al., 1996), similarly to trisomy 16 mice and DS patients. The overexpression of ETS2 induces neuronal apoptosis, suggesting that overexpression of ETS2 may contribute to the increased rate of apoptosis of neurons in DS (Wolvetang, Bradfield, Hatzistavrou et al., 2003). It has been found that ETS2 protein transactivates APP gene and that fibroblasts overexpressing ETS2 show molecular abnormalities seen in DS such as elevated expression of APP gene and increased beta-amyloid proteins (Wolvetang, Bradfield, Tymms et al., 2003). These findings suggest that ETS2 overexpression in DS determines overexpression of APP and may play a role in the pathogenesis of the brain abnormalities in Alzheimer disease and DS.

7.4 S100 Calcium Binding Protein B (S100B) Gene S100B is a calcium-binding protein synthesised and released by astrocytes in response to serotonin (5-HT)-mediated stimulation of 5-HT1A receptors, and is an important extracellular neurotrophic agent during normal foetal brain development, with effects on neuroblasts and glia, involving the neuronal cytoskeleton (Azmitia, Griffin, Marshak, Van Eldik, & Whitaker-Azmitia, 1992; Morii et al., 1991). It has been found that S100b null mice develop normally, with no evident alterations in the cytoarchitecture of the brain. However, they have enhanced LTP in the hippocampal CA1 region and also enhanced spatial memory in the Morris water maze tests and fear memory in the contextual fear conditioning. These results indicate that S100b is a glial modulator of neuronal synaptic plasticity and of information processing in the brain (Nishiyama, Knopfel, Endo, & Itohara, 2002). The S100B RNA and protein are overexpressed in DS brain and Alzheimer disease (Epstein, 2001; Griffin et al., 1998). Transgenic mice overexpressing mouse

50

M. Rachidi and C. Lopes

S100b or human S100B show changes in cytoskeletal markers, such as the ­dendritic-associated protein, MAP-2, the growth-associated protein-43 and the dendritic spine marker, drebrin, leading to an increased density of dendrites within the hippocampus (Shapiro and Whitaker-Azmitia, 2004). Interestingly, drebrin protein is decreased in DS and AD brain regions (Kojima & Shirao, 2007; Shin & Lubec, 2002). Alterations have also been found in astrocyte morphology and axonal sprouting, especially in the dentate gyrus of the S100B transgenic mice (Bell, Shokrian, Potenzieri, & Whitaker-Azmitia, 2003; Reeves et  al., 1994; Shapiro & WhitakerAzmitia, 2004). Mice overexpressing S100B show decreased spatial learning and memory in the Morris water maze, radial-arm maze and Y-maze (Bell et al., 2003; Gerlai & Roder, 1996; Whitaker-Azmitia et al., 1997; Winocur, Roder, & Lobaugh, 2001). These results suggest that S100B overexpression contributes to glial-neuronal interactions, dendritic abnormalities and MR in DS.

7.5 Dual-Specificity Tyrosine Y Kinase 1 Subunit A (DYRK1A) Gene DYRK1A has been initially identified as the human homolog of the Drosophila minibrain gene, MNB, and is involved in neuroblast proliferation and reduction of the adult Drosophila brain (Tejedor et  al., 1995). DYRK1A encodes a serinethreonine kinase (Kentrup et  al., 1996). DYRK1A is expressed in the cortex, hippocampus and cerebellum (Guimera, Casas, Estivill, & Pritchard, 1999; Guimera et al., 1996; Rahmani, Lopes, Rachidi, & Delabar, 1998) and is overexpressed in mouse trisomic model Ts65Dn (Guimera et al., 1999), in DS foetal brain and in other trisomic tissues (Lyle et al., 2004; Saran et al., 2003). The Dyrk1A mutant mice are lethal during gestation. The heterozygote mice (Dyrk1A+/−) show a decreased size in several brain regions, a decreased neuronal cell number in the superior colliculus, an increased neuronal density in the cortex and in the thalamus, and exhibit neurobehavioural delays and defects (Fotaki et al., 2002). At cellular level, Dyrk1A+/− mice show smaller size of the pyramidal cell somata, shorter dendritic length, lower spine number, and altered spine distribution, suggesting the implication of Dyrk1A in the capability of the pyramidal cells to integrate information (Benavides-Piccione et al., 2005). Two transgenic mouse models overexpressing DYRK1A have been generated. The first one carried a human YAC 152F7, containing DYRK1A (Smith et  al., 1995), while the second carried the full-length DYRK1A cDNA (Altafaj et  al., 2001). In the Morris water test, the transgenic lines carrying the YAC 152F7 showed lower performance in the probe test, in which the platform is removed. In the reverse learning paradigm, the transgenic mice showed the most severe deficits with no significant learning of the new platform position, indicating deficits in learning flexibility (Smith et  al., 1997). A mouse line carrying a 152F7 YAC fragment (152F7tel) containing only the DYRK1A gene showed the same phenotype to the original YAC lines demonstrating that the overexpression of DYRK1A

Mental Retardation and Human Chromosome 21 Gene Overdosage

51

is responsible for the learning and memory defects in these mice (Smith et  al., 1997). Moreover, these transgenic mice showed increased brain size and neuronal size (Branchi et al., 2004; Rachidi et al., 2007). The significant impairment in spatial learning and memory observed in the two mouse models overexpressing DYRK1A indicates that the correct dosage of DYRK1A gene is crucial for brain hippocampal and prefrontal cortex functions, particularly concerning a cognitive dysfunction of the reference memory (Altafaj et  al., 2001; Smith et  al., 1997). Moreover, the transgenic mice overexpressing Dyrk1A exhibit neurodevelopmental defects, delayed craniocaudal maturation and motor dysfunction (Altafaj et  al., 2001; Fotaki et  al., 2002). Recently, DYRK1A bacterial artificial chromosome (BAC) transgenic mice have also shown learning and memory defects (Ahn et al., 2006). In addition, these mice showed abnormal LTP and long-term depression (LTD), suggesting synaptic plasticity alteration (Ahn et al., 2006). These phenotypes are comparable with those found in murine models of DS with trisomy of chromosome 16, and suggest a causative role of DYRK1A in MR in DS patients. Dyrk1A proteins are transported through the neuron dendrites and regulate their development (Hammerle et al., 2003), as also demonstrated by the overexpression of a kinase-deficient DYRK1A that impedes neurite outgrowth (Yang, Ahn, & Chung, 2001). Moreover, DYRK1A is co-localised in dendrites with Dynamin 1 (DYN1), a GTPase putative substrate of DYRK1A, involved in synaptic vesicle recycling, membrane trafficking and neurite outgrowth (Chen-Hwang, Chen, Elzinga, & Hwang, 2002; Hammerle et al., 2003). Dyrk1A proteins also modulate the activity of the CREB, which participates in signal transduction pathways involved in synaptic plasticity and neuronal differentiation (Hammerle et al., 2003). DYRK1A is involved in several pathways and it has recently been demonstrated that it can influence the NFATc pathways through its kinase activity (Arron et al., 2006; Gwack et al., 2006).

7.6 Single-Minded (SIM2) Gene SIM2, the first gene identified in the DSCR region (Dahmane et al., 1995), shows a high homology with the Drosophila single minded gene, sim, encoding a transcription factor/helix-loop-helix protein (Crews, Thomas, & Goodman, 1988). The Drosophila sim is a master gene of the midline development in the central nervous system, functioning as transcriptional regulator in cell fate determination (Crews et  al., 1988; Nambu, Lewis, Wharton, & Crews, 1991; Thomas, Crews, & Goodman, 1988). The mammalian Sim2 is expressed in the embryonic brain in delimited regions of the neuroepithelium of D1 and D2 neuromeric regions and along the neural tube (Dahmane et al., 1995; Ema et al., 1996; Fan et al., 1996; Rachidi, Lopes, Charron et al., 2005). In later human foetal stages, SIM2 gene expression is found at different levels in discrete human brain regions, including the cortical layers, the hippocampus and the cerebellum (Rachidi, Lopes, Charron et al., 2005), which are key regions

52

M. Rachidi and C. Lopes

involved in learning and memory, and are also altered in DS patients (Golden & Hyman, 1994; Ito, 2002; Milner et al., 1998; Miyashita, 2004; Raz et al., 1995). Sim2 is overexpressed about 1.5-fold in Ts1Cje mouse foetuses (Vialard et  al., 2000) and in trisomic tissues (Lyle et al., 2004). Transgenic mice overexpressing Sim2 display reduced sensitivity to pain and mild impairment of learning (Chrast, Scott, Papasavvas et al., 2000; Ema et al., 1999). These behavioural anomalies found in the Sim2 transgenic mice recall some phenotypes observed in trisomic mouse models for DS, Ts65Dn and Ts1Cje (Coussons-Read & Crnic, 1996; Martinez-Cue et al., 1999; Sago et al., 1998), and in DS patients (Hennequin, Morin, & Feine, 2000). Sim2 mutant mice are lethal in the early post-natal days and show skeletal alteration due probably to cell proliferation defects (Goshu et al., 2002). Functional studies indicated that SIM2 protein control the Shh expression in the brain (Epstein et al., 2000), involved in cell growth and differentiation in the brain. Moreover, SIM2 can inhibit cell cycle by inhibition of cyclin E expression (Meng, Shi, Peng, Zou, & Zhang, 2006) suggesting a key role of SIM2 in neulogical alterations seen in DS.

7.7 Regulator of the Calcineurin (RCAN1) Gene Also called Down syndrome critical region 1 gene (DSCR1) or myocyte-enriched calcineurin-interacting protein 1 (MCIP1) or calcipressin 1 (CSP1), the regulator of the calcineurin 1 protein (RCAN1) gene directly modulates the activity of the protein phosphatase, calcineurin. The CSP1, the protein encoded by DSCR1, interacts with calcineurin A (Fuentes et al., 2000) to inhibit calcineurin activity (Rothermel, Vega, & Williams, 2000). DSCR1 is highly expressed in brain and heart (Fuentes et  al., 1995). It is overexpressed in the brain of DS foetuses (Fuentes et al., 2000), in brains from DS patients with AD symptoms (Ermak, Morgan, & Davies, 2001), and in the brain of DS mouse models (Amano et al., 2004; Dauphinot et al., 2005; Lyle et al., 2004). It has been observed that the calcineurin activity is decreased in AD (Ladner, Czech, Maurice, Lorens, & Lee, 1996), in DS foetal brain tissue, and in Drosophila mutants that overexpress DSCR1 (Chang, Shi, & Min, 2003). As DSCR1 is an inhibitor of calcineurin activity, it is possible that these changes could be caused by increased levels of DSCR1, as occurs in DS, promoting the development of AD. Indeed, overexpression of DSCR1 in rat primary neurones causes formation of aggresome-like inclusion bodies similar to those observed in DS and AD brains, as well as reducing the expression of the synaptic vesicle protein, synaptophysin, in neural processes (Ma et al., 2004). Interestingly, loss-of-function and overexpression of mutants of nebula, the Drosophila orthologue of DSCR1, both display severe learning defects in several basic learning assays, indicating that DSCR1 may affect regulatory pathways of synaptic transmission (Chang et al., 2003). In agreement with this indication, forebrain-specific knock-out of calcineurin in mice results in impaired hippocampal-dependent memory tasks and synaptic plasticity (Zeng et  al., 2001). DSCR1/Mcip1−/− mice have an impaired cardiac hypertrophic response to

Mental Retardation and Human Chromosome 21 Gene Overdosage

53

pressure overload, suggesting that this gene may also function as a calcineurin facilitator in  vivo (Vega et  al., 2003). Mice deficient in Mcip1/2 show more dramatic impairment in cardiac hypertrophy than the DSCR1−/−. Moreover, these DSCR1 knock-out mice displayed a neurological phenotype and showed faster overall movements in an open field test and a significant impairment in working memory, as assessed by novel and familiar object recognition analysis (Sanna et al., 2006). Recently, DSCR1 has also been demonstrated to play a role in memory and synaptic plasticity by examining the behavioural and electrophysiological properties of DSCR1 knock-out mice (Hoeffer et  al., 2007). These mice exhibit deficits in spatial learning and memory, reduced associative cued memory, and impaired latephase long-term potentiation (L-LTP), phenotypes similar to those of transgenic mice with increased calcineurin activity. Consistent with this, the DSCR1 knock-out mice display increased enzymatic calcineurin activity, increased abundance of a cleaved calcineurin fragment, and decreased phosphorylation of the calcineurin substrate dopamine and cAMP-regulated phosphoprotein-32. These findings suggest that DSCR1 regulates LTP and memory via inhibition of phosphatase signalling (Hoeffer et al., 2007).

7.8 DOPEY2 Gene C21orf5 gene, that we recently renamed DOPEY2 following HUGO nomenclature, is a member of the Dopey family containing leucine zipper-like domains involved in multiple protein–protein interactions (Rachidi, Lopes, Costantine et al., 2005). DOPEY2 is more highly expressed in the differentiating zones than in the proliferating zones in embryonic human and mouse brain (Lopes et  al., 2003; Rachidi et al., 2006), suggesting a role of DOPEY2 in cell differentiation and developmental patterning. This potential role is also supported by the high homology of DOPEY2 with the Caenorhabditis elegans Pad-1, required for embryonic patterning during gastrulation (Guipponi et al., 2000), the yeast Dop1 and DopA, required for normal growth patterning, and cell differentiation and organogenesis in fungi (Dujon, 1996; Pascon & Miller, 2000). DOPEY2 expression becomes restricted to cerebellum, cortex and medial temporal-lobe system during foetal development and in adult brain (Lopes et al., 2003; Rachidi et al., 2006), in which this gene shows different transcriptional intensities, as demonstrated by an improved new optic technology allowing comparison of the cell density and the expression intensity (Rachidi et al., 2006). These findings are of the most interest because the medial temporal-lobe system, including the hippocampal formation and perirhinal cortex, works as a control centre of the memory circuits and storage (Krasuski, 2002; Milner et al., 1998) and also, the cortex and the cerebellum participate in elaboration of memory (Ito, 2002; Miyashita, 2004). DOPEY2 is expressed in brain regions that play key roles in learning and memory and present neuronal alterations in DS patients (Golden & Hyman, 1994; Raz et al., 1995), suggesting a role of this gene in the learning and memory.

54

M. Rachidi and C. Lopes

DOPEY2 is overexpressed in DS lymphoblasts 1.5–2-fold compared to normal lymphoblasts (Lopes et  al., 2003) and in trisomic tissues (Lyle et  al., 2004), suggesting that DOPEY2 is a dosage-sensitive gene. Transgenic mice carrying the human YAC 230E8 (Smith et al., 1995) contain the entire DOPEY2 gene (Lopes et al., 2003; Rachidi, Lopes, Costantine et al., 2005), and overexpress it at less than twofold (Lopes et al., 2003; Rachidi, Lopes, Costantine et al., 2005). These transgenic mice show increased cortical cell density (Rachidi, Lopes, Costantine et al., 2005; Smith et al., 1997) that overexpresses DOPEY2 (Rachidi, Lopes, Costantine et al., 2005). This phenotype corresponds well to the abnormal lamination pattern found in the cortex of DS patients (Golden & Hyman, 1994). Recently, a new cerebellar phenotype has been discovered in two independent mouse lines carrying the YAC 230E8 characterised by elongation of the anteroposterior axis, increased length of rostral folia of the vermis, and abnormal culmen and declivus lobules (Rachidi et  al., 2007). These phenotypes in the cortex and cerebellum may also explain the learning and memory deficits of these mouse models (Rachidi, Lopes, Costantine et al., 2005; Rachidi et al., 2007; Smith et al., 1997) suggesting a role of the DOPEY2 gene in neuropathological defects and MR in DS.

7.9 Potassium Inwardly Rectifying Channel (KCNJ6) Gene Potassium inwardly rectifying channel J6 (KCNJ6) or G-protein coupled inward rectifying potassium channel subunit 2 (GIRK2) (Lesage et al., 1994; Ohira et al., 1997) encodes the GIRK2, a member of the ATP-sensitive potassium channels, involved in increase of the intracellular ATP concentration, linking cellular metabolism to the electrical excitability of the plasma membrane. GIRK2 is highly expressed in the brain, particularly in the cerebellar granule cell, suggesting a role of Girk2 in granule cell differentiation (Goldowitw & Smeyne, 1995). Girk2 mutation is responsible of the weaver phenotype in mouse, characterised by a drastically reduced cerebellum due to the depletion of granular cell neurons (Patil et al., 1995). GIRK2 is overexpressed in trisomic tissues (Kahlem et  al., 2004; Lyle et  al., 2004; Saran et al., 2003) and in the brain of DS mouse model Ts65Dn, particularly in the hippocampus (Harashima, Jacobowitz, Witta et al., 2006), and determines an increase of GIRK channel density in Ts65Dn neurons and a twofold increase of GABAB-mediated GIRK current (Best, Siarey, & Galdzicki, 2006). In addition, the GIRK2 overexpression seems to alter the GIRK1/GIRK2 ratio, which likely affects the balance between excitatory and inhibitory neuronal transmission in Ts65Dn, and thus overexpression of GIRK2 could contribute to DS neurophysiological phenotypes (Best et  al., 2006; Harashima, Jacobowitz, Stoffel et  al., 2006; Harashima, Jacobowitz, Witta et al., 2006). Girk2 heterozygous animals show behavioural changes intermediate between wild-type and null mutants only in the elevated plus-maze test after social isolation (Harashima, Jacobowitz, Stoffel et  al., 2006), in agreement with gene-dosage

Mental Retardation and Human Chromosome 21 Gene Overdosage

55

dependence of RNA and protein expression (Blednow, Stoffel, Chang, & Harris, 2001). Girk2 homozygous mice show a higher level of locomotion and a higher number of rearing in the light area in the light/dark box. In the elevated plus-maze test, these mice spent a higher percentage of time in the open arms and showed a higher number of total entries. These results suggest hyperactivity and reduced anxiety in Girk2 null mice (Harashima, Jacobowitz, Stoffel et al., 2006).

7.10 Tetratricopeptide Down Syndrome (TPRD) Gene TPRD, also called TTC3, gene (Ohira et al., 1996; Tsukahara, Hattori, Muraki, & Sakaki, 1996), is localised in the DSCR region. TPRD protein contains 2–3 units of a 34-amino acid repeat similar to the tetratricopeptide (TPR) motif, in the different splicing forms (Dahmane et al., 1998), that are involved in protein–protein interactions (Das, Cohen PW, & Barford, 1998; Groves & Barford, 1999). TPRD shows regional and cellular specificity during mouse and human brain development, and its expression is higher in the differentiating areas than in the proliferating ones, suggesting a role of TPRD in neuronal cell differentiation (Lopes et  al., 1999; Rachidi et  al., 2000). The strong TPRD expression in the human foetal cortex corresponds to the crucial developmental stage when the size of the cortical mantle doubles in thickness and the cortical lamination begins, suggesting a role of TPRD in cortical lamination. Interestingly, TPRD shows a differential expression in the human foetal hippocampus and cerebellum with variable intensities in specific neuronal cell types as estimated by a novel microscopy technology allowing gene transcription quantification (Rachidi et al., 2000). This specific expression pattern corresponds well to abnormal brain regions seen in DS patients (Golden & Hyman, 1994; Raz et al., 1995). TPRD is overexpressed at more than 1.5-fold in trisomic tissues including brain (Amano et  al., 2004; Lyle et al., 2004; Saran et al., 2003). To date, two transgenic mice have been produced carrying human YACs, 141G6 and 152F7, each containing TPRD gene (Fig. 1) (Smith et al., 1997). The neurological and behavioural phenotypes observed in the 152F7 mice are determined by the DYRK1A gene overexpression. The 141G6 mice showed a lower performance in fear-conditioning against sound as acoustic conditional stimulus (Chabert et al., 2004), although any evident neuroanatomical and cognitive defects have not yet been demonstrated in the 141G6 mice (Smith et al., 1997). In addition, the 141G6 mice showed aberrant protein expression compared to control mice. Interestingly, these expression protein alterations may potentially lead to impairment of cognitive functions. In particular, these mice showed decreased CaMKII protein in hippocampus (Shin et  al., 2006), and it is known that alteration of the CaMKII-pathway lead to a downstream alteration of the CREB pathway associated with impairment of fear memory (Bourtchuladze et al., 1994). Recently, it has been demonstrated that TPRD/TTC3 protein interacts with citron kinase (CIT-K) and citron N (CIT-N), two effectors of the RhoA small

56

M. Rachidi and C. Lopes

GTPase, involved in neuronal proliferation and differentiation (Berto et al., 2007). Interestingly, these authors demonstrated that TPRD/TTC3 overexpression strongly inhibits neurite extension, while its knock-down stimulates the neurite extension. In agreement with the previous results (Lopes et al., 1999; Rachidi et al., 2000), these dose-dependent effects are Rho-dependent, suggesting an important role of the TPRD-RhoA-CIT-K in neuronal differentiation (Berto et al., 2007). Overall, the expression pattern, the high overexpression of TPRD in the brain, and its role in neuronal differentiation suggest that this gene could be involved in fine neurological alterations in DS patients, and could participate to MR pathogenesis.

7.11 Down Syndrome Cell Adhesion Molecule (DSCAM) Gene Down syndrome cell adhesion molecule (DSCAM), an axon guidance molecule mapping to HSA21 (Schmucker et al., 2000; Yamakawa et al., 1998), is expressed in the developing spinal cord and cortex. This gene may be involved in definition of the dorsal–ventral axis in the developing spinal cord at the time of neurite extension and may participate in the determination of regional neuronal fates in the developing cortex (Barlow, Micales, Chen, Lyons, & Korenberg, 2002). DSCAM proteins are expressed in the cerebral and cerebellar white matter, in accordance with the temporal and spatial sequence of myelination. In DS brains, DSCAM protein level is increased in the Purkinje cells at all ages, and in the cortical neurons during adulthood, compared to that for controls. In demented DS patients, DSCAM protein appeared in the core and periphery of senile plaques. This DSCAM expression pattern suggests that this gene may play a role as an adhesion molecule regulating myelination. In addition, the overexpression of DSCAM may also play a role in the MR and the precocious dementia of DS patients (Saito et al., 2000). The DSCAM Drosophila melanogaster homolog, dDscam, has been well characterised, and an important role has been demonstrated in neuronal wiring specificity (Chen et al., 2006; Hattori et al., 2007). During nervous system development, commissural axons project towards and across the ventral midline, a process mediated by netrin-1 and the netrin-1 receptor. It has been demonstrated recently that DSCAM is also required for commissural axon guidance. DSCAM is expressed on spinal commissural axons, binds netrin-1, and is necessary for commissural axons to grow towards and across the midline. Thus, overexpression of DSCAM, by causing enhanced netrin-DSCAM interactions, could contribute to the axonal wiring defects seen in DS (Ly et al., 2008).

7.12 Synaptoganin 1 (SYNJ1) Gene Synaptojanin 1 (SYNJ1) is a polyphosphoinositide phosphatase that dephosphorylates the phosphatidylinositol-4,5-bisphosphate, a lipid that regulates membrane

Mental Retardation and Human Chromosome 21 Gene Overdosage

57

transduction and membrane trafficking in the endocytic pathway at synapses. The phosphatidylinositol is a signalling phospholipid implicated in a wide variety of cellular functions. At synapses, where normal phosphatidylinositol balance is required for proper neurotransmission, the phosphoinositide phosphatase synaptojanin 1 is a key regulator of its metabolism. Synaptojanin is highly enriched in the brain and is located at nerve terminals, and is associated with synaptic vesicles and coated endocytic intermediates (Haffner et al., 1997; McPherson, Takei, Schmid, & De Camilli, 1994). Moreover, the distribution of synaptojanin is coincident with that of other endocytic proteins, such as amphiphysin and dynamin (McPherson, Garcia, Slepnev, David, & Zhang, 1996; McPherson et al., 1994). For these reasons, synaptojanin may play a role in the endocytosis and could be involved in the recycling of synaptic vesicles. Synaptojanin 1 mutant mice die early after birth and exhibit accumulation of clathrin-coated vesicles at nerve terminals and increased synaptic depression in hippocampus, supporting a role for synaptojanin in the uncoating step of the recycling pathway (Cremona et al., 1999). The role that synaptojanin 1 plays in specific steps of the synaptic vesicle cycle has been studied by combined quantitative imaging with electron microscopy on cultured cortical neurons from synaptojanin 1 knock-out mice (Kim et al., 2002). The findings indicate that the rapid degradation of phosphatidylinositol by synaptojanin 1 is of critical importance for efficient synaptic vesicle regeneration and for the recovery of normal presynaptic function after an exocytic burst. In the absence of synaptojanin 1, sustained activity leads to a kinetic delay in synaptic vesicle reformation and to an increased, transient backup of synaptic membrane. This study provides direct evidence for the hypothesis that synaptojanin 1 plays a key physiologic role in the transition from early endocytic compartments to newly reformed synaptic vesicles fully incorporated into the functional pool. These results provide new evidence for a critical role of phosphoinositides in synaptic physiology and for their importance in regulating membrane traffic in the presynaptic terminal (Kim et al., 2002). It has been found that phosphoinositide metabolism is altered in the brain of Ts65Dn mice and in transgenic mice overexpressing Synj1 from BAC constructs. This defect is rescued by restoring Synj1 to disomy in Ts65Dn mice. The Synj1 transgenic mice also exhibit deficits in performance of the Morris water maze task, suggesting that phosphoinositide dyshomeostasis caused by gene dosage imbalance for Synj1 may contribute to brain dysfunction and cognitive disabilities in DS (Vonorov et al., 2008).

7.13 Intersectin 1 (ITSN1) Gene The human ITSN1 gene spans 250 kb of genomic DNA and maps to HSA21 (Pucharcos et al., 1999). ITSN1 protein has five consecutive SH3 domains (SH3A-E), commonly found in signal transduction and cytoskeletal proteins (Pawson, 1995),

58

M. Rachidi and C. Lopes

and which interact with proline-rich domain-containing proteins involved in ­clathrin-mediated synaptic vesicle endocytosis (McPherson, 1999). ITSN1 encodes two isoforms, a long and a short, and both are expressed in the brain, but the long form is neuronal-specific and the short form is expressed in glial cells (Hussain et  al., 1999; Ma et  al., 2003). In addition, the long isoform is expressed in zones of proliferating and differentiating neurones, in both adult and foetal mouse brain, and this long ITSN1 transcript is overexpressed in the brains of individuals with DS (Pucharcos et al., 1999). ITSN1 seems to function as a scaffolding protein, providing a link between the components of endocytosis and the actin cytoskeleton, and has a role in signal transduction (Hussain et al., 1999; Jenna et al., 2002; Martina, Bonangelino, Aguilar, & Bonifacino, 2001; Roos & Kelly, 1998; Yamabhai et al., 1998). Interestingly, ITSN1 overexpression blocks clathrin-mediated endocytosis (Sengar, Wang, Bishay, Cohen, & Egan, 1999), presumably through disruption of higher order protein complexes between ITSN1 and its binding partners. Also, ITSN1 overexpression was found to block epidermal growth factor (EGF)-mediated MAPK activation by inhibiting Ras activation directly, most probably by preventing the Ras/mSos interaction, suggesting a role for ITSN1 in Ras activation (Tong et al., 2000). Mice with a null mutation in Intersectin 1 (Itsn1) showed alterations in endocyic and vesicle trafficking, including reduced number of exocytosis events in chromaffin cells, slowing of endocytosis in neurons, increased endosome size in neurons and reduced nerve growth factor (NGF) levels and decreased levels of choline acetyl transferase (ChAT) positive cells in the septal region of the brain (Yu, Chu, Bowser, Keating, & Dubach, 2008). Interestingly, the presence of enlarged endosomes in the neurons of DS is an early sign of AD, suggesting that ITSN1 could contribute to the disturbance in endocytosis in early AD pathogenesis in DS (Yu et al., 2008).

7.14 Contribution of MicroRNAs in Down Syndrome Mental Retardation MicroRNAs (miRNAs) are small, non-protein coding RNAs that base pair with specific mRNA targets leading to translational repression or mRNA cleavage (Bartel, 2004; Bushati & Cohen, 2007; Wang et al., 2007). MiRNAs are expressed as long primary transcripts that are subsequently processed into mature miRNAs (about 22 nucleotides) by several nuclear and cytoplasmic enzymatic steps (Bushati & Cohen, 2007; Wang et al., 2007). MiRNAs have been shown to play a fundamental role in diverse biological and pathological processes (Bushati & Cohen, 2007; Wang et al., 2007). Recently, five miRNA genes harboured on HSA21, including miR-99a, let-7c, miR-125b-2, miR-155, and miR-802, have been identified (Kuhn et  al., 2008). Among these miRNAs, the bic/miR-155 gene is well characterised and its ­expression is regulated by lipopolysaccharide (LPS) and cytokines (O’Connell, Taganov, Boldin, Cheng, & Baltimore, 2007; Taganov, Boldin, Chang, & Baltimore, 2006).

Mental Retardation and Human Chromosome 21 Gene Overdosage

59

As described above, alteration of the expression of the five HSA21-derived miRNAs in DS has been studied, demonstrating that they are overexpressed in DS tissues when compared with normal tissues (Kuhn et  al., 2008; Sethupathy et  al., 2007). Moreover, other miRNAs have also been identified showing expression alteration (over- or downexpression) in DS tissues when compared with normal tissues (Kuhn et al., 2008). These data suggest that trisomic 21 gene dosage overexpression of HSA21-derived miRNAs results in the decreased expression of specific target proteins and contributes, in part, to features of the neuronal and cardiac DS phenotype. Importantly, HSA21derived miRNAs may provide novel therapeutic targets in the treatment of individuals with DS.

8 Potential Molecular Pathways and Mechanisms Involved in Mental Retardation of Down Syndrome The important update of the genomic sequence and the identification of HSA21 genes, determination of candidate genes and their mouse orthologs for DS phenotypes, particularly those involved in brain alterations, learning and memory deficits, and also the development and improvement of public databases, datamining and algorithms to perform genome wide analyses (Gardiner, Fortna, Bechtel, & Davisson, 2003; Hattori et  al., 2000; International Human Genome Sequencing Consortium, 2004; Kapranov et  al., 2002; Nikolaienko, Nguyen, Crinc, Cios, & Gardiner, 2005), has produced important knowledge and tools to study the molecular effects of the expression variation of gene products from the triplicated genes and their functional variations predisposing to specific cognitive deficits in the goal to better understand the molecular pathophysiology of MR in the DS.

8.1 Molecular and Cellular Mechanisms Leading to Mental Retardation in Down Syndrome Recently, we have proposed a global mechanism model explaining the molecular and cellular origin of MR in DS (Rachidi & Lopes, 2007). In this model, we considered the complexity of gene interactions allowing the gene expression variations caused by the gene overdosage. These expression variations may firstly induce functional alterations at cellular level in the brain, that we called the primary phenotypes, and the final combination of these neuronal alterations could determine brain morphological defects, behavioural alterations and the MR in DS, that we called the secondary phenotypes (Fig. 2). We proposed that three principal genetic mechanisms could participate in concert to determine the final transcriptome and proteome alteration in DS. First,  some

3 copies of HSA21 genes Dosage sensitive genes 1.5-fold expression level

Other HSA21 genes + no HSA21 genes Secondary gene dosage effect

Amplified developmental instability

Genes expression different of 1.5-fold

First brain phenotypes

Neuronal differentiation

Neurosignaling pathways Neurometabolic pathways

Dendritogenesis Myelination

Neuronal identity

Synaptogenesis

LTP/LTD

Axonal growth

Neuronal migration

Secondary brain phenotypes

Cerebellar alterations

Cortex alterations

BFCN alterations

Learning defects

Hippocampal alterations Memory defects

Behavioural alterations

MENTAL RETARDATION

Fig. 2  Molecular and cellular mechanism leading to neurological phenotypes and mental retardation in Down syndrome. In this model, the overdosage of the HSA21 genes induces global alterations of gene expression via different genetic mechanisms. On the one hand, a primary effect of the gene overdosage determines a 1.5-fold gene overexpression. On the other hand, the overexpressed genes on the HSA21 could variably modify the expression level of other trisomic genes on HSA21 and of disomic genes on other chromosomes, by a secondary gene effect or by a more general alteration of the transcriptional homeostasis (amplified developemental instability). These gene expression variations may firstly induce functional alterations at cellular level in the brain, that we called the primary phenotypes. These neuronal alterations could determine the neuromorphological, neurological, and behavioural alterations, which we called the secondary phenotypes. The final combination of these neurological alterations leads to the mental retardation in DS

Mental Retardation and Human Chromosome 21 Gene Overdosage

61

HSA21 genes, the dosage-dependent genes, are expressed on average 1.5-fold the normal level as a direct consequence of the 1.5 gene overdosage. In some cases, their overexpression could directly determine a primary phenotype in DS brain. Second, in addition, some of these dosage-dependent genes could modulate directly or directly the expression of target genes on both the HSA21 and the other chromosomes. In these cases, the modulation of the target gene expression could be different of 1.5-fold up-regulation and could also be down-regulated, depending on the nature of the dosage-dependent gene products, and thus determine a secondary gene dosage effect. And three, all or some HSA21 genes with altered expression could participate in the amplified developmental instability of the genome in trisomy 21, determining a more general gene expression alteration and disequilibrium (Fig. 2). These gene expression changes in the brain determine alterations at cellular level, which we globally call primary phenotypes (Fig.  2), include metabolic pathways, regulatory cascades and cellular processes, such as proliferation, differentiation and apoptosis. Recently, alterations in synaptogenesis and dendritogenesis haver received increasing interest for their implication in MR. Abnormal ultrastructure and number of synapses and dendrites are observed both in DS patients and mouse models (Becker et al., 1986, 1991; Belichenko et al., 2004; Benavides-Piccione et al., 2004; Dierssen & Ramakers, 2006; Hanson et al., 2007; Kurt, Davies, Kidd, Dierssen, & Florez, 2000; Kurt, Kafa, Dierssen, & Davies, 2004; Takashima et  al., 1994). Moreover, the synapses of trisomic brains also show functional alterations, such as LTP and LTD in the hippocampus and dentate gyrus (Kleschevnikov et al., 2004; Siarey et al., 1997, 1999, 2005), reduced noradrenergic function in the hippocampus (Dierssen et al., 1996, 1997), and reduced excitatory and inhibitory inputs to pyramidal neurons in CA3 of Ts65Dn hippocampus (Hanson et al., 2007). These primary phenotypes and their combinations may determine the more complex secondary phenotypes, in particular functional alterations of the cognitive network and brain plasticity that participate in the MR in DS (Fig. 2). As an elucidated example, during cerebellar development in Ts65Dn mice, a primary cellular phenotype has been identified to be the molecular origin of a DS secondary neurological phenotype (Roper et al., 2006). In Ts65Dn mice, a cerebellar hypoplasia is observed, due to decreased cerebellar granular cells and their precursors. It has been demonstrated that reduced mitosis is determined by a deficit in response to the Sonic hedgehog (Shh) mitogenic signals (Roper et al., 2006). This suggests that a dosage-sensitive gene or genes make cells less sensitive to Shh when overexpressed, corresponding to the primary phenotype, and that determines the cerebellar hypoplasia corresponding to the secondary phenotype.

8.2 Molecular Pathways Contributing to Mental Retardation in Down Syndrome The genetic disruption caused by trisomy 21 in neural patterning and signal transduction pathways during development leads to alteration of the neuronal circuitry

62

M. Rachidi and C. Lopes

and could be the biological mechanism responsible for the pathogenesis of MR in DS. Thus, it is of the most interest to elucidate the molecular pathways involving the HSA21 genes function and to discover which of them are perturbed in trisomy 21 and are relevant to neurological disorders, cognition, learning and memory in DS. Importantly, the first altered genetic pathway involved in some DS phenotypes has been identified (Arron et al., 2006) in which two HSA21 dosage-sensitive genes are involved, DYRK1A and RCAN1/DSCR1, both located in the critical DSCR region, and impact nuclear factor of activated T cells (NFATc) activity. Transgenic mice harbouring different mutations in the NFATc transcription factors exhibit phenotypes similar to features seen in DS (Arron et  al., 2006). Moreover, it has been demonstrated that DYRK1A and DSCR1 regulate NFATc and their overexpression dysregulates the NFATc pathways (Arron et al., 2006). The NFATc pathways play critical roles in vertebrate development and organogenesis of several organs (Crabtree & Olson, 2002; Graef, Che, & Crabtree, 2001), in particular the central nervous system. The signalling pathways (Fig. 3) are activated by the entry of calcium into the cell and results in activation of calcineurin, the catalytic subunit of the Ca2+/calmodulindependent protein phosphatase PP2B. In the cytoplasm, activated calcineurin removes phosphate groups from NFATc factors. Dephosphorylated NFATc proteins enter the Ca2+ Cell membrane

P

DSCR1 NFATc Calcineurin

P P DYRK1A NFATc

Nucleus

Target gene

Fig.  3  Cooperation of gene-dosage imbalance of DYRK1A and DSCR1 dysregulate NFATc signalling pathway. The phosphatase calcineurin is activated by the calcium and removes phosphate groups (P) from NFATc transcription factors in the cytoplasm that allows the dephosphorylated NFATc proteins to enter into the nucleus and activate their target genes. DSCR1 is a cytoplasmic inhibitor of calcineurin and, thus, decreases NFATc dephosphorylation. In the nucleus, NFATc may be phospharylated by DYRK1A kinase and the phosphorylated NFATc returns into the cytoplasm, decreasing gene transcription activity. In DS, overexpression of DSCR1 leads to a decrease in NFATc dephosphorylation and, consequently, to a reduction in nuclear NFATc and target gene transcription. Similarly, in the nucleus, overexpression of DYRK1A leads to increase of the phosphorylated NFATc proteins and their cytoplasmic translocation, and thus to additional decrease of target gene transcription. Thus, DYRK1A and DSCR1 proteins regulate the levels of NFATc phosphorylation and, in DS, their overexpression determines the dysregulation of NFATc-dependent gene expression and their associated phenotypic features

Mental Retardation and Human Chromosome 21 Gene Overdosage

63

nucleus and activate their target genes. DSCR1 encodes an inhibitor of calcineurin (Fuentes et al., 2000; Rothermel et al., 2000) and decreases NFATc dephosphorylation in the cytoplasm. Once in the nucleus, NFATc may be phosphorylated by DYRK1A serine-threonine protein kinase, and the phosphorylated forms of NFATc return into the cytoplasm, decreasing gene transcription activity. In normal conditions, DYRK1A and DSCR1 act synergistically to control phosphorylation levels of the NFATc and NFATc-regulated gene transcription (Fig.  3). In transgenic mice overexpressing Dyrk1A or DSCR1 alone, or both Dyrk1A and DSCR1, NFATc is mostly phosphorylated and found in the cytoplasm, suggesting that overexpression of Dyrk1A and DSCR1 reduces NFATc transcriptional activity by increase of the phosphorylated forms of NFATc proteins, which leads to their cytoplasmic localisation. Interestingly, the mice lacking NFATc2 and NFATc4, such as transgenic mice overexpressing DYRK1A and/or DSCR1, have similar phenotypes to those seen in trisomic mouse models, Ts65Dn and Ts1Cje, and DS patients, including neuronal and behavioural phenotypes (Arron et al., 2006). In agreement with this molecular pathway, significant reduced calcineurin activity is detected in DS foetal brain tissue as well as in Drosophila mutants that overexpress DSCR1 (Chang et al., 2003). As described above, both DYRK1A and DSCR1 are involved in synaptic development, maturation and plasticity. Moreover, DYRKIA also appears involved in splicing control. In the nuclei, DYRK1A is localised to the nuclear speckles that represent the splicing compartment (Alvarez, Estivill, & de la Luna, 2003), and several splicing factors and proteins involved in splicing regulation are DYRK1A substrates (de Graaf et  al., 2004). The splicing is a fundamental step of mRNA maturation and correct protein production, and almost all genes show alternative splicing forms. Overexpression of DYRK1A in DS may dysregulate the splicing control and may determine several developmental and functional alterations, particularly in the brain. Interestingly, several DYRK1A substrates, including the transcription factors GLI1, ARIP4, GR, FKHR and CREB (Mao et al., 2002; Sitz, Tigges, Baumgartel, Khaspekov, Lutz, 2004; Woods et al., 2001; Yang et al., 2001), are involved in the MAPK pathway. In addition, DSCR1 inhibites calcineurin and, indirectly, affects its substrates, which include dynamin and SYNJ1 (Cousin, Tan, & Robinson, 2001), and is a critical point of connection between the calcineurin and the MAPK pathways (Rothermel et  al., 2003). Other genes of the chromosome 21 are involved in the MAPK pathway, such as SUMO3, involved in sumoylation, a post-translational process regulating protein function and activity level, and NRIP1 (also known as RIP140), a steroid hormone co-repressor. The MAPK pathway is involved in regulation of synaptic plasticity and memory (Sweatt, 2001; Sweatt & Weeber, 2003; Thomas & Huganir, 2004). Dysregulation of MAPK affects LTP, spatial learning, and context discrimination, which correspond to some defects observed in Ts65Dn mice (Hyde, Frisone, & Crnic, 2001; Kleschevnikov et al., 2004; Stasko & Costa, 2004) and in children with DS (Pennington, Moon, Edgin, Stedron, & Nadel, 2003). Dysregulation of calcineurin activity is associated with cognitive and behavioural deficits, including spatial learning and context discrimination (Lee & Ahnn, 2004;

64

M. Rachidi and C. Lopes

Mansuy, 2003) that are related to DS. Calcineurin interacts with dynamin in a Ca2+dependent fashion during depolarisation-induced vesicle recycling (Liu, Sim, & Robinson, 1994) and, in Drosophila, disruption of this interaction blocks endocytosis and impairs neurotransmission (Kuromi, Yoshihara, & Kidokoro, 1997). Inhibition of calcineurin reduces the level of synaptic vesicle recycling as well as the total vesicle pool size in synaptic terminals (Kumashiro et al., 2005), indicating a potential role for endogenous calcineurin inhibitors in regulating synaptic transmission. It is plausible, given its role in the regulation of calcineurin activity, that overexpression of DSCR1, as observed in DS and AD, may adversely affect at least two calcineurin-dependent pathways by blocking calcineurin activity. Firstly, elevated levels of DSCR1 may disrupt endocytosis and vesicle recycling due to the inhibition of calcineurin-dependent dephosphin dephosphorylation, and secondly, may contribute to the hyperphosphorylation of Tau by reducing calcineurin phosphatase activity. Indeed, while short-term induction of the DSCR1 protein can provide stress protection in neurons, it has been proposed that long-term induction causes gradual accumulation of hyperphosphorylated tau protein, leading to AD (Ermak et al., 2001). DSCR1 knock-out mice showed increased enzymatic activity of both calcineurin and protein phosphatase 1 (PP1) and decreased phosphorylation of the calcineurin substrate DARPP-32, consistent with an elevation in calcineurin activity in the hippocampus of DSCR1 knock-out mice (Hoeffer et al., 2007), demonstrating a critical role for DSCR1 in the proper manifestation of memory and in MR in DS.

9 Potential Directions for Mental Retardation Therapeutics in Down Syndrome The functional genomic advances for generating gene–phenotype correlations are of the most interest towards the identification of potential targets in the molecular pathways involved in DS phenotype. The efforts are focused particularly on the pathways involved in learning and memory processes in mouse models of DS towards identification of targets for therapeutics that will correct the MR features in DS patients. It is known that individuals with DS present neuropathology indistinguishable from those with AD (Mann & Eisiri, 1989), including loss of acetylcholine and related enzymes in the hippocampus and throughout the neocortex. The cholinergic degeneration found in AD and DS led to the suggestion that pharmacological treatments directed at cholinergic systems might attenuate the degree of dementia in AD (Smith & Swash, 1978). Today, the most widely used treatment for dementia in AD is the administration of acetylcholinesterase inhibitors (AChEI), which enhance cholinergic transmission. Donepezil is a selective AChEI, which produces clinical improvement in patients with dementia in AD (Kaduszkiewicz, Zimmermann, Beck-Bornholdt, & van den Bussche, 2005). Donepezil administration improved cognition in several animal models of impaired learning including aged rodents, in animals with experimentally induced cholinergic deficits and in mouse models of AD (Yoo, Valdovinos, & Williams, 2007). Several studies (Prasher, 2004) have

Mental Retardation and Human Chromosome 21 Gene Overdosage

65

addressed the treatment of cognitive decline related to dementia in DS. These studies have reported limited improvements after donepezil treatment in global functioning, cognitive skills and adaptive behaviour in people with DS. Donepezil administration also produces some improvement in language skills in adults (Heller et  al., 2003) and children (Heller et al., 2004) with DS. Therapeutic interventions have been tried out to improve learning in Ts65Dn mice. Oestrogen administration improved learning performance in the T-maze and reversed cholinergic impairment in 11–15-month Ts65Dn females (Granholm et al., 2002). Several studies have also suggested that the deficits in learning and memory seen in the Ts65Dn mouse might be partially due to increased inhibition at the synaptic level. Ts65Dn mice show a decrease of excitatory synapses (Kurt et al., 2000, 2004) and in synapse connectivity (Belichenko et al., 2004; Hanson et al., 2007). Furthermore, two studies provided evidence of increases in GABAAreceptor-mediated inhibition in Ts65Dn mice. In Ts65Dn mice, it has been shown that evoked LTP in granule cells of the dentate gyrus is reduced due to an increased GABA-dependent inhibition of these neurons (Kleschevnikov et  al., 2004). Moreover, it has been found that there is a significant reduction of the amount of theta-burst stimulation (TBS)-induced LTP in Ts65Dn mice that could be rescued via picrotoxin application. Therefore, an increase in GABAA-mediated inhibition or in plasticity of the inhibitory circuitry in Ts65Dn mice may underlie the cognitive deficits found in these mice (Costa & Grybko, 2005). Recently, it has been demonstrated that administering the GABAA antagonists, picrotoxin, bilobalide or pentylenetetrazol (PTZ), restored cognition and LTP in the Ts65Dn mouse, and suggested that this positive effect could me mediated by reducing inhibition in this mouse (Fernandez et al., 2007). The chronic systemic administration of noncompetitive GABAA antagonists leads to a persistent, post-drug recovery of cognition in Ts65Dn mice, as well as recovery of deficits in LTP. Only 10 days of drug treatment resulted in improved performance in the novel object recognition and spontaneous alternation in the T-maze tasks, an improvement that persisted for several months after drug treatment ended (Fernandez et al., 2007). More recently, these studies were confirmed using the non-competitive GABAA antagonist PTZ which rescued Ts65Dn performance in the Morris water maze (Rueda, Florez, & Martinez-Cué, 2008). These findings suggest that excessive GABAergic inhibition of specific brain circuits is a potential cause of MR in DS, and that GABAA antagonists may be useful therapeutic tools to facilitate functional changes that can ameliorate cognitive impairment in children and young adults with the disorder. More recently, it has been demonstrated that acute injections of the NMDA receptor antagonist memantine rescue performance deficits in the Ts65Dn mouse model of DS on a conditioning fear test (Costa, Scott-McKean, & Stasko, 2008). One target of memantine is the NMDA receptor, whose function is predicted to be perturbed by the integrated effects of increased expression of several HSA21 genes, including RCAN1, APP, ITSN1 and DYRK1A, all of which are being intensively studied for relevance to DS. It is known that NMDA receptors are among the targets of calcineurin. It has been demonstrated that the pharmacological inhibition of calcineurin activity leads

66

M. Rachidi and C. Lopes

to increased NMDA receptors mean open time and opening probability (Lieberman & Mody, 1994). Theoretically, such modulation of kinetic parameters should lead to an increase in inhibition of NMDA receptors-mediated currents by open channel blockers, including the noncompetitive NMDA receptors antagonist MK-801. Accordingly, conditional calcineurin null-mutant mice display increased responses to the locomotor-stimulating effects of MK-801 (Miyakawa et al., 2003). DSCR1 knock-out mice have pronounced spatial learning and memory deficits (Hoeffer et  al., 2007). These deficits were similar to those found in mice with inducible, hippocampal-restricted overexpression of constitutively active calcineurin (Mansuy, Mayford et al., 1998; Mansuy, Winder et al., 1998) and the direct opposite of learning behaviours of animals in which calcineurin was inhibited by either transgenic expression of a calcineurin inhibitory domain or application of antisense oligonucleotides (Ikegami & Inokuchi, 2000; Malleret et al., 2001). This suggests that DSCR1 provides a constraint on calcineurin activity during learning and memory and that this constraint is absent in the DSCR1 knock-out mice. Moreover, recent findings indicate that acute blockade of calcineurin activity improves memory and cognitive function in AD model mice (Dineley, Hogan, Zhang, & Taglialatela, 2007). These findings strongly suggest that DSCR1 facilitates synaptic plasticity and memory by constraining phosphatase signalling via inhibition of calcineurin and its downstream target PP1. Thus, DSCR1 represents an important potential therapeutic target for the treatment of numerous neurological disorders whose pathologies involve the dysregulation of calcineurin (Hoeffer et al., 2007). In addition, efforts could also be concentrated on the microRNAs of the HSA21 to elucidate their biological role in DS, because it has been demonstrated recently that microRNAs show an increasing importance in the gene expression control and seem to play a fundamental role in diverse biological and pathological processes, including cell proliferation, differentiation, apoptosis, carcinogenesis, and cardiovascular disease (Bushati & Cohen, 2007; Wang et  al., 2007). Considering the hypothesis that trisomic 21 gene-dosage overexpression of HSA21-derived miRNAs results in the decreased expression of specific target proteins and contributes, in part, to features of the neuronal and cardiac DS phenotype, HSA21-derived miRNAs may provide novel therapeutic targets in the treatment of individuals with DS (Kuhn et al., 2008). This indicates that studies of variation of gene expression and its genomic regulation, and functional studies of these genes and other conserved DNA elements, are two fundamental research priorities that may also provide potential future directions for MR therapeutics in DS.

10 Conclusion and Perspectives Partial trisomies of HSA21 or MMU16 allowed genetic dissection of DS phenotypes, particularly the neurological ones, the goal of which is the identification of candidate genes contributing to neurological and behavioural phenotypes, and to MR in DS.

Mental Retardation and Human Chromosome 21 Gene Overdosage

67

The gene expression studies in mouse models and human have shown similar genotype/phenotype correlations. The highly parallel outcomes that result when the same evolutionarily conserved genetic programmes are perturbed in mice and human validate the recent studies that focus on DS phenotypes, particularly neurological alterations. Thus, these studies indicate that brain dysfunctions and MR may be due to over-dosage of genes involved, directly or indirectly, in brain developmental processes throughout neurogenesis, neuronal growth and neuronal differentiation. All these studies identified some genes which are overexpressed in the brain and involved in brain development, learning and memory as candidate genes for MR in DS. The increased information about function of the proteins encoded by these genes, their interaction with other proteins and their involvement in regulatory and metabolic pathways is giving a clearer view of the origin of the MR in DS. This leads to the identification of potential targets in the molecular altered pathways involved in MR pathogenesis that may be potentially corrected, in the perspective of new therapeutic approaches. Furthermore, the regulation of gene expression by microRNAs or small interfering RNAs provide exciting possibilities for exogenous correction of the aberrant gene expression in DS and also provide potential directions for clinical therapeutics of MR. Given the new pharmacotherapies for cognitive impairment in a mouse model of DS and the relevance of these findings to the treatment of cognitive deficits in the human DS population, substantial interest emerges in clinical settings and trials in DS and has created substantial interest of the scientific and medical community towards a novel biomedical era for therapeutics of MR in DS. Acknowledgments  We are grateful to J. M. Delabar (University Paris 7) for his continuous support. We thank our colleagues of the Department of Molecular Biology-Jacques Monod at the Pateur Insitute (Paris) for their advice and support. We also thank L. Peltzer (University of French Polynesia) and C. Tetaria (Hospital Centre of French Polynesia) for their continuous support.

References Ahn, K. J., Jeong, H. K., Choi, H. S., Ryoo, S. R., Kim, Y. J., et al. (2006). DYRK1A BAC transgenic mice show altered synaptic plasticity with learning and memory defects. Neurobiology of Disease, 22, 463–472. Aït Yahya-Graison, E., Aubert, J., Dauphinot, L., Rivals, I., Prieur, M., et al. (2007). Classification of human chromosome 21 gene-expression variations in Down syndrome: Impact on disease phenotypes. American Journal of Human Genetics, 81, 475–491. Aitken, D. A., McCaw, G., Crossley, J. A., Berry, E., Connor, J. M., et al. (1993). First-trimester biochemical screening for fetal chromosome abnormalities and neural tube defects. Prenatal Diagnosis, 13, 681–689. Akeson, E. C, Lambert J. P., Narayanswami, S., Gardiner, K., Bechtel, L. J., et al. (2001). Ts65Dn: localization of the translocation breakpoint and trisomic gene content in a mouse model for Down syndrome. Cytogenetics and Cellular Genetics, 93, 270–276. Aldridge, K., Reeves, R. H., Olson, L. E., & Richtsmeier, J. T. (2007). Differential effects of trisomy on brain shape and volume in related aneuploid mouse models. American Journal of Medical Genetics, 143A, 1060–1070.

68

M. Rachidi and C. Lopes

Altafaj, X., Dierssen, M., Baamonde, C., Marti, E., Visa, J., et al. (2001). Neurodevelopmental delay, motor abnormalities and cognitive deficits in transgenic mice overexpressing Dyrk1A (minibrain), a murine model of Down’s syndrome. Human Molecular Genetics, 10, 1915–1923. Alvarez, M., Estivill, X., & de la Luna, S. (2003). DYRK1A accumulates in splicing speckles through a novel targeting signal and induces speckle disassembly. Journal of Cell Science, 116, 3099–3107. Amano, K., Sago, H., Uchikawa, C., Suzuki, T., Kotliarova, S. E., et al. (2004). Dosage-dependent over-expression of genes in the trisomic region of Ts1Cje mouse model for Down syndrome. Human Molecular Genetics, 13, 1333–1340. Antonarakis, S. E. (1998). 10 years of genomics, chromosome 21, and Down syndrome. Genomics, 51, 1–16. Antonarakis, S. E. (2001). Chromosome 21: From sequence to applications. Current Opinion in Genetics & Development, 11, 241–246. Antonarakis, S. E., & Epstein, C. J. (2006). The challenge of Down syndrome. Trends in Molecular Medicine, 12, 473–479. Antonarakis, S. E., Lyle, R., Dermitzakis, E. T., Reymond, A., & Deutsch, S. (2004). Chromosome 21 and down syndrome: From genomics to pathophysiology. Nature Review Genetics, 5, 725–738. Arai, Y., Ijuin, T., Takenawa, T., Becker, L. E., & Takashima, S. (2002). Excessive expression of synaptojanin in brains with Down syndrome. Brain & Development, 24, 67–72. Arron, J. R., Winslow, M. M., Polleri, A., Chang, C. P., Wu, H., et al. (2006). NFAT dysregulation by increased dosage of DSCR1 and DYRK1A on chromosome 21. Nature, 441, 595–600. Avraham, K. B., Sugarman, H., Rotshenker, S., & Groner, Y. (1991). Down’s syndrome: Morphological remodelling and increased complexity in the neuromuscular junction of transgenic CuZn-superoxide dismutase mice. Journal of Neurocytology, 20, 208–215. Aylward, E. H., Li, Q., Honeycutt, N. A., Warren, A. C., Pulsifer, M. B., et al. (1999). MRI volumes of the hippocampus and amygdala in adults with Down’s syndrome with and without dementia. The American Journal of Psychiatry, 156, 564–568. Ayraham, K. B., Sugarman, H., Rotshenker, S., & Groner, Y. (1991). Down’s syndrome: Morphological remodelling and increased complexity in the neuromuscular junction of transgenic CuZn-superoxide dismutase mice. Journal of Neurocytology, 20, 208–215. Azmitia, E. C., Griffin, W. S. T., Marshak, D. R., Van Eldik, L. J., & Whitaker-Azmitia, P. M. (1992). S100b and serotonin: A possible astrocytic-neuronal link to neuropathology of Alzheimer’s disease. Progress in Brain Research, 92, 459–473. Bahn, S., Mimmack, M., Ryan, M., Caldwell, M. A., Jauniaux, E., et al. (2002). Neuronal target genes of the neuron-restrictive silencer factor in neurospheres derived from fetuses with Down’s syndrome: A gene expression study. Lancet, 359, 310–315. Balcz, B., Kirchner, L., Cairns, N., Fountoulakis, M., & Lubec, G. (2001). Increased brain protein levels of carbonyl reductase and alcohol dehydrogenase in Down syndrome and Alzheimer’s disease. Journal of Neural Transmission, 61, 193–201. Barlow, G. M., Micales, B., Chen, X. N., Lyons, G. E., & Korenberg, J. R. (2002). Mammalian DSCAMs: Roles in the development of the spinal cord, cortex, and cerebellum? Biochemical and Biophysical Research Communications, 293, 881–891. Bartel, D. P. (2004). MicroRNAs: Genomics, biogenesis, mechanism, and function. Cell, 116, 281–297. Baxter, L. L., Moran, T. H., Richtsmeier, J. T., Troncoso, J., & Reeves, R. H. (2000). Discovery and genetic localization of Down syndrome cerebellar phenotypes using the Ts65Dn mouse. Human Molecular Genetics, 9, 195–202. Becker, L., Mito, T., Takashima, S., & Onodera, K. (1991). Growth and development of the brain in Down syndrome. In C. H. Epstein (Ed.), The morphogenesis of Down syndrome (pp. 133–152). New York: Wiley-Liss. Becker, L. E., Armstrong, D. L., & Chan, F. (1986). Dendritic atrophy in children with Down’s syndrome. Annals of Neurology, 20, 520–532. Behar, T. N., & Colton, C. A. (2003). Redox regulation of neuronal migration in a Down syndrome model. Free Radical Biology & Medicine, 35, 566–575.

Mental Retardation and Human Chromosome 21 Gene Overdosage

69

Belichenko, P. V., Masliah, E., Kleschevnikov, A. M., Villar, A. J., Epstein, C. J., et al. (2004). Synaptic structural abnormalities in the Ts65Dn mouse model of Down syndrome. The Journal of Comparative Neurology, 480, 281–298. Bell, K., Shokrian, D., Potenzieri, C., & Whitaker-Azmitia, P. M. (2003). Harm avoidance, anxiety, and response to novelty in the adolescent S-100beta transgenic mouse: Role of serotonin and relevance to Down syndrome. Neuropsychopharmacology, 28, 1810–1816. Benavides-Piccione, R., Ballesteros-Yanez, I., Martinez de Lagran, M., Elston, G., Estivill, X., et al. (2004). On dendrites in Down syndrome and DS murine models: A spiny way to learn. Progress in Neurobiology, 74, 111–126. Benavides-Piccione, R., Dierssen, M., Ballesteros-Yaöez, I., Martinez de Lagran, M., Arbones, M. L., et al. (2005). Alterations in the phenotype of neocortical pyramidal cells in the Dyrk1A+/− mouse. Neurobiology of Disease, 20, 115–122. Berto, G., Camera, P., Fusco, C., Imarisio, S., Ambrogio, C., et al. (2007). The Down syndrome critical region protein TTC3 inhibits neuronal differentiation via RhoA and Citron kinase. Journal of Cell Science, 120, 1859–1867. Best, T. K., Siarey, R. J., & Galdzicki, Z. (2006). Ts65Dn, a mouse model of Down syndrome, exhibits increased GABAB induced potassium current. Journal of Neurophysiology, 97, 892–900. Black, A. H., Nadel, L., & O’Keefe, J. (1977). Hippocampal function in avoidance learning and punishment. Psychological Bulletin, 84, 1107–1129. Blednow, Y. A., Stoffel, M., Chang, S. R., & Harris, R. A. (2001). GIRK2 deficient mice. Evidence for hyperactivity and reduced anxiety. Physiology & Behavior, 74, 109–117. Bourtchuladze, R., Frenguelli, B., Blendy, J., Cioffi, D., Schutz, G., et al. (1994). Deficient longterm memory in mice with a targeted mutation of the cAMP-responsive element-binding protein. Cell, 79, 59–68. Branchi, I., Bichler, Z., Minghetti, L., Delabar, J. M., Malchiodi-Albedi, F., et  al. (2004). Transgenic mouse in vivo library of human Down syndorme Critical Region 1: Association between DYRK1A overexpession, brain development abnormalities, and cell cycle protein alteration. Journal of Neuropathology and Experimental Neurology, 63, 429–440. Brem, R. B., Yvert, G., Clinton, R., & Kruglyak, L. (2002). Genetic dissection of transcriptional regulation in budding yeast. Science, 296, 752–755. Brown, F. R., Greer, M. K., Aylward, E. H., & Hunt, H. H. (1990). Intellectual and adaptive functioning in individuals with Down syndrome in relation to age and environmental placement. Pediatrics, 85, 450–452. Brown, J. H., Johnson, M. H., Patterson, S. J., Gilmore, R., Longhi, E., et al. (2003). Spatial representation and attention in toddlers with Williams syndrome and Down syndrome. Neuropsychologia, 41, 1037–1046. Bushati, N., & Cohen, S. M. (2007). MicroRNA functions. Annual Review of Cell and Developmental Biology, 3, 175–205. Carlesimo, G. A., Marotta, L., & Vicari, S. (1997). Long-term memory in mental retardation: Evidence for a specific impairment in subjects with Down’s syndrome. Neuropsychologia, 35, 71–79. Casanova, M. F., Walker, L. C., Whitehouse, P. J., & Price, D. L. (1985). Abnormalities of the nucleus basalis in Down’s syndrome. Annals of Neurology, 18, 310–313. Chabert, C., Jamon, M., Cherfouh, A., Duquenne, V., Smith, D. J., et al. (2004). Functional analysis of genes implicated in Down syndrome: 1. Cognitive abilities in mice transpolygenic for Down syndrome chromosomal region-1 (DCR-1). Behavior Genetics, 34, 559–569. Chang, K. T., Shi, Y. J., & Min, K. T. (2003). The Drosophila homolog of Down’s syndrome critical region 1 gene regulates learning: Implications for mental retardation. Proceedings of the National Academy of Sciences of the United States of America, 100, 15794–15799. Chapman, R. S., & Hesketh, L. J. (2000). Behavioral phenotype of individuals with Down syndrome. Mental Retardation and Developmental Disabilities Research Reviews, 6, 84–95. Chapman, R. S., Seung, H. K., Schwartz, S. E., & Kay-Raining Bird, E. (1998). Language skills of children and adolescents with Down syndrome II: Production deficits. Journal of Speech, Language, and Hearing Research, 41, 861–873.

70

M. Rachidi and C. Lopes

Chen, B. E., Kondo, M., Garnier, A., Watson, F. L., Puettmann-Holgado, R., et al. (2006). The molecular diversity of Dscam is functionally required for neuronal wiring specificity in Drosophila. Cell, 125, 607–620. Chen-Hwang, M. C., Chen, H. R., Elzinga, M., & Hwang, Y. W. (2002). Dynamin is a minibrain kinase/dual specificity Yak1-related kinase 1A substrate. The Journal of Biological Chemistry, 277, 17597–17604. Cheon, M. S., Bajo, M., Kim, S. H., Claudio, J. O., Stewart, A. K., et al. (2003). Protein levels of genes encoded on chromosome 21 in fetal Down syndrome brain: Challenging the gene dosage effect hypothesis (Part II). Amino Acids, 24, 119–125. Cheon, M. S., Fountoulakis, M., Dierssen, M., Ferreres, J. C., & Lubec, G. (2001). Expression profiles of proteins in fetal brain with Down syndrome. Journal of Neural Transmission. Supplementum, 61, 311–319. Cheon, M. S., Kim, S. H., Ovod, V., Kopitar Jerala, N., Morgan, J. I., et al. (2003). Protein levels of genes encoded on chromosome 21 in fetal Down syndrome brain: Challenging the gene dosage effect hypothesis (Part III). Amino Acids, 24, 127–134. Cheon, M. S., Kim, S. H., Yaspo, M. L., Blasi, F., Aoki, Y., et al. (2003). Protein levels of genes encoded on chromosome 21 in fetal Down syndrome brain: Challenging the gene dosage effect hypothesis (Part I). Amino Acids, 24, 111–117. Cheon, M. S., Shim, K. S., Kim, S. H., Hara, A., & Lubec, G. (2003). Protein levels of genes encoded on chromosome 21 in fetal Down syndrome brain: Challenging the gene dosage effect hypothesis (Part IV). Amino Acids, 25, 41–47. Cheung, V. G., Conlin, L. K., Weber, T. M., Arcaro, M., Jen, K. Y., et al. (2003). Natural variation in human gene expression assessed in lymphoblastoid cells. Nature Genetics, 33, 422–425. Chou, C. Y., Liu, L. Y., Chen, C. Y., Tsai, C. H., Hwa, H. L., et al. (2008). Gene expression variation increase in trisomy 21 tissues. Mammalian Genome. doi: 10.1007/s00335-008-9121-1. Chrast, R., Scott, H. S., Madani, R., Huber, L., Wolfer, D. P., et al. (2000). Mice trisomic for a bacterial artificial chromosome with the single-minded 2 gene (Sim2) show phenotypes similar to some of those present in the partial trisomy 16 mouse models of Down syndrome. Human Molecular Genetics, 9, 1853–1864. Chrast, R., Scott, H. S., Papasavvas, M. P., Rossier, C., Antonarakis, E. S., et  al. (2000). The mouse brain transcriptome by SAGE: Differences in gene expression between P30 brains of the partial trisomy 16 mouse model of Down syndrome (Ts65Dn) and normals. Genome Research, 10, 2006–2021. Clark, D., & Wilson, G. N. (2003). Behavioral assessment of children with Down syndrome using the Reiss psychopathology scale. American Journal of Medical Genetics, 118, 210–216. Collacot, R. A., Cooper, S. A., Branford, D., & McGrother, C. (1998). Behaviour phenotype for Down’s syndrome. The British Journal of Psychiatry, 172, 85–89. Connolly, B. H., Morgan, S. B., Russell, F. F., & Fulliton, W. L. (1993). A longitudinal study of children with Down syndrome who experienced early intervention programming. Physical Therapy, 73, 170–179. Cooper, J. D., Salehi, A., Delcroix, J. D., Howe, C. L., Belichenko, P. V., et  al. (2001). Failed retrograde transport of NGF in a mouse model of Down’s syndrome: Reversal of cholinergic neurodegenerative phenotypes following NGF infusion. Proceedings of the National Academy of Sciences of the United States of America, 98, 10439–10444. Costa, A. C. S., & Grybko, M. J. (2005). Deficits in hippocampal CA1 LTP induced by TBS but not HFS in the Ts65Dn mouse: A model of Down syndrome. Neuroscience Letters, 382, 317–322. Costa, A. C. S., Scott-McKean, J. J., & Stasko, M. R. (2008). Acute injections of the NMDA receptor antagonist memantine rescue performance deficits of the Ts65Dn mouse model of Down syndrome on a conditioning fear test. Neuropsychopharmacology, 33, 1624–1632. Cousin, M. A., Tan, T. C., & Robinson, P. J. (2001). Protein phosphorylation is required for endocytosis in nerve terminals: Potential role for the dephosphins dynamin I and synaptojanin, but not AP180 or amphiphysin. Journal of Neurochemistry, 76, 105–116. Coussons-Read, M. E., & Crnic, L. S. (1996). Behavioral assessment of the Ts65Dn mouse, a model for Down syndrome: Altered behavior in the elevated plus maze and open field. Behavior Genetics, 26, 7–13.

Mental Retardation and Human Chromosome 21 Gene Overdosage

71

Cox, D. R., Smith, S. A., Epstein, L. B., & Epstein, C. J. (1984). Mouse trisomy 16 as an animal model of human trisomy 21 (Down syndrome): Production of viable trisomy 16 diploid mouse chimeras. Developmental Biology, 101, 416–424. Coyle, J. T., Oster-Granite, M. L., & Gearhart, J. D. (1986). The neurobiologic consequences of Down syndrome. Brain Research Bulletin, 16, 773–787. Crabtree, G. R., & Olson, E. N. (2002). NFAT signalling: Choreagraphing the social lives of cells. Cell, 109(Suppl.), S67–S79. Cremona, O., Di Paolo, G., Wenk, M. R., Luthi, A., Kim, W. T., et al. (1999). Essential role of phosphoinositide metabolism in synaptic vesicle recycling. Cell, 99, 179–188. Crews, S. T., Thomas, J. B., & Goodman, C. S. (1988). The Drosophila singleminded gene encodes a nuclear protein with sequence similarity to the per gene product. Cell, 52, 143–151. Dahmane, N., Ait Ghezala, G. A., Gosset, P., Chamoun, Z., Dufresne-Zacharia, M. C., et  al. (1998). Transcriptional map of the 2.5-Mb CBR–ERG region of chromosome 21 involved in Down syndrome. Genomics, 48, 12–23. Dahmane, N., Charron, G., Lopes, C., Yaspo, M. L., Maunoury, C., et al. (1995). Down syndrome critical region contains a gene homologous to Drosophila sim expressed during rat and human central nervous system development. Proceedings of the National Academy of Sciences of the United States of America, 92, 9191–9195. Dalton, A. J., & Crapper-McLachlan, D. R. (1986). Clinical expression of Alzheimer’s disease in Down syndrome. The Psychiatric Clinics of North America, 9, 959–970. Das, A. K., Cohen, P. W., & Barford, D. (1998). The structure of the tetratricopeptide repeats of protein phosphatase 5: Implications for TPR-mediated protein-protein interactions. The EMBO Journal, 17, 1192–1199. Dauphinot, L., Lyle, R., Rivals, I., Dang, M. T., Moldrich, R. X., et  al. (2005). The cerebellar transcriptome during postnatal development of the Ts1Cje mouse, a segmental trisomy model for Down syndrome. Human Molecular Genetics, 14, 373–384. Davisson, M. T., Schmidt, C., & Akeson, E. C. (1990). Segmental trisomy of murine chromosome 16: A new model system for studying Down syndrome. Progress in Clinical and Biological Research, 360, 263–280. de Graaf, K., Hekerman, P., Spelten, O., Herrman, A., Packman, L. C., et al. (2004). Characterization of cyclin L2, a novel cyclin with an arginine/serine-rich domain: Phosphorylation by DYRK1A and colocalization with splicing factors. The Journal of Biological Chemistry, 279, 4612–4624. de Kerchove, D’ Exaerde A., Cartaud, J., Ravel-Chapuis, A., Seroz, T., Pasteau, F., et al. (2002). Expression of mutant Ets protein at the neuromuscular synapse causes alterations in morphology and gene expression. EMBO Reports, 3, 1075–1081. Delabar, J. M., Theophile, D., Rahmani, Z., Chettouh, Z., Blouin, J. L., et al. (1993). Molecular mapping of twenty-four features of Down syndrome on chromosome 21. European Journal of Human Genetics, 1, 114–124. Deutsch, S., Lyle, R., Dermitzakis, E. T., Attar, H., Subrahmanyan, L., et al. (2005). Gene expression variation and expression quantitative trait mapping of human chromosome 21 genes. Human Molecular Genetics, 14, 3741–3749. Devenny, D. A., Silverman, W. P., Hill, A. L., Jenkins, E., Sersen, E. A., et al. (1996). Normal ageing in adults with Down’s syndrome: A longitudinal study. Journal of Intellectual Disability Research, 40, 208–221. Dierssen, M., & Ramakers, G. J. A. (2006). Dendritic pathology in mental retardation; from molecular genetics to neurobiology. Genes, Brain, and Behavior, 5(Suppl 2), 48–60. Dierssen, M., Vallina, I. F., Baamonde, C., Lumbreras, M. A., Martinez-Cue, C., et  al. (1996). Impaired cyclic AMP production in the hippocampus of a Down syndrome murine model. Brain Research. Developmental Brain Research, 95, 122–124. Dierssen, M., Vallna, I. F., Baamonde, C., Garcia-Calatayud, S., & Lumbreras, M. A. (1997). Alterations of central noradrenergic transmission in Ts65Dn mouse, a model for Down syndrome. Brain Research, 749, 238–244. Dineley, K. T., Hogan, D., Zhang, W. R., & Taglialatela, G. (2007). Acute inhibition of calcineurin restores associative learning and memory in Tg2576 APP transgenic mice. Neurobiology of Learning and Memory, 88, 217–224.

72

M. Rachidi and C. Lopes

Dufresne-Zacharia, M. C., Dahmane, N., Theophile, D., Orti, R., Chettouh, Z., et al. (1994). 3.6-Mb genomic and YAC physical map of the Down syndrome chromosome region on chromosome 21. Genomics, 19, 462–469. Dujon, B. (1996). The yeast genome project: What did we learn? Trends in Genetics, 12, 263–270. Ema, M., Ikegami, S., Hosoya, T., Mimura, J., Ohtani, H., et al. (1999). Mild impairment of learning and memory in mice overexpressing the mSim2 gene located on chromosome 16: An animal model of Down’s syndrome. Human Molecular Genetics, 8, 1409–1415. Ema, M., Suzuki, M., Morita, M., Hirose, K., Sogawa, K., et al. (1996). cDNA cloning of a murine homologue of Drosophila single-minded, its mRNA expression in mouse development, and chromosome localization. Biochemical and Biophysical Research Communications, 218, 588–594. Engidawork, E., Balic, N., Fountoulakis, M., Dierssen, M., Graber-Platzer, S., et  al. (2001). b-Amyloid precursor protein, ETS-2 and collagen alpha 1 (VI) chain precursor, encoded on chromosome 21, are not overexpressed in fetal Down syndrome: Further evidence against gene dosage effect. Journal of Neural Transmission. Supplementum, 61, 335–346. Engidawork, E., Gulesserian, T., Balic, N., Cairns, N., & Lubec, G. (2001). Changes in nicotinic acetylcholine receptor subunits expression in brain of patients with Down syndrome and Alzheimer’s disease. Journal of Neural Transmission, 61, 211–222. Engidawork, E., Gulesserian, T., Fountoulakis, M., & Lubec, G. (2003). Aberrant protein expression in cerebral cortex of fetus with Down syndrome. Neuroscience, 122, 145–154. Engidawork, E., & Lubec, G. (2003). Molecular changes in fetal Down syndrome brain. Journal of Neurochemistry, 84, 895–904. Epstein, C. J. (1986). The consequences of chromosome imbalance: Principles, mechanisms, and models. New York: Cambridge University Press. Epstein, C. J. (1988). Mechanisms of the effects of aneuploidy in mammals. Annual Review of Genetics, 22, 51–75. Epstein, C. J. (1990). The consequences of chromosome imbalance. American Journal of Human Genetics, 7(Suppl), 31–37. Epstein, C. J., Avraham, K. B., Lovett, M., Smith, S., Elroy-Stein, O., et al. (1987). Transgenic mice with increased Cu/Zn-superoxide dismutase activity: Animal model of dosage effects in Down syndrome. Proceedings of the National Academy of Sciences of the United States of America, 84, 8044–8048. Epstein, C. J., Hofmeister, B. G., Yee, D., Smith, S. A., Philip, R., et al. (1985). Stem cell deficiencies and thymic abnormalities in fetal mouse trisomy 16. The Journal of Experimental Medicine, 162, 695–712. Epstein, D. J., Matinu, L., Michaud, J. L., Losos, K. M., Fan, C. M., et al. (2000). Members of the bHLH-PAS family regulate Shh transcription in forebrain regions of the mouse CNS. Development, 127, 1701–1709. Epstein, J. (2001). Down syndrome. In C. R. Scriver, A. L. Beaudet, W. S. Sly, & D. Valle (Eds.), The metabolic and molecular bases of inherited disease (7th ed., pp. 1223–1256). New York: McGraw-Hill. Ermak, G., Morgan, T. E., & Davies, K. J. (2001). Chronic overexpression of the calcineurin inhibitory gene DSCR1 (Adapt78) is associated with Alzheimer’s disease. The Journal of Biological Chemistry, 276, 38787–38794. Fabbro, F., Alberti, A., Gagliardi, C., & Borgatti, R. (2002). Differences in native and foreign language repetition task between subjects with Williams and Down syndromes. Journal of Neurolinguistics, 15, 1–10. Fan, C. M., Kuwana, E., Bulfone, A., Fletcher, C. F., Copeland, N. G., et al. (1996). Expression patterns of two murine homologs of Drosophila single-minded suggest possible roles in embryonic patterning and in the pathogenesis of Down syndrome. Molecular and Cellular Neurosciences, 7, 1–16. Fernandez, F., Morishita, W., Zuniga, E., Nguyen, J., Blank, M., et al. (2007). Pharmacotherapy for cognitive impairment in a mouse model of Down syndrome. Nature Neuroscience, 10, 411–413.

Mental Retardation and Human Chromosome 21 Gene Overdosage

73

Ferrando-Miguel, R., Cheon, M. S., & Lubec, G. (2004). Protein levels of genes encoded on chromosome 21 in fetal Down syndrome brain (Part V); overexpression of phosphatidylinositol-glycan class P protein (DSCR5). Amino Acids, 26, 255–261. FitzPatrick, D. R., Ramsay, J., McGill, N. I., Shade, M., Carothers, A. D., et al. (2002). Transcriptome analysis of human autosomal trisomy. Human Molecular Genetics, 11, 3249–3256. Fotaki, V., Dierssen, M., Alcantara, S., Martinez, S., Marti, E., et al. (2002). Dyrk1A ­haploinsufficiency affects viability and causes developmental delay and abnormal brain ­morphology in mice. Molecular and Cellular Biology, 22, 6636–6647. Frangou, S., Aylward, E., Warren, A., Sharma, T., Barta, P., et al. (1997). Small planum temporale volume in Down’s syndrome: A volumetric MRI study. The American Journal of Psychiatry, 154, 1424–1429. Freidl, M., Gulesserian, T., Lubec, G., Fountoulakis, M., & Lubec, B. (2001). Deterioration of the transcriptional, splicing and elongation machinery in brain of fetal Down syndrome. Journal of Neural Transmission. Supplementum, 61, 47–57. Fuentes, J. J., Genesca, L., Kingsbury, T. J., Cunningham, K. W., Perez-Riba, M., et al. (2000). DSCR1, overexpressed in Down syndrome, is an inhibitor of calcineurin-mediated signaling pathways. Human Molecular Genetics, 9, 1681–1690. Fuentes, J. J., Pritchard, M. A., Planas, A. M., Bosch, A., Ferrer, I., et al. (1995). A new human gene from the Down syndrome critical region encodes a proline-rich protein highly expressed in fetal brain and heart. Human Molecular Genetics, 4, 1935–1944. Funahashi, S., Takeda, K., & Watanabe, Y. (2004). Neural mechanisms of spatial working memory: Contributions of the dorsolateral pre-frontal cortex and the thalamic mediodorsal nucleus. Cognitive, Affective & Behavioral Neuroscience, 4, 409–420. Gahtan, E., Auerbach, J. M , Groner, Y., & Segal, M. (1998). Reversible impairment of long-term potentiation in transgenic Cu/Zn-SOD mice. The European Journal of Neuroscience, 10, 538–544. Gardiner, K., Fortna, A., Bechtel, L., & Davisson, M. T. (2003). Mouse models of Down syndrome: How useful can they be? Comparison of the gene content of human chromosome 21 with orthologous mouse genomic regions. Gene, 318, 137–147. Gearhart, J. D., Davisson, M. T., & Oster-Granite, M. L. (1986). Autosomal aneuploidy in mice: Generation and developmental consequences. Brain Research Bulletin, 16, 789–801. Gerlai, R., & Roder, J. (1996). Spatial and nonspatial learning in mice: Effects of S100 beta overexpression and age. Neurobiology of Learning and Memory, 66, 143–154. Giannone, S., Strippoli, P., Vitale, L., Casadei, R., Canaider, S., et  al. (2004). Gene expression profile analysis in human T lymphocytes from patients with Down syndrome. Annals of Human Genetics, 68, 546–554. Gitton, Y., Dahmane, N., Baik, S., Ruiz, I., Altaba, A., et al. (2002). A gene expression map of human chromosome 21 orthologues in the mouse. Nature, 420, 586–590. Golden, J. A., & Hyman, B. T. (1994). Development of the superior temporal neocortex is anomalous in trisomy 21. Journal of Neuropathology and Experimental Neurology, 53, 513–552. Goldowitw, D., & Smeyne, R. J. (1995). Tune into the weaver channel. Nature Genetics, 11, 107–109. Goshu, E., Jin, H., Fasnacht, R., Sepenski, M., Michaud, J. L., et al. (2002). Sim2 mutant have developmental defects not overlapping with those of the Sim1 mutants. Molecular and Cellular Biology, 22, 4147–4157. Graef, I. A., Che, F., & Crabtree, G. R. (2001). NFAT signaling in vertebrate development. Current Opinion in Genetics & Development, 11, 505–512. Granholm, A. C., Ford, K. A., Hyde, L. A., Bimonte, H. A., Hunter, C. L., et al. (2002). Estrogen restores cognition and cholinergic phenotype in an animal model of Down syndrome. Physiology & Behavior, 77, 371–385. Greber-Platzer, S., Schatzmann-Turhani, D., Cairns, N., Balcz, B., & Lubec, G. (1999). Expression of the transcription factor ETS2 in brain of patients with Down syndrome-evidence against the overexpression-gene dosage hypothesis. Journal of Neural Transmission, 57(Suppl), 269–281. Griffin, W. S. T., Sheng, J. G., McKenzie, L. E., Royston, M. C., Gentleman, S. M., et al. (1998). Life-long over-expression of S100b in Down’s syndrome: Implications for Alzheimer’s pathogenesis. Neurobiology of Aging, 19, 401–405.

74

M. Rachidi and C. Lopes

Groves, M. R., & Barford, D. (1999). Topological characteristics of helical repeat proteins. Current Opinion in Structural Biology, 9, 383–389. Guimera, J., Casas, C., Estivill, X., & Pritchard, M. (1999). Human minibrain homologue (MNBH/DYRK1): Characterization, alternative splicing, differential tissue expression, and overexpression in Down syndrome. Genomics, 57, 407–418. Guimera, J., Casas, C., Pucharcos, C., Solans, A., Domenech, A., et  al. (1996). A human ­homologue of Drosophila minibrain (MNB) is expressed in the neuronal regions affected in Down syndrome and maps to the critical region. Human Molecular Genetics, 5, 1305–1310. Guipponi, M., Brunschwig, K., Chamoun, Z., Scott, H. S., Shibuya, K., et al. (2000). C21orf5, a novel human chromosome 21 gene, has a Caenorhabditis elegans ortholog (pad-1) required for embryonic patterning. Genomics, 68, 30–40. Gurney, M. E., Pu, H., Chiu, A. Y., CantoMC, D., Polchow, C. Y., et al. (1994). Motor neuron degeneration in mice that express a human Cu, Zn superoxide dismutase mutation. Science, 264, 1772–1775. Gwack, Y., Sharma, S., Nardone, J., Tanasa, B., Iuga, A., et  al. (2006). A genome-wide Drosophila RNAi screen identifes DYRK-family kinases as regulators of NFAT. Nature, 441, 646–650. Haffner, C., Takei, K., Chen, H., Ringstad, N., Hudson, A., et  al. (1997). Synaptojanin 1: Localization on coated endocytic intermediates in nerve terminals and interaction of its 170 kDa isoform with Eps15. FEBS Letters, 419, 175–180. Hammerle, B., Carnicero, A., Elizalde, C., Ceron, J., Martinez, S., & Tejedor, F. J. (2003). Expression patterns and subcellular localization of the Down syndrome candidate protein MNB/DYRK1A suggest a role in late neuronal differentiation. The European Journal of Neuroscience, 17, 2277–2286. Hanson, J. E., Blank, M., Valenzuela, R. A., Garner, C. C., & Madison, D. V. (2007). The functional nature of synaptic circuitry is altered in area CA3 of the hippocampus in a mouse model of Downs’s syndrome. Journal of Physiology, 579, 53–67. Harashima, C., Jacobowitz, D. M., Stoffel, M., Chakrabarti, L., Haydar, T. F., et  al. (2006). Elevated expression of the G-protein-activated inwardly rectifying potassium channel 2 (GIRK2) in cerebellar unipolar brush cells of a Down syndrome mouse model. Cellular and Molecular Neurobiology, 26, 719–734. Harashima, C., Jacobowitz, D. M., Witta, J., Borke, R. C., Best, T. K., et  al. (2006). Abnormal expression of the G-protein-activated inwardly rectifying potassium channel 2 (GIRK2) in hippocampus, frontal cortex, and substantia nigra of Ts65Dn mouse: A model of Down syndrome. The Journal of Comparative Neurology, 494, 815–833. Harris-Cerruti, C., Kamsler, A., Kaplan, B., Lamb, B., Segal, M., et  al. (2004). Functional and mophological alterations in compound transgenic mice overexpressing Cu/Zn superoxide dismutase and amyloid precursor protein. The European Journal of Neuroscience, 19, 1174–90. Hattori, D., Demir, E., Kim, H. W., Viragh, E., Zipursky, S. L., et al. (2007). Dscam diversity is essential for neuronal wiring and self-recognition. Nature, 449, 223–227. Hattori, M., Fujiyama, A., Taylor, T. D., Watanabe, H., Yada, T., et al. (2000). The DNA sequence of human chromosome 21. Nature, 405, 311–319. Heller, J. H., Spiridigliozzi, G. A., Doraiswamy, P. M., Sullivan, J. A., Crissman, B. G., et  al. (2004). Donepezil effects on language in children with Down syndrome: Results of the first 22-week pilot clinical trial. American Journal of Medical Genetics, 130, 325–326. Heller, J. H., Spiridigliozzi, G. A., Sullivan, J. A., Doraiswamy, P. M., Krishnan, R. R., et  al. (2003). Donepezil for the tratment of language deficits in adults with Down syndrome: A preliminary 24-week open trial. American Journal of Medical Genetics, 116, 111–116. Hennequin, M., Morin, C., & Feine, J. S. (2000). Pain expression and stimulus localisation in individuals with Down’s syndrome. Lancet, 356, 1882–1887. Hodapp, R. M., Ewans, D. E., & Gray, F. L. (1999). Intellectual development in children with Down syndrome. In J. A. Rondal, J. Perera, & L. Nadel (Eds.), Down syndrome: A review of current knowledge (pp. 124–132). London: Whurr publisher.

Mental Retardation and Human Chromosome 21 Gene Overdosage

75

Hodapp, R. M., & Zigler, E. (1990). Applying the developmental perspective to individuals with Down syndrome. In D. Cicchetti & M. Beeghly (Eds.), Children with Down syndrome: A developmental perspective (pp. 1–28). New York: Cambridge University Press. Hoeffer, C. A., Dey, A., Sachan, N., Wong, H., Patterson, R. J., et al. (2007). The Down syndrome critical region protein RCAN1 regulates long-term potentiation and memory via inhibition of phosphatase signaling. The Journal of Neuroscience, 27, 13161–13172. Holtzman, D. M., Santucci, D., Kilbridge, J., Chua-Couzens, J., Fontana, D. J., et al. (1996). Developmental abnormalities and age-related neurodegeneration in a mouse model of Down syndrome. Proceedings of the National Academy of Sciences of the United States of America, 93, 13333–13338. Hulme, C., & Mackenzie, S. (1992). Working memory and severe learning difficulties. Hove: Erlbaum. Hunter, C. L., Isacson, O., Nelson, M., Bimonte-Nelson, H., Seo, H., et al. (2003). Regional alterations in amyloid precursor protein and nerve growth factor across age in a mouse model of Down’s syndrome. Neuroscience Research, 45, 437–445. Hussain, N. K., Yamabhai, M., Ramjaun, A. R., Guy, A. M., Baranes, D., et  al. (1999). Splice variants of intersectin are components of the endocytic machinery in neurons and nonneuronal cells. The Journal of Biological Chemistry, 274, 15671–15677. Huttenlocher, P. R. (1974). Dendritic development in neocortex of children with mental defect and infantile spasms. Neurology, 24, 203–210. Hyde, L. A., Frisone, D. F., & Crnic, L. S. (2001). Ts65Dn mice, a model for Down syndrome, have deficits in context discrimination learning suggesting impaired hippocampal function. Behavioural Brain Research, 118, 53–60. Ikeda, M., & Arai, Y. (2002). Longitudinal changes in brain CT scans and development of dementia in Down’s syndrome. European Neurology, 47, 205–208. Ikegami, S., & Inokuchi, K. (2000). Antisense DNA against calcineurin facilites memory in contextual fear conditioning by lowering the threshold for hippocampal long-term potentiation induction. Neuroscience, 98, 637–646. International Human Genome Sequencing Consortium. (2004). Finishing the euchromatic sequence of the human genome. Nature, 431, 931–945. Ito, M. (2002). The molecular organization of cerebellar long-term depression. Nature Review Neuroscience, 3, 896–902. Janus, C. (2004). Search strategies used by APP transgenic mice during navigation in the Morris water maze. Learning & Memory, 11, 337–346. Jenna, S., Hussain, N. K., Danek, E. I., Triki, I., Wasiak, S., et  al. (2002). The activity of the GTPase-activating protein CdGAP is regulated by the endocytic protein intersectin. The Journal of Biological Chemistry, 277, 6366–6373. Jernigan, T. L., Bellugi, U., Sowell, E., Doherty, S., & Hesselink, J. R. (1993). Cerebral morphologic distinctions between Williams and Down syndromes. Archives of Neurology, 50, 186–191. Kadota, M., Nishigaki, R., Wang, C. C., Toda, T., Shirayoshi, Y., et al. (2004). Proteomic signatures and aberrations of mouse embryonic stem cells containing a single human chromosome 21 in neuronal differentiation: An in vitro model of Down syndrome. Neuroscience, 129, 325–335. Kaduszkiewicz, H., Zimmermann, T., Beck-Bornholdt, H. P., & van den Bussche, H. (2005). Cholinesterase inhibitors for patients with Alzheimer’s disease: Systematic review of randomised clinical trials. BMJ, 331, 321–327. Kahlem, P., Sultan, M., Herwig, R., Steinfath, M., Balzereit, D., et al. (2004). Transcript level alterations reflect gene dosage effects across multiple tissues in a mouse model of Down syndrome. Genome Research, 14, 1258–1267. Kang, J., Lemaire, H. G., Unterbeck, A., Salbaum, J. M., Masters, C. L., et al. (1987). The precursor of Alzheimer’s disease amyloid A4 protein resembles a cell-surface receptor. Nature, 325, 733–736. Kapranov, P., Cawley, S. E., Drenkow, J., Bekiranov, S., Strausberg, R. L., et al. (2002). Largescale transcriptional activity in chromosomes 21 and 22. Science, 296, 916–919. Kentrup, H., Becker, W., Heukelbach, J., Wilmes, A., Schurmann, A., et al. (1996). Dyrk, a dual specificity protein kinase with unique structural features whose activity is dependent on tyrosine residues between subdomains VII and VIII. The Journal of Biological Chemistry, 271, 3488–3495.

76

M. Rachidi and C. Lopes

Kesslak, J. P., Nagata, S. F., Lott, I., & Nalcioglu, O. (1994). Magnetic resonance imaging analysis of age-related changes in the brains of individuals with Down’s syndrome. Neurology, 44, 1039–1045. Kim, S. H., & Lubec, G. (2001). Decreased alpha-endosulfine, an endogenous regulator of ATPsensitive potassium channels, in brains from adult Down syndrome patients. Journal of Neural Transmission. Supplementum, 61, 1–9. Kim, W. T., Chang, S., Daniell, L., Cremona, O., Di Paolo, G., et al. (2002). Delayed reentry of recycling vesicles into the fusion-competent synaptic vesicle pool in synaptojanin 1 knockout mice. Proceedings of the National Academy of Sciences of the United States of America, 99, 17143–17148. King, M. C., & Wilson, A. C. (1975). Evolution at two levels in humans and chimpanzees. Science, 188, 107–116. Kish, S., Karlinsky, H., Becker, L., Gilbert, J., Rebbetoy, M., et  al. (1989). Down’s syndrome individuals begin life with normal levels of brain cholinergic markers. Journal of Neurochemistry, 2, 1183–1187. Kleschevnikov, A. M., Belichenko, P. V., Villar, A. J., Epstein, C. J., & Malenka, R. C. (2004). Hippocampal long-term potentiation suppressed by increased inhibition in the Ts65Dn mouse, a genetic model of Down syndrome. The Journal of Neuroscience, 24, 8153–8160. Kojima, N., & Shirao, T. (2007). Synaptic dysfunction and disruption of postsynaptic drebrinactin complex: A study of neurological disorders accompanied by cognitive deficits. Neuroscience Research, 58, 1–5. Korenberg, J. R., Chen, X. N., Schipper, R., Sun, Z., Gonsky, R., et al. (1994). Down syndrome phenotypes: The consequences of chromosomal imbalance. Proceedings of the National Academy of Sciences of the United States of America, 91, 4997–5001. Korenberg, J. R., Kawashima, H., Pulst, S. M., Ikeuchi, T., Ogasawara, N., et al. (1990). Molecular definition of a region of chromosome 21 that causes features of the Down syndrome phenotype. American Journal of Human Genetics, 47, 236–246. Krasuski, J. S., Alexander, G. E., Horwitz, B., Rapoport, S. I., & Schapiro, M. B. (2002). Relation of medial temporal lobe volumes to age and memory function in nondemented adults with Down’s syndrome: Implications for the prodromal phase of Alzheimer’s disease. The American Journal of Psychiatry, 159, 74–81. Kuhn, D. E., Nuovo, G. J., Martin, M. M., Malana, G. E., Pleister, A. P., et al. (2008). Human chromosome 21-derived miRNAs are overexpressed in Down syndrome brains and hearts. Biochemical and Biophysical Research Communications, 370, 473–477. Kumashiro, S., Lu, Y. F., Tomizawa, K., Matsushita, M., Wei, F. Y., et al. (2005). Regulation of synaptic vesicle recycling by calcineurin in different vesicle pools. Neuroscience Research, 51, 435–443. Kurnit, D. M. (1979). Down syndrome: Gene dosage at the transcriptional level in skin fibroblasts. Proceedings of the National Academy of Sciences of the United States of America, 76, 2372–2375. Kurnit, D. M., Aldridge, J. F., Matsuoka, R., & Matthysse, S. (1985). Increased adhesiveness of trisomy 21 cells and atrioventricular canal malformations in Down syndrome: A stochastic model. American Journal of Medical Genetics, 20, 385–399. Kurnit, D. M., Layton, W. M., & Matthysse, S. (1987). Genetics, chance, and morphogenesis. American Journal of Human Genetics, 41, 979–995. Kuromi, H., Yoshihara, M., & Kidokoro, Y. (1997). An inhibitory role of calcineurin in endocytosis of synaptic vesicles at nerve terminals of Drosophila larvae. Neuroscience Research, 27, 101–113. Kurt, M. A., Davies, D. C., Kidd, M., Dierssen, M., & Florez, J. (2000). Synaptic deficit in the temporal cortex of partial trisomy 16 (Ts65Dn) mice. Brain Research, 858, 191–197. Kurt, M. A., Kafa, I. M., Dierssen, M., & Davies, C. D. (2004). Deficits of neuronal density in CA1 and synaptic density in the dentat gyrus, CA3 and CA1, in a mouse model of Down syndrome. Brain Research, 1022, 101–109. Lacey-Casem, M. L., & Oster-Granite, M. L. (1994). The neuropathology of the trisomy 16 mouse. Critical Reviews in Neurobiology, 8, 293–322.

Mental Retardation and Human Chromosome 21 Gene Overdosage

77

Ladner, C. J., Czech, J., Maurice, J., Lorens, S. A., & Lee, J. M. (1996). Reduction of calcineurin enzymatic activity in Alzheimer’s disease: Correlation with neuropathologic changes. Journal of Neuropathology and Experimental Neurology, 55, 924–931. Lai, F., & Williams, R. S. (1989). A prospective study of Alzheimer disease in Down syndrome. Archives of Neurology, 46, 849–853. Lalonde, R., Dumont, M., Paly, E., London, J., & Strazielle, C. (2004). Characterization of hemizygous SOD1/wild-type transgenic mice with the SHIRPA primary screen and tests of sensorimotor function and anxiety. Brain Research Bulletin, 64, 251–258. Lalonde, R., Le Pecheur, M., Strazielle, C., & London, J. (2005). Exploratory activity and motor coordination in wild-type SOD1/SOD1 transgenic mice. Brain Research Bulletin, 66, 155–162. Lamb, B. T., Sisodia, S. S., Lawler, A. M., Slunt, H. H., Kitt, C. A., et al. (1993). Introduction and expression of the 400 kilobase amyloid precursor protein gene in transgenic mice. Nature Genetics, 5, 22–30. Le Pecheur, M., Bourdon, E., Paly, E., Farout, L., Friguet, B., et al. (2005). Oxidized SOD1 alters proteasome activities in vitro and in the cortex of SOD1 overexpressing mice. FEBS Letters, 579, 3613–3618. Lee, J. I., & Ahnn, J. (2004). Calcineurin in animal behavior. Molecules and Cells, 17, 390–396. Lejeune, J. (1990). Pathogenesis of mental deficiency in trisomy 21. American Journal of Medical Genetics, 7(Suppl), 20–30. Lejeune, J., Gautier, M., & Turpin, R. (1959). Etude des chromosomes somatiques de neufs enfants Mongoliens. Comptes rendus de l’Académie des sciences. Série III, Sciences de la vie Paris, 248, 1721–1722. Lesage, F., Duprat, F., Fink, M., Guillemare, E., Coppola, T., et  al. (1994). Cloning provides evidence for a family of inward rectifier and G-protein coupled K+ channels in the brain. FEBS Letters, 353, 37–42. Li, C. M., Guo, M., Salas, M., Schupt, N., Silvermann, W., et al. (2006). Cell type-specific overexpression of chromosome 21 genes in fibroblasts and fetal hearts with trisomy 21. BMC Medical Genetics, 7, 24. Li, Z., Yu, T., Morishima, M., Pao, A., LaDuca, J., et al. (2007). Duplication of the entire 22.9 Mb human chromosome 21 syntenic region on mouse chromosome 16 causes cardiovascular and gastrointestinal abnormalities. Human Molecular Genetics, 16, 1359–1366. Lieberman, D. N., & Mody, I. (1994). Regulation of NMDA channel function by endogenous Ca(2+)-dependent phosphatase. Nature, 369, 235–239. Liu, J. P., Sim, A. T., & Robinson, P. J. (1994). Calcineurin inhibition of dynamin I GTPase activity coupled to nerve terminal depolarization. Science, 265, 970–973. Lopes, C., Chettouh, Z., Delabar, J. M., & Rachidi, M. (2003). The differentially expressed C21orf5 gene in the medial temporal-lobe system could play a role in mental retardation in Down syndrome and transgenic mice. Biochemical and Biophysical Research Communications, 305, 915–924. Lopes, C., Rachidi, M., Gassanova, S., Sinet, P. M., & Delabar, J. M. (1999). Developmentally regulated expression of mtprd, the murine ortholog of tprd, a gene from the Down syndrome chromosomal region 1. Mechanisms of Development, 84, 189–193. Lott, I. T., & Head, E. (2005). Alzheimer disease and Down syndrome: Factors in pathogenesis. Neurobiology of Aging, 26, 383–389. Ly, A., Nikolaev, A., Suresh, G., Zheng, Y., Tessier-Lavigne, M., et al. (2008). DSCAM is a netrin receptor that collaborates with DCC in mediating turning responses to netrin-1. Cell, 133, 1241–1254. Lyle, R., Gehrig, C., Neergaard-Henrichsen, C., Deutsch, S., & Antonarakis, S. E. (2004). Gene expression from the aneuploid chromosome in a trisomy mouse model of Down syndrome. Genome Research, 14, 1268–1274. Ma, H., Xiong, H., Liu, T., Zhang, L., Godzik, A., et al. (2004). Aggregate formation and synaptic abnormality induced by DSCR1. Journal of Neurochemistry, 88, 1485–1496. Ma, Y. J., Okamoto, M., Gu, F., Obata, K., Matsuyama, T., et al. (2003). Neuronal distribution of EHSH1/intersectin: Molecular linker between clathrin-mediated endocytosis and signalling pathways. Journal of Neuroscience Research, 71, 468–477.

78

M. Rachidi and C. Lopes

Malago, W., Jr., Sommer, C. A., Del Cistia, A. C., Soares-Costa, A., Abrao Possik, P., et al. (2005). Gene expression profile of human Down syndrome leukocytes. Croatian Medical Journal, 46, 647–656. Malleret, G., Haditsch, U., Genoux, D., Jones, M. W., Bliss, T. V., et al. (2001). Inducible and reversible enhancement of learning, memory, and long-term potentiation by genetic inhibition of calcineurin. Cell, 104, 675–686. Mann, E. M. A., & Eisiri, M. M. (1989). The pattern of acquisition of plaques and tangles in the brains of patients under 50 years of age with Down’s syndrome. Journal of the Neurological Sciences, 89, 169–179. Mansuy, I. M. (2003). Calcineurin in memory and bidirectional plasticity. Biochemical and Biophysical Research Communications, 311, 1195–1208. Mansuy, I. M., Mayford, M., Jacob, B., Kandel, E. R., & Bach, M. E. (1998). Restricted and regulated overexpression reveals calcineurin as a key component in the transition from short-term to long-term memory. Cell, 92, 39–49. Mansuy, I. M., Winder, D. G., Moallem, T. M., Osman, M., Mayford, M., et al. (1998). Inducible and reversible gene expression with the rtTA system for the study of memory. Neuron, 1998(21), 257–265. Mao, J., Maye, P., Kogerman, P., Tejedor, F. J., Toftgard, R., et  al. (2002). Regulation of Gli1 transcriptional activity in the nucleus by Dyrk1. The Journal of Biological Chemistry, 277, 35156–35161. Mao, L., Zabel, C., Herrmann, M., Nolden, T., Mertes, F., et al. (2007). Proteomic shifts in embryonic stem cells with gene dose modifications suggest the presence of balancer proteins in protein regulatory networks. PLoS One, 2, e1218. Mao, R., Wang, X., Spitznagel, E. L., Jr., Frelin, L. P., Ting, J. C., et  al. (2005). Primary and secondary transcriptional effects in the developing human Down syndrome brain and heart. Genome Biology, 6, R107. Mao, R., Zielke, C. L., Zielke, H. R., & Pevsner, J. (2003). Global up-regulation of chromosome 21 gene expression in the developing Down syndrome brain. Genomics, 81, 457–467. Marin-Padilla, M. (1972). Structural abnormalities of the cerebral cortex in human chromosomal aberrations: A Golgi study. Brain Research, 44, 625–629. Marin-Padilla, M. (1976). Pyramidal cell abnormalities in the motor cortex of a child with Down’s syndrome. The Journal of Comparative Neurology, 167, 63–75. Martina, J. A., Bonangelino, C. J., Aguilar, R. C., & Bonifacino, J. S. (2001). Stonin 2: An adaptor-like protein that interacts with components of the endocytic machinery. The Journal of Cell Biology, 153, 1111–1120. Martinez-Cue, C., Baamonde, C., Lumbreras, M. A., Vallina, I. F., Dierssen, M., et al. (1999). A murine model for Down syndrome shows reduced responsiveness to pain. NeuroReport, 10, 1119–1122. McPherson, P. S. (1999). Regulatory role of SH3 domain-mediated protein-protein interactions in synaptic vesicle endocytosis. Cellular Signalling, 11, 229–238. McPherson, P. S., Garcia, E. P., Slepnev, V. I., David, C., Zhang, X., et al. (1996). A presynaptic inositol-5-phosphatase. Nature, 379, 353–357. McPherson, P. S., Takei, K., Schmid, S. L., & De Camilli, P. (1994). p145, a major Grb2-binding protein in brain, is co-localized with dynamin in nerve terminals where it undergoes activitydependent dephosphorylation. The Journal of Biological Chemistry, 269, 30132–30139. Meng, X., Shi, J., Peng, B., Zou, X., & Zhang, C. (2006). Effect of mouse Sim2 gene on the cell cycle of PC12 cells. Cell Biology International, 30, 349–353. Milner, B., Squire, L. R., & Kandel, E. R. (1998). Cognitive neuroscience and the study of memory. Neuron, 20, 445–468. Miyakawa, T., Leiter, L. M., Gerber, D. J., Gainetdinov, R. R., Sotnikova, T. D., et  al. (2003). Conditional calcineurin knockout mice exhibit multiple abnormal behaviours related to schizophrenia. Proceedings of the National Academy of Sciences of the United States of America, 100, 8987–8992. Miyashita, Y. (2004). Cognitive memory: Cellular and network machineries and their top-down control. Science, 306, 435–440.

Mental Retardation and Human Chromosome 21 Gene Overdosage

79

Monks, S. A., Leonardson, A., Zhu, H., Cundiff, P., Pietrusiak, P., et al. (2004). Genetic inheritance of gene expression in human cell lines. American Journal of Human Genetics, 75, 1094–1105. Morice, E., Andreae, L. C., Cooke, S. F., Vanes, L., Fisher, E. M., et al. (2008). Preservation of long-term memory and synaptic plasticity despit short-term impairments in theTc1 mouse model of Down syndrome. Learning & Memory, 15, 492–500. Morii, K., Tanaka, R., Takahashi, Y., Minoshima, S., Fukuyama, R., et al. (1991). Structure and chromosome assignment of human S100 alpha and beta subunit genes. Biochemical and Biophysical Research Communications, 175, 185–191. Morley, M., Molony, C. M., Weber, T. M., Devlin, J. L., Ewens, K. G., et al. (2004). Genetic analysis of genome-wide variation in human gene expression. Nature, 430, 743–747. Mural, R. J., Adams, M. D., Myers, E. W., Smith, H. O., Miklos, G. L., et al. (2002). A comparison of whole-genome shotgun-derived mouse chromosome 16 and the human genome. Science, 296, 1661–1671. Nadel, L. (2003). Down’s syndrome: A genetic disorder in biobehavioral perspective. Genes, Brain, and Behavior, 2, 156–166. Nadel, L., & Willner, J. (1980). Context and conditioning: A place for space. Physiol Psychol, 8, 218–228. Nambu, J. R., Lewis, J. O., Wharton, K. A., & Crews, S. T. (1991). The Drosophila single-minded gene encodes a helix-loop-helix protein that acts as a master regulator of CNS midline development. Cell, 67, 1157–1167. Newby, D., Aitken, D. A., Crossley, J. A., Howatson, A. G., Macri, J. N., et al. (1997). Biochemical markers of trisomy 21 and the pathophysiology of Down’s syndrome pregnancies. Prenatal Diagnosis, 17, 941–951. Nikolaienko, O., Nguyen, C., Crinc, L. S., Cios, K. J., & Gardiner, K. (2005). Human chromosome 21/Down syndrome gene function and pathway database. Gene, 364, 90–98. Nishiyama, H., Knopfel, T., Endo, S., & Itohara, S. (2002). Glial protein S100B modulates longterm neuronal synaptic plasticity. Proceedings of the National Academy of Sciences of the United States of America, 99, 4037–4042. O’Connell, R. M., Taganov, K. D., Boldin, M. P., Cheng, G., & Baltimore, D. (2007). MicroRNA155 is induced during the macrophage inflammatory response. Proceedings of the National Academy of Sciences of the United States of America, 104, 1604–1609. O’Doherty, A., Ruf, S., Mullingan, C., Hildreth, V., Errington, M. L., et al. (2005). An aneuploid mouse strain carrying human chromosome 21 with Down syndrome phenotypes. Science, 309, 2033–2037. Ohira, M., Ootsuyama, A., Suzuki, E., Ichikawa, H., Seki, N., et  al. (1996). Identification of a novel human gene containing the tetratricopeptide repeat domain from the Down syndrome region of chromosome 21. DNA Research, 3, 9–16. Ohira, M., Seki, N., Nagase, T., Suzuki, E., Nomura, N., et  al. (1997). Gene identification in 1.6-Mb region of the Down syndrome region on chromosome 21. Genome Research, 7, 47–58. Oleksiak, M. F., Churchill, G. A., & Crawford, D. L. (2002). Variation in gene expression within and among natural populations. Nature Genetics, 32, 261–266. Olson, L. E., Richtsmeier, J. T., Leszl, J., & Reeves, R. H. (2004). A chromosome 21 critical region does not cause specific Down syndrome phenotypes. Science, 306, 687–690. Olson, L. E., Roper, R. J., Baxter, L. L., Carlson, E. J., Epstein, C. J., et al. (2004). Down syndrome mouse models Ts65Dn, Ts1Cje, and Ms1Cje/Ts65Dn exhibit variable severity of cerebellar phenotypes. Developmental Dynamics, 230, 581–589. Olson, L. E., Roper, R. J., Sengstaken, C. L., Peterson, E. A., Aquino, V., et al. (2007). Trisomy for the Down syndrome “critical region” is necessary but not sufficient for brain phenotypes of trisomic mice. Human Molecular Genetics, 16, 774–782. Pascon, R. C., & Miller, B. L. (2000). Morphogenesis in Aspergillus nidulans requires Dopey (DopA), a member of a novel family of leucine zipper-like proteins conserved from yeast to humans. Molecular Microbiology, 36, 1250–1264.

80

M. Rachidi and C. Lopes

Patil, N., Cox, D. R., Bhat, D., Faham, M., Myers, R. M., et al. (1995). A potassium channel mutation in weaver mice implicates membrane excitability in granule cell differentiation. Nature Genetics, 11, 126–129. Pawson, T. (1995). Protein modules and signalling networks. Nature, 373, 573–580. Pearlson, G. D., Breiter, S. N., Aylward, E. H., Warren, A. C., Grygorcewicz, M., et al. (1998). MRI brain changes in subjects with Down syndrome with and without dementia. Developmental Medicine and Child Neurology, 40, 326–334. Pennington, B. F., Moon, J., Edgin, J., Stedron, J., & Nadel, L. (2003). The neuropsychology of Down syndrome: Evidence for hippocampal dysfunction. Child Development, 74, 75–93. Pinter, J. D., Brown, W. E., Eliez, S., Schmitt, J. E., Capone, G. T., et al. (2001). Amygdala and hippocampal volumes in children with Down syndrome: A high- resolution MRI study. Neurology, 56, 972–974. Pinter, J. D., Eliez, S., Schmitt, J. E., Capone, G. T., & Reiss, A. L. (2001). Neuroanatomy of Down’s syndrome: A high-resolution MRI study. The American Journal of Psychiatry, 158, 1659–1665. Potier, M. C., Rivals, I., Mercier, G., Ettwiller, L., Moldrich, R. X., et al. (2006). Transcriptional disruptions in Down syndrome: A case study in the Ts1Cje mouse cerebellum during postnatal development. Journal of Neurochemistry, 97(Suppl 1), 1004–109. Prandini, P., Deutsch, S., Lyle, R., Gagnebin, M., Delucinge Vivier, C., et al. (2007). Natural geneexpression variation in Down syndrome modulates the outcome of gene-dosage imbalance. American Journal of Human Genetics, 81, 252–263. Prasher, V. P. (2004). Review of donepezil, rivastigmine, galantamine and memantine for the treatment of dementia in Alzheimer’s disease in adults with Down syndrome: Implications for the intellectual disability population. International Journal of Geriatric Psychiatry, 19, 509–515. Priller, C., Bauer, T., Mitteregger, G., Krebs, B., Kretzschmar, H. A., et al. (2006). Synapse formation and function is modulated by the amyloid precursor protein. The Journal of Neuroscience, 26, 7212–7221. Pritchard, M. A., & Kola, I. (1999). The gene dosage effect hypothesis versus the amplified development instability hypothesis in Down syndrome. Journal of Neural Transmission, 57(Suppl), 293–303. Pucharcos, C., Fuentes, J. J., Casas, C., de la Luna, S., Alcàntara, S., et  al. (1999). Alu-splice cloning of the human intersectin (ITSN), a putative multivalent binding protein expressed in proliferating and differentiating neurons and over-expressed in Down syndrome. European Journal of Human Genetics, 7, 704–712. Pulsifer, M. B. (1996). The neuropsychology of mental retardation. Journal of the International Neuropsychological Society, 2, 159–176. Purpura, D. P. (1974). Dendritic spine dysgenesis and mental retardation. Science, 186, 1126–1128. Rachidi, M., & Lopes, C. (2007). Mental retardation in Down syndrome: From gene dosage imbalance to molecular and cellular mechanisms. Neuroscience Research, 59, 349–369. Rachidi, M., Lopes, C., Charron, G., Delezoide, A. L., Paly, E., et al. (2005). Spatial and temporal localization during embryonic and fetal human development of the transcription factor SIM2 in brain regions altered in Down syndrome. International Journal of Developmental Neuroscience, 23, 475–484. Rachidi, M., Lopes, C., Costantine, M., & Delabar, J. M. (2005). C21orf5, a new member of dopey family involved in morphogenesis, could participate in neurological alterations and mental retardation in Down syndrome. DNA Research, 12, 203–210. Rachidi, M., Lopes, C., Delezoide, A. L., & Delabar, J. M. (2006). C21orf5, a human candidate gene for brain abnormalities and mental retardation in Down syndrome. Cytogenetic and Genome Research, 112, 16–22. Rachidi, M., Lopes, C., Gassanova, S., Sinet, P. M., Vekemans, M., Attie, T., et al. (2000). Regional and cellular specificity of the expression of TPRD, the tetratricopeptide Down syndrome gene, during human embryonic development. Mechanisms of Development, 93, 189–193. Rachidi, M., Lopes, C., Vayssettes, C., Smith, D. J., Rubin, E. M., et al. (2007). New cerebellar phenotypes in YAC transgenic mouse in  vivo library of human Down syndrome critical region-1. Biochemical and Biophysical Research Communications, 364, 488–494.

Mental Retardation and Human Chromosome 21 Gene Overdosage

81

Rahmani, Z., Blouin, J. L., Creau-Goldberg, N., Watkins, P. C., Mattei, J. F., et al. (1989). Critical role of the D21S55 region on chromosome 21 in the pathogenesis of Down syndrome. Proceedings of the National Academy of Sciences of the United States of America, 86, 5958–5962. Rahmani, Z., Lopes, C., Rachidi, M., & Delabar, J. M. (1998). Expression of the mnb (dyrk) protein in adult and embryonic mouse tissues. Biochemical and Biophysical Research Communications, 253, 514–518. Raz, N., Torres, I. J., Briggs, S. D., Spencer, W. D., Thornton, A. E., et al. (1995). Selective neuroanatomic abnormalities in Down’s syndrome and their cognitive correlates: Evidence from MRI morphometry. Neurology, 45, 356–366. Reeves, R. H. (2006). Down syndrome mouse models are looking up. Trends in Molecular Medicine, 12, 237–240. Reeves, R. H., Irving, N. G., Moran, T. H., Wohn, A., Kitt, C., et al. (1995). A mouse model for Down syndrome exhibits learning and behaviour deficits. Nature Genetics, 11, 177–184. Reeves, R. H., Yao, J., Crowley, M. R., Buck, S., Zhang, X., et al. (1994). Astrocytosis and axonal proliferation in the hippocampus of S100b transgenic mice. Proceedings of the National Academy of Sciences of the United States of America, 91, 5359–5363. Reymond, A., Marigo, V., Yaylaoglu, M. B., Leoni, A., Ucla, C., et al. (2002). Human chromosome 21 gene expression atlas in the mouse. Nature, 420, 582–586. Risch, N., Merikangas, K. (1996). The future of genetic studies of complex human diseases. Science, 273, 1516–1517. Risser, D., Lubec, G., Cairns, N., & Herrera-Marschitz, M. (1997). Excitatory amino acids and monoamines in parahippocampal gyrus and frontal cortical pole of adults with Down syndrome. Life Sciences, 60, 1231–1237. Roizen, N. J., & Patterson, D. (2003). Down’s syndrome. Lancet, 361, 1281–1289. Roos, J., & Kelly, R. B. (1998). Dap160, a neural-specific Eps15 homology and multiple SH3 domain-containing protein that interacts with Drosophila dynamin. The Journal of Biological Chemistry, 273, 19108–19119. Roper, R. J., Baxter, L. L., Saran, N. G., Klinedinst, D. K., Beachy, P. A., et al. (2006). Defective cerebellar response to mitogenic Hedgehog signaling in Down syndrome mice. Proceedings of the National Academy of Sciences of the United States of America, 103, 1452–1456. Rothermel, B., Vega, R. B., Yang, J., Wu, H., Bassel-Duby, R., et al. (2000). A protein encoded within the Down syndrome critical region is enriched in striated muscles and inhibits calcineurin signaling. The Journal of Biological Chemistry, 275, 8719–8725. Rothermel, B. A., Vega, R. B., & Williams, R. S. (2003). The role of modulatory calcineurininteracting proteins in calcineurin signaling. Trends in Cardiovascular Medicine, 13, 15–21. Rueda, N., Florez, J., & Martinez-Cué, C. (2008). Chronic pentylenetetrazole but not donepezil treatment rescues spatial cognition in Ts65Dn mice, a model for Down syndrome. Neuroscience Letters, 433, 22–27. Sago, H., Carlson, E. J., Smith, D. J., Kilbridge, J., Rubin, E. M., et al. (1998). Ts1Cje, a partial trisomy 16 mouse model for Down syndrome, exhibits learning and behavioural abnormalities. Proceedings of the National Academy of Sciences of the United States of America, 95, 6256–6261. Sago, H., Carlson, E. J., Smith, D. J., Rubin, E. M., Crnic, L. S., et al. (2000). Genetic dissection of region associated with behavioral abnormalities in mouse models for Down syndrome. Pediatric Research, 48, 606–613. Saito, Y., Oka, A., Mizuguchi, M., Motonaga, K., Mori, Y., et al. (2000). The developmental and aging changes of Down’s syndrome cell adhesion molecule expression in normal and Down’s syndrome brains. Acta Neuropathologica, 100, 654–664. Sanna, B., Brandt, E. B., Kaiser, R. A., Pfluger, P., Witt, S. A., et al. (2006). Modulatory calcineurin-interacting proteins 1 and 2 function as calcineurin facilitators in vivo. Proceedings of the National Academy of Sciences of the United States of America, 103, 7327–7332. Saran, N. G., Pletcher, M. T., Natale, J. E., Cheng, Y., & Reeves, R. H. (2003). Global disruption of the cerebellar transcriptome in a Down syndrome mouse model. Human Molecular Genetics, 12, 2013–2019.

82

M. Rachidi and C. Lopes

Schadt, E. E., Monks, S. A., Drake, T. A., Lusis, A. J., Che, N., et al. (2003). Genetics of gene expression surveyed in maize, mouse and man. Nature, 422, 297–302. Schapiro, M. B., Haxby, J. V., & Grady, C. L. (1992). Nature of mental retardation and dementia in Down syndrome: Study with PET, CT, and neuropsychology. Neurobiology of Aging, 13, 723–734. Schapiro, M. B., Luxenberg, J. S., Kaye, J. A., Haxby, J. V., Friedland, R. P., et al. (1989). Serial quantitative CT analysis of brain morphometrics in adult Down’s syndrome at different ages. Neurology, 39, 1349–1353. Schimmel, M. S., Hammerman, C., Bromiker, R., & Berger, I. (2006). Third ventricle enlargement among newborn infants with trisomy 21. Pediatrics, 117, e928–e931. Schmidt-Sidor, B., Wisniewski, K. E., Shepard, T. H., & Sersen, E. A. (1990). Brain growth in Down syndrome subjects 1–22 weeks of gestation age and birth to 60 months. Clinical Neuropathology, 4, 181–190. Schmucker, D., Clemens, J. C., Shu, H., Worby, C. A., Xiao, J., et al. (2000). Drosophila Dscam is an axon guidance receptor exhibiting extraordinary molecular diversity. Cell, 101, 671–684. Schneider, C., Risser, D., Kirchner, L., Kitzmuller, E., Cairns, N., et al. (1997). Similar deficits of central histaminergic system in patients with Down syndrome and Alzheimer disease. Neuroscience Letters, 222, 183–186. Seabrook, G. R., Smith, D. W., Bowery, B. J., Easter, A., Reynolds, T., et al. (1999). Mechanisms contributing to the deficits in hippocampal synaptic plasticity in mice lacking amyloid precursor protein. Neuropharmacology, 38, 349–359. Sengar, A. S., Wang, W., Bishay, J., Cohen, S., & Egan, S. E. (1999). The EH and SH3 domain Ese proteins regulate endocytosis by linking to dynamin and Eps15. The EMBO Journal, 18, 1159–1171. Sethupathy, P., Borel, C., Gagnebin, M., Grant, G. R., Deutsch, S., et al. (2007). Human microRNA-155 on chromosome 21 differentially interacts with its polymorphic target in the AGTR1 3¢ untranslated region: A mechanism for functional single-nucleotide polymorphisms related to phenotypes. American Journal of Human Genetics, 81, 405–413. Shapiro, B. L. (1983). Down syndrome – A disruption of homeostasis. American Journal of Medical Genetics, 14, 241–269. Shapiro, B. L. (1997). Down syndrome critical regions? Human Genetics, 99, 421–423. Shapiro, B. L., & Whither-Azmitia, P. M. (1997). Down syndrome critical regions? Human Genetics, 99, 421–423. Shapiro, L. A., & Whitaker-Azmitia, P. M. (2004). Expression levels of cytoskeletal proteins indicate pathological aging of S100B transgenic mice: An immunohistochemical study of MAP-2, drebrin and GAP-43. Brain Research, 1019, 39–46. Shefner, J. M., Reaume, A. G., Flood, D. G., Scott, R. W., Kowall, N. W., et al. (1999). Mice lacking cytosolic copper/zinc superoxide dismutase display a distinctive motor axonopathy. Neurology, 53, 1239–1246. Shin, J. H., Gulesserian, T., Verger, E., Delabar, J. M., & Lubec, G. (2006). Protein dysregulation in mouse hippocampus polytransgenic for chromosome 21 structures in the Down syndrome critical region. Journal of Proteome Research, 5, 44–53. Shin, K. S., & Lubec, G. (2002). Drebrin, a dendritic spine protein, is manifold decreased in brains of patients with Alzheimer’s disease and Down syndrome. Neuroscience Letters, 324, 209–212. Shinohara, T., Tomizuka, K., Miyabara, S., Takehara, S., Kazuki, Y., et al. (2001). Mice containing a human chromosome 21 model behavioral impairment and cardiac anomalies of Down’s syndrome. Human Molecular Genetics, 10, 1163–1175. Siarey, R. J., Carlson, E. J., Epstein, C. J., Balbo, A., Rapoport, S. I., et  al. (1999). Increased synaptic depression in the Ts65Dn mouse, a model for mental retardation in Down syndrome. Neuropharmacology, 38, 1917–1920. Siarey, R. J., Stoll, S. I., Rapoport, S. I., & Galdzicki, Z. (1997). Altered long-term potentiation in the young and old Ts65Dn mouse, a model for Down syndrome. Neuropharmacology, 36, 1549–1554.

Mental Retardation and Human Chromosome 21 Gene Overdosage

83

Siarey, R. J., Villar, A. J., Epstein, C. J., & Galdzicki, Z. (2005). Abnormal synaptic plasticity in the Ts1Cje segmental trisomy 16 mouse model of Down syndrome. Neuropharmacology, 49, 122–128. Sitz, J. H., Tigges, M., Baumgartel, K., Khaspekov, L. G., & Lutz, B. (2004). Dyrk1A potentiates steroid hormone-induced transcription via the chromatin remodeling factor Arip4. Molecular and Cellular Biology, 24, 5821–5834. Smith, C. M., & Swash, M. (1978). Possible biochemical basis of memory disorder in Alzheimer disease. Annals of Neurology, 3, 471–473. Smith, D. J., Stevens, M. E., Sudanagunta, S. P., Bronson, R. T., Makhinson, M., et al. (1997). Functional screening of 2Mb of human chromosome 21q22.2 in transgenic mice implicates minibrain in learning defects associated with Down syndrome. Nature Genetics, 16, 28–36. Smith, D. J., Zhu, Y., Zhang, J., Cheng, J. F., & Rubin, E. M. (1995). Construction of a panel of transgenic mice containing a contiguous 2-Mb set of YAC/P1 clones from human chromosome 21q22.2. Genomics, 27, 425–434. Spielman, R. S., Bastone, L. A., Burdick, J. T., Morley, M., Ewens, W. J., et al. (2007). Common genetic variants account for differences in gene expression among ethnic groups. Nature Genetics, 39, 226–231. Stasko, M. R., & Costa, A. C. (2004). Experimental parameters affecting the Morris water maze performance of a mouse model of Down syndrome. Behavioural Brain Research, 154, 1–17. Stoll, C., Alembik, Y., Dott, B., & Roth, M. P. (1990). Epidemiology of Down syndrome in 118, 265 consecutive births. American Journal of Medical Genetics, 7(Suppl), 79–83. Storey, J. D., Madeoy, J., Strout, J. L., Wurfel, M., Ronald, J., et al. (2007). Gene-expression variation within and among human populations. American Journal of Human Genetics, 80, 502–509. Stranger, B. E., Forrest, M. S., Clark, A. G., Minichiello, M. J., Deutsch, S., et  al. (2005). Genome-wide associations of gene expression variation in humans. PLoS Genetics, 1, e78. Sturchler-Pierrat, C., Abramowski, D., Duke, M., Wiederhold, K. H., Mistl, C., et al. (1997). Two amyloid precursor protein transgenic mouse models with Alzheimer disease-like pathology. Proceedings of the National Academy of Sciences of the United States of America, 94, 13287–13292. Sultan, M., Piccini, I., Balzereit, D., Herwig, R., Saran, N. G., et al. (2007). Gene expression variation in Down’s syndrome mice allows prioritization of candidate genes. Genome Biology, 8, R91. Sumarsono, S. H., Wilson, T. J., Tymms, M. J., Venter, D. J., Corrick, C. M., et al. (1996). Down’s syndrome-like skeletal abnormalities in Ets2 transgenic mice. Nature, 379, 534–537. Sweatt, J. D. (2001). The neuronal MAP kinase cascade: A biochemical signal integration system subserving synaptic plasticity and memory. Journal of Neurochemistry, 76, 1–10. Sweatt, J. D., & Weeber, E. J. (2003). Genetics of childhood disorders: LII. Learning and memory, part 5: Human cognitive disorders and the ras/ERK/CREB pathway. Journal of the American Academy of Child and Adolescent Psychiatry, 42, 873–876. Taganov, K. D., Boldin, M. P., Chang, K. J., & Baltimore, D. (2006). NF-kappaB-dependent induction of microRNA miR-146, an inhibitor targeted to signaling proteins of innate immune responses. Proceedings of the National Academy of Sciences of the United States of America, 103, 12481–12486. Takashima, S., Becker, L. E., Armstrong, D. L., & Chan, F. (1981). Abnormal neuronal development in the visual cortex of the human fetus and infant with Down’s syndrome. A quantitative and qualitative Golgi study. Brain Research, 225, 1–21. Takashima, S., Ieshima, A., Nakamura, H., & Becker, L. E. (1989). Dendrites, dementia and the Down syndrome. Brain & Development, 11, 131–133. Takashima, S , Iida, K., Mito, T., & Arima, M. (1994). Dendritic and histochemical development and aging in patients with Down’s syndrome. Journal of Intellectual Disability Research, 38, 265–273. Teipel, S. J., Alexander, G. E., Schapiro, M. B., Moller, H. J., Rapoport, S. I., et al. (2004). Agerelated cortical grey matter reductions in non-demented Down’s syndrome adults determined by MRI with voxel-based morphometry. Brain, 127, 811–824. Teipel, S. J., & Hampel, H. (2006). Neuroanatomy of Down syndrome in vivo: A model of preclinical Alzheimer’s disease. Behavior Genetics, 36, 405–415.

84

M. Rachidi and C. Lopes

Teipel, S. J., Schapiro, M. B., Alexander, G. E., Krasuski, J. S., Horwitz, B., et al. (2003). Relation of corpus callosum and hippocampal size to age in non-demented adults with Down’s syndrome. The American Journal of Psychiatry, 160, 1870–1878. Tejedor, F., Zhu, X. M., Kaltenbach, E., Ackermann, A., Baumann, A., et al. (1995). Minibrain: A new protein kinase family involved in postembryonic neurogenesis in Drosophila. Neuron, 14, 287–301. Thomas, G. M., & Huganir, R. L. (2004). MAPK cascade signalling and synaptic plasticity. Nature Reviews. Neuroscience, 5, 173–183. Thomas, J. B., Crews, S. T., & Goodman, C. S. (1988). Molecular genetics of the single-minded locus: A gene involved in the development of the Drosophila nervous system. Cell, 52, 133–141. Tong, X. K., Hussain, N. K., de Heuvel, E., Kurakin, A., Abi-Jaoude, E., et al. (2000). The endocytic protein intersectin is a major binding partner for the Ras exchange factor mSos1 in rat brain. The EMBO Journal, 19, 1263–1271. Toyoda, A., Noguchi, H., Taylor, T. D., Ito, T., Pletcher, M. T., et al. (2002). Comparative genomic sequence analysis of the human chromosome 21 Down syndrome critical region. Genome Research, 12, 1323–1332. Tsukahara, F., Hattori, M., Muraki, T., & Sakaki, Y. (1996). Identification and cloning of a novel cDNA belonging to tetratricopeptide repeat gene fa mily from Down syndrome-critical region 21q22.2. Journal of Biochemistry, 120, 820–827. Vacik, T., Ort, M., Gregorova, S., Strnad, P, Blatny, R., et al. (2005). Segmental trisomy of chromosome 17: A mouse model of human aneuploidy syndromes. Proceedings of the National Academy of Sciences of the United States of America, 102, 4500–4505. Vega, R. B., Rothermel, B. A., Weinheimer, C. J., Kovacs, A., Naseem, R. H., et al. (2003). Dual roles of modulatory calcineurin-interacting protein 1 in cardiac hypertrophy. Proceedings of the National Academy of Sciences of the United States of America, 100, 669–674. Vialard, F., Toyama, K., Vernoux, S., Carlson, E. J., Epstein, C. J., et al. (2000). Over-expression of mSim2 gene in the zona limitans of the diencephalon of segmental trisomy 16 Ts1Cje fetuses, a mouse model for trisomy 21: A novel whole-mount RNA hybridization assay. Developmental Brain Research, 121, 73–78. Vicari, S. (2004). Memory development and intellectual disabilities. Acta Paediatrica, 93(Suppl), 60–63. Vicari, S. (2006). Motor development and neuropsychological patterns in persons with trisomy 21. Behavior Genetics, 36, 355–364. Villar, A. J., Belichenko, P. V., Gillespie, A. M., Kozy, H. M., Mobley, W. C., et al. (2005). Identification and characterization of a new Down syndrome model, Ts[Rb(12.1716)]2Cje, resulting from a spontaneous Robertsonian fusion between T(1716)65Dn and mouse chromosome 12. Mammalian Genome, 16, 79–90. Vonorov, S. V., Frere, S. G., Giovedi, S., Pollina, E. A., Borel, C., et  al. (2008). Synaptojanin 1-linked phosphoinositide dyshomeostasis and cognitive deficits in mouse models of Down’s syndrome. Proceedings of the National Academy of Sciences of the United States of America, 105, 9415–9420. Vuksic, M., Petanjek, Z., Rasin, M., & Kostovic, I. (2002). Perinatal growth of prefrontal layer III pyramidal in Down syndrome. Pediatric Neurology, 27, 36–38. Wang, C. C., Kadota, M., Nishigaki, R., Kazuki, Y., Shirayoshi, Y., et al. (2004). Molecular hierarchy in neurons differentiated from mouse ES cells containing a single human chromosome 21. Biochemical and Biophysical Research Communications, 314, 335–350. Wang, P. P., Doherty, S., Hesselink, J. R., & Bellugi, U. (1992). Callosal morphology concurs with neurobehavioral and neuropathological findings in two neurodevelopmental disorders. Archives of Neurology, 49, 407–411. Wang, Y., Stricker, H. M., Gou, D., & Liu, L. (2007). MicroRNA: Past and present. Frontiers Bioscience, 12, 2316–2329. Watson, D. K., McWilliams-Smith, M. J., Nunn, M. F., Duesberg, P. H., O’Brien, S. J., et al. (1985). The ets sequence from the transforming gene of avian erythroblastosis virus, E26, has unique domains on human chromosomes 11 and 21: Both loci are transcriptionally active. Proceedings of the National Academy of Sciences of the United States of America, 82, 7294–7298.

Mental Retardation and Human Chromosome 21 Gene Overdosage

85

Weis, S., Weber, G., Neuhold, A., & Rett, A. (1991). Down syndrome: MR quantification of brain structures and comparison with normal control subjects. American Journal of Neuroradiology, 12, 1207–1211. Weitzdoerfer, R., Stolzlechner, D., Dierssen, M., Ferreres, J., Fountoulakis, M., et  al. (2001). Reduction of nucleoside diphosphate kinase B, Rab GDP-dissociation inhibitor beta and histidine triad nucleotidebinding protein in fetal Down syndrome brain. Journal of Neural Transmission, 61, 347–359. Whitaker-Azmitia, P. M., Wingate, M., Borella, A., Gerlai, R., Roder, J., et al. (1997). Transgenic mice overexpressing the neurotrophic factor S-100 beta show neuronal cytoskeletal and behavioral signs of altered aging processes: Implications for Alzheimer’s disease and Down’s syndrome. Brain Research, 776, 51–60. Winocur, G., Roder, J., & Lobaugh, N. (2001). Learning and memory in S100-beta transgenic mice: An analysis of impaired and preserved function. Neurobiology of Learning and Memory, 75, 230–243. Wishart, J. G. (1995). Cognitive abilities in children with Down syndrome: Developmental instability and motivational deficits. In C. J. Epstein, T. Hassold, I. T. Lott, et al. (Eds.), Etiology and pathogenesis of Down syndrome (pp. 57–91). New York: Wiley-Liss. Wisniewski, K. E. (1990). Down syndrome children often have brain with maturation delay, retardation of growth, and cortical dysgenesis. American Journal of Medical Genetics, 7, 274–281. Wisniewski, K. E., & Bobinski, M. (1991). Hypothalamic abnormalities in Down syndrome. Progress in Clinical and Biological Research, 373, 153–167. Wisniewski, K. E., & Schmidt-Sidor, B. (1989). Postnatal delay of myelin formation in brains from Down syndrome infants and children. Clinical Neuropathology, 8, 55–62. Wolvetang, E. J., Bradfield, O. M., Hatzistavrou, T., Crack, P. J., Busciglio, J., et  al. (2003). Overexpression of the chromosome 21 transcription factor Ets2 induces neuronal apoptosis. Neurobiology of Disease, 14, 349–356. Wolvetang, E. J., Bradfield, O. M., Tymms, M., Zavarsek, S., Hatzistavrou, T., et al. (2003). The chromosome 21 transcription factor ETS2 transactivates the beta-APP promoter: Implications for Down syndrome. Biochimica et Biophysica Acta, 1628, 105–110. Woods, Y. L., Rena, G., Morrice, N., Barthel, A., Becker, W., et al. (2001). The kinase DYRK1A phosphorylates the transcription factor FKHR at Ser329 in vitro, a novel in vivo phosphorylation site. The Biochemical Journal, 355, 597–607. Yamabhai, M., Hoffman, N. G., Hardison, N. L., McPherson, P. S., Castagnoli, L., et al. (1998). Intersectin, a novel adaptor protein with two Eps15 homology and five Src homology 3 domains. The Journal of Biological Chemistry, 273, 31401–31407. Yamakawa, K., Huot, Y. K., Haendelt, M. A., Hubert, R., Chen, X. N., et al. (1998). DSCAM: A novel member of the immunoglobulin superfamily maps in a Down syndrome region and is involved in the development of the nervous system. Human Molecular Genetics, 7, 227–237. Yamamoto, H., Flannery, M. L., Kupriyanov, S., Pearce, J., McKercher, S. R., et  al. (1998). Defective trophoblast function in mice with a targeted mutation of Ets2. Genes & Development, 12, 1315–1326. Yang, E. J., Ahn, Y. S., & Chung, K. C. (2001). Protein kinase Dyrk1 activates cAMP response element-binding protein during neuronal differentiation in hippocampal progenitor cells. The Journal of Biological Chemistry, 276, 39819–39824. Yates, C. M., Simpson, J., Gordon, A., Maloney, A. F., Allison, Y., et al. (1983). Catecholamines and cholinergic enzymes in pre-senile and senile Alzheimer-type dementia and Down’s syndrome. Brain Research, 280, 119–126. Yoo, J. H., Valdovinos, M. G., & Williams, D. C. (2007). Relevance of donepezil in enhancing learning and memory in spetial populations. A review of the literature. Journal of Autism and Developmental Disorders, 37, 1883–1901. Yu, Y., Chu, P. Y., Bowser, D. N., Keating, D. J., Dubach, D. (2008). Mice deficient for the chromosome 21 ortholog Itsn1 exhibit vesicle trafficking abnormalities. Human Molecular Genetics. doi:10.1093/hng/ddn224. Yvert, G., Brem, R. B., Whittle, J., Akey, J. M., Foss, E., et al. (2003). Trans-acting regulatory variation in Saccharomyces cerevisiae and the role of transcription factors. Nature Genetics, 35, 57–64.

86

M. Rachidi and C. Lopes

Zang, D. W., Yang, Q., Wang, H. X., Egan, G., Lopes, E. C., et al. (2004). Magnetic resonance imaging reveals neuronal degeneration in the brainstem of the superoxide dismutase 1 transgenic mouse model of amyotrophic lateral sclerosis. The European Journal of Neuroscience, 20, 1745–1751. Zeng, H., Chattarji, S., Barbarosie, M., Rondi-Reig, L., Philpot, B. D., et al. (2001). Forebrainspecific calcineurin knochout selectively impairs bidirectional synaptic plasticity and working/ episodic-like memory. Cell, 107, 617–629.

Epigenetic Programming of Stress Responses and Trans-Generational Inheritance Through Natural Variations in Maternal Care A Role for DNA Methylation in Experience-Dependent (Re)programming of Defensive Responses Ian C.G. Weaver Abstract  Human epidemiological and animal studies show that many chronic adult conditions have their antecedents in compromised fetal and early postnatal development. Mother–infant interactions are the primary source of social stimulation and influence physiological and cellular defense mechanisms resulting in persistent changes in offspring phenotype. These effects appear to be mediated, in part, by neonatal programming of the hypothalamic–pituitary–adrenal (HPA) axis and glucocorticoid function in the neuroendocrine system. Rodent models provide evidence that epigenetic mechanisms are involved: natural variations in maternal care influence HPA stress reactivity in offspring via long-term changes in tissue-specific gene expression. Both in vivo and in vitro studies show that maternal licking and grooming increases glucocorticoid receptor expression in the offspring through increased hippocampal serotonergic tone accompanied by increased histone acetyltransferase activity, histone acetylation and DNA demethylation mediated by the transcription factor NGFI-A. These effects are reversed by early postnatal cross-fostering and by pharmacological manipulations in adulthood, including Trichostatin A (TSA) and l-methionine administration. Maternal care influences the maternal behavior of female offspring, an effect that appears to be related to epigenetic regulation of estrogen receptor-a expression in the medial preoptic area, providing a mechanism for trans-generational inheritance of maternal behavior from mother to offspring. Experience-induced changes in the epigenotype provide a potential mechanism to explain how physiological adaptations to changes in the early environment result in permanent programming and affect risk to disease later in life. Keywords  Maternal Behavior • Medial Preoptic Area • Oxytocin • Estrogen Receptor • Stress • Hippocampus • Glucocorticoid Receptor • Chromatin Plasticity • DNA Methylation I.C.G. Weaver (*) The Krembil Family Epigenetics Laboratory, Centre for Addiction and Mental Health (CAMH), 250 College Street, Toronto, ON, Canada M5T 1R8 e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_3, © Springer Science+Business Media, LLC 2011

87

88

I.C.G. Weaver

1 Introduction: Environmental Influences and the Origins of Adult Health and Disease When asked what worried him most, the British Prime Minister (1957–1963) Harold Macmillan famously replied: “Events, my dear boy, events” (Sandbrook, 2005). Our ability to endure the prevailing demands of life, however, appears to be associated with early-life events that influence health in adolescence and adulthood. The importance of early environmental influences on lifelong health has emerged from observations including those by Geoffrey Rose in the 1960s, describing a familial pattern of coronary heart disease (CHD), still birth and infant mortality (Rose, 1964). Infant mortality has since been geographically correlated with cardiovascular disease (Forsdahl, 1977). Retrospective studies found an inverse association between birth weight and adult CHD mortality, and concluded that these effects are mediated in part by intrauterine deprivation (Barker & Osmond, 1986). Consistent with this hypothesis, low birth weight is associated with an increased incidence of heart disease (Rich-Edwards et al., 1997), hypertension (Law & Shiell, 1996), and type 2 diabetes (Hales et al., 1991), as well as markers of abnormal glucose-insulin metabolism (Hales et  al., 1991) and serum cholesterol concentrations (Barker, Martyn, Osmond, Hales, & Fall, 1993). On the other hand, high birth weight is associated with childhood leukemia, testicular cancer and an increased risk in breast cancer (Hjalgrim et al., 2003; Michels et al., 1996; Michels & Xue, 2006). Consequently, reduced or heightened overall fetal body growth is seen as constitutive marker of a coordinated fetal response to the intrauterine environment, resulting in changes in tissue and organ development that condition the risk of disease later in life (Gluckman & Hanson, 2004). The persistent effects of early-life environmental cues and events (including maternal nutrition) on cellular function and physiology gave rise to the term “programming” (Hales & Barker, 2001; Lucas, 1991). It is now widely appreciated that both prenatal and early postnatal conditions are key in developmental programming and the emergence of certain metabolicdisorder phenotypes in the later stages of life (Gluckman, Hanson, Cooper, & Thornburg, 2008). The generation of different phenotypes from a single genome based on the conditions during early development forms the basis for “phenotypic plasticity.” Plasticity provides an organism with an evolutionary advantageous ability to hone specific biological defensive systems for survival in the prevailing environmental demand (stressors). However, such adaptations can be pathological if there is a mismatch between perinatal adaptation and later-life environmental conditions (Felitti et al., 1998; Lissau & Sorensen, 1994; McCauley et al., 1997). Therefore, the ability to cope (stress resilience) plays a vital role in mediating the effects of adverse conditions on health outcome. Stress-diathesis models suggest that maternal influences on the development of neuroendocrine systems that underlie the hypothalamic–pituitary–adrenal (HPA) axis and behavioral responses to stress mediate the relation between early-life environment and health in the adult offspring (Francis & Meaney, 1999; Heim et al., 2000; Nemeroff, 1996; Repetti, Taylor, & Seeman, 2002; Seckl & Meaney, 1993;

Epigenetic Programming of Stress Responses and Trans-Generational Inheritance

89

Sroufe, 1997). Importantly, the nature of the mother–offspring interaction influences the expression of genes that regulate the neuronal circuitry of behavioral responses in the offspring throughout life. This raises several key questions: (1) the identity of the relevant genomic targets; (2) the mechanisms that sustain gene expression; (3) whether the programmed state is reversible later in life; and (4) the mechanism for trans-generational inheritance. In this review, we discuss the results from studies using rodent models that suggest that maternal care in the first week of postnatal life establishes diverse and stable phenotypes in the offspring through the epigenetic modification of genes expressed in the brain that regulate neuroendocrine and behavioral responses throughout life. In particular, we focus on maternal programming of the rat hippocampal glucocorticoid receptor (GR) and HPA stress responses in the offspring. We also suggest a possible mechanism for transmission of maternal care across generations and conclude by considering the implications of our findings within the context of recently published studies in humans.

2 Maternal Care in the Rat and HPA and Behavioral Responses to Stress in Adulthood Small mammals such as the rat produce altricial, relatively helpless, offspring – they are born blind, deaf, naked (except for small vibrissae on the snout), are unable to urinate and defecate properly, have poor motor coordination and cannot maintain their own body temperature (Krinke, 2000). Neonate survival is dependent on their ability to process sensory information continuously and adapt to changing ambient conditions. Since the mother comprises the prime source of environmental input (nutrition, warmth, stimulation, protection, companionship, etc.), she essentially provides the nurturing environment crucial for survival of the pups. The results of observational studies of mother-pup interactions provide evidence for natural variations in maternal care including stable individual differences in licking/grooming (LG) and arched-back nursing (ABN) posture, over the first week of lactation (Caldji et al., 1998; Francis, Diorio, Liu, & Meaney, 1999; Liu et al., 1997; Myers, Brunelli, Shair, Squire, & Hofer, 1989; Stern, 1997). Exposure to different levels of maternal LG-ABN during the postnatal period is associated with HPA blunting and changes in forebrain GR expression levels that persist into adulthood (Francis et al., 1999; Liu et al., 1997) (Fig. 1). The adult offspring of mothers that exhibit increased levels of pup licking/grooming and ABN (i.e., High LG-ABN mothers) over the first week of life show increased hippocampal GR expression and enhanced glucocorticoid feedback sensitivity, decreased hypothalamic corticotropinreleasing factor (CRF) expression, and more modest HPA responses to stress by comparison to adult animals reared by Low LG-ABN mothers (Francis et al., 1999; Liu et al., 1997). In support of this idea that early “nurturing touch” regulates the development of stress reactivity of the neonate, artificially reared offspring that receive tactile stimulation with a warm, wet paintbrush for 15 min per day during

90

I.C.G. Weaver

Fig. 1  Hypothalamic–pituitary–adrenal (HPA) axis. In response to a physical or psychological stressor, parvocellular cells in the paraventricular nucleus of the hypothalamus produce corticotrophinreleasing factor (CRF) and arginine-vasopressin (AVP), which are then released into the hypophysial portal circulation. CRF and AVP stimulate the release and synthesis of adrenocorticotroph hormone (ACTH) from pro-opiomelanocortin (POMC) in the anterior pituitary gland. ACTH potently induces the adrenal cortex to secrete glucocorticoids (GC), cortisol in humans and corticosterone in rodents. GCs feed back by binding and activating hippocampal glucocorticoid receptors (GR) to inhibit further HPA activity, which prevents tissue damage from extended exposure to GCs, facilitating stress resistance and behavioral adaptation

the first week of postnatal life show increased hippocampal GR expression and decreased serum glucocorticoid release (Jutapakdeegul, Casalotti, Govitrapong, & Kotchabhakdi, 2003). Eliminating the difference in hippocampal GR levels eliminates the effects of early experience on HPA responses to stress in adulthood (Meaney, Aitken, Viau, Sharma, & Sarrieau, 1989), suggesting that the difference in hippocampal GR expression serves as a mechanism for the effects of early experience on the development of individual differences in HPA responses to stress (Meaney, 2001). Consistent with this, mice with forebrain-specific disruption of the GR gene and loss of hippocampal GR expression show impairments to HPA axis regulation and an increase in anxiety-related behavior (Boyle, Kolber, Vogt, Wozniak, & Muglia, 2006). Interestingly, the adult offspring of High LG-ABN mothers are behaviorally less fearful under conditions of stress than are animals reared by Low LG-ABN dams (Caldji et al., 1998), and subsequent gene expression profile analysis of these animals suggests that effects of early-life experience have a stable and broad effect on the hippocampal transcriptome, which may play a role in the development

Epigenetic Programming of Stress Responses and Trans-Generational Inheritance

91

of anxiety-mediated behaviors through life (Weaver, Meaney, & Szyf, 2006). Importantly, cross-fostering paradigms have provided evidence for direct effects of maternal behavior on the behavioral and neuroendocrine responses to stress (Francis et al., 1999; Weaver et al., 2004). These findings suggest that the developing rodent forebrain is exquisitely sensitive to tactile stimulation provided by the mother during the first week of life and that different frequencies of LG-ABN provided during this period program neurodevelopment with long-lasting consequences on hippocampal GR function and HPA responses to stress. However, since the rearing mother and not the biological mother defines the behavioral and neuroendocrine responses to stress, these studies support an epigenetic mechanism for long-term programming (Fleming, O’Day, & Kraemer, 1999; Francis et al., 1999; Liu et al., 1997; Meaney, 2001).

3 Epigenetic Gene Regulation: Heritable Changes in Gene Expression Potential Conrad H. Waddington originally defined “epigenetics” in the early 1940s as the study of “the interactions between genes and their products which bring phenotype into being” (Waddington, 1968). Epigenetics is now understood as the study of heritable changes in gene expression that are not caused by changes in the DNA sequence (Jaenisch & Bird, 2003). While the DNA sequence defines the primary structure of the proteins, epigenetic mechanisms control the quantity, location and timing of gene expression and are stably and mitotically heritable, providing a possible mechanism for the maintenance of cell type-specific gene expression in the developing brain (Mill & Petronis, 2007). Epigenetic regulation in mammalian cells is mediated principally through changes in DNA methylation (Razin, 1998), chromatin structure (Kadonaga, 1998), and non-coding RNAs (microRNA) (Bergmann & Lane, 2003; Chuang & Jones, 2007; Saito & Jones, 2006). The “epigenome” refers to the epigenetically modified genome, and the “epigenotype” refers to mitotically heritable patterns of DNA methylation and modifications to chromatin proteins that package DNA.

3.1 Chromatin Structure In the nucleus of eukaryotic cells, chromosomal DNA is packaged into chromatin fibers in repeating protein–DNA complexes called nucleosomes, comprised of approximately 146 bp of DNA wound 1.8 times around an octomer consisting of two copies each of histone proteins H2A, H2B, H3 and H4 (Kornberg, 1974). The attraction between the positively-charged histones and negatively-charged DNA maintains the histone–DNA interaction (Grunstein, 1997).

92

I.C.G. Weaver

The amino-terminal tails of the nucleosomal core histones are subjected to post-translational modifications such as acetylation (Wade, Pruss, & Wolffe, 1997), poly-ADP-ribosylation, carbonylation (Wondrak, Cervantes-Laurean, Jacobson, & Jacobson, 2000), glycosylation, methylation (Jenuwein, 2001), phosphorylation, SUMOylation (Shiio & Eisenman, 2003) and ubiquitination (Shilatifard, 2006). Typically, acetylation occurs at lysine residues and all acetylation modifications are associated with transcriptional activation, as are phosphorylation and arginine methylation modifications (Bernstein et  al., 2005). On the other hand, lysine methylation can be associated with either transcriptional activation or repression depending on the lysine residue. The effect of ubiquitination on transcription likewise is dependent on location, with ubiquitination on H2A and H2B associated with transcriptional repression and activation, respectively. Sumoylation has thus far been associated solely with transcriptional repression. Other chromatin remodeling systems that have been implicated in epigenetic changes include nucleosome sliding (mediated by ATP-dependent chromatin remodeling proteins) and histone substitution (exchange of histones from nucleosome with external histones) (Tsankova, Renthal, Kumar, & Nestler, 2007). Importantly, the relationship between regional patterns of histone modifications and locus-specific transcriptional activity provides evidence for the existence of a “histone code” for dictating cell-specific gene expression programs (Jenuwein & Allis, 2001). For the purpose of this review, we will focus on H3 acetylation of lysine 9 residues (H3K9Ac) at the 5¢ regions and how it pertains to regulating gene activation. The amount of acetylation on histone tails is controlled by the opposing enzymatic activities of histone acetyltransferases (HATs) and histone deacetylases (HDACs) (Kuo & Allis, 1998). Histone acetylation neutralizes the positive charge of the histone tail and decreases its affinity to negatively charged DNA generating a more open DNA conformation (euchromatin) (Hong, SchrothMatthews, Yau, & Bradbury, 1993; Sealy & Chalkley, 1978). Transcription factors, regulatory complexes and RNA polymerase transcription factors then have access to the DNA, and expression of the corresponding genes is facilitated. Thus, H3K9Ac is a marker of active gene transcription. On the other hand, removal of the acetyl group by HDAC enzymes restores the positive charge to the lysine residue, fostering stronger interactions between histones and DNA, reducing the accessibility of transcription factors and the transcription apparatus to their cognate binding sites, resulting in gene silencing (heterochromatin) (Davie & Chadee, 1998). Permissive and repressed intermediate chromatin states also provide an additional level of epigenetic regulation (Tsankova et  al., 2007). This suggests that the active targeting of histonemodifying enzymes and chromodomain proteins determines the state of histone modification and level of expression of the underlying genes.

3.2 DNA Methylation Euchromatin is generally associated with hypomethylated DNA, whereas heterochromatin is associated with hypermethylated DNA (Holliday & Pugh, 1975).

Epigenetic Programming of Stress Responses and Trans-Generational Inheritance

93

In mammalian genomes, methylation mostly occurs at the carbon-5 position of cytosine residues of 5¢-cytosine-phosphodiester-guanine (CpG)-3¢ dinucleotide motifs, converting cytosine to 5-methylcytosine (5-mC) (Razin & Szyf, 1984). In human cells 60–80% of CpG dinucleotide sequences are methylated (Razin & Szyf, 1984). Many studies of epigenetic regulation have focused on CpG-rich regions (CpG islands) often found at gene promoters (Antequera & Bird, 1993; Bird, 1996; GardinerGarden & Frommer, 1987). Recent studies, however, illustrate the importance of 5-mC markings within intragenic and intergenic regions in mediating tissue-specific gene expression (Ching et al., 2005; Fazzari & Greally, 2004; Khulan et al., 2006). In vertebrates, there is a cell-specific pattern of methylation on CpG dinucleotides (Razin & Szyf, 1984). These methylation patterns are actively maintained by a family of enzymes called DNA methyltransferases (DNMTs) that catalyze the transfer of a methyl-group (CH3 or C1 group) from the methyl donor S-adenosylmethionine (SAM) to cytosine residues to form 5-mC (Adams, McKay, Craig, & Burdon, 1979). The DNMT-1 maintenance methyltransferase has a preference for hemi-methylated DNA (Bestor & Verdine, 1994; Leonhardt & Bestor, 1993; Smith, 1994) and restores methylation to the CpG dinucleotides of the nascent DNA strand following DNA replication, providing a mechanism for the safeguarding of epigenetic programs in proliferating cells (Bestor, 1988, 1992). DNMT-3a and DNMT-3b, on the other hand, are responsible for the wave of de novo methylation during early embryogenesis and actively methylate CpG dinucleotides within non-dividing somatic cells, such as neurons (Okano, Bell, Haber, & Li, 1999). DNA methylation directly and indirectly regulates gene silencing through sequential effects on chromatin structure (Bird, 2001; Bird & Wolffe, 1999; Hashimshony, Zhang, Keshet, Bustin, & Cedar, 2003; Kadonaga, 1998; Li, 2002; Nan et al., 1998). In the direct mechanism, DNA methylation within transcription factor binding sites interferes with the binding of methylation sensitive transcription factors to their cognate binding sites (Tate & Bird, 1993; Watt & Molloy, 1988). Herein, DNA methylation serves as an epi-mutation of the transcription factor binding site and repels the transcription factor. Indirectly, methylated DNA attracts methyl-CpG binding domain (MBD)-containing protein family members such as methyl-CpG binding protein (MeCP)-2, which bind to methylated DNA to suppress gene expression (Bird, 2001; Bird & Wolffe, 1999; Hashimshony et al., 2003; Kadonaga, 1998; Li, 2002; Nan et  al., 1998). These MBD proteins are themselves transcriptional repressors, and are further coupled to other co-repressor proteins and histone modification enzymes, leading to repressive chromatin remodeling and gene silencing (Jones et al., 1998; Ng et al., 1999; Wade et al., 1999; Zhang et al., 1999).

3.3 DNA Demethylation Although DNA methylation patterns are removed passively during DNA replication in primordial germ cells (Morgan, Santos, Green, Dean, & Reik, 2005), the precise mechanisms by which active demethylation occurs in mammals remains a subject of debate (Ooi & Bestor, 2008). One proposal is that the cell is able to remove and replace

94

I.C.G. Weaver

mutations in the DNA by nucleotide excision repair (Weiss, Keshet, Razin, & Cedar, 1996). In contrast, base excision repair involves the removal of a mutated or chemically altered base and its replacement with the correct base (David & Williams, 1998). However, the removal of nucleotides would seriously compromise the integrity of the genome when undergoing complete demethylation, as observed in the paternal genome following fertilization of the embryo (Oswald et al., 2000). In addition, demethylation events are very rapid and it appears unlikely that an excision-repair system would be able to complete genome-wide demethylation and nucleotide replacement within this time period. MBD2b (a shorter isoform of MBD2) has been reported to trigger active DNA demethylation by removal of methyl groups directly from the cytosine base and to induce gene expression in mammalian cells (Bhattacharya, Ramchandani, Cervoni, & Szyf, 1999; Cervoni, Bhattacharya, & Szyf, 1999; Cervoni & Szyf, 2001; Detich, Bovenzi, & Szyf, 2003; Detich, Hamm et al. 2003; Detich, Theberge, & Szyf, 2002; Hamm et al., 2008; Ramchandani, Bhattacharya, Cervoni, & Szyf, 1999; Szyf & Bhattacharya, 2002a, 2002b). The reaction requires a water molecule and transfers the methyl group off the cytosine to form methanol (Bhattacharya et al., 1999; Cervoni et al., 1999; Ramchandani et al., 1999). Although the assignment of demethylase activity to MBD2b was contested (Boeke, Ammerpohl, Kegel, Moehren, & Renkawitz, 2000; Ng et al., 1999; Wade et al., 1999), MBD2b levels are inversely correlated to the levels of DNA methylation of certain genes in hepatocytes (Goel, Mathupala, & Pedersen, 2003) and lymphocytes from lupus patients (Balada, Ordi-Ros, Serrano-Acedo, Martinez-Lostao, & Vilardell-Tarres, 2007), and depletion of MBD2b results in hypermethylation of unmethylated genes in metastatic cancer (Pakneshan, Tetu, & Rabbani, 2004; Shukeir, Pakneshan, Chen, Szyf, & Rabbani, 2006). It has been suggested that MBD2b might carry the potential for bidirectional enzymatic activity (Detich et  al., 2002). Likewise, a recent publication provides evidence for DNA demethylation induced by MBD3 (Brown, Suderman, Hallett, & Szyf, 2008), while two other reports propose that DNMT-3a and DNMT-3b possess deaminase activity and are involved in a dynamic demethylation-methylation pathway that operates during gene transcription (Kangaspeska et al., 2008; Metivier et al., 2008). Herein, MBD2 knockout female mice are significantly slower at retrieving pups to their nests in comparison to the wild-type dams (Hendrich, Guy, Ramsahoye, Wilson, & Bird, 2001), signifying the potential involvement of DNA (de)methylation mechanisms in long-term programming by maternal behavior.

4 DNA Methylation and the Maternal Programming of Stress Responses The 5¢ non-coding variable exon 1 region of the rat hippocampal GR gene contains multiple alternate sequences, including the exon 17 sequence, which appears to be a brain-specific promoter (McCormick et  al., 2000). In adult rats, hippocampal expression of GR mRNA splice variants containing exon 17 is increased by maternal

Epigenetic Programming of Stress Responses and Trans-Generational Inheritance

95

LG-ABN behavior. High and Low LG-ABN mothers differ in the frequency of pup LG-ABN only during the first postnatal week. We found group differences in the methylation status of individual CpG dinucleotides in the exon 17 promoter sequence emerged during the same time period, involving a process of demethylation (Weaver et al., 2004). The exon 17 promoter contains a binding site for the transcription factor nerve growth factor-inducible protein A (NGFI-A, also known as Egr-1, Zif268 or Krox-24) (Habener, Meyer, Yun, Waeber, & Hoeffler, 1990; Hyman et  al., 1988; Milbrandt, 1987, 1988; Mitchell, Rowe, Boksa, & Meaney, 1990; Montminy, Sevarino, Wagner, Mandel, & Goodman, 1986; Self & Nestler, 1995; Sheng & Greenberg, 1990; Smythe, Rowe, & Meaney, 1994; Tsukada, Fink, Mandel, & Goodman, 1987). By the end of the first postnatal week, the 5¢ CpG dinucleotide of the NGFI-A binding site is demethylated in the High LG-ABN but not in the Low LG-ABN group, and the maternal effect persists through to adulthood. Crossfostering reverses the differences in methylation of the 5¢ CpG dinucleotide, proposing a direct relationship between maternal behavior and changes in DNA methylation of the GR exon 17 promoter (Weaver et al., 2004). The methylation differences described above suggest an alteration of NGFI-A binding. In support of this, chromatin immunoprecipitation (ChIP) assays (CraneRobinson, Myers, Hebbes, Clayton, & Thorne, 1999) indicate that binding of NGFI-A protein to the hippocampal GR exon 17 promoter in adult pups is three-fold higher in offspring of High LG-ABN mothers than in offspring of Low LG-ABN dams (Weaver et al., 2004). Also, transient transfection studies show that DNA methylation reduces the ability of NGFI-A to activate the GR exon 17 promoter (Weaver et al., 2007). NGFI-A activates the genes by recruiting a transcriptional coactivator and HAT called cyclic adenosine 3¢, 5¢ monophosphate (cAMP) response element binding proteinbinding protein (CBP) to the promoter region. ChIP assays on the same tissue samples used in the NGFI-A studies, with an antibody against the acetylated form of H3, show increased association of acetylated H3K9 with the GR exon 17 promoter in offspring of the High LG-ABN mothers (Weaver et al., 2007). These findings are consistent with the hypothesis that increased NGFI-A binding to the GR exon 17 promoter and recruitment of HATs results in increased transcriptional activation. Taken together, these findings suggest that an epi-mutation at a single cytosine residue within the NGFI-A response element alters NGFI-A binding and might explain the sustained effect of maternal care on hippocampal GR expression and HPA responses to stress.

4.1 Epigenetic Programming by Maternal Care Is Reversible in the Adult Animal Given the bidirectional relationship between DNA methylation and chromatin structure (Razin & Cedar, 1977), chromatin remodeling is potentially reversible and therefore amenable to therapeutic intervention (Szyf, 2001). Evidence for this is provided by tissue culture studies. The HDAC inhibitor (HDACi) Trichostatin A (TSA) can

96

I.C.G. Weaver

induce histone acetylation (Yoshida, Kijima, Akita, & Beppu, 1990) and trigger active, replication-independent DNA demethylation (Cervoni & Szyf, 2001). We have proposed that increased acetylation results in increased accessibility of a gene to the demethylation machinery (Cervoni & Szyf, 2001) and that transcription is required for active DNA demethylation in genes silenced through methylation (D’Alessio, Weaver, & Szyf, 2007). While small changes in DNA methylation could be attributable in part to the small number of new neurons that arise from the subgranulare zone of the dentate gyrus and subventricular zone of the lateral ventricles throughout life (Cameron & Gould, 1994), widespread changes in the epigenetic state of GR in the mammalian brain could only occur if the DNA demethylation and methylation machinery remain present in the cell. We therefore addressed the question of whether epigenetic programming early in life could be modulated during adulthood. Central infusion of TSA in the adult offspring of Low LG-ABN mothers increased H3K9Ac, cytosine demethylation, NGFI-A binding and GR exon 17 promoter activation and reduced HPA responses and anxiety-related behavior to levels comparable with those observed in the offspring of High LG-ABN dams (Weaver et al., 2004, 2006). Subsequent expression profiling of TSA-treated rats reveal specific effects of TSA on the hippocampal transcriptome (Weaver et al., 2006). These results suggest causal relationships between maternal care, histone acetylation, DNA methylation of the exon 17 promoter, GR expression and HPA responses to stress. Accordingly, we reasoned that, if DNA methylating and demethylating enzymes dynamically maintain the DNA methylation pattern in adult neurons, then it should also be possible to reverse the demethylated state of the GR exon 17 promoter. Dietary l-methionine is converted by methionine adenosyltransferase into SAM, which serves as a methyl donor for DNA methylation (described above) (Cantoni, 1975; Mudd & Cantoni, 1958). In addition, SAM inhibits DNA demethylation either by stimulating DNA methylation enzymes (Pascale et  al., 1991) or by inhibiting demethylases (Szyf, 2001), and systemic injection of the methyl-donor l-methionine has been previously shown to increase the levels of SAM and DNA methylation (Tremolizzo et al., 2002). Chronic central infusion of adult offspring of High or Low LG-ABN mothers with l-methionine increased DNA methylation within the NGFI-A binding site and reduced NGFI-A binding to the exon 17 promoter selectively in the offspring of High LG-ABN dams, abolishing group differences in both hippocampal GR expression and HPA responses to stress (Weaver et al., 2005). These studies illustrate that maternal epigenetic programming early in life can be reversed later in life, suggesting that DNA methylation patterns are dynamic and potentially reversible even in adult neurons, which presumably contain the machinery required for de novo DNA demethylation or methylation that mediate CpG methylation rheostasis in the mature mammalian brain. Fig. 2 (continued) which mediates HPA and behavioral responses. Asterisk: In the adult rat, the epigenetic state of the hippocampal GR exon 17 promoter is reversible. Injection of the histone deacetylase inhibitor Trichostatin A (TSA) into the hippocampus of offspring from Low LG-ABN mothers increases histone acetylation, facilitating demethylation and increased activation of the GR exon 17 promoter. On the other hand, l-methionine (MET) inhibits DNA demethylation and increases DNA methylation, which inhibits NGFI-A binding and reduces GR exon 17 promoter activity in the offspring of High LG-ABN dams (see text for details)

Epigenetic Programming of Stress Responses and Trans-Generational Inheritance

97

4.2 Mechanisms Leading from Maternal Care to Chromatin Plasticity We propose that maternal behavior stimulates a signaling pathway, which activates certain transcription factors directing the epigenetic machinery (chromatin and DNA modifying enzymes) to specific targets within the genome. Our in  vivo and in  vitro studies suggest that maternal LG in early life elicits a thyroid hormone-dependent increase in serotonin (5-HT) activity at 5-HT7 receptors, and the subsequent activation of cAMP and cAMP-dependent protein kinase A (PKA) accompanied by increased hippocampal expression of the transcription factor NGFI-A and CBP (Chawla, Hardingham, Quinn, & Bading, 1998) (Meaney, Aitken, Sapolsky, 1987; Meaney et al., 2000) (Fig. 2a). In synergy, NGFI-A also directly regulates transcription of CBP, which has HAT activity (Yu et al., 2004). NGFI-A and CBP are recruited to the GR exon 17

Fig. 2  Our model of epigenetic (re)programming of hippocampal glucocorticoid receptor (GR) gene expression and stress responses by maternal behavior. (a) Maternal licking/grooming and arched-back nursing (LG-ABN) of the offspring increases hippocampal serotonin (5-HT) turnover and activation of a 5-HT7 receptor, which is positively coupled to cyclic adenosine 3¢, 5¢ monophosphate (cAMP). Increased cAMP activity results in activation of protein kinase-A (PKA) and cAMP response elementbinding protein (CREB). Subsequent phosphorylated-CREB (pCREB) activity drives expression of the transcription factor NGFI-A, which targets its cognate binding site on the GR exon 17 promoter. (b) NGFI-A recruits a histone acetyltransferase called CREB binding protein (CBP) that increases acetylation and accessibility to the DNA demethylase MBD2b and stable GR promoter activation. (c)  Increased stress levels stimulate GR activation in the hippocampus (described in Fig.  1),

98

I.C.G. Weaver

promoter in response to maternal care (Weaver et al., 2007). The GR exon 17 promoter region contains a binding site for NGFI-A (McCormick et  al., 2000). Interestingly, NGFI-A can actively target methylated-DNA binding proteins to genomic targets (Carvin, Parr, & Kladde, 2003), and we previously showed that a pharmacological increase in acetylation in vivo using TSA resulted in demethylation of the GR promoter in the hippocampus (Weaver et al., 2004). Equally, tissue culture experiments demonstrate that recruitment of NGFI-A to the GR exon 17 promoter results in replicationindependent demethylation (Cervoni & Szyf, 2001). We have recently found that both NGFI-A and MBD2b proteins simultaneously bind the same GR exon 17 promoter molecule and that binding of NGFI-A to its cognate binding site was required for MBD2b to function as a demethylating enzyme (Weaver et al., unpublished data). While the exact mechanisms by which NGFI-A recruits MBD2b to the GR exon 17 promoter remain unknown, these findings are consistent with the idea that NGFI-A facilitates the accessibility of the DNA sequence to MBD2b leading to targeted active demethylation (Fig. 2b). Accordingly, we propose that NGFI-A plays a bimodal role in the regulation of GR expression. During early development, high physiological levels of NGFI-A induced by maternal behavior interact with the methylated GR exon 17 promoter and trigger demethylation of the sequence, whereas later in life physiological levels of NGFI-A discriminate between the methylated and unmethylated GR exon 17 promoters and selectively activate the unmethylated sequences. Therefore, the different methylation states of the GR exon 17 promoter from the offspring of High and Low LG-ABN results in different levels of hippocampal GR expression (Fig. 2c). This suggests that the neonatal brain of altricial species such as the rat is not an immature version of the adult brain but is uniquely designed to optimize epigenetic programming by the mother.

5 Trans-Generational Inheritance of Epigenetic Programming by Maternal Behavior Another important aspect of epigenetic traits is their potential heritability. Transgenerational epigenetic inheritance has been demonstrated in plants (Richards, 2006) and mammals (Morgan, Sutherland, Martin, & Whitelaw, 1999). The question here is how these maternal effects remain stably transmitted across generations. Evidence from both human and non-human primates suggest that individual differences in infant-directed behaviors are transmitted from mother to daughter (Fairbanks, 1989; Miller, Kramer, Warner, Wickramaratne, & Weissman, 1997). In the rat, the adult female offspring of High LG-ABN dams are themselves high in maternal LG-ABN behavior towards their pups and, likewise, the offspring of Low LG-ABN mothers are low in maternal LG-ABN behavior towards their pups (Francis et  al., 1999). Consistent with this, cross-fostering paradigms provide evidence for direct effects of maternal behavior on the transmission of maternal

Epigenetic Programming of Stress Responses and Trans-Generational Inheritance

99

LG-ABN behavior (Francis et al., 1999), supporting an epigenetic mechanism for transmission of individual differences in maternal behavior. The central neuroanatomical circuitry in the female rat brain thought to mediate maternal-responsive behaviors is termed the maternal circuit (for review, see Fleming, 1986; Numan & Numan, 1994). Mechanisms regulating estrogen receptor (ER)-a function in the medial preoptic area (MPOA) and the connections to the maternal circuit and the mesolimbic dopamine system appear to be crucial in regulating maternal LG-ABN behavior towards the offspring (Champagne, Diorio, Sharma, & Meaney, 2001, 2003; Champagne et  al., 2004; Numan & Callahan, 1980; Stack, Balakrishnan, Numan, & Numan, 2002). In infancy, the female offspring of High LG-ABN mothers show increased MPOA ER-a expression compared to Low LG-ABN mothers, which is maintained into adulthood (Champagne et al., 2003). Consistent with this, cross-fostering paradigms provided evidence for direct effects of maternal behavior on ER-a expression in the MPOA (Champagne, Weaver, Diorio, Dymov, Szyf, & Meaney, 2006). The question then concerns the mechanism whereby variations in maternal behavior result in long-term changes of MPOA ER-a expression in the offspring.

5.1 Evidence for Epigenetic Mechanisms Linking Maternal Care of the Mother with Maternal Behavior in the Female Offspring Differential promoter usage is an important mechanism for tissue-specific and developmentally regulated ER-a gene expression (Schibler & Sierra, 1987). The 5¢ non-coding variable exon 1 region of the rat ER-a gene contains multiple alternate sequences, including the exon 1b sequence, which appears to be a neuronal tissuespecific promoter (Schibler & Sierra, 1987). We argued that the maternal programming of ER-a expression was associated with differences in cytosine methylation of the exon 1b promoter. We found group differences in the methylation status of individual CpG dinucleotides in the exon 1b promoter sequence; levels of DNA methylation were decreased in the adult offspring of High compared to Low LG-ABN mothers (Champagne et al., 2006). Activation of the Janus kinase (JAK) and signal transducers and activators of transcription (STAT) pathway increases ER-a expression (Frasor & Gibori, 2003). Exon 1b contains a binding site for the growth hormone- and prolactin-activated STAT-5b protein. The difference in methylation of the CpG dinucleotides in the STAT-5b binding site suggests an alteration of STAT-5b binding. In support of this, ChIP assays indicate that STAT-5b binding to the MPOA ER-a exon 1b promoter in adult females is increased in offspring of High LG-ABN mothers compared to Low LG-ABN mothers. Finally, we found increased expression of the transcription factor STAT-5b in the MPOA of neonates reared by High compared with Low LG-ABN mothers, with no differences in STAT-5b expression in the adult offspring

100

I.C.G. Weaver

(Champagne et  al., 2006). The role of STAT-5b in epigenetic regulation of the MPOA ER-a exon 1b promoter may therefore be bimodal, similar to the proposed role NGFI-A plays in regulation of hippocampal GR expression (described above). From these studies, our working hypothesis is that increased maternal LG-ABN behavior activates STAT-5b expression and targets demethylation of the STAT-5b binding site on the ER-a exon 1b promoter, which increases exon 1b promoter activation and expression of ER-a in the MPOA (Fig.  3a–c). Increased ER-a expression in the MPOA might serve to increase estrogen sensitivity in response to the rising hormone levels experienced in late gestation. ER-a is essential for estradiol-mediated induction of oxytocin receptor (OTR) gene expression and receptor binding (Bale & Dorsa, 1995; de Kloet, Voorhuis, Boschma, & Elands, 1986; Johnson et  al., 1989; Young, Wang, Donaldson, & Rissman, 1998). The nonapeptide OT binds its cognate receptor to modulate gene transcription and is a key mediator of complex emotional and social behaviors, including maternal responsivity (Fahrbach, Morrell, & Pfaff, 1984; Pedersen & Prange, 1979) and levels of licking and grooming of the offspring (Fahrbach et al., 1984). Furthermore, increases in hypothalamic OTR binding may potentially activate mesolimbic dopaminergic neurons, which serve to increase the duration and frequency of LG provided towards the pups (Fig. 3d). Although future studies will obviously be critical in providing evidence of causality in this general cascade of events, we believe these studies form the basis for a mechanism for the programming of individual differences in ER-a expression and the transmission of maternal behavior in the female offspring.

5.2 Environmental Influences and the Mother–Offspring Dyad Exposure of the mother to environmental adversity alters the nature of the mother– offspring interaction, which in turn influences the development of defensive responses to threat as well as reproductive strategies in the offspring (Agrawal, 2001; Rossiter, 1999) (Fig. 3 asterisk). For example, gestational stress during the third trimester reduces OTR levels and LG behavior in High LG-ABN mothers, an effect which is sustained through the second and third litter (Champagne & Meaney, 2006). The adult offspring of the gestationally stressed High LG-ABN mothers resemble those of Low LG-ABN mothers on behavioral measures of anxiety and maternal behavior (Champagne & Meaney, 2006). Furthermore, the prenatallystressed High LG-ABN offspring also have reduced hippocampal GR protein expression comparable to non-stressed Low LG-ABN offspring (Weaver, Champagne, & Meaney, unpublished data). Consistent with this, Mueller & Bale (2008) recently showed using mice that early gestational stress results in increased hypothalamic CRF and GR exon 17 promoter methylation and decreased central CRF and GR expression, which in turn increases HPA responsivity in the offspring (Mueller & Bale, 2008). This is the first illustration of epigenetic reprogramming

Epigenetic Programming of Stress Responses and Trans-Generational Inheritance

101

Fig. 3  Our model of epigenetic (re)programming of estrogen receptor-alpha (ER-a) gene expression in the medial preoptic area (MPOA) and the transmission of maternal behavior in the female offspring. (a) Maternal licking and grooming (LG) toward the offspring might activate the Janus kinase (JAK) and signal transducers and activators of transcription (STAT) pathway that drives expression of the transcription factor STAT-5b, which targets its cognate binding site on the ER-a exon 1b promoter. (b) STATs recruit histone acetyltransferases that increase acetylation and accessibility to DNA demethylase enzyme(s) and stable ER-a promoter activation. (c) Increased ER-a expression leads to increased estrogen (E) sensitivity in response to the rising hormone levels experienced in late gestation. The hormone-bound form of ER-a is a transcription factor that drives oxytocin receptor (OTR) gene expression. Oxytocin (OT) binds to its cognate receptor and regulates gene expression in neural systems mediating maternal LG behavior. (d) Connections between the mesolimbic dopamine (DA) neurons and hypothalamic OT neurons mediate these effects. Before the onset of maternal LG, the magnitude of DA release is greater amongst High LG mothers, which in turn LG their offspring for longer bouts by comparison to Low LG dams. Asterisk: Points of inflection: gestationally-stress High LG-ABN mothers show reduced MPOA OTR expression and LG behavior; whereas postweaning environmental enrichment and impoverishment reverse the maternal effects of low and high LG behavior on MPOA OTR expression and LG behavior in female offspring, respectively. These effects are also transmitted to the next generation of offspring (see text for details)

by stress-mediated modulation of the epigenome during early gestation and suggests that the enzyme(s) required for DNA methylation are involved and might also contribute to the timing and vulnerability of the developing fetus to maternal perturbations during pregnancy (Mueller & Bale, 2008). Interestingly, post-weaning environmental enrichment and impoverishment can reverse behavioral differences between female High and Low LG-ABN offspring on measures of OTR expression, maternal behavior and anxiety (Champagne & Meaney, 2007), implying that the

102

I.C.G. Weaver

social environment both within and beyond the postnatal period can (re)program the neural pathways that regulate maternal LG-ABN behavior. Together, these findings suggest that mother–offspring interactions early in life enhance the capacity for defensive responses in the progeny by programming emotional, cognitive and endocrine systems toward increased sensitivity to adversity, and that these programs can be transmitted across generations via an epigenetic mode of inheritance involving maternal behavior, which in turn is influenced by the ambient conditions.

6 Epigenetic Programming Early in Life and Inter-Individual Differences in Human Behavior and Health Studies in humans suggest that the forebrain GR function is complicit in the regulation of the HPA axis and the development of affective disorders and other sequelae (DeRijk & Sternberg, 1997; Holsboer, 2000; Invitti, Redaelli, Baldi, & Cavagnini, 1999). This raises the question of whether epigenetic modification in response to early environmental conditions can explain the effects of early infant adversity on adult health in humans [for detailed review, see Weaver (2009)]. Due to the important ethical, social and legal implications arising from any attempt to harvest living cells from the CNS, the epigenetic alterations in the peripheral blood mononuclear cells (PBMCs) are commonly compared in human studies. Interestingly, during their lifetime, monozygotic twins increasingly differ in their epigenotype (life-long drift) (Fraga et al., 2005), which might explain the frequent discordance of neuropsychiatric disorders such as schizophrenia and bipolar disorder (Kato, Iwamoto, Kakiuchi, Kuratomi, & Okazaki, 2005). These studies raise the possibility that suboptimal epigenetic modifications arise over time resulting in late onset mental pathologies. In support of this, ribosomal RNA (rRNA) promoter methylation (Brown & Szyf, 2007, 2008) was shown to be increased in suicide victims compared to controls (McGowan et al., 2008), suggesting a reduced capacity for protein synthesis in suicide brains (Brown & Szyf, 2007, 2008). Protein synthesis has long been known to be required for associative learning to consolidate into long-term memory (Agranoff, Davis, Casola, & Lim, 1967), which involves epigenetic regulation (Korzus, Rosenfeld, & Mayford, 2004), and a decline in cognitive plasticity is commonly observed with age (Kadar, Silbermann, Brandeis, & Levy, 1990). More recently, however, we found altered expression levels of DNMT enzymes and aberrant gene silencing by DNA methylation in suicide brains (Poulter et al., 2008), suggesting that a shift in the steady-state balance between DNA methylating and demethylating machinery might influence specific neural pathways and account for inter-individual differences in emotional reactivity and mental health in humans. The effects of tactile stimulation through mother–infant interactions on interindividual differences in cognitive development and stress responses in rodents is consistent with work in humans (Feldman, Eidelman, Sirota, & Weller, 2002) and non-human primates (Harlow & Zimmermann, 1959). This raises the question of

Epigenetic Programming of Stress Responses and Trans-Generational Inheritance

103

whether comparable epigenetic labile regions to the GR exon 17 promoter exist in the human genome. Alignment of splice sites reveals that the distally located exon 1F promoter of human type II GR (hGR, OMIM +138040; NR3C1) shows high homology to the GR exon 17 promoter in the rat, and contains an NGFI-A binding site (Turner & Muller, 2005). Interestingly, studies in healthy human subjects show that CpG methylation patterns of conserved transcription factor binding sites on the NR3C1 exon 1F promoter are both stochastic and unique to the individual (Turner, Pelascini, Macedo, & Muller, 2008). Furthermore, neonatal methylation at the 5¢ CpG dinucleotide within the NGFI-A binding site on the NR3C1 exon 1F promoter has been suggested as an early epigenetic marker of maternal mood and risk of altered HPA function in the developing infant (Oberlander et al., 2008). Although future studies are required to examine the functional consequence of the methylated 5¢ CpG dinucleotide, these findings are consistent with our studies in the neonate and adult offspring of Low LG-ABN mothers that show hypermethylation of the 5¢ CpG dinucleotide within the NGFI-A binding site on the exon 17 promoter, decreased GR expression and increased HPA responsivity (Weaver et al., 2004). In support of this paradigm, a recent study shows that the NGFI-A binding sites on the NR3C1 exon 1F promoter are hypermethylated in the hippocampus of suicide victims with a history of childhood abuse, and that mRNA transcript expression from the exon 1F promoter is decreased in comparison to controls (victims of sudden, accidental death with no history of abuse) (McGowan et al., 2009), suggesting that the transmission of vulnerability for depression from parent to offspring could occur through the epigenetic modification of genomic regions that are implicated in the regulation of stress responses.

7 Concluding Remarks Epigenetic mechanisms are likely to play an underlying role in the developmental origins of health and disease, whereby transient environmental cues and events during early development can persistently alter gene regulation resulting in metabolic imprinting affecting disease susceptibility. The studies presented in this review provide support for the effect of maternal behavior on endocrine and behavioral responses to stress in the offspring, and that these effects are biologically embedded throughout life by epigenetic programming. However, reprogramming can take place at several points throughout the life-span in response to changes in environmental conditions. Notably, maternal care influences the maternal behavior of female offspring, an effect that also appears to be related to epigenetic regulation of endocrine function, providing a mechanism for trans-generational inheritance of maternal behavior from mother to offspring. Because environmental stressors influence the nature of maternal behavior, maternal care remains a key mediator of epigenetic programming of neurodevelopment, and in turn the expression of biological defense systems that respond to environmental adversity.

104

I.C.G. Weaver

Acknowledgments  I would like to thank Dr. Shelley E. Brown for her helpful comments and numerous constructive suggestions throughout the preparation of this manuscript. Competing Interests Statement. The author declares that he has no competing financial interests.

References Adams, R. L., McKay, E. L., Craig, L. M., & Burdon, R. H. (1979). Mouse DNA methylase: Methylation of native DNA. Biochimica et Biophysica Acta, 561, 345–357. Agranoff, B. W., Davis, R. E., Casola, L., & Lim, R. (1967). Actinomycin D blocks formation of memory of shock-avoidance in goldfish. Science, 158, 1600–1601. Agrawal, A. A. (2001). Phenotypic plasticity in the interactions and evolution of species. Science, 294, 321–326. Antequera, F., & Bird, A. (1993). CpG islands. EXS, 64, 169–185. Balada, E., Ordi-Ros, J., Serrano-Acedo, S., Martinez-Lostao, L., & Vilardell-Tarres, M. (2007). Transcript overexpression of the MBD2 and MBD4 genes in CD4+ T cells from systemic lupus erythematosus patients. Journal of Leukocyte Biology, 81, 1609–1616. Bale, T. L., & Dorsa, D. M. (1995). Regulation of oxytocin receptor messenger ribonucleic acid in the ventromedial hypothalamus by testosterone and its metabolites. Endocrinology, 136, 5135–5138. Barker, D. J., Martyn, C. N., Osmond, C., Hales, C. N., & Fall, C. H. (1993). Growth in utero and serum cholesterol concentrations in adult life. BMJ, 307, 1524–1527. Barker, D. J., & Osmond, C. (1986). Infant mortality, childhood nutrition, and ischaemic heart disease in England and Wales. Lancet, 1, 1077–1081. Bergmann, A., & Lane, M. E. (2003). HIDden targets of microRNAs for growth control. Trends in Biochemical Sciences, 28, 461–463. Bernstein, B. E., Kamal, M., Lindblad-Toh, K., Bekiranov, S., Bailey, D. K., Huebert, D. J., et al. (2005). Genomic maps and comparative analysis of histone modifications in human and mouse. Cell, 120, 169–181. Bestor, T. (1988). Structure of mammalian DNA methyltransferase as deduced from the inferred amino acid sequence and direct studies of the protein. Biochemical Society Transactions, 16, 944–947. Bestor, T. H. (1992). Activation of mammalian DNA methyltransferase by cleavage of a Zn binding regulatory domain. The EMBO Journal, 11, 2611–2617. Bestor, T. H., & Verdine, G. L. (1994). DNA methyltransferases. Current Opinion in Cell Biology, 6, 380–389. Bhattacharya, S. K., Ramchandani, S., Cervoni, N., & Szyf, M. (1999). A mammalian protein with specific demethylase activity for mCpG DNA. Nature, 397, 579–583. Bird, A. (2001). Molecular biology. Methylation talk between histones and DNA. Science, 294, 2113–2115. Bird, A. P. (1996). The relationship of DNA methylation to cancer. Cancer Surveys, 28, 87–101. Bird, A. P., & Wolffe, A. P. (1999). Methylation-induced repression–belts, braces, and chromatin. Cell, 99, 451–454. Boeke, J., Ammerpohl, O., Kegel, S., Moehren, U., & Renkawitz, R. (2000). The minimal repression domain of MBD2b overlaps with the methyl-CpG-binding domain and binds directly to Sin3A. The Journal of Biological Chemistry, 275, 34963–34967. Boyle, M. P., Kolber, B. J., Vogt, S. K., Wozniak, D. F., & Muglia, L. J. (2006). Forebrain glucocorticoid receptors modulate anxiety-associated locomotor activation and adrenal responsiveness. The Journal of Neuroscience, 26, 1971–1978. Brown, S. E., Suderman, M. J., Hallett, M., & Szyf, M. (2008). DNA demethylation induced by the methyl-CpG-binding domain protein MBD3. Gene, 420, 99–106. Brown, S. E., & Szyf, M. (2007). Epigenetic programming of the rRNA promoter by MBD3. Molecular and Cellular Biology, 27, 4938–4952.

Epigenetic Programming of Stress Responses and Trans-Generational Inheritance

105

Brown, S. E., & Szyf, M. (2008). Dynamic epigenetic states of ribosomal RNA promoters during the cell cycle. Cell Cycle, 7, 382–390. Caldji, C., Tannenbaum, B., Sharma, S., Francis, D., Plotsky, P. M., & Meaney, M. J. (1998). Maternal care during infancy regulates the development of neural systems mediating the expression of fearfulness in the rat. Proceedings of the National Academy of Sciences of the United States of America, 95, 5335–5340. Cameron, H. A., & Gould, E. (1994). Adult neurogenesis is regulated by adrenal steroids in the dentate gyrus. Neuroscience, 61, 203–209. Cantoni, G. L. (1975). Biological methylation: Selected aspects. Annual Review of Biochemistry, 44, 435–451. Carvin, C. D., Parr, R. D., & Kladde, M. P. (2003). Site-selective in vivo targeting of cytosine-5 DNA methylation by zinc-finger proteins. Nucleic Acids Research, 31, 6493–6501. Cervoni, N., Bhattacharya, S., & Szyf, M. (1999). DNA demethylase is a processive enzyme. The Journal of Biological Chemistry, 274, 8363–8366. Cervoni, N., & Szyf, M. (2001). Demethylase activity is directed by histone acetylation. The Journal of Biological Chemistry, 276, 40778–40787. Champagne, F., Diorio, J., Sharma, S., & Meaney, M. J. (2001). Naturally occurring variations in maternal behavior in the rat are associated with differences in estrogen-inducible central oxytocin receptors. Proceedings of the National Academy of Sciences of the United States of America, 98, 12736–12741. Champagne, F. A., Chretien, P., Stevenson, C. W., Zhang, T. Y., Gratton, A., & Meaney, M. J. (2004). Variations in nucleus accumbens dopamine associated with individual differences in maternal behavior in the rat. The Journal of Neuroscience, 24, 4113–4123. Champagne, F. A., & Meaney, M. J. (2006). Stress during gestation alters postpartum maternal care and the development of the offspring in a rodent model. Biological Psychiatry, 59, 1227–1235. Champagne, F. A., & Meaney, M. J. (2007). Transgenerational effects of social environment on variations in maternal care and behavioral response to novelty. Behavioral Neuroscience, 121, 1353–1363. Champagne, F. A., Weaver, I. C., Diorio, J., Dymov, S., Szyf, M., & Meaney, M. J. (2006). Maternal care associated with methylation of the estrogen receptor-alpha1b promoter and estrogen receptor-alpha expression in the medial preoptic area of female offspring. Endocrinology, 147, 2909–2915. Champagne, F. A., Weaver, I. C., Diorio, J., Sharma, S., & Meaney, M. J. (2003). Natural variations in maternal care are associated with estrogen receptor alpha expression and estrogen sensitivity in the medial preoptic area. Endocrinology, 144, 4720–4724. Chawla, S., Hardingham, G. E., Quinn, D. R., & Bading, H. (1998). CBP: A signal-regulated transcriptional coactivator controlled by nuclear calcium and CaM kinase IV. Science, 281, 1505–1509. Ching, T. T., Maunakea, A. K., Jun, P., Hong, C., Zardo, G., Pinkel, D., et al. (2005). Epigenome analyses using BAC microarrays identify evolutionary conservation of tissue-specific methylation of SHANK3. Nature Genetics, 37, 645–651. Chuang, J. C., & Jones, P. A. (2007). Epigenetics and microRNAs. Pediatric Research, 61, 24R–29R. Crane-Robinson, C., Myers, F. A., Hebbes, T. R., Clayton, A. L., & Thorne, A. W. (1999). Chromatin immunoprecipitation assays in acetylation mapping of higher eukaryotes. Methods Enzymol, 304, 533–547. D’Alessio, A. C., Weaver, I. C., & Szyf, M. (2007). Acetylation-induced transcription is required for active DNA demethylation in methylation-silenced genes. Molecular and Cellular Biology, 27, 7462–7474. David, S. S., & Williams, S. D. (1998). Chemistry of glycosylases and endonucleases involved in base-excision repair. Chemical Reviews, 98, 1221–1262. Davie, J. R., & Chadee, D. N. (1998). Regulation and regulatory parameters of histone modifications. Journal of Cellular Biochemistry Supplement, 30–31, 203–213. de Kloet, E. R., Voorhuis, D. A., Boschma, Y., & Elands, J. (1986). Estradiol modulates density of putative ‘oxytocin receptors’ in discrete rat brain regions. Neuroendocrinology, 44, 415–421. DeRijk, R., & Sternberg, E. M. (1997). Corticosteroid resistance and disease. Annali Medici, 29, 79–82.

106

I.C.G. Weaver

Detich, N., Bovenzi, V., & Szyf, M. (2003). Valproate induces replication-independent active DNA demethylation. The Journal of Biological Chemistry, 278, 27586–27592. Detich, N., Hamm, S., Just, G., Knox, J. D., & Szyf, M. (2003). The methyl donor S-adenosylmethionine inhibits active demethylation of DNA: A candidate novel mechanism for the pharmacological effects of S-Adenosylmethionine. The Journal of Biological Chemistry, 278, 20812–20820. Detich, N., Theberge, J., & Szyf, M. (2002). Promoter-specific activation and demethylation by MBD2/demethylase. The Journal of Biological Chemistry, 277, 35791–35794. Fahrbach, S. E., Morrell, J. I., & Pfaff, D. W. (1984). Oxytocin induction of short-latency maternal behavior in nulliparous, estrogen-primed female rats. Hormones and Behavior, 18, 267–286. Fairbanks, L. A. (1989). Early experience and cross-generational continuity of mother-infant contact in vervet monkeys. Developmental Psychobiology, 22, 669–681. Fazzari, M. J., & Greally, J. M. (2004). Epigenomics: Beyond CpG islands. Nature Reviews. Genetics, 5, 446–455. Feldman, R., Eidelman, A. I., Sirota, L., & Weller, A. (2002). Comparison of skin-to-skin (kangaroo) and traditional care: Parenting outcomes and preterm infant development. Pediatrics, 110, 16–26. Felitti, V. J., Anda, R. F., Nordenberg, D., Williamson, D. F., Spitz, A. M., Edwards, V., et  al. (1998). Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults. The Adverse Childhood Experiences (ACE) Study. American Journal of Preventive Medicine, 14, 245–258. Fleming, A. (1986). Psychobiology of rat maternal behavior: How and where hormones act to promote maternal behavior at parturition. Annals of the New York Academy of Sciences, 474, 234–251. Fleming, A. S., O’Day, D. H., & Kraemer, G. W. (1999). Neurobiology of mother-infant interactions: Experience and central nervous system plasticity across development and generations. Neuroscience and Biobehavioral Reviews, 23, 673–685. Forsdahl, A. (1977). Are poor living conditions in childhood and adolescence an important risk factor for arteriosclerotic heart disease? British Journal of Preventive & Social Medicine, 31, 91–95. Fraga, M. F., Ballestar, E., Paz, M. F., Ropero, S., Setien, F., Ballestar, M. L., et al. (2005). Epigenetic differences arise during the lifetime of monozygotic twins. Proceedings of the National Academy of Sciences of the United States of America, 102, 10604–10609. Francis, D., Diorio, J., Liu, D., & Meaney, M. J. (1999). Nongenomic transmission across generations of maternal behavior and stress responses in the rat. Science, 286, 1155–1158. Francis, D. D., & Meaney, M. J. (1999). Maternal care and the development of stress responses. Current Opinion in Neurobiology, 9, 128–134. Frasor, J., & Gibori, G. (2003). Prolactin regulation of estrogen receptor expression. Trends in Endocrinology and Metabolism, 14, 118–123. Gardiner-Garden, M., & Frommer, M. (1987). CpG islands in vertebrate genomes. Journal of Molecular Biology, 196, 261–282. Gluckman, P. D., & Hanson, M. A. (2004). Developmental origins of disease paradigm: A mechanistic and evolutionary perspective. Pediatric Research, 56, 311–317. Gluckman, P. D., Hanson, M. A., Cooper, C., & Thornburg, K. L. (2008). Effect of in utero and early-life conditions on adult health and disease. The New England Journal of Medicine, 359, 61–73. Goel, A., Mathupala, S. P., & Pedersen, P. L. (2003). Glucose metabolism in cancer. Evidence that demethylation events play a role in activating type II hexokinase gene expression. The Journal of Biological Chemistry, 278, 15333–15340. Grunstein, M. (1997). Histone acetylation in chromatin structure and transcription. Nature, 389, 349–352. Habener, J. F., Meyer, T. E., Yun, Y., Waeber, G., & Hoeffler, J. P. (1990). Characterization of a cAMP-regulated enhancer-binding protein. Metabolism, 39, 13–16. Hales, C. N., & Barker, D. J. (2001). The thrifty phenotype hypothesis. British Medical Bulletin, 60, 5–20.

Epigenetic Programming of Stress Responses and Trans-Generational Inheritance

107

Hales, C. N., Barker, D. J., Clark, P. M., Cox, L. J., Fall, C., Osmond, C., et al. (1991). Fetal and infant growth and impaired glucose tolerance at age 64. BMJ, 303, 1019–1022. Hamm, S., Just, G., Lacoste, N., Moitessier, N., Szyf, M., & Mamer, O. (2008). On the mechanism of demethylation of 5-methylcytosine in DNA. Bioorganic & Medicinal Chemistry Letters, 18, 1046–1049. Harlow, H. F., & Zimmermann, R. R. (1959). Affectional responses in the infant monkey; orphaned baby monkeys develop a strong and persistent attachment to inanimate surrogate mothers. Science, 130, 421–432. Hashimshony, T., Zhang, J., Keshet, I., Bustin, M., & Cedar, H. (2003). The role of DNA methylation in setting up chromatin structure during development. Nature Genetics, 34, 187–192. Heim, C., Newport, D. J., Heit, S., Graham, Y. P., Wilcox, M., Bonsall, R., et al. (2000). Pituitaryadrenal and autonomic responses to stress in women after sexual and physical abuse in childhood. JAMA, 284, 592–597. Hendrich, B., Guy, J., Ramsahoye, B., Wilson, V. A., & Bird, A. (2001). Closely related proteins MBD2 and MBD3 play distinctive but interacting roles in mouse development. Genes & Development, 15, 710–723. Hjalgrim, L. L., Westergaard, T., Rostgaard, K., Schmiegelow, K., Melbye, M., Hjalgrim, H., et al. (2003). Birth weight as a risk factor for childhood leukemia: A meta-analysis of 18 epidemiologic studies. American Journal of Epidemiology, 158, 724–735. Holliday, R., & Pugh, J. E. (1975). DNA modification mechanisms and gene activity during development. Science, 187, 226–232. Holsboer, F. (2000). The corticosteroid receptor hypothesis of depression. Neuropsychopharmacology, 23, 477–501. Hong, L., Schroth, G. P., Matthews, H. R., Yau, P., & Bradbury, E. M. (1993). Studies of the DNA binding properties of histone H4 amino terminus. Thermal denaturation studies reveal that acetylation markedly reduces the binding constant of the H4 “tail” to DNA. The Journal of Biological Chemistry, 268, 305–314. Hyman, S. E., Comb, M., Lin, Y. S., Pearlberg, J., Green, M. R., & Goodman, H. M. (1988). A common trans-acting factor is involved in transcriptional regulation of neurotransmitter genes by cyclic AMP. Molecular and Cellular Biology, 8, 4225–4233. Invitti, C., Redaelli, G., Baldi, G., & Cavagnini, F. (1999). Glucocorticoid receptors in anorexia nervosa and Cushing’s disease. Biological Psychiatry, 45, 1467–1471. Jaenisch, R., & Bird, A. (2003). Epigenetic regulation of gene expression: How the genome integrates intrinsic and environmental signals. Nature Genetics, 33(suppl), 245–254. Jenuwein, T. (2001). Re-SET-ting heterochromatin by histone methyltransferases. Trends in Cell Biology, 11, 266–273. Jenuwein, T., & Allis, C. D. (2001). Translating the histone code. Science, 293, 1074–1080. Johnson, A. E., Ball, G. F., Coirini, H., Harbaugh, C. R., McEwen, B. S., & Insel, T. R. (1989). Time course of the estradiol-dependent induction of oxytocin receptor binding in the ventromedial hypothalamic nucleus of the rat. Endocrinology, 125, 1414–1419. Jones, P. L., Veenstra, G. J., Wade, P. A., Vermaak, D., Kass, S. U., Landsberger, N., et al. (1998). Methylated DNA and MeCP2 recruit histone deacetylase to repress transcription. Nature Genetics, 19, 187–191. Jutapakdeegul, N., Casalotti, S. O., Govitrapong, P., & Kotchabhakdi, N. (2003). Postnatal touch stimulation acutely alters corticosterone levels and glucocorticoid receptor gene expression in the neonatal rat. Developmental Neuroscience, 25, 26–33. Kadar, T., Silbermann, M., Brandeis, R., & Levy, A. (1990). Age-related structural changes in the rat hippocampus: Correlation with working memory deficiency. Brain Research, 512, 113–120. Kadonaga, J. T. (1998). Eukaryotic transcription: An interlaced network of transcription factors and chromatin-modifying machines. Cell, 92, 307–313. Kangaspeska, S., Stride, B., Metivier, R., Polycarpou-Schwarz, M., Ibberson, D., Carmouche, R. P., et al. (2008). Transient cyclical methylation of promoter DNA. Nature, 452, 112–115.

108

I.C.G. Weaver

Kato, T., Iwamoto, K., Kakiuchi, C., Kuratomi, G., & Okazaki, Y. (2005). Genetic or epigenetic difference causing discordance between monozygotic twins as a clue to molecular basis of mental disorders. Molecular Psychiatry, 10, 622–630. Khulan, B., Thompson, R. F., Ye, K., Fazzari, M. J., Suzuki, M., Stasiek, E., et al. (2006). Comparative isoschizomer profiling of cytosine methylation: The HELP assay. Genome Research, 16, 1046–1055. Kornberg, R. D. (1974). Chromatin structure: A repeating unit of histones and DNA. Science, 184, 868–871. Korzus, E., Rosenfeld, M. G., & Mayford, M. (2004). CBP histone acetyltransferase activity is a critical component of memory consolidation. Neuron, 42, 961–972. Krinke, G. J. (2000). The laboratory rat. San Diego: Academic. Kuo, M. H., & Allis, C. D. (1998). Roles of histone acetyltransferases and deacetylases in gene regulation. Bioessays, 20, 615–626. Law, C. M., & Shiell, A. W. (1996). Is blood pressure inversely related to birth weight? The strength of evidence from a systematic review of the literature. Journal of Hypertension, 14, 935–941. Leonhardt, H., & Bestor, T. H. (1993). Structure, function and regulation of mammalian DNA methyltransferase. EXS, 64, 109–119. Li, E. (2002). Chromatin modification and epigenetic reprogramming in mammalian development. Nature Reviews. Genetics, 3, 662–673. Lissau, I., & Sorensen, T. I. (1994). Parental neglect during childhood and increased risk of obesity in young adulthood. Lancet, 343, 324–327. Liu, D., Diorio, J., Tannenbaum, B., Caldji, C., Francis, D., Freedman, A., et al. (1997). Maternal care, hippocampal glucocorticoid receptors, and hypothalamic-pituitary-adrenal responses to stress [see comments]. Science, 277, 1659–1662. Lucas, A. (1991). Programming by early nutrition in man. Ciba Found Symp, 156, 38–50. discussion 50–55. McCauley, J., Kern, D. E., Kolodner, K., Dill, L., Schroeder, A. F., DeChant, H. K., et al. (1997). Clinical characteristics of women with a history of childhood abuse: Unhealed wounds. JAMA, 277, 1362–1368. McCormick, J. A., Lyons, V., Jacobson, M. D., Noble, J., Diorio, J., Nyirenda, M., et al. (2000). 5¢-Heterogeneity of glucocorticoid receptor messenger RNA is tissue specific: Differential regulation of variant transcripts by early-life events. Molecular Endocrinology, 14, 506–517. McGowan, P. O., Sasaki, A., D’Alessio, A. C., Dymov, S., Labonte, B., Szyf, M., et al. (2009). Epigenetic regulation of the glucocorticoid receptor in human brain associates with childhood abuse. Nature Neuroscience, 12, 342–348. McGowan, P. O., Sasaki, A., Huang, T. C., Unterberger, A., Suderman, M., Ernst, C., et al. (2008). Promoter-wide hypermethylation of the ribosomal RNA gene promoter in the suicide brain. PLoS One, 3, e2085. Meaney, M. J. (2001). Maternal care, gene expression, and the transmission of individual differences in stress reactivity across generations. Annual Review of Neuroscience, 24, 1161–1192. Meaney, M. J., Aitken, D. H., & Sapolsky, R. M. (1987). Thyroid hormones influence the development of hippocampal glucocorticoid receptors in the rat: A mechanism for the effects of postnatal handling on the development of the adrenocortical stress response. Neuroendocrinology, 45, 278–283. Meaney, M. J., Aitken, D. H., Viau, V., Sharma, S., & Sarrieau, A. (1989). Neonatal handling alters adrenocortical negative feedback sensitivity and hippocampal type II glucocorticoid receptor binding in the rat. Neuroendocrinology, 50, 597–604. Meaney, M. J., Diorio, J., Francis, D., Weaver, S., Yau, J., Chapman, K., et al. (2000). Postnatal handling increases the expression of cAMP-inducible transcription factors in the rat hippocampus: The effects of thyroid hormones and serotonin. The Journal of Neuroscience, 20, 3926–3935. Metivier, R., Gallais, R., Tiffoche, C., Le Peron, C., Jurkowska, R. Z., Carmouche, R. P., et  al. (2008). Cyclical DNA methylation of a transcriptionally active promoter. Nature, 452, 45–50.

Epigenetic Programming of Stress Responses and Trans-Generational Inheritance

109

Michels, K. B., Trichopoulos, D., Robins, J. M., Rosner, B. A., Manson, J. E., Hunter, D. J., et al. (1996). Birthweight as a risk factor for breast cancer. Lancet, 348, 1542–1546. Michels, K. B., & Xue, F. (2006). Role of birthweight in the etiology of breast cancer. International Journal of Cancer, 119, 2007–2025. Milbrandt, J. (1987). A nerve growth factor-induced gene encodes a possible transcriptional regulatory factor. Science, 238, 797–799. Milbrandt, J. (1988). Nerve growth factor induces a gene homologous to the glucocorticoid receptor gene. Neuron, 1, 183–188. Mill, J., & Petronis, A. (2007). Molecular studies of major depressive disorder: The epigenetic perspective. Molecular Psychiatry, 12, 799–814. Miller, L., Kramer, R., Warner, V., Wickramaratne, P., & Weissman, M. (1997). Intergenerational transmission of parental bonding among women. Journal of the American Academy of Child and Adolescent Psychiatry, 36, 1134–1139. Mitchell, J. B., Rowe, W., Boksa, P., & Meaney, M. J. (1990). Serotonin regulates type II corticosteroid receptor binding in hippocampal cell cultures. The Journal of Neuroscience, 10, 1745–1752. Montminy, M. R., Sevarino, K. A., Wagner, J. A., Mandel, G., & Goodman, R. H. (1986). Identification of a cyclic-AMP-responsive element within the rat somatostatin gene. Proceedings of the National Academy of Sciences of the United States of America, 83, 6682–6686. Morgan, H. D., Santos, F., Green, K., Dean, W., & Reik, W. (2005). Epigenetic reprogramming in mammals. Human Molecular Genetics, 14 Spec No 1, R47–R58. Morgan, H. D., Sutherland, H. G., Martin, D. I., & Whitelaw, E. (1999). Epigenetic inheritance at the agouti locus in the mouse. Nature Genetics, 23, 314–318. Mudd, S. H., & Cantoni, G. L. (1958). Activation of methionine for transmethylation. III. The methionine-activating enzyme of Bakers’ yeast. The Journal of Biological Chemistry, 231, 481–492. Mueller, B. R., & Bale, T. L. (2008). Sex-specific programming of offspring emotionality after stress early in pregnancy. The Journal of Neuroscience, 28, 9055–9065. Myers, M. M., Brunelli, S. A., Shair, H. N., Squire, J. M., & Hofer, M. A. (1989). Relationships between maternal behavior of SHR and WKY dams and adult blood pressures of crossfostered F1 pups. Developmental Psychobiology, 22, 55–67. Nan, X., Ng, H. H., Johnson, C. A., Laherty, C. D., Turner, B. M., Eisenman, R. N., et al. (1998). Transcriptional repression by the methyl-CpG-binding protein MeCP2 involves a histone deacetylase complex. Nature, 393, 386–389. Nemeroff, C. B. (1996). The corticotropin-releasing factor (CRF) hypothesis of depression: New findings and new directions. Molecular Psychiatry, 1, 336–342. Ng, H. H., Zhang, Y., Hendrich, B., Johnson, C. A., Turner, B. M., Erdjument-Bromage, H., et al. (1999). MBD2 is a transcriptional repressor belonging to the MeCP1 histone deacetylase complex. Nature Genetics, 23, 58–61. Numan, M., & Callahan, E. C. (1980). The connections of the medial preoptic region and maternal behavior in the rat. Physiology & Behavior, 25, 653–665. Numan, M., & Numan, M. J. (1994). Expression of Fos-like immunoreactivity in the preoptic area of maternally behaving virgin and postpartum rats. Behavioral Neuroscience, 108, 379–394. Oberlander, T. F., Weinberg, J., Papsdorf, M., Grunau, R., Misri, S., & Devlin, A. M. (2008). Prenatal exposure to maternal depression, neonatal methylation of human glucocorticoid receptor gene (NR3C1) and infant cortisol stress responses. Epigenetics, 3, 97–106. Okano, M., Bell, D. W., Haber, D. A., & Li, E. (1999). DNA methyltransferases Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian development. Cell, 99, 247–257. Ooi, S. K., & Bestor, T. H. (2008). The colorful history of active DNA demethylation. Cell, 133, 1145–1148. Oswald, J., Engemann, S., Lane, N., Mayer, W., Olek, A., Fundele, R., et al. (2000). Active demethylation of the paternal genome in the mouse zygote. Current Biology, 10, 475–478. Pakneshan, P., Tetu, B., & Rabbani, S. A. (2004). Demethylation of urokinase promoter as a prognostic marker in patients with breast carcinoma. Clinical Cancer Research, 10, 3035–3041. Pascale, R. M., Simile, M. M., Satta, G., Seddaiu, M. A., Daino, L., Pinna, G., et  al. (1991). Comparative effects of L-methionine, S-adenosyl-L-methionine and 5¢-methylthioadenosine

110

I.C.G. Weaver

on the growth of preneoplastic lesions and DNA methylation in rat liver during the early stages of hepatocarcinogenesis. Anticancer Research, 11, 1617–1624. Pedersen, C. A., & Prange, A. J., Jr. (1979). Induction of maternal behavior in virgin rats after intracerebroventricular administration of oxytocin. Proceedings of the National Academy of Sciences of the United States of America, 76, 6661–6665. Poulter, M. O., Du, L., Weaver, I. C., Palkovits, M., Faludi, G., Merali, Z., et al. (2008). GABA(A) receptor promoter hypermethylation in suicide brain: Implications for the involvement of epigenetic processes. Biological Psychiatry, 64, 645–652. Ramchandani, S., Bhattacharya, S. K., Cervoni, N., & Szyf, M. (1999). DNA methylation is a reversible biological signal. Proceedings of the National Academy of Sciences of the United States of America, 96, 6107–6112. Razin, A. (1998). CpG methylation, chromatin structure and gene silencing-a three-way connection. The EMBO Journal, 17, 4905–4908. Razin, A., & Cedar, H. (1977). Distribution of 5-methylcytosine in chromatin. Proceedings of the National Academy of Sciences of the United States of America, 74, 2725–2728. Razin, A., & Szyf, M. (1984). DNA methylation patterns. Formation and function. Biochimical et Biophysica Acta, 782, 331–342. Repetti, R. L., Taylor, S. E., & Seeman, T. E. (2002). Risky families: Family social environments and the mental and physical health of offspring. Psychological Bulletin, 128, 330–366. Rich-Edwards, J. W., Stampfer, M. J., Manson, J. E., Rosner, B., Hankinson, S. E., Colditz, G. A., et al. (1997). Birth weight and risk of cardiovascular disease in a cohort of women followed up since 1976. BMJ, 315, 396–400. Richards, E. J. (2006). Inherited epigenetic variation–revisiting soft inheritance. Nature Reviews. Genetics, 7, 395–401. Rose, G. (1964). Familial Patterns in Ischaemic Heart Disease. British Journal of Preventive & Social Medicine, 18, 75–80. Rossiter, M. C. (1999). Maternal effects as adaptations. Oxford University Press: London. Saito, Y., & Jones, P. A. (2006). Epigenetic activation of tumor suppressor microRNAs in human cancer cells. Cell Cycle, 5, 2220–2222. Sandbrook, D. (2005). Never had it so good: A history of Britain from Suez to the “Beatles” – Britain in the sixties (1st ed.). Boston: Little Brown. Schibler, U., & Sierra, F. (1987). Alternative promoters in developmental gene expression. Annual Review of Genetics, 21, 237–257. Sealy, L., & Chalkley, R. (1978). DNA associated with hyperacetylated histone is preferentially digested by DNase I. Nucleic Acids Research, 5, 1863–1876. Seckl, J. R., & Meaney, M. J. (1993). Early life events and later development of ischaemic heart disease. Lancet, 342, 1236. Self, D. W., & Nestler, E. J. (1995). Molecular mechanisms of drug reinforcement and addiction. Annual Review of Neuroscience, 18, 463–495. Sheng, M., & Greenberg, M. E. (1990). The regulation and function of c-fos and other immediate early genes in the nervous system. Neuron, 4, 477–485. Shiio, Y., & Eisenman, R. N. (2003). Histone sumoylation is associated with transcriptional repression. Proceedings of the National Academy of Sciences of the United States of America, 100, 13225–13230. Shilatifard, A. (2006). Chromatin modifications by methylation and ubiquitination: Implications in the regulation of gene expression. Annual Review of Biochemistry, 75, 243–269. Shukeir, N., Pakneshan, P., Chen, G., Szyf, M., & Rabbani, S. A. (2006). Alteration of the methylation status of tumor-promoting genes decreases prostate cancer cell invasiveness and tumorigenesis in vitro and in vivo. Cancer Research, 66, 9202–9210. Smith, S. S. (1994). Biological implications of the mechanism of action of human DNA (cytosine-5) methyltransferase. Progress in Nucleic Acid Research and Molecular Biology, 49, 65–111. Smythe, J. W., Rowe, W. B., & Meaney, M. J. (1994). Neonatal handling alters serotonin (5-HT) turnover and 5-HT2 receptor binding in selected brain regions: Relationship to the handling

Epigenetic Programming of Stress Responses and Trans-Generational Inheritance

111

effect on glucocorticoid receptor expression. Brain Research. Developmental Brain Research, 80, 183–189. Sroufe, L. A. (1997). Psychopathology as an outcome of development. Development and Psychopathology, 9, 251–268. Stack, E. C., Balakrishnan, R., Numan, M. J., & Numan, M. (2002). A functional neuroanatomical investigation of the role of the medial preoptic area in neural circuits regulating maternal behavior. Behavioural Brain Research, 131, 17–36. Stern, J. M. (1997). Offspring-induced nurturance: Animal-human parallels. Developmental Psychobiology, 31, 19–37. Szyf, M. (2001). Towards a pharmacology of DNA methylation. Trends in Pharmacological Sciences, 22, 350–354. Szyf, M., & Bhattacharya, S. K. (2002a). Extracting DNA demethylase activity from mammalian cells. Methods Mol Biol, 200, 163–176. Szyf, M., & Bhattacharya, S. K. (2002b). Measuring DNA demethylase activity in vitro. Methods Mol Biol, 200, 155–161. Tate, P. H., & Bird, A. P. (1993). Effects of DNA methylation on DNA-binding proteins and gene expression. Current Opinion in Genetics & Development, 3, 226–231. Tremolizzo, L., Carboni, G., Ruzicka, W. B., Mitchell, C. P., Sugaya, I., Tueting, P., et al. (2002). An epigenetic mouse model for molecular and behavioral neuropathologies related to schizophrenia vulnerability. Proceedings of the National Academy of Sciences of the United States of America, 99, 17095–17100. Tsankova, N., Renthal, W., Kumar, A., & Nestler, E. J. (2007). Epigenetic regulation in psychiatric disorders. Nature Reviews. Neuroscience, 8, 355–367. Tsukada, T., Fink, J. S., Mandel, G., & Goodman, R. H. (1987). Identification of a region in the human vasoactive intestinal polypeptide gene responsible for regulation by cyclic AMP. The Journal of Biological Chemistry, 262, 8743–8747. Turner, J. D., & Muller, C. P. (2005). Structure of the glucocorticoid receptor (NR3C1) gene 5¢ untranslated region: Identification, and tissue distribution of multiple new human exon 1. Journal of Molecular Endocrinology, 35, 283–292. Turner, J. D., Pelascini, L. P., Macedo, J. A., & Muller, C. P. (2008). Highly individual methylation patterns of alternative glucocorticoid receptor promoters suggest individualized epigenetic regulatory mechanisms. Nucleic Acids Research, 36, 7207–7218. Waddington, C. H. (1968). Towards a theoretical biology. Nature, 218, 525–527. Wade, P. A., Gegonne, A., Jones, P. L., Ballestar, E., Aubry, F., & Wolffe, A. P. (1999). Mi-2 complex couples DNA methylation to chromatin remodelling and histone deacetylation. Nature Genetics, 23, 62–66. Wade, P. A., Pruss, D., & Wolffe, A. P. (1997). Histone acetylation: Chromatin in action. Trends in Biochemical Sciences, 22, 128–132. Watt, F., & Molloy, P. L. (1988). Cytosine methylation prevents binding to DNA of a HeLa cell transcription factor required for optimal expression of the adenovirus major late promoter. Genes & Development, 2, 1136–1143. Weaver, I. C., Cervoni, N., Champagne, F. A., D’Alessio, A. C., Sharma, S., Seckl, J. R., et al. (2004). Epigenetic programming by maternal behavior. Nature Neuroscience, 7, 847–854. Weaver, I. C., Champagne, F. A., Brown, S. E., Dymov, S., Sharma, S., Meaney, M. J., et  al. (2005). Reversal of maternal programming of stress responses in adult offspring through methyl supplementation: Altering epigenetic marking later in life. The Journal of Neuroscience, 25, 11045–11054. Weaver, I. C., D’Alessio, A. C., Brown, S. E., Hellstrom, I. C., Dymov, S., Sharma, S., et al. (2007). The transcription factor nerve growth factor-inducible protein a mediates epigenetic programming: Altering epigenetic marks by immediate-early genes. The Journal of Neuroscience, 27, 1756–1768. Weaver, I. C., Meaney, M. J., & Szyf, M. (2006). Maternal care effects on the hippocampal transcriptome and anxiety-mediated behaviors in the offspring that are reversible in adulthood.

112

I.C.G. Weaver

Proceedings of the National Academy of Sciences of the United States of America, 103, 3480–3485. Weaver, I. C. G. (2009). Life at the interface between a dynamic environment and a fixed genome: epigenetic programming of stress responses by maternal behavior. In D. Janigro (Ed) Mammalian brain development, contemporary neuroscience (pp. 17–39). Totowa, NJ: Humana Press. Weiss, A., Keshet, I., Razin, A., & Cedar, H. (1996). DNA demethylation in vitro: Involvement of RNA. Cell, 86, 709–718. Wondrak, G. T., Cervantes-Laurean, D., Jacobson, E. L., & Jacobson, M. K. (2000). Histone carbonylation in vivo and in vitro. Biochemical Journal, 351 Pt 3, 769–777. Yoshida, M., Kijima, M., Akita, M., & Beppu, T. (1990). Potent and specific inhibition of mammalian histone deacetylase both in  vivo and in  vitro by trichostatin A. The Journal of Biological Chemistry, 265, 17174–17179. Young, L. J., Wang, Z., Donaldson, R., & Rissman, E. F. (1998). Estrogen receptor alpha is essential for induction of oxytocin receptor by estrogen. NeuroReport, 9, 933–936. Yu, J., de Belle, I., Liang, H., & Adamson, E. D. (2004). Coactivating factors p300 and CBP are transcriptionally crossregulated by Egr1 in prostate cells, leading to divergent responses. Molecular Cell, 15, 83–94. Zhang, Y., Ng, H. H., Erdjument-Bromage, H., Tempst, P., Bird, A., & Reinberg, D. (1999). Analysis of the NuRD subunits reveals a histone deacetylase core complex and a connection with DNA methylation. Genes & Development, 13, 1924–1935.

Prenatal Viral Infection in Mouse: An Animal Model of Schizophrenia S. Hossein Fatemi and Timothy D. Folsom

Abstract  Schizophrenia is a major debilitating disease with a lifetime prevalence of 1% throughout the world. There is robust epidemiologic evidence indicating that environmental contributions, such as prenatal infections, may lead to the genesis of schizophrenia. Our laboratory has developed an animal model using human influenza virus to infect pregnant Balb/c and C57BL/6 mice intranasally at selected time points during pregnancy to investigate the role of prenatal viral infection on brain development. In this chapter, we review our research using this model and the changes in brain structure, gene expression, neurochemistry, and behavior that are observed in the offspring of infected dams. Our observations are consistent with findings observed in subjects with schizophrenia, providing additional evidence for the role of prenatal viral infection in the etiology of this disease. Keywords  Schizophrenia • Prenatal viral infection • Brain • Mouse • Microarray

1 An Introduction to Prenatal Viral Infection and Schizophrenia The potential role of prenatal viral infection as a cause of schizophrenia dates back to Menninger (1928), who described 67 cases of schizophrenia in a large cohort of patients who contracted influenza during the pandemic of 1919 (caused by the same S.H. Fatemi (*) Department of Psychiatry, Division of Neuroscience Research, University of Minnesota, Medical School, MMC 392, 420 Delaware Street S.E, Minneapolis, MN 55455, USA and Department of Pharmacology, University of Minnesota Medical School, Minneapolis, MN 55455, USA and Department of Neuroscience, University of Minnesota Medical School, Minneapolis, MN 55455, USA e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_4, © Springer Science+Business Media, LLC 2011

113

114

S.H. Fatemi and T.D. Folsom

strain of virus used in our animal model). Later, an excess of schizophrenic patients were found to be born during late winter and spring, suggesting that influenza infections may be responsible for these cases (Hare, Price, & Slater, 1972; Machon, Mednick, & Schulsinger, 1983). Moreover, there is a 5–15% of the excess schizophrenic births in the Northern Hemisphere which occur during the months of January and March (Boyd, Pulver, & Stewart, 1986; Pallast, Jongbloet, & Straatman, 1994; Susser, Brown, & Gorman, 1999). This excess winter birth is not a methodological artifact, nor is it due to unusual patterns of conception (Pulver, Liang, & Wolyniec, 1992; Susser et  al., 1999). Additional studies have shown increased risk of schizophrenia in individuals whose mothers were exposed to influenza during pregnancy (positive studies: Adams, Kendell, & Hare, 1993; Barr, Mednick, & Munk-Jorgensen, 1990; Fahy, Jones, & Sham, 1993; Kunugi, Nanko, & Takei, 1995; McGrath, Pemberton, & Welham, 1994; O’Callaghan, Gibson, & Colohan, 1991; Sham, O’Callaghan, & Takei, 1992; Takei et  al., 1994; Takei, Sham, & O’Callaghan, 1994). Other studies, however, have found equivocal or no effect (Adams et  al., 1993; Erlenmeyer-Kimling et  al., 1994; Kendell & Kemp, 1989; Selten & Sleats, 1994; Susser, Lin, & Brown, 1994; Takei, Van Os, & Murray, 1995; Torrey, Bowler, Taylor, & Gottesman, 1994). The fourth through seventh months of gestation has been identified to be a period during which the risk of developing schizophrenia is especially high (Brown et al., 2004; Susser et al., 1997). Additional cohort studies have shown increased risk for schizophrenia following prenatal exposure to influenza (Mednick, Huttunen, & Macon, 1994; Stober, Franzek, & Beckmann, 1992; Wright, Rakei, Rifkin, & Murray, 1995). A review by Brown (2006) summarized that there was a 10- to 20-fold risk of developing schizophrenia following prenatal exposure to rubella; a sevenfold risk of developing schizophrenia following prenatal exposure to influenza in the first trimester and a threefold increased risk following infection in early to mid-gestation; presence of maternal antibodies against Toxoplasma gondii lead to a 2.5-fold increased risk (Brown, 2006). Nucleotide sequences homologous to retroviral polymerase genes have been identified in the cerebrospinal fluid (CSF) of subjects with schizophrenia while no such sequences were found in individuals with noninflammatory neurological illnesses or in normal subjects (Karlsson et al. 2001; Lewis, 2001). These studies suggest that the development of schizophrenia involves the interaction of genetic and environmental risks on brain development (Lewis). Moreover, identification of potential environmental risk factors, such as influenza virus or retroviruses such as endogenous retroviral-9 family and the human endogenous retrovirus-W family (HERV-W, grouped on the basis of a tryptophan (W) tRNA motif identified in HERV-W sequences; Blond et al., 1999) observed by Karlsson et al. (2001), will help in targeting early interventions at repressing the expression of these transcripts. An alternate approach would be to vaccinate against influenza, thus influencing the course and outcome of schizophrenia in the susceptible individuals (Lewis, 2001). Several groups, including our laboratory, have shown evidence for viral infections and/or immune challenges being responsible for production of abnormal

Prenatal Viral Infection in Mouse: An Animal Model of Schizophrenia

115

brain structure and function in rodents where mothers were exposed to viral insults throughout pregnancy (Fatemi, Pearce, Brooks, & Sidwell, 2005; Meyer et  al., 2006). Application of the viral mimic polyribocytidilic acid (PolyI:C) at embryonic day 9 (E9) (which corresponds to late first trimester of pregnancy in humans) and E17 (which corresponds to late second trimester of pregnancy in humans), resulted in distinct behavioral deficits, neuropathological differences and acute cytokine responses (Meyer et  al.). Date of exposure resulted in differences of behavioral deficits with those exposed on E9 showing impaired exploratory behavior while those exposed on E17 displayed perseverative behavior (Meyer et al.). Moreover, time of exposure had differing effects on Reelin expression with a greater reduction following exposure on E9 (Meyer et al.). In contrast, following PolyI:C exposure at E17, there was an upregulation of caspase-3 in the dorsal dentate gyrus (Meyer et al., 2006) signifying increased apoptosis (Rami et al., 2003). Finally, exposure at E17 resulted in increased IL-10 and TNF-a in fetal brain (Meyer et al.). Taken together, these results provide evidence that the time of prenatal insult results in important differences that are persistent through adulthood.

2 Mouse Brain Development Mouse brain development progresses through four prenatal stages that exhibit various susceptibilities to prenatal insults. These stages consist of: (1) cleavage and blastulation, equal to E0–E5; (2) implantation, gastrulation, and early organogenesis, equal to E5–E10; (3) organogenesis, equal to E10–E14; and (4) fetal growth and development, equal to E14–E19 or 20 (Hogan, Beddington, Constantini, & Lacy, 1994). These stages in mouse brain development can be roughly compared to human brain development at the first (E0–E10) and second trimesters (E10–E19/20), respectively. Mouse postnatal brain development (P1–P10) corresponds roughly to the third trimester of brain growth in man (Susser et al., 1999). Recent evidence points to several prenatal sensitive periods in each brain area of the mouse, when neurogenesis, gliogenesis, or neuronal migration is at its peak (Morgane, Austin-LaFrance, Bronzino, Tonkiss, & Galler, 1992). Thus, brain areas, such as the cerebral cortex, hippocampus, and cerebellum, each follow specific timetables for brain growth (Acuff-Smith & Vorhees, 1999; Avishai-Eliner, Brunson, Sandman, & Baram, 2002; Boksa & Luheshi, 2003; Kaufmann, 2000; Morgane et  al., 1992; Romijn, Hofman, & Gramsbergen, 1991; Susser et  al., 1999). For example, critical periods of neurogenesis for various brain areas in mouse consist of E9 (cerebellar Purkinje cells), E11–15.5 (hippocampal pyramidal cells), E11.5–16 (cerebral cortex), and P2–13 (cerebellar granule cells) (Rodier, 1980). There are additional critical periods for neurogenesis that affect specific brain cell populations. For example, neural crest-derived cells migrating to craniofacial structures are most susceptible to various insults during E8.5–10.5 in the mouse (Yamagishi, Garg, Matsuoka, Thomas, & Srivastava, 1999). Accordingly,

116

S.H. Fatemi and T.D. Folsom

retinoic acid causes deleterious effects on development of brain and craniofacial structures during the E8–10 period (Holson, Adams, & Ferguson, 1999). Previous studies show an increased risk of growth retardation and other morbidity/mortality in mice that were prenatally exposed an H1N1 strain of influenza virus in the E8–10 period (Molanova & Blaskovic, 1975). We have also shown that administration of the same strain of virus on E7, E9, E16, and E18 of pregnancy results in multiple brain abnormalities in postnatal mice (Fatemi et  al., 1999; Fatemi, Folsom, Reutiman, & Sidwell, 2008; Fatemi, Pearce, et  al., 2005; Fatemi, Reutiman, Folsom, Huang, et  al., 2008; Fatemi, Reutiman, Folsom, & Sidwell, 2008; Fatemi, Sidwell, Akhter, et al., 1998; Fatemi, Sidwell, Kist, et al., 1998; Su et  al., 2008; Winter et  al., 2008). Indeed, the onset of teratogenic susceptibility begins at relatively similar temporal windows in mouse and man (E5 and E11–12, respectively) (Wilson, 1964). However, exact extrapolation from mouse to human is not possible based on specific days of pregnancy, but can be construed qualitatively from prenatal and postnatal peaks of neurogenesis, gliogenesis, and/or neuron migration/differentiation (Morgane et al., 1992). The accumulation of various immunologic, epidemiologic and case study data indicate that prenatal maternal exposure to human influenza infection around late first to mid-second trimester can increase the risk of births that lead to schizophrenia (Brown et  al., 2004; Susser et al., 1999).

3 Impact of Prenatal Viral Infection on Brain Development 3.1 Brain Structural Abnormalities Following Prenatal Viral Infection Numerous reports have documented the presence of various neuropathologic findings in postmortem brains of patients with schizophrenia (Boyer, Phillips, Rousseau, & Ilivitsky, 2007; Crespo-Facorro, Barbadillo, Pelayo-Terán, & Rodríguez-Sánchez, 2007). These findings include: cortical atrophy; ventricular enlargement; reduced volumes of hippocampus, amygdala and parahippocampal gyrus; disturbed cytoarchitecture in hippocampus; and reduced cell size in Purkinje cells of the cerebellum (Andreasen, 1999; Arnold & Trojanowski, 1996). Prenatal viral infection of Balb/c and C57BL/6 mice at E9 (late first trimester), E16 (middle second trimester) and E18 (late second trimester) lead to deleterious effects on brain morphology including changes in volume, fractional anisotropy, and pyramidal cell density consistent with these observed changes in subjects with schizophrenia (Fatemi, Earle, et al., 2002; Fatemi et al., 1999; Fatemi, Reutiman, Folsom, Huang, et al., 2008). Following infection of Balb/c mice at E9 with influenza, brains from infected P0 offspring displayed decreases in neocortical [39% reduction in layer I (p <0.0001); 27.2% reduction in layers II–VI and intermediate zone (p <0.00015)]

Prenatal Viral Infection in Mouse: An Animal Model of Schizophrenia

117

and hippocampal (18.1% reduction, non-significant) thickness when compared with sham-infected controls (Fatemi et al., 1999). Brains of infected offspring at P0 also displayed increased pyramidal cell density (170%, p <0.0038) and significantly reduced pyramidal cell nuclear size (29%, p <0.047) (Fatemi, Earle, et al., 2002). By adulthood (P98), there continued to be an increase in pyramidal cell density (130%, p <0.038) and nonpyramidal cell density (146%, p <0.026), and a significant ­reduction in pyramidal cell nuclear size (37%, p <0.025) (Fatemi, Earle, et al.). Recently, morphometric analysis of brain following infection of C57BL/6 mice at E16 revealed changes in brain volume and fractional anisotropy (FA). The areas for hippocampus at P35 and cerebellum at P14 were both reduced (p <0.014 and p <0.0298, respectively). Overall brain volume was reduced at P14 (p <0.0185) and ventricle volume was reduced at P0 (p <0.025) (Table 1; Fatemi et al., 2009a, b). Measurement of FA showed decreased FA of the internal capsule (IC; right) at P0 (0.51 ± 0.022 for control vs. 0.44 ± 0.044 for infected, p <0.033); increased FA in the corpus callosum at P14 (0.61 ± 0.036 for control vs. 0.68 ± 0.017, p <0.024); and increased FA of the middle cerebellar peduncle (MCP; right) at P56 (0.71 ± 0.031 for control vs. 0.78 ± 0.016 for infected, p <0.006) (Fatemi et  al., 2009a, b).

Table 1  Changes in brain and ventricular areas following prenatal infection at E16 PD Brain area Control Infected Change p-Value P0 Cerebellum1 3.47 ± 0.50 3.06 ± 0.44 – 0.267 Hippocampus2 2.29 ± 0.37 2.27 ± 0.12 – 0.916 Total brain2 88.03 ± 4.03 89.67 ± 2.87 – 0.532 Ventricle1 0.03 ± 0.01 0.010 ± 0.006 ↓66% 0.0253 P14 Cerebellum1 46.06 ± 2.06 41.18 ± 2.23 ↓11% 0.0298 Hippocampus2 15.24 ± 0.60 14.26 ± 0.77 – 0.116 Total brain2 359.94 ± 11.28 332.92 ± 8.55 ↓7.5 0.0185 Ventricle1 0.077 ± 0.067 0.030 ± 0.014 – 0.293 P35 Cerebellum1 56.09 ± 2.43 56.39 ± 2.31 – 0.862 Hippocampus2 19.43 ± 0.60 18.24 ± 0.34 ↓6% 0.0141 Total brain2 416.56 ± 18.28 408.14 ± 16.67 – 0.521 Ventricle1 0.34 ± 0.19 0.17 ± 0.13 – 0.2 Cerebellum1 56.26 ± 2.24 54.51 ± 3.20 – 0.403 P56 Hippocampus2 18.81 ± 0.81 18.52 ± 0.32 – 0.534 Total brain2 436.55 ± 14.82 423.44 ± 15.95 – 0.274 0.17 ± 0.11 0.95 ± 1.63 – 0.379 Ventricle1 1 Data taken from: Reprinted from Fatemi, S. H., Folsom, T. D., Reutiman, T. J., Abu-Odeh, D., Mori, S., Huang, H., et al. (2009). Abnormal expression of myelination genes and alterations in white matter fractional anisotropy following prenatal viral influenza infection at E16 in mice. Schizophr Res, 112(1–3), 49, Table 4, Copyright (2009), with permission from Elsevier; 2Reprinted from Fatemi, S. H., Folsom, T. D., Reutiman, T. J., Huang, H., Oishi, K., & Mori, S. (2009). Prenatal viral infection of mice at E16 causes changes in gene expression in hippocampi of the offspring. Eur Neuropsychopharmacol, 19(9), 651, Table  3, Copyright (2009), with permission from Elsevier

118

S.H. Fatemi and T.D. Folsom

Fig.  1  MRI imaging reveals significant (p <0.05) brain atrophy in multiple brain areas of the 35-day-old virally infected mouse offspring (right) as compared to sham infected mice (left). Reprinted from Fatemi, S. H., Folsom, T. D., Reutiman, T. J., Huang, H., Oishi, K., Mori, S., et al. (2008). Maternal infection leads to abnormal gene regulation and brain atrophy in mouse offspring: Implications for genesis of neurodevelopmental disorders. Schizophr Res, 99(1–3), 61, Figure 1, Copyright (2008), with permission from Elsevier

Following infection of C57BL/6 mice at E18, analysis revealed a significant reduction of brain volume by »4% (p <0.014) in P35 offspring of exposed mice (Fig. 1; Table 2; Fatemi, Reutiman, Folsom, Huang, et al., 2008). There were significant reductions in volume for the cerebellum (p <0.001) and hippocampus (p <0.00005) at P35 (Fatemi et al., unpublished observations). The ventricular values did not vary significantly between the two groups (Table 2). Analysis of FA of corpus callosum revealed reduced FA on P35 offspring (p <0.0082) of exposed mice (Fatemi, Reutiman, Folsom, Huang, et al.). Thus, while E9 infection led to early thinning and later increase in brain size (Fatemi, Earle, et  al., 2002; Fatemi et  al., 1999), E16 infection lead to variable white matter changes and atrophy of cerebellum, hippocampus, and reduced total brain size of 11, 6, and 7.5%, respectively in P14 and P35. Finally, E18 infection led to atrophy of cerebellum (6.7%), hippocampus (14.3%), and reduced total brain size (4.2%) in P35 offspring only. These results indicate that infection at different time points lead to varying degrees of abnormal brain development.

3.2 Brain Gene Expression Abnormalities Following Prenatal Viral Infection Several reports using DNA microarray technology implicate various gene families as being involved in pathology of schizophrenia, i.e., genes involved in signal

Prenatal Viral Infection in Mouse: An Animal Model of Schizophrenia

119

Table 2  Changes in brain and ventricular areas following prenatal infection at E18 PD Brain area Control Infected Change p-Value P0 Cerebellum 2.58 ± 0.78 2.05 ± 0.25 – 0.321 Hippocampus 2.53 ± 0.21 2.52 ± 0.08 – 0.948 Total brain1 88.73 ± 1.93 89.09 ± 4.94 – 0.91 P14 Cerebellum 46.36 ± 1.22 46.26 ± 0.33 – 0.871 Hippocampus 18.90 ± 1.14 18.64 ± 1.16 – 0.78 Total brain1 366.95 ± 8.44 357.18 ± 15.32 – 0.37 P35 Cerebellum 60.13 ± 0.49 56.12 ± 0.93 ↓6.70% 0.001 Hippocampus 21.80 ± 0.16 18.80 ± 0.37 ↓14.30% 0.00005 Total brain1 425.72 ± 5.63 407.22 ± 7.08 ↓4.20% 0.014 Ventricle 0.229 ± 0.020 0.222 ± 0.075 – 0.88 Cerebellum 59.05 ± 2.81 60.19 ± 1.10 – 0.749 P56 Hippocampus 19.84 ± 1.70 20.23 ± 0.26 – 0.959 Total brain1 413.30 ± 26.57 420.39 ± 2.52 – 0.61 Ventricle 0.265 ± 0.125 0.202 ± 0.137 – 0.56 1 Reprinted from Fatemi, S. H., Folsom, T. D., Reutiman, T. J., Huang, H., Oishi, K., Mori, S., et al. (2008). Maternal infection leads to abnormal gene regulation and brain atrophy in mouse offspring: Implications for genesis of neurodevelopmental disorders. Schizophr Res, 99(1–3), 62, Table 3, Copyright (2008), with permission from Elsevier

transduction (Hakak et al., 2001; Mirnics & Lewis, 2001; Vawter et al., 2002), cell growth and migration (Chung, Tallerico, & Seeman, 2003), myelination (Hakak et  al., 2001; Tkachev et  al., 2003), regulation of presynaptic membrane function (Mirnics, Middleton, Marquez, Lewis, & Levitt, 2000; Vawter et  al., 2002), and GABAergic function (Hakak et al., 2001; Hashimoto et al., 2003). We have employed microarray technology to measure changes in brain genes following infection at E9, E16, and E18. Microarray analysis of E9 virally-exposed mouse brains from Balb/c mice showed significant (p <0.05) at least twofold upregulation of 21 genes and downregulation of 18 genes in brain homogenates at P0 (Fatemi, Pearce, et al., 2005). Examples of upregulated genes included ryanodine receptor 2, insulin-like growth factor binding protein 4, and synaptosomal complex protein SC65 (Fatemi, Pearce, et al.). Examples of downregulated genes included aquaporin 4 (Aqp4), carbonic anhydrase 3, myelin basic protein (Mbp), nucleolin, and proteolipid protein (Plp1) (Fatemi, Pearce, et  al.). Both myelin basic protein and proteolipid protein have previously been shown to display altered expression in subjects with schizophrenia (Hakak et al., 2001; Tkachev et al., 2003). At P35, 50 genes were significantly upregulated and of 21 genes were downregulated in neocortex vs. controls as a result of infection on E9 (Fatemi, Folsom, et al., 2008). In cerebellum at P35, 103 genes were upregulated while 102 genes were downregulated (Fatemi, Folsom, et al.). A number of these genes are associated with schizophrenia including dopa decarboxylase, forkhead box P2, glutamate receptor ionotropic, kainate 1, cholinergic receptor muscarinic CNS, neurogranin, oxytocin, and regulator of G protein signaling 4 (Rgs4) (Table 3).

120

S.H. Fatemi and T.D. Folsom

Table  3  Microarray results for selected schizophrenia candidate genes significantly affected following infection at E9a Gene Symbol Direction Area Day Function 1 Nucleolin Ncl Up Cer 35 Regulation of nucleobase, nucleoside, nucleotide, and nucleic acid metabolism Heat shock protein 1A2 Hspa1a Down NC 35 Protein metabolism Glutamate receptor, Grik1 Up NC 35 Transport ionotropic, kainate 12 Apoa1 Up NC 56 Transport Apolipoprotein A12 Microcephalin2 Mcph1 Down NC 56 Unknown Forkhead Box P21 Foxp2 Up Cer 35 Regulation of nucleobase, nucleoside, nucleotide, and nucleic acid metabolism Regulator of G Protein Rgs4 Down Cer 35 Cell communication; Signaling 41 signal transduction Cholinergic receptor, Chrm1 Down Cer 35 Membrane receptor muscarinic CNS1 Ngrn Down Cer 35 Neurobiology Neurogranin1 Dopa decarboxylase1 Ddc Up Cer 35 Metabolism, energy pathways Oxytocin1 Oxt Up Cer 35 Growth factor/ hormone B-cell translocation gene 1, Btgl Down Cer 56 Cell growth and/or anti-proliferative1 maintenance Transthyretin1 Ttr Down Cer 56 Growth factor/hormone B brain; Cer cerebellum; NC neocortex a All changes are at least twofold with p <0.05 (Fatemi, Folsom, et al., 2008; Fatemi, Reutiman, Folsom, & Sidwell, 2008) 1 With kind permission from Springer Science+Business Media: Fatemi, S. H., Reutiman, T. J., Folsom, T. D., & Sidwell, R. W. (2008). The role of cerebellar genes in pathology of autism and schizophrenia. Cerebellum, 7, 282–284; 285–287, Tables 1 and 2 2 Reprinted from Fatemi, S. H., Folsom, T. D., Reutiman, T. J., & Sidwell, R. W. (2008). Viral regulation of aquaporin 4, connexin 43, microcephalin, and nucleolin. Schizophr Res, 98(1–3), 164–166, Tables 1 and 2, and Appendix A, Copyright (2009), with permission from Elsevier

Microarray analysis showed that infection at E9 led to the significant (p <0.05) upregulation of 13 genes and downregulation of 11 genes in neocortex vs. controls at P56 (Fatemi, Folsom, et al., 2008). There was significant (p <0.05) upregulation of 27 genes and downregulation of 23 genes at P56 in cerebellum (Fatemi, Reutiman, Folsom, & Sidwell, 2008). Genes associated with schizophrenia that were significantly altered at this time point include apolipoprotein A1, B-cell translocation gene 1, anti-proliferative, and transthyretin (Fatemi, Folsom, et al., 2008; Fatemi, Reutiman, Folsom, & Sidwell, 2008; Table 3).

Prenatal Viral Infection in Mouse: An Animal Model of Schizophrenia

121

Microarray analysis of brains from exposed C57BL/6 mice following infection at E16 showed a significant (p <0.05) altered expression of genes in frontal cortex (151 upregulated and 35 downregulated at P0; 106 upregulated and 86 downregulated at P14; and 107 upregulated and 118 downregulated at P56), hippocampus (299 upregulated and 191 downregulated at P0; 34 upregulated and 46 downregulated at P14; and 87 upregulated and 30 downregulated at P56), and cerebellum (27 upregulated and 73 downregulated at P0; 205 upregulated and 16 downregulated at P14; and 405 upregulated and 205 downregulated at P56) of virally-exposed mouse offspring (Fatemi et al., 2009a, b, unpublished observations), suggesting a more deleterious effect than what was observed following infection at E9. Numerous genes are related to schizophrenia, including adrenomedulin, pro-melanin concentrating hormone, Rgs4, v-erbb2 erythroblastic leukemia viral oncogene homolog 3 (avian) (Erbb4), neural cell adhesion molecule 1 (Ncam1) and very low density lipoprotein receptor (Vldlr) displayed altered expression (Fatemi et al., unpublished observations). These genes are involved in multiple processes, including signal transduction, synaptogenesis, transport, and metabolism (Table  4). Interestingly, a number of myelination genes associated with schizophrenia were also significantly altered: Myelin and lymphocyte-associated protein (Mal), Mbp, Myelin oligodendrocyte protein (Mog), Myelin-associated oligodendrocytic basic protein (Mobp), and Plp1, (Fatemi et al., 2009a, b). Infection of C57BL/6 mice on E18 resulted in an overall decrease in the number of genes that showed altered expression when compared with infection at E16. Gene expression data showed a significant (p <0.05) altered expression of genes in frontal cortex (43 upregulated and 29 downregulated at P0; 16 upregulated and 17 downregulated at P14; and 86 upregulated and 24 downregulated at P56), hippocampus (129 upregulated and 46 downregulated at P0; 9 upregulated and 12 downregulated at P14; and 45 upregulated and 17 downregulated at P56), and cerebellum (120 upregulated and 37 downregulated at P0; 11 upregulated and 5 downregulated at P14; and 74 upregulated and 22 downregulated at P56) of mouse offspring (Fatemi, Reutiman, Folsom, Huang, et al., 2008). Several genes implicated in the etiopathology of schizophrenia, (e.g., Erbb4, myelin transcription factor 1-like, semaphorin 3A) were shown to be affected significantly (p <0.05) by DNA microarray in the same direction and magnitude of change were validated by qRT-PCR (Fatemi, Reutiman, Folsom, Huang, et al.; Table 5). Several genes such as NS1A influenza binding protein and aryl hydrocarbon receptor nuclear translocator, which are involved in influenza-mediated RNA processing, were upregulated in all three brain regions at P0 (Table 5). Our microarray results following infection at E9, E16, and E18 are consistent with findings from groups using postmortem tissues from subjects with schizophrenia.

3.3 Altered Protein Expression in Offspring Following Prenatal Viral Infection Prenatal viral infection at E9 leads to altered expression of a number of important brain proteins, many of which have been implicated in schizophrenia. Nitric oxide

122

S.H. Fatemi and T.D. Folsom

Table  4  Microarray results for selected schizophrenia candidate following infection at E16 Fold Gene Symbol Area PD change Neuropeptide Y receptor Y1 Nyp1r Cer  0 1.56 Adrenomedulin Adm Cer 14 2.82

genes significantly affected

Myelin basic protein Myelin oligodendrocyte protein Myelin-associated glycoprotein Myelin-associated oligodendrocytic basic protein Myelin and lymphocyteassociated protein Neural cell adhesion molecule 1 Proteolipid protein (myelin) 1 Regulator of G protein 4 Synapsin 2 Synaptotagmin 3 Gamma-aminobutyric acid (GABA) receptor, subunit alpha 4 Cadherin 8 v-erb-a erythroblastic leukemia viral oncogene homolog 4 (avian) Forkhead box P2 Glutamic acid decarboxylase 2

pa 0.043 0.043

14 14 56 14

1.71 1.65 −1.99 1.98

0.032 0.032 0.049 0.023

Cer

14 56

2.11 −2.83

0.043 0.022

Mal

Cer

14

2.04

0.036

Function STCC Regulation of physiological processes Myelination Immune response Immune response Cell growth and/or maintenance Transport

Ncam1

Cer

56

−1.97

0.010

STCC

Plp1

Cer

56

−2.06

0.047

Rgs4

Cer

56

−3.58

0.005

Cell growth and/or maintenance STCC

Syn2 Syt3 Gabra4

PFC PFC PFC

 0  0 14

1.56 −1.73 −1.61

0.016 0.001 0.036

Synaptogenesis Synaptogenesis Transport

Cdh8 Erbb4

PFC PFC

56 56

−1.56 1.74

0.025 0.006

STCC STCC

Foxp2 Gad2

Hipp Hipp

 0  0

−6.38 −1.81

0.046 0.018

RNNNM Metabolism, energy pathways STCC

Mbp Mog

Cer Cer

Mag

Cer

Mobp

Neuron-glia-CAM-related Nrcam Hipp 14 −1.70 0.014 cell adhesion molecule 5-hydroxytriptamine Htr1a Hipp 14 −1.60 0.025 STCC (serotonin) receptor 1A Neurotensin Nts Hipp 56 2.36 0.037 STCC Reprinted from (1) Fatemi, S. H., Folsom, T. D., Reutiman, T. J., Abu-Odeh, D., Mori, S , Huang, H., et  al. (2009). Abnormal expression of myelination genes and alterations in white matter fractional anisotropy following prenatal viral influenza infection at E16 in mice. Schizophr Res, 112(1–3), 48, Table  1, and Appendix A Copyright (2009), with permission from Elsevier; and (2) Fatemi, S. H , Folsom, T. D., Reutiman, T. J., Huang, H., Oishi, K., & Mori, S. (2009). Prenatal viral infection of mice at E16 causes changes in gene expression in hippocampi of the offspring. Eur Neuropsychopharmacol, 19(9), 650, Table 1, and Appendix A, Copyright (2009), with permission from Elsevier

Prenatal Viral Infection in Mouse: An Animal Model of Schizophrenia

123

synthase, neuronal (nNOS) expression in hippocampus of infected offspring is increased at P0 (Fatemi, Sidwell, Akhter, et al., 1998). The increased nNOS expression likely reflects an increased potential for glutamate excitotoxicity, apoptosis, growth arrest, N-nitrosylation, and increased glutamate release (Nicotera, Brune, & Bagetta, 1997). Synaptosomal-associated protein 25 kDa (SNAP-25) was significantly altered at P0 in hippocampus of infected offspring (Fatemi, Sidwell, Kist, et al., 1998). Expression varied by layer with increased expression in the septal-dorsal hippocampus (except for subplate), a smaller increase in mid-septo-temporal hippocampus, and a reduction in the temporal-ventral areas (Fatemi, Sidwell, Kist, et  al.). Glial fibrillary acidic protein (GFAP) also showed significant increases in expression in cortical (P14), hippocampal cells (P35), ependymal cell layer (P14), and whole brain (P0) (Fatemi, Emamian, et  al., 2002; Fatemi, Pearce, Brooks, & Sidwell, 2003), which implicates a significant role of infection by human influenza on subsequent gliosis in brains of exposed mice. GFAP protein and mRNA have been shown to be altered in subjects with schizophrenia (Rajkowska et  al., 2002; Toro, Hallak, Dunham, & Deakin, 2006; Webster, O’Grady, Kleinman, & Weickert, 2005). Glutamic acid decarboxylase 65 and 67  kDa proteins (GAD65 and GAD67), the enzymes that catalyze the conversion of glutamate to gamma-aminobutyric acid (GABA), show altered expression in the brains of infected mice suggesting GABAergic dysfunction resulting from prenatal viral infection. Specifically, GAD65 (P14, P35) and GAD67 (P35, P56) proteins show significant increases in whole brain (Fatemi, Stary, Earle, Araghi-Niknam, & Eagan, 2005), a significant finding, as some reports indicate decreases in both GAD65 and 67 kDa proteins and mRNA (Fatemi et al., 2004; Heckers et al., 2002), with others showing increases in both (Hakak et al., 2001) in subjects with schizophrenia. Significant reductions in Reelin in cerebral cortex layers II–IV and hippocampus at P0 have been observed following prenatal viral infection at E9 (Fatemi et  al., 1999). Levels of Reelin protein are reduced in several brain areas of adult schizophrenics or in heterozygous reeler mutant mice, both groups exhibiting defective sensorimotor gating abnormalities (Fatemi, Earle, & McMenomy, 2000; Impagnatiello et al., 1998; Tueting et al., 1999). We sought to verify our microarray results following infection at E9 for microcephalin (downregulated at P56, neocortex), nucleolin (downregulated at P0, brain; upregulated at P35, cerebellum; downregulated at P35 in neocortex), AQP4 (downregulated at P0 in brain), and connexin 43 (Cx43) by measuring protein levels via SDS PAGE and western blotting (Fatemi, Folsom, et al., 2008; Fatemi, Pearce, et al., 2005; Fatemi, Reutiman, Folsom, & Sidwell, 2008). We found that nucleolin was increased by 40% (p <0.032) at P35 and 53% (p <0.014) at P56 in neocortex, and by 138% (p <0.031) at P56 in cerebellum. AQP4 was decreased by 22% (p <0.041) at P35. At P56, reductions in both microcephalin (32%, p <0.016) in cerebellum, and CX43 (20%, p <0.025) in neocortex were observed (Fatemi, Folsom, et al., 2008). Cer Cerebellum; Hipp hippocampus; PD postnatal day; PFC prefrontal cortex; RNNNNM regulation of nucleobase, nucleotide, nucleoside, nucleic acid metabolism; STCC signal transduction, cell communication (Fatemi, unpublished observations) a Significant p <0.05

Death associated protein kinase 1 DEAD (Asp-Glu-AlaAsp) box polypeptide 3, Y-linked Ephrin B2 v-erb-a erythroblastic leukemia viral oncogene homolog 4 (avian)

Gene Cdc42 guanine nucleotide exchange factor (GEF) 9 (Collybistin) Aryl hydrocarbon receptor nuclear translocator

Area PFC

CER HIP PFC CER CER

HIP HIP

Symbol Arhgef9

Arnt Arnt Arnt Dapk1

Dby

Efnb2 Erbb4

P0 P0

P56

P0 P0 P0 P0

Day P0

2.334 3.434

3.996

* * * 2.180

Microarray fold change 2.787

Table 5  Microarray and qRT-PCR results for selected affected genes in E18 infected mice

0.0345 0.0055

0.0436

* * * 0.0051

Microarray (p-value) 0.0229

2.413 2.097

6.711

1.505 2.082 1.447 1.220

Gene relative to normalizer (qRT-PCR) 1.251

0.021 0.048

0.005

0.024 0.021 0.006 0.013

QPCR (p-value) 0.026

124 S.H. Fatemi and T.D. Folsom

Ivns1abp Ivns1abp Ivns1abp Myt1l

CER HIP PFC HIP

P0 P0 P0 P0

* * * 2.355

* * * 0.0387

1.215 1.91 1.02 2.034

0.0004 0.079 0.8603 0.051

Myelin transcription factor 1-like Neurexophilin 2 Nxph2 HIP P0 4.146 0.0006 3.504 0.010 HIP P0 3.517 0.0184 3.489 0.013 Sema domain, immunoglobulin Sema3a domain (Ig), short basic domain, secreted, (semaphorin) 3A SRY-box containing gene 2 Sox2 CER P0 2.164 0.0083 1.299 0.010 Transferrin receptor 2 Trfr2 CER P0 2.266 0.0449 1.565 0.026 Uty CER P56 3.681 0.0333 5.345 0.018 Ubiquitously transcribed tetratricopeptide repeat gene, Y chromosome Reprinted from Fatemi, S. H., Folsom, T. D., Reutiman, T. J., Huang, H., Oishi, K., Mori, S., et al. (2008). Maternal infection leads to abnormal gene regulation and brain atrophy in mouse offspring: Implications for genesis of neurodevelopmental disorders. Schizophr Res, 99(1–3), 60, Table 1, Copyright (2008), with permission from Elsevier *Not altered in microarray analysis

Influenza virus NS1A binding protein

Prenatal Viral Infection in Mouse: An Animal Model of Schizophrenia 125

126

S.H. Fatemi and T.D. Folsom

Fig. 2  Mean Foxp2/b-actin ratios for the progeny of infected (filled histogram bars) and sham-infected mice are shown. Levels of Foxp2/b-actin were significantly increased in P0 (a) and P35 (b) cerebellum (p = 0.050 and 0.032, respectively) and in P35 (c) and P56 (d) hippocampus (p = 0.036 and 0.022 respectively). Foxp2 bands from P0 and P35 cerebellum and P35 and P56 hippocampus homogenates (60 mg/ lane) of representative progeny from sham-infected and infected mice are shown. **p <0.05 vs. control. Reprinted from Fatemi, S. H., Folsom, T. D., Reutiman, T. J., Huang, H., Oishi, K., Mori, S., et al. (2008). Maternal infection leads to abnormal gene regulation and brain atrophy in mouse offspring: Implications for genesis of neurodevelopmental disorders. Schizophr Res, 99(1–3), 62, Figure  2, Copyright (2008), with permission from Elsevier

Taken together, these results suggest that infection at E9 leads to changes in multiple brain markers for enhanced ribosome genesis (nucleolin), increased production of immature neurons (microcephalin), and abnormal glial-neuronal communication and neuronal migration (CX43 and AQP4) (Fatemi, Folsom, et al.). Due to a number of myelination genes showing altered expression following infection at E16, we measured protein levels of myelin basic protein (MBP), myelin associated glycoprotein (MAG), proteolipid protein (PLP1), and a splice variant of PLP1 known as DM20 in the cerebella of offspring from infected mice vs. sham-infected controls. We found significant reductions in MBP isoforms at P14 (MBP 14 kDa protein; p < 0.022) and at P56 (Mbp 17.2 and 18 kDa proteins measured together; p < 0.026) in the offspring of infected mice when compared with controls (Fatemi et al., 2009b).

Prenatal Viral Infection in Mouse: An Animal Model of Schizophrenia

127

Cerebella of infected mice also showed reductions in MAG at P14 (p < 0.044) and at P35 (p < 0.0025) (Fatemi et al., 2009b). Finally, DM20 expression was increased at P35 (p < 0.0034) and P56 (p < 0.016) while PLP1 showed no difference in expression (Fatemi et al., 2009b). Forkhead box P2 (Foxp2), a gene implicated in both autism (Gong et al., 2004) and schizophrenia (Sanjuan et  al., 2006), has previously been demonstrated to display upregulated mRNA in P35 cerebellum of exposed offspring of Balb/c dams infected on E9 (Fatemi, Folsom, et  al., 2008). Western blotting experiments revealed that Foxp2 protein was upregulated significantly in cerebella of P0 (p < 0.050) and P35 (p < 0.032; which verifies our microarray data for Foxp2 at this time point) exposed progeny of mice infected on E18 (Fig. 2; Fatemi, Reutiman, Folsom, Huang, et al., 2008) and hippocampi in P35 (p < 0.036) and P56 (p < 0.022) exposed offspring (Fig. 2; Fatemi, Reutiman, Folsom, Huang, et al.).

3.4 Brain Neurotransmitters are Altered in Exposed Offspring Following Viral Infection at E16 and E18 Previous work has demonstrated altered dopaminergic and serotonergic function in patients with schizophrenia (Abi-Dargham et al., 2002, 2002; Juckel et al., 2003, 2008). Levels of several monoamines, monoamine metabolites, and amino acids [glutamate, glycine, GABA, 3,4-dihydroxyphenylacetic acid (DOPAC), 5-Hydroxyindoleacetic acid (5-HIAA), taurine, dopamine (DA), and serotonin (5-HT)] were determined in the cerebella of offspring from C57BL/6 mice virallyexposed at E16 and controls (Winter et  al., 2008). When compared to controls, there was a significant decrease in 5-HT levels in the cerebella of virally exposed mice at P14 (23.8 ± 2.5 vs. 37.2 ± 2.0). No differences were observed in levels of other neurotransmitters in exposed and control mice, although there was a significant decrease in dopamine at P14 and P56 when compared to P0 (Winter et al.). Quite interestingly, measurement of levels of neurotransmitters in the cerebella of offspring from C57BL/6 mice virally-exposed at E18 showed significant (p < 0.05) reductions between infected and controls in 5-HT (37.8 ± 2.6 vs. 51.8 ± 1.0) and its metabolite 5-HIAA (74.6 ± 4.8 vs. 103.6 ± 5.9) at P14 and significant (p < 0.05) reductions in 5-HT (62.0 ± 6.9 vs. 73.2 ± 2.3) and taurine (47.6 ± 2.8 vs. 59.5 ± 3.0) at P35 (Fatemi, Reutiman, Folsom, Huang, et al., 2008). Overall, serotonin appears to be reduced following infection at E16 and E18. The reduction in serotonin, which has neurotrophic effects during embryogenesis (Sodhi & Sanders-Bush, 2004), could contribute to the brain atrophy observed following infection at E18 (Fatemi, Reutiman, Folsom, Huang, et  al., 2008). Reduced levels of taurine, which has a role in development and survival of neurons, have previously been observed in cerebral spinal fluid of subjects with schizophrenia (Do et al., 1995).

128

S.H. Fatemi and T.D. Folsom

3.5 Abnormal Behavior of Offspring Following Prenatal Viral Infection Behavioral abnormalities associated with schizophrenia manifest themselves during adolescence and early adulthood (Andreasen, 1999). These abnormalities include disturbances in attention, sensorimotor gating, perception, and social interaction, reflecting establishment of developmental misconnections between several important areas of the brain (prefrontal cortex, hippocampus, thalamus, and cerebellum) in schizophrenic individuals. Specific neurodevelopmental insults and various mutations affecting levels of important neuroregulatory proteins, could account for the emergence of similar behaviors in rodent models (Lipska, Jaskiw, & Weinberger, 1993; Lipska, Khaing, & Weinberger, 1999; Shi, Fatemi, Sidwell, & Patterson, 2003). Prenatal viral infection on E9 resulted in various behavioral abnormalities in exposed offspring of infected mice consistent with subjects with schizophrenia (Shi et  al., 2003). These alterations included abnormal exploratory behavior, reflecting difficulty handling stress, which is observed in schizophrenia. The offspring of exposed mice showed significantly less time exploring their environment vs. control mice (Shi et al.). Moreover, the offspring of exposed mice contacted each other less frequently than the control mice, suggesting altered social behavior (Shi et al.). Finally, the offspring of exposed mice displayed an abnormal acoustic startle response (Shi et  al.). Acoustic startle response has been used effectively in humans and rodents to measure sensorimotor gating (Koch, 1999; Swerdlow & Geyer, 1998). When a prepulse, which is too small to cause a startle itself, precedes the startle stimulus, the response is reduced, a phenomenon that is termed prepulse inhibition (PPI) (Shi et al., 2003) occurs. Exposed BALB/c and C57BL/6 mice showed significant PPI deficits at prepulses of 75 and 80 dB vs. controls (Shi et al.). This is similar to PPI deficits in untreated schizophrenic subjects (Geyer, Braff, & Swerdlow, 1999). Administration of antipsychotic agents chlorpromazine (a typical agent) and clozapine (an atypical agent), which treat schizophrenic symptoms and correct PPI deficits in patients, caused significant increases in PPI in the exposed mice vs. controls, correcting the PPI deficits (Shi et al., 2003). The response by offspring of exposed mice to both antipsychotics shows that our animal model has predictive validity for positive symptoms of schizophrenia (Shi et al.).

4 Conclusions An association between prenatal viral infection and the development of schizophrenia is strongly suggested based on extensive epidemiological data. Animal models provide a means to study the effects of prenatal viral infection on the developing brain. Our laboratory has demonstrated extensive changes in brain

Behavior

Microarray

P42–P56a,b

b

P56     

P0b P35b

↑21 ↑50 ↑103 ↑13 ↑27

↓18 ↓21 ↓102 ↓11 ↓23

↓PPI, ↓SR, ↓EB, ↓SI

WB NC Cer NC Cer

NC

P56a

P14a

P0a

PFC Hipp Cer PFC Hipp Cer PFC Hipp Cer

↑151 ↑300 ↑26 ↑106 ↑33 ↑204 ↑107 ↑86 ↑449

↓31 ↓190 ↓72 ↓86 ↓45 ↓15 ↓118 ↓17 ↓204

Table 6  Detailed summary of changes following prenatal viral infection at E9, E16, and E18 E9 E16 Morphology P0a ↓PC density and P0a ↓Ventricular volume; nuclear size ↓white matter at IC P14a ↓Brain and Cer volumes; ↑white matter at CC P35a P98a ↑PC and non PC density, ↓Hipp volume ↓PC nuclear size, ↑brain area, ↓ventricular area ↑White matter at MCP P56a

ND

P56a

P14a

P0a

E18 P35a

PFC Hipp Cer PFC Hipp Cer PFC Hipp Cer

↑43 ↑129 ↑120 ↑16 ↑9 ↑11 ↑86 ↑45 ↑74

(continued)

↓29 ↓46 ↓37 ↓17 ↓12 ↓5 ↓24 ↓17 ↓22

↓Cer, Hipp, total brain volume; ↓white matter at CC

Prenatal Viral Infection in Mouse: An Animal Model of Schizophrenia 129

↑nNOS, ↑SNAP-25, ↓Reelin, ↑GFAP

↑GFAP, ↑GAD65 ↑nNOSa, ↑GFAPa, ↑GAD65b, ↑GAD67b, ↑Foxp2b ↓nNOSa, ↑GAD67b

P14a P35a

P56a

P0

P14a

ND

E16

↓5-HT

P56a

P35a

P0a

E18

↑Dby, ↑Uty, ↑Foxp2

↑Arhgef9, ↑Arnt, ↑Dapk1, ↑Efnb2, ↑Erbb4, ↑Ivns1abp, ↑Myt1l, ↑Nxph2, ↑Sema3a, ↑Sox2, ↑Trfr2, ↑Foxp2 ↑Foxp2

Neuroc­hemistry

ND

P14a ↓5-HT, ↓Taurine P35a ↓5-HT, ↓5-HIAA Reprinted from (1) Fatemi, S. H., Folsom, T. D., Reutiman, T. J., Huang, H., Oishi, K., Mori, S., et al. (2008). Maternal infection leads to abnormal gene regulation and brain atrophy in mouse offspring: Implications for genesis of neurodevelopmental disorders. Schizophr Res, 99(1–3), 67, Table 6, Copyright (2008), with permission from Elsevier; (2) Fatemi, S. H., Folsom, T. D., Reutiman, T. J., Abu-Odeh, D., Mori, S., Huang, H., et al. (2009). Abnormal expression of myelination genes and alterations in white matter fractional anisotropy following prenatal viral influenza infection at E16 in mice. Schizophr Res, 112(1–3), 51, Table 7, Copyright (2009), with permission from Elsevier; and (3) Fatemi, S. H., Folsom, T. D., Reutiman, T. J., Huang, H., Oishi, K., & Mori, S. (2009). Prenatal viral infection of mice at E16 causes changes in gene expression in hippocampi of the offspring. Eur Neuropsychopharmacol, 19(9), 651, Table 5, Copyright (2009), with permission from Elsevier CC corpus callosum; Cer cerebellum; EB exploratory behavior; Hipp hippocampus; IC internal capsule; MCP middle cerebellar peduncle; NC neocortex; ND not determined; PC pyramidal cell; PPI prepulse inhibition; SI social interaction; SR startle response; WB whole brain ↑ increase; ↓ decrease a C57BL6J mice b Balb/c mice

Genes/proteins

a

Table 6  (continued) E9

130 S.H. Fatemi and T.D. Folsom

Prenatal Viral Infection in Mouse: An Animal Model of Schizophrenia

131

morphology, gene and protein expression, neurochemistry, and behavior as a result of prenatal viral infection (Table  6). This model displays many of the ­morphological, gene expression, and behavioral dysfunctions found in subjects with schizophrenia. Our model has the potential to provide important information on treatments or therapies to reverse or prevent the deleterious effects of prenatal infection. Acknowledgments  Grant support by National Institute of Child Health and Human Development (#5R01-HD046589-04 and 3R01-HD046589-04S1) to SHF is gratefully acknowledged. Portions of this article are reprinted from: (1) Fatemi, S. H., Folsom, T. D., Reutiman, T. J., Abu-Odeh, D., Mori, S., Huang, H., et al. (2009). Abnormal expression of myelination genes and alterations in white matter fractional anisotropy following prenatal viral influenza infection at E16 in mice. Schizophr Res, 112(1–3), Copyright (2009), with permission from Elsevier; (2) Fatemi, S. H., Folsom, T. D., Reutiman, T. J., Huang, H., Oishi, K., & Mori, S. (2009). Prenatal viral infection of mice at E16 causes changes in gene expression in hippocampi of the offspring. Eur Neuropsychopharmacol, 19(9), Copyright (2009), with permission from Elsevier; (3) Fatemi, S. H., Folsom, T. D., Reutiman, T. J., Huang, H., Oishi, K., Mori, S., et al. (2008). Maternal infection leads to abnormal gene regulation and brain atrophy in mouse offspring: Implications for genesis of neurodevelopmental disorders. Schizophr Res, 99(1–3), Copyright (2008), with permission from Elsevier; (4) with kind permission from Springer Science+Business Media: Fatemi, S. H., Reutiman, T. J., Folsom, T. D., & Sidwell, R. W. (2008). The role of cerebellar genes in pathology of autism and schizophrenia. Cerebellum, 7; and (5) Fatemi, S. H., Folsom, T. D., Reutiman, T. J., & Sidwell, R. W. (2009). Viral regulation of aquaporin 4, connexin 43, microcephalin, and nucleolin. Schizophr Res, 98(1–3), Copyright (2009), with permission from Elsevier.

References Abi-Dargham, A., Mawlawi, O., Lombardo, I., Gil, R., Martinez, D., Huang, Y., et  al. (2002). Prefrontal dopamine D1 receptors and working memory in schizophrenia. J Neurosci, 22, 3708–3719. Abi-Dargham, A., Rodenhiser, J., Printz, D., Zea-Ponce, Y., Gil, R., Kegeles, L. S., et al. (2000). Increased baseline occupancy of D2 receptors by dopamine in schizophrenia. Proc Natl Acad Sci U S A, 97, 8104–8109. Acuff-Smith, K. D., & Vorhees, C. V. (1999). Neurobehavioral teratology. In R. J. M. Niesink, R. M. A. Jaspers, L. M. W. Kornet, J. M. VanRee, & H. A. Tilson (Eds.), Introduction to neurobehavioral toxicology, food and environment (pp. 26–69). Boca Raton, FL: CRC. Adams, W., Kendell, R. E., & Hare, E. T. (1993). Epidemiological evidence that maternal influenza contributes to the aetiology of schizophrenia: An analysis of Scottish, English, and Danish data. Br J Psychiatry, 163, 522–534. Andreasen, N. C. (1999). A unitary model of schizophrenia. Bleuler’s “Fragmented phrene” as schizencephaly. Arch Gen Psychiatry, 56, 781–793. Arnold, S. E., & Trojanowski, J. Q. (1996). Recent advanced in defining the neuropathology of schizophrenia. Acta Neuropath, 92, 217–231. Avishai-Eliner, S., Brunson, K. L., Sandman, C. A., & Baram, T. Z. (2002). Stressed-out or in (utero)? Trends Neurosci, 25, 518–524. Barr, C. E., Mednick, S. A., & Munk-Jorgensen, P. (1990). Exposure to influenza epidemics during gestation and adult schizophrenia: A 40-year study. Arch Gen Psychiatry, 47, 869–874. Blond, J. L., Besème, F., Duret, L., Bouton, O., Bedin, F., Perron, H., et  al. (1999). Molecular characterization and placental expression of HERV-W, a new human endogenous retrovirus family. J Virol, 73, 1175–1185.

132

S.H. Fatemi and T.D. Folsom

Boksa, P., & Luheshi, G. N. (2003). On the use of animal modeling to study maternal infection during pregnancy and prenatal cytokine exposure as risk factors for schizophrenia. Clin Neurosci, 3, 339–346. Boyd, J. H., Pulver, A. E., & Stewart, W. (1986). Season of birth: Schizophrenia and bipolar disorder. Schizophr Bull, 12, 173–186. Boyer, P., Phillips, J. L., Rousseau, F. L., & Ilivitsky, S. (2007). Hippocampal abnormalities and memory deficits: New evidence of a strong pathophysiological link in schizophrenia. Brain Res Rev, 54, 92–112. Brown, A. S. (2006). Prenatal infection as a risk factor for schizophrenia. Schizophr Bull, 32, 200–202. Brown, A. S., Begg, M. D., Gravenstein, S., Schaefer, C. A., Wyatt, R. J., Bresnahan, M., et al. (2004). Serologic evidence of prenatal influenza in the etiology of schizophrenia. Arch Gen Psychiatry, 61, 774–780. Chung, C., Tallerico, T., & Seeman, P. (2003). Schizophrenia hippocampus has elevated expression of chondrex glycoprotein gene. Synapse, 50, 29–34. Crespo-Facorro, B., Barbadillo, L., Pelayo-Terán, J. M., & Rodríguez-Sánchez, J. M. (2007). Neuropsychological functioning and brain structure in schizophrenia. Int Rev Psychiatry, 19, 325–336. Do, K. Q., Lauer, C. J., Schreiber, W., Zollinger, M., Gutteck-Amsler, U., Cuenod, M., et  al. (1995). Gamma-glutamylglutamine and taurine concentrations are decreased in the cerebrospinal fluid of drug-naive patients with schizophrenic disorders. J Neurochem, 65, 2652–2662. Erlenmeyer-Kimling, L., Folnegovic, Z., Hrabak-Zerjavic, V., Borcic, B., Folnegovic-Smalc, V., & Susser, E. (1994). Schizophrenia and prenatal exposure to the 1957 A2 influenza epidemic in Croatia. Am J Psychiatry, 151, 1496–1498. Fahy, T. A., Jones, P. B., & Sham, P. C. (1993). Schizophrenia in Afro-Caribbeans in the UK ­following prenatal exposure to the 1957 A2 influenza epidemic. Schizophr Res, 6, 98–99. Fatemi, S. H., Earle, J. A., Kanodia, R., Kist, D., Emamian, E. S., Patterson, P. H., et al. (2002). Prenatal viral infection leads to pyramidal cell atrophy and macrocephaly in adulthood: Implications for genesis of autism and schizophrenia. Cell Mol Neurobiol, 22, 25–33. Fatemi, S. H., Earle, J., & McMenomy, T. (2000). Reduction in Reelin immunoreactivity in ­hippocampus of subjects with schizophrenia, bipolar disorder, and major depression. Mol Psychiatry, 5, 654–663. Fatemi, S. H., Emamian, E. S., Kist, D., Sidwell, R. W., Nakajima, K., Akhter, P., et al. (1999). Defective corticogenesis and reduction in Reelin immunoreactivity in cortex and hippocampus of prenatally infected neonatal mice. Mol Psychiatry, 4, 145–154. Fatemi, S. H., Emamian, E. S., Sidwell, R. W., Kist, D. A., Stary, J. M., Earle, J. A., et al. (2002). Human influenza viral infection in utero alters glial fibrillary acidic protein immunoreactivity in developing brains of neonatal mice. Mol Psychiatry, 7, 633–640. Fatemi, S. H., Folsom, T. D., Reutiman, T. J., & Sidwell, R. W. (2008a). Viral regulation of ­aquaporin 4, connexin 43, microcephalin and nucleolin. Schizophr Res, 98, 163–177. Fatemi, S. H., Laurence, J. A., Araghi-Niknam, M., Stary, J. M., Schulz, S. C., Lee, S., et  al. (2004). Glial fibrillary acidic protein is reduced in cerebellum of subjects with major depression, but not schizophrenia. Schizophr Res, 69, 317–323. Fatemi, S. H., Pearce, D. A., Brooks, A., & Sidwell, R. (2003). cDNA microarray studies of prenatal virally induced brain disorders in mouse. Stanley Proteomics and Genomes Meeting. Fatemi, S. H., Pearce, D. A., Brooks, A. I., & Sidwell, R. W. (2005). Prenatal viral infection in mouse causes differential expression of genes in brains of mouse progeny: A potential animal model for schizophrenia and autism. Synapse, 57, 91–99. Fatemi, S. H., Reutiman, T. J., Folsom, T. D., Huang, H., Oishi, K., Mori, S., et  al. (2008b). Maternal infection leads to abnormal gene regulation and brain atrophy in mouse offspring: Implications for genesis of neurodevelopmental disorders. Schizophr Res, 99, 56–70. Fatemi, S. H., Reutiman, T. J., Folsom, T. D., & Sidwell, R. W. (2008c). The role of cerebellar genes in pathology of autism and schizophrenia. Cerebellum, 7, 279–294.

Prenatal Viral Infection in Mouse: An Animal Model of Schizophrenia

133

Fatemi, S.H., Folsom, T.D., Reutiman, T.J., Huang, H., Oishi, K., Mori, S., 2009a. Prenatal viral infection of mice at E16 causes changes in gene expression in the hippocampi of the offspring. Eur Neuropsychopharmacol, 19, 648–653. Fatemi, S.H., Folsom, T.D., Reutiman, T.J., Abu-Odeh, D., Mori, S., Huang, H., Oishi, K., 2009b. Abnormal expression of myelination genes and alterations in white matter fractional anisotropy following prenatal viral influenza infection at E16 in mice. Schizophr Res, 112, 46–53. Fatemi, S. H., Sidwell, R., Akhter, P., Sedgwick, J., Thuras, P., & Barley, K. (1998). Human influenza viral infection in utero increases nNOS expression in hippocampi of neonatal mice. Synapse, 29, 84–88. Fatemi, S. H., Sidwell, R., Kist, D., Akhter, P., Bailey, K., Thuras, P., et al. (1998). Differential expression of synaptosome-associated protein 25 kDa (SNAP-25) in hippocampi of neonatal mice following exposure to human influenza virus in utero. Brain Res, 800, 1–9. Fatemi, S. H., Stary, J. M., Earle, J. A., Araghi-Niknam, M., & Eagan, E. (2005). GABAergic dysfunction in schizophrenia and mood disorders as reflected by decreased levels of glutamic acid decarboxylase 65 and 67 kDa and Reelin proteins in cerebellum. Schizophr Res, 72, 109–122. Geyer, M. A., Braff, D. L., & Swerdlow, N. R. (1999). Startle-response measures of information processing in animals: Relevance to schizophrenia. In M. Haug & R. E. Whalen (Eds.), Animal models of human emotion and cognition (pp. 103–116). Washington, DC: APA. Gong, X., Jia, M., Ruan, Y., Shuang, M., Liu, J., Wu, S., et al. (2004). Association between the FOXP2 gene and autistic disorder in Chinese population. Am J Med Genet B Neuropsychiatr Genet, 127, 113–116. Hakak, Y., Walker, J. R., Li, C., Wong, W. H., Davis, K. L., Buxbaum, J. D., et  al. (2001). Genome-wide expression analysis reveals dysregulation of myelination-related genes in chronic schizophrenia. Proc Natl Acad Sci U S A, 98, 4746–4751. Hare, E. H., Price, J. S., & Slater, E. (1972). Schizophrenia and season of birth. Br J Psychiatry, 120, 125–126. Hashimoto, T., Volk, D. W., Eggan, S. M., Mirnics, K., Pierri, J. N., Sun, Z., et al. (2003). Gene expression deficits in a subclass of GABA neurons in the prefrontal cortex of subjects with schizophrenia. J Neurosci, 23, 6315–6326. Heckers, S., Stone, D., Walsh, J., Shick, J., Koul, P., & Benes, F. M. (2002). Differential hippocampal expression of glutamic acid decarboxylase 67&67 messenger RNA in bipolar disorder and schizophrenia. Arch Gen Psychiatry, 59, 521–529. Hogan, B., Beddington, R., Constantini, F., & Lacy, E. (1994). Manipulating the mouse embryo. A laboratory manual. New York: CSHL. Holson, R. R., Adams, J., & Ferguson, S. A. (1999). Gestational stage-specific effects of retinois acid exposure in the rat. Neurotoxicol Teratol, 21, 393–402. Impagnatiello, F., Guidotti, A. R., Pesold, C., Dwivedi, Y., Caruncho, H., Pisu, M., et al. (1998). A decrease of Reelin expression as a putative vulnerability factor in schizophrenia. Proc Natl Acad Sci U S A, 95, 15718–15723. Juckel, G., Gallinat, J., Riedel, M., Sokullu, S., Schulz, C., Moller, H. J., et al. (2003). Serotonergic dysfunction in schizophrenia assessed by the loudness dependence measure of primary auditory cortex evoked activity. Schizophr Res, 64, 115–124. Juckel, G., Gudlowski, Y., Müller, D., Özgürdal, S., Brüne, M., Gallinat, J., et  al. (2008). Loudness dependence of the auditory evoked N1/P2-component as an indicator of serotonergic dysfunction in patients with schizophrenia – a replication study. Psychiatry Res, 158, 79–82. Karlsson, H., Bachmann, S., Schroder, J., McArthur, J., Torrey, E. F., & Yolken, R. H. (2001). Retroviral RNA identified in the cerebrospinal fluids and brains of individuals with schizophrenia. Proc Natl Acad Sci U S A, 98, 4634–4639. Kaufmann, W. (2000). Developmental neurotoxicity. In G. J. Krinke (Ed.), The laboratory rat (pp. 227–250). New York: Academic. Kendell, R. E., & Kemp, I. W. (1989). Maternal influenza in the etiology of schizophrenia. Arch Gen Psychiatry, 46, 878–882.

134

S.H. Fatemi and T.D. Folsom

Koch, M. (1999). The neurobiology of startle. Prog Neurbiol, 59, 107–128. Kunugi, H., Nanko, S., & Takei, N. (1995). Schizophrenia following in utero exposure to the 1957 influenza epidemics in Japan. Am J Psychiatry, 152, 450–452. Lewis, D. A. (2001). Retroviruses and the pathogenesis of schizophrenia. Proc Natl Acad Sci USA, 94, 4293–4294. Lipska, B. K., Jaskiw, G. E., & Weinberger, D. R. (1993). Postpubertal emergency of hyperresponsiveness to stress and to amphetamine after neonatal excitotoxic hippocampal damage: A potential animal model of schizophrenia. Neuropsychopharmacology, 9, 67–75. Lipska, B. K., Khaing, Z. Z., & Weinberger, D. R. (1999). Neonatal hippocampal damage in the rat: A neuristic model of schizophrenia. Psychiatr Ann, 29, 157–160. Machon, R. A., Mednick, S. A., & Schulsinger, F. (1983). The interaction of seasonality, place of birth, genetic risk and subsequent schizophrenia in a high risk sample. Br J Psychiatry, 143, 383–388. McGrath, J. J., Pemberton, M., & Welham, J. L. (1994). Schizophrenia and the influenza epidemics of 1954, 1957 and 1959: A Southern hemisphere study. Schizophr Res, 14, 1–8. Mednick, S. A., Huttunen, M. O., & Macon, R. A. (1994). Prenatal influenza infections and adult schizophrenia. Schizophr Bull, 20, 263–267. Mednick, S. A., Machon, R. A., & Huttunen, M. O. (1988). Adult schizophrenia following prenatal exposure to an influenza epidemic. Arch Gen Psychiatry, 45, 189–192. Menninger, K. A. (1928). The schizophrenic syndromes as a product of acute infectious disease. Arch Neurol Psychiatry, 20, 464–481. Meyer, U., Nyffeler, M., Engler, A., Urwyler, A., Schedlowski, M., Knuesel, I., et al. (2006). The time of prenatal immune challenge determines the specificity of inflammation-mediated brain and behavioral pathology. J Neurosci, 26, 4752–4762. Mirnics, K., & Lewis, D. A. (2001). Genes and subtypes of schizophrenia. Trends Mol Med, 7, 169–174. Mirnics, K., Middleton, F. A., Marquez, A., Lewis, D. A., & Levitt, P. (2000). Molecular characterization of schizophrenia viewed by microarray analysis of gene expression in prefrontal cortex. Neuron, 28, 53–67. Molanova, A., & Blaskovic, D. (1975). The influence of influenza infection of pregnant mice on the development of their fetuses. Acta Virol, 19, 259. Morgane, P. J., Austin-LaFrance, R. J., Bronzino, J. D., Tonkiss, J., & Galler, J. R. (1992). Malnutrition and the developing central nervous system. In R. L. Isaacson & R. F. Jensen (Eds.), The vulnerable brain and environmental risks, volume 1: Malnutrition and hazard assessment (pp. 3–44). New York: Plenum. Nicotera, P., Brune, B., & Bagetta, G. (1997). Nitric oxide: Inducer or suppresser of apoptosis? Trends Pharmacol Sci, 189–190. O’Callaghan, E., Gibson, T., & Colohan, H. A. (1991). Season of birth in schizophrenia: Evidence for confinement of an excess of winter births to patients without a family history of mental disorder. Br J Psychiatry, 158, 764–769. Pallast, E. G., Jongbloet, P. H., & Straatman, H. M. (1994). Excess seasonality of births among patients with schizophrenia and seasonal ovopathy. Schizophr Bull, 20, 269–276. Pulver, A. E., Liang, K. Y., & Wolyniec, P. S. (1992). Season of birth among siblings of schizophrenic patients. Br J Psychiatry, 160, 71–75. Rajkowska, G., Miguel-Hidalgo, J. J., Makkos, Z., Meltzer, H., Overholser, J., & Stockmeier, C. (2002). Layer-specific reductions in GFAP-reactive astroglia in the dorsolateral prefrontal cortex in schizophrenia. Schizophr Res, 57, 127–138. Rami, A., Jansen, S., Giesser, I., & Winckler, J. (2003). Post-ischemic activation of caspase-3 in the rat hippocampus: Evidence of an axonal and dendritic localization. Neurochem Int, 43, 211–223. Rodier, P. M. (1980). Chronology of neuron development: Animal studies and their clinical ­implications. Dev Med Child Neurol, 22, 525–545.

Prenatal Viral Infection in Mouse: An Animal Model of Schizophrenia

135

Romijn, H. J., Hofman, M. A., & Gramsbergen, A. (1991). At what age is the developing cerebral cortex of the rat comparable to that of the full-term newborn human baby? Early Hum Dev, 26, 61–67. Sanjuan, J., Tolosa, A., Gonzalez, J. C., Aguilar, E. J., Perez-Tur, J., Najera, C., et  al. (2006). Association between FOXP2 polymorphisms and schizophrenia with auditory hallucinations. Psychiatr Genet, 16, 67–72. Selten, J. P. C. J., & Sleats, J. P. J. (1994). Evidence against maternal influenza as a risk factor for schizophrenia. Br J Psychiatry, 164, 674–676. Sham, P. C., O’Callaghan, E., & Takei, N. (1992). Schizophrenia following prenatal exposure to influenza epidemics between 1939 and 1960. Br J Psychiatry, 160, 461–466. Shi, L., Fatemi, S. H., Sidwell, R. W., & Patterson, P. H. (2003). Maternal influenza infection causes marked behavioral and pharmacological changes in the offspring. J Neurosci, 23, 297–302. Sodhi, M. S., & Sanders-Bush, E. (2004). Serotonin and brain development. Int Rev Neurobiol, 59, 111–174. Stober, G., Franzek, E., & Beckmann, J. (1992). The role of maternal infectious diseases during pregnancy in the aetiology of schizophrenia in offspring. Eur Psychiatry, 7, 147–152. Su, J., Wang, L., Yuan, L., Reutiman, T. J., Folsom, T. D., Smee, D., et al. (2008). Impairment of synaptic plasticity in offspring of virally infected mouse progeny. Int J Neuropsychopharmacol, 11(S1):264. Susser, E. S., Brown, A. S., & Gorman, J. M. (1999). Prenatal exposures in schizophrenia. Washington, DC: American Psychiatric Press. Susser, E., Lin, S. P., & Brown, A. S. (1994). No relation between risk of schizophrenia and prenatal exposure to influenza in Holland. Am J Psychiatry, 151, 922–924. Susser, E., Neugebauer, R., Hoek, H. W., Brown, A. S., Lin, S., Labovitz, D., et  al. (1997). Schizophrenia after prenatal famine. Arch Gen Psychiatry, 53, 25–31. Swerdlow, N. R., & Geyer, M. A. (1998). Using an animal model of deficient sensorimotor gating to study the pathophysiology and new treatments of schizophrenia. Schizophr Bull, 24, 285–301. Takei, N., Mortensen, P. B., Klaening, U., Murray, R. M., Sham, P. C., O’Callaghan, E., et  al. (1994). Relationship between in utero exposure to influenza epidemic and risk of schizophrenia in Denmark (letter). Schizophr Res, 11, 95. Takei, N., Sham, P. C., & O’Callaghan, E. (1994). Prenatal exposure to influenza and the development of schizophrenia: Is the effect confined to females? Am J Psychiatry, 151, 117–119. Takei, N., Van Os, J., & Murray, R. M. (1995). Maternal exposure to influenza and risk of schizophrenia: A 22-year study from The Netherlands. J Psychiatr Res, 29, 435–445. Tkachev, D., Mimmack, M. L., Ryan, M. M., Wayland, M., Freeman, T., Jones, P. B., et al. (2003). Oligodendrocyte dysfunction in schizophrenia and bipolar disorder. Lancet, 362, 798–805. Toro, C. T., Hallak, J. E., Dunham, J. S., & Deakin, J. F. (2006). Glial fibrillary acidic protein and glutamine synthetase in subregions of prefrontal cortex in schizophrenia and mood disorder. Neurosci Lett, 404, 276–281. Torrey, E. F., Bowler, A. E., Taylor, E. H., & Gottesman, I. I. (1994). Schizophrenia and manic depression disorders: The biological roots of mental illness as revealed by a landmark study of identical twins. New York: Basic Books. Tueting, P., Costa, E., Dwivedi, Y., Guidotti, A., Impagnatiello, F., Manev, R., et al. (1999). The phenotypic characteristics of heterozygous reeler mouse. Neuroreport, 10, 1329–1334. Vawter, M. P., Crook, J. M., Hyde, T. M., Kleinman, J. E., Weinberger, D. R., Becker, K. G., et al. (2002). Microarray analysis of gene expression in the prefrontal cortex in schizophrenia: A preliminary study. Schizophr Res, 58, 11–20. Webster, M. J., O’Grady, J., Kleinman, J. E., & Weickert, C. S. (2005). Glial fibrillary acidic protein mRNA levels in the cingulate cortex of individuals with depression, bipolar disorder and schizophrenia. Neuroscience, 133, 453–461.

136

S.H. Fatemi and T.D. Folsom

Wilson, J. G. (1964). General principles of experimental teratology. In: Congenital Malformations Proceedings of the First International Conference. Philadelphia: Lippincott. Winter, C., Reutiman, T. J., Folsom, T. D., Sohr, R., Wolf, R. J., Juckel, G., et al. (2008). Dopamine and serotonin levels following prenatal viral infection in mouse – implications for psychiatric disorders such as schizophrenia and autism. Eur Neuropsychopharmacol, 18, 712–716. Wright, P., Rakei, N., Rifkin, L., & Murray, R. (1995). Maternal influenza, obstetric complications, and schizophrenia. Am J Psychiatry, 152, 1714–1720. Yamagishi, H., Garg, V., Matsuoka, R., Thomas, T., & Srivastava, D. (1999). A molecular pathway revealing a genetic basis for human cardiac and cranio-facial defects. Science, 283, 1158–1160.

Proteomic Actions of Growth Hormone in the Nervous System Steve Harvey and Marie-Laure Baudet

Abstract  The brain is an endocrine target site for pituitary growth hormone (GH) and an autocrine or paracrine target site for GH produced within nervous tissue. Growth hormone receptor (GHR)-mediated GH actions in the nervous system promote neural growth and differentiation, neuroprotection, neurotransmission, neuroendocrine function and behavior. Growth hormone signaling in the nervous system involves intracellular cascades and changes in gene transcription that often result in proteomic changes in the central and peripheral nervous systems. Neural GH actions mediated through changes in protein synthesis are summarized in this brief review. Keywords  Growth hormone • Neurogenesis • Neuroprotection • Neurotransmission • Neuromodulation • Behavior • Proteins • Proteomics • Brain • Nervous systems • Nerves Abbreviations 5HT AIF BASP-1 BI-1 CIS CNPase CREB-1 COX-2 eNOs Erg-1 ERK

Serotonin Apoptosis-inducing factor Brain abundant membrane attached signal protein-1 Bax inhibitor 1 Cytokine inducible SH2 protein 2¢3¢-Cyclic nucleotide 3¢-phosphohydrolase Cyclic AMP response element binding protein-1 Cycloxygenase-2 Endothelial nitric oxide synthase Early growth response-1 Extracellular signal regulated kinase

S. Harvey (*) Department of Physiology, University of Alberta, 7-41 Medical Sciences Building, Edmonton, Alberta T6G 2H7, Canada e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_5, © Springer Science+Business Media, LLC 2011

137

138

GFAP GGT GHBP Ghitm GHR GHRH IGF-1 IGF-BP3 iNOs JAK2 LDH MAP kinase MBP MEK NAA Ngn1 NMDA nNos NPY NR1 NR2A NR2B ODC PARP1 PI3 kinase PKC PKC inhibitor-1 PSD-95 RGC RNAP SLM-2 SOCS SRE SRIF STAT trk VEGF

S. Harvey and M.-L. Baudet

Glial fibrillary acidic protein Gamma glutamyltranspeptidase Growth hormone binding protein GH-inducible transmembrane protein Growth hormone receptor Growth hormone releasing hormone Insulin-like growth factor-1 IGF-1 binding protein 3 Inducible nitric oxide synthase Janus kinase-2 Lactate dehydrogenase Mitogen-activated protein kinase Myelin basic protein MAP kinase kinase N-acetyl aspartate Neurogenin N-methyl-d-aspartic acid Neuronal nitric oxide synthase Neuropeptide Y NMDA receptor 1 NMDA receptor 2A NMDA receptor 2B Ornithine decarboxylase Poly (ADP-ribose) polymerase 1 Phosphatase inisitol-3 kinase Protein kinase C Protein kinase C inhibitor 1 Postsynaptic density protein-95 Retinal ganglion cell RNA polymerase Sam 68-like mammalian protein-2 Suppressor of cytokine signaling Serum response element Somatostatin Signal transducer and activator of transcription Tyrosine kinase Vascular endothelial growth factor

1 Introduction It is now well established that the brain is a target site for growth hormone (GH) action (see Aberg, Brywe, & Isgaard 2006; Harvey & Hull, 2003; Harvey, Hull, & Fraser, 1993; Nyberg, 2000; Nyberg & Burman, 1996; Scheepens, Moderscheim,

Proteomic Actions of Growth Hormone in the Nervous System

139

& Gluckman, 2005, for reviews). This reflects the widespread presence of GH binding sites (Attardo & Harvey, 1990; Fraser, Attardo, & Harvey, 1990; Mustafa, Adem, Roos, & Nyberg, 1994; Mustafa, Nyberg, et  al., 1994) and GH receptor (GHR) proteins and mRNA (Burton, Kabigting, Clifton, & Steiner, 1992; Fraser et al., 1990; Hasegawa, Minami, Sugihara, & Wakabayashi, 1993; Kastrup, Le Greves, Nyberg, & Blomqvist, 2005; Lai, Emtner, Roos, & Nyberg, 1991; Lai et al., 1993; Lobie et al., 1993; Minami et al., 1993) in the brain, the ability of circulating GH to cross the blood–brain barrier (Coculescu, 1999; Pan et al., 2005; Walsh, Slaby, & Posner, 1987) and the local synthesis of GH in the nervous system (Donahue, Kosik, & Shors, 2006; Gossard, Dihl, Pelletier, Dubois, & Morel, 1987; Martinoli, Ouellet, Rheaume, Pelletier, 1991; Render, Hull, & Harvey, 1995; Sun, Al-Regaiey, et al., 2005; Sun, Evans, et al., 2005; Yoshizato, Fujikawa, Soya, Tanaka, & Nakashima, 1998; Zearfoss, Alarcon, Trifilieff, Kandel, & Richter, 2008). Within the brain, the actions of GH include an increase in neural precursor cells, brain growth, an increase in the number and length of dendrites and axons, an increase in neuron myelination and arborization, and an increase in glial differentiation (e.g., Ajo, Cacicedo, Navarro, & Sanchez-Franco, 2003; Laron & Galatzer, 1985; Scheepens et al., 2005). GH also increases blood flow in the brain (Creyghton, van Dam, & Koppeschaar, 2004) and increases its microvascular density (Sonntag et al., 2000). GH similarly increases the growth of the spinal cord and motoneuron size (Chen, Lund, Burgess, Rudisch, & McIlwain, 1996). GH is also involved in neurotransmitter regulation (Burman et al., 1996; Lea & Harvey, 1993; McGauley, 1996) and the regulation of sleep (Obal, Bodosi, Szilagyi, Kacsoh, & Krueger, 1997; Van Cauter et al., 2004), mood (Schneider, Pagotto, & Stalla, 2003), appetite (Wang, Day, Zhou¸Beard, & Vasilatos-Younken, 2000), cognitive function (Sartorio et  al., 1996; Sonntag et  al., 2000), memory, mental alertness and motivation (Donahue et al., 2002; Frago et  al., 2002; Le Greves et al., 2000; Pavel, Lohmann,Hahn, & Hoffmann, 2003) and locomotor activity (Alvarez & Cacabelos, 1993; Johansson, Winberg, & Bjornsson, 2005). Many of these actions reflect changes in the brain proteome (Table 1), as detailed in this brief review.

2 GH and Neural Action The actions of GH in the nervous system reflect the widespread presence of GHRs in the brain, spinal cord and peripheral nerves (Aberg et al., 2006; Harvey & Hull, 2003). As a result of GH signaling, its receptor is dimerized and autophophorylated, activating a JAK (janus kinase)/STAT (signal transducer and activator of transcription) signaling cascade. This cascade may involve the phosphorylation of a number of JAKs and STATs; for instance, STAT3 phospho-Y205 and JAK2 phospho-y1007/1008 are reduced in the hippocampus of CPEB-1 (cytoplasmic polyadenylation element binding protein-1) knockout mice that have a hippo­campal GH-deficiency, whereas JAK2 phosphorylation is increased in the hippo­ campus of wild-type mice following GH treatment in vitro (Zearfoss et al., 2008).

140

S. Harvey and M.-L. Baudet

Table 1  GH-responsive proteins in the nervous system Protein References Agouti-related protein Bohlooly et al. (2005) Akt Sanders, Parker, and Harvey (2006); Sun, Al-Regaiey, et al. (2005) b adrengenic receptor Popova, Ivanova, et al. (1991); Popova, Robeva, et al. (1991) b endorphin Johansson et al. (1995) b tubulin Ajo et al. (2003) bad Shin et al. (2004) BASP-1 Baudet et al. (2008) bax Shin et al. (2004) Bcl2 Frago et al. (2002) Cabindin Ransome, Goldshmit, Bartlett, Waters, and Turnley (2004) Calretinin Ransome et al. (2004) Calpain Sanders et al. (2006) Caspase 3 Han et al. (2007); Sanders, Parker, Aramburo, and Harvey (2005); Svensson, Bucht, Hallbert, and Nyberg (2008) Caspase 9` Sanders et al. (2005) c-fos Burton , Kabigting, Steiner, & Clifton (1995); Cui, Kwok, & Schwartz (2008); Kamegai, Minami, Sugihara, Higuchi, and Wakabayashi (1994); Kato et al. (2009); Zearfoss et al. (2008) CIS Kasagi, Tokita, Nakata, Imaki, and Minami (2004) CNPase Noguchi, Sugisaki, & Tsukada (1983); Sugisaki, Noguchi, and Tsukada (1985) connexin-43 Aberg et al. (2000) COX-2 Martinez-Canabal, Angoa-Perez, Rugerio-Vargas, Borgonio-Perez, and Rivas-Arancibia (2008) CREB Cui et al., (2008); Sanders, Parker and Harvey (2008); Sun, Al-Regaiey, et al. (2005) Cyclophilin A Baudet et al. (2008) Egr-1 Kato et al. (2009) ERK Aberg et al. (2006); Sanders et al. (2008) GGT Brown-Borg, Rakoczy, and Uthus (2004) Hull and Harvey (1998b) GHBP Ghitm Reimers et al (2007) GHR Ajo et al. (2003); Hull and Harvey (1998a, 1998b); Le Greves et al. (2006) GHRH Chen et al. (1996); Minami et al. (1993); Sato & Frohman (1993) GFAP Ajo et al. (2003) Glutathione peroxidase Brown-Borg et al. (2004); Brown-Borg & Rakoczy (2003) Glutathione Brown-Borg et al. (2004) hippocalcin Noguchi (1996) IGF-1 Johansson et al. (1995); Lopez-Fernandez et al. (1996); Frago et al. (2002) IGF-1 receptor Ajo et al. (2003), Baudet, Sanders, and Harvey (2003); Frago et al. (2002); Le Greves et al. (2006); Lopez-Fernandez et al. (1996) (continued)

Proteomic Actions of Growth Hormone in the Nervous System Table 1  (continued) Protein IGF-BP3 JAK2 Jun B LDH MEK MAP kinase MBP eNOs iNOs nNOs NAA Nestin Ngn1 NPY NR1 NR2A NR2B ODC Opioid receptor proteins PI3 kinase PARP1 PKC inhibitor-1 PSD-95 RNAP I, II 5HT receptor SLM-2 SOCS 2 SOCS 3 SRIF STAT3 STAT5 STAT5 STAT5b Thy-1

141

References Ajo et al. (2003); Johansson et al. (1995) Zearfoss et al. (2008) Kato et al. (2009) Le Greves, Steensland, Le Greves, & Nyberb (2002); Minami et al. (1993); Svensson et al. (2008) Mahmoud and Grover (2006) Mahmoud and Grover (2006) Almazan, Honegger, Matthieu, and Guentert-Lauber (1985) Shin et al. (2004) Kamegai et al. (1994); Shin et al. (2004) Huang, Hu, and Tian (2005) Han et al. (2007); Van Dam et al. (2005) Van Marle, Antony, Silva, Sullivan, and Power (2005) Bertherat et al. (1993); Turnley (2005); Turnley, Faux, Rietz, Coonan, and Bartlett (2002) Bohlooly et al. (2005) Le Greves et al. (2002, 2006) Le Greves et al. (2002, 2006) Le Greves et al. (2002, 2006) Almazan et al. (1985); Yang, Raizada, and Fellows (1981) Persson, Thorlin, and Eriksson (2005) Mahmoud & Grover (2006); Sun, Al-Regaiey, et al. (2005) Sanders et al. (2005) Baudet et al. (2008) Ajo et al. (2003); Le Greves et al. (2006) Berti-Mattera, Gomez, and Krawiec (1983) Popova, Ivanova et al. 1991; Popova, Robeva et al. 1991 Baudet et al. (2008) Turnley et al. (2002) Kasagi et al. (2004) Lopez-Fernandez et al. (1996); Sato & Frohman (1993) Zearfoss et al. (2008) McLenachan, Lum, Waters, Turnley (2009) Kato et al. (2009) Bennett et al. (2005) Noguchi (1996)

GH similarly increases the expression and phosphorylation of STAT5b in somatostatin neurons in the periventricular nucleus of the rat brain (Bennett et al., 2005). The increased GH concentration in the hippocampus of pituitary GH-deficient Ames dwarf mice is similarly thought to account for their increased activation of P13 kinase and the phosphorylation of Akt and cyclic AMP response element binding protein (CREB) (Sun, Al-Regaiey, et al., 2005). In CA1 pyramidal neurons of the rat hippocampus, GH action has been correlated with the activation of janus kinase 2 (JAK2), phosphatidylinositol (PI) 3-kinase, and mitogen-activated protein

142

S. Harvey and M.-L. Baudet

(MAP) kinase kinase (MEK) (Mahmoud & Grover, 2006). Likewise, GH action in the hippocampus results in increased phosphorylation of STAT5 and the expression of the SRE (serum response element) genes, Egr-l (early growth response-1), Fos, and Jun-B (Kato et al., 2009). As a result of GH signaling in the CNS, autoregulatory responses in GHR expression occur. For instance, the abundance of the GHR transcript in the chicken brain (whole brain and hypothalamus) is rapidly downregulated within 2 h of intravenous or intracerebroventricular GH administration (Hull & Harvey, 1998a). Chronic GH treatment (daily injections for 6 days) similarly downregulates the abundance of GHR transcripts in the actuate nucleus of the rat hypothalamus (Bennett, Levy, Sophokleous, Robinson, & Lightman, 1995). A single bolus injection of GH was, however, found to increase (by 25–30%) the GHR transcripts in the rat hypothalamus, in which GH binding protein (GHBP) mRNA was also increased, although the ratio of the GHR:GHBP transcripts was decreased (Hull & Harvey 1998b). Indeed, the GHR and GHBP transcripts in the brain are thought to be regulated differentially to those in peripheral tissues (Lobie, Zhu, Graichen, & Goh, 2000). The actions of GH in the CNS may be direct or mediated by other GH-responsive genes. For instance, GH-response gene (GHRG)-1 is a specific marker of GH action in chickens, in which it is widely distributed in the brain (Harvey, Johnson, & Sanders, 2001; Harvey, Lavelin, & Pines, 2002). Insulin-like growth factor (IGF)-1 is similarly increased by GH in most species in the peripheral nervous system (Kanje, Skottner, & Lundborg, 1988) and in the CNS (Frago et al., 2002; Lopez-Fernandez et al., 1996) and is thought to mediate many of the actions of GH in neural function (Aberg et al., 2000).

3 GH and Neural Protein Synthesis It is well established that GH increases the content of DNA and RNA in the brain (Noguchi, 1996; Noguchi, Sugisaki, & Tsukada 1982; Noguchi, Sugisaki, Watanabe, et al., 1982). Indeed, exogenous GH increases the nuclear transcription rate (Kato, 2002), the activity of RNA polymerase I and II and the rate of RNA synthesis (BertiMattera et al. 1983). GH similarly increases the rate of protein synthesis (g protein synthesized/g RNA) in the brain (Ohsumi et  al., 2007; Ohsumi, Tujioka, Hayase, Nagata, & Yokogoshi, 2008). This increased rate of protein synthesis is indicated by the increased accumulation of c-fos (a marker of neural activity and protein synthesis) in the brain after GH treatment (Burton et al. 1995; Kamegai et  al., 1994; Minami, Kamegai, Sugihara, Hasegawa, & Wakabayashi, 1992). Conversely, in the absence of GH or GH signaling, hippocampal c-fos levels are reduced (Zearfoss et al., 2008). GH-induced brain protein synthesis is thought to be of physiological significance and is thought to reflect GH-induced sleep, since one function of sleep is to facilitate brain protein synthesis (Methippara et  al., 2008; Van Cauter & Copinschi, 2000).

Proteomic Actions of Growth Hormone in the Nervous System

143

4 GH and Neural Growth and Differentiation Trophic actions of GH in neural development were among the first neural actions of GH to be discovered. Yang et al. (1981) used ornithine decarboxylase (ODC) activity as a marker of cell growth and differentiation and found it to be increased in the brain after GH treatment, as did Almazan et al. (1985). CNPase (2¢3¢-cyclic nucleotide 3¢-phosphohydrolase) activity, a myelin-marker enzyme, was similarly increased in the brains of GH-treated postnatal mice and correlated with increased brain weight, cerebral DNA content, spontaneous locomotor activity and increased concentrations of myelin basic protein (MBP) (Noguchi, Sugisaki, Watanabe, et al., 1982; Sarlieve, Bouchon, Koehl, & Neskovic, 1983; Sugisaki et al., 1985). Exogenous GH similarly increased cerebral CNPase activity in hydrocortisoneintoxicated rats and increased cerebral thymidine kinase (TK) activity and DNA content (Noguchi, Sugisaki, & Tsukada, 1982). These actions of GH are likely direct, since GH also increased ODC, CNP and MBP in brain cell aggregates in vitro (Almazan et al., 1985). GH also increased the activity of neuron-specific enolase (a glycolytic isoenzyme) in the brains of GH-deficient mice (Noguchi, 1996; Noguchi, Sugisaki, & Tsukada, 1982) and the neuronal content of ganglioside proteins (Noguchi, 1996; Noguchi & Sugisaki 1986). The GH-induced promotion of oligodendrocyte development and myelination in the CNS of fetal rats occurs as a result of the activation of a MAP kinase (Aberg et al., 2006; Scheepens et al., 2001). The GH-induced proliferation and differentiation of rat embryonic cerebrocortical cells is similarly associated with increased extracellular signal-regulated kinase (ERK) activity. GH-deficient mice are also deficient in the Thy-1 antigen and hippocalcin (Noguchi, 1996), suggesting poor synaptogenesis, although this may reflect their hypothyroid state as much as their GH deficiency. Transgenic mice that are GHR deficient (−/−) also have sparser dendritic branching in the cerebral cortex than GHR (+/+) mice, further suggesting GH involvement in synaptic function. This possibility is supported by the stimulatory effect of exogenous GH on the expression of the genes for the N-methyl-d-aspartate (NMDA) receptor subunits NR1 and NR2 (Le Greves et al., 2002; Le Greves et al., 2006). GH-induced synaptic function is also indicated by the increased expression of post-synaptic density protein (PSD)-95 after GH treatment (Le Greves et al., 2006). The differentiation of neural stem cells into neurons or astrocytes occurs during early embryonic development in response to interactions between intrinsic and environmental cues (Turnley et al., 2002). The role of GH in fetal CNS maturation is still poorly understood, but GH is thought to be a factor involved in inhibiting the differentiation of neural progenitor cells, by a mechanism blocked by SOCS (suppressor of cytokine signaling) proteins. Indeed, exogenous GH inhibits neuronal differentiation in mouse brain progenitor cells, in which the immunoneutralization of endogenous GH or an increase in the expression of the SOCS-2 gene increases neural differentiation (Turnley 2005; Turnley et al., 2002). SOCS-2 inhibition of GH action appears to be mediated through its inhibition of STAT5 activity downstream of GH signaling,

144

S. Harvey and M.-L. Baudet

blocking the inhibitory effect of GH on the expression of neurogenin (Ngn1), a neurogenic transcription factor. As this action of GH on Ngn1 production inhibits neurogenesis, this may account for the increased density of cortical projecting neurons in the brains of GHR (−/−) mice, in comparison with their wild-type counterparts and consequent increased levels of the neural markers calbindin and calretinin (Ransome et al., 2004). GH similarly delays the differentiation of neural progenitor cells derived from the subventricular zone of the adult mouse forebrain (Turnley et al., 2002), although it promotes the proliferation of adult neurosphere cultures (McLenachan et al., 2009), in which it increases STAT5 phosphorylation. In contrast to its role in inhibiting neural differentiation, GH may be involved in astrocyte proliferation. For instance, exogenous GH promotes the proliferation of fetal rat cerebral cortical cells in primary culture, as assessed by cell counting, PCNA analysis and the incorporation of [3H]-thymidine and BrdU (Ajo et  al., 2003). This effect was observed at physiological GH levels (0.5 and 5  ng/ml), although higher doses (50 and 500 ng/ml) were less effective, probably as a result of receptor saturation and downregulation. The phenotype of the proliferating cells in early embryogenesis was initially of cells expressing neuronal nestin, but was of astrocytic b-tubulin expressing cells in later embryogenesis. The observation that GH induces astrocyte proliferation was confirmed by the induction of GFAP (glial fibrillary acidic protein) by GH treatment (Ajo et  al., 2003). The fact that these neurotrophic actions of GH were blocked by IGF-1 antiserum strongly suggests that they were mediated by IGF-1, the concentration of which was also increased by GH treatment. GH-induced neurotrophic actions mediated by IGF-1 may also have been augmented by the GH-induced increase in IGF-binding protein 3 (IGF-BP3), IGF-1 receptor protein and its phosphorylation (Ajo et al., 2003). The importance of GH in neural development is also demonstrated by the small brain size in GHR (−/−) mice (Ransome et al., 2004). This decrease in brain size is not due to neuronal loss, although it does reflect a decrease in soma size in some brain regions and a paucity of astrocytes (in both number and size) in others. These defects might also be due to decreased IGF-1 expression, although the increased neural:glia ratio reflects decreased GFAP expression and poor astrocyte development. The reduced dendritic branching in these mice may also reflect the smaller size of their calbindin- and calretinin-expressing cells, as calbindin and calretinin are associated with synapse formation and appear to be regulated by GH in a similar fashion to that induced by other neurotrophic factors (Fiumelli, Kiraly, Ambrus, Magistretti, & Martin, 2000). In addition to synapses, communication in the brain is facilitated by the formation of gap junctions. These are formed by two hexameric protein complexes, one from each cell membrane, that form aqueous pores, allowing ions and small molecules to move from one cell (neuron or glia) to another. These proteins are connexins, of which connexin-43 is the most abundant. The abundance of connexin-43 mRNA and protein is increased in the cerebral cortex and hypothalamus of rats after systemic treatment (1 mg/kg/day/19 days, s.c.) with bovine GH, although not in the brainstem or hippocampus (Aberg et al., 2000). This effect was due to direct GH action, since similar treatment with IGF-1 had no effect on connexin-43 mRNA or protein levels.

Proteomic Actions of Growth Hormone in the Nervous System

145

Other GH-responsive proteins in the nervous system that have roles in neural proliferation, neurite outgrowth and vascularization include brain abundant membrane attached signal protein-1 (BASP-1), protein-kinase C (PKC) inhibitor 1 (PKC inhibitor-1), cyclophilin transduction-associated protein (a member of the signal transduction and activation of RNA family of proteins), which is alternatively called Sam68-like mammalian protein 2 (SLM-2). The abundance of these three proteins differed in the retinas of GHR (−/−) dwarf mice in comparison with their abundance in the retinas of wild-type mice. In the absence of GH-signaling, BASP-1 was downregulated, whereas PKC inhibitor-1, cyclophilin / SLM-2 were upregulated. A suppression of PKC signaling abolishes ATP-induced proliferation in the embryonic chick retina (Sanders, Baudet, Parker, & Harvey, 2009). GH may thus promote retinal cell proliferation in wildtype mice by suppressing PKC inhibitor-1 levels, especially as the neuroblastic layer of retina (composed of mitotic cells) is reduced in size in GHR (−/−) animals. Retinal vascularization is also initiated in newborn mice by the development of the superficial vascular plexus and PKC inhibition prevents the growth of retinal vascular endothelial cells (Cai, Rook, Jiang, Takahara, & Aiello, 2000). It is thus possible that GH induces retinal angiogenesis through its suppression of PKC inhibitor-1 expression. This effect might be induced by locally produced GH, as GH and GHR are found within the retinal blood vessels of the superficial vascular plexus (Baudet et al., 2008). PKC inhibitor activity also reduces laminin-induced axon outgrowth of amacrine cells in neonatal rat neural retina cells in  vitro (Politi, Insua, & Buzzi, 1998) suggesting GH involvement in neuron development. GH has also been suggested to increase neurite length and number in fetal rat cortical cell culture (Ajo et  al., 2003), to promote axon outgrowth in VSC4.1 cells (Lyuh et al., 2007), to induce penile nerve fiber growth and carvenous nerve and sciatic nerve regeneration in adult rats (Cui, Zhang, Pei, Wei, & Hu, 2006; Kanje et al., 1988). and to induce outgrowths from embryonic chick retinal ganglion cells (RGCs) in vitro (Baudet, Rattray, Martin, & Harvey, 2009). It is therefore possible that GH signaling might induce axon growth in the mouse retina via PKC inhibitor-1 downregulation. Cyclophilin A, the cytosolic isoform of the cyclophilin family of proteins, levels were reduced in the retinas of GHR (−/−) mice. This is of interest, as a cyclophilin A defect has been correlated with a loss of visual function in mice (Muramatsu & Miyauchi, 2003; Philp, Ochrietor, Rudoy, Muramatsu, & Linser, 2003) and cyclophilin A expression in the visual cortex is reduced following retinal lesions (Arckens et al., 2003). Exogenous cyclophilin A promotes proliferation, migration, invasive capacity and tubulogenesis of endothelial cells at low concentrations, whereas at high concentrations it decreases endothelial cell migration and viability (Kim, Lessner, Sakurai, & Galis, 2004). GH signaling might thus mediate retinal vascularization, by preserving low levels of cyclophilin in the neural retina of newborn wild-type mice. SLM-2 regulates the expression of VEGF (vascular endothelial growth factor) and of tau (Arikan et  al., 2002; Li, Arikan, & Andreadis, 2003), two proteins involved, respectively, in the ontogeny of the superficial retinal vasculature and of

146

S. Harvey and M.-L. Baudet

axon growth (Andreadis, 2005; Dorrell & Friedlander, 2006). It is thus possible that GH regulates the expression of VEGF and tau and thus indirectly regulates retinal vascularization and axon growth. BASP proteins are amongst the most abundant genes expressed in the mouse neural retina from embryonic day 13.5 of development (Biffo et al., 1990; Ivanov, Dvoriantchikova, Nathanson, McKinnon, & Shestopalov, 2006). These proteins are well-established axon growth promoters and neuronal pathfinding modulators (Mosevitsky, 2005). The downregulation of BASP-1 in GHR (−/−) mice may thus be responsible, in part, for the decreased width of the inner plexiform layer and optic fiber layer in these animals (Baudet et al., 2008).

5 GH and Neuroprotection GH promotes cell survival by anti-apoptotic actions in many tissues, including the brain. Specifically, GH inhibits neuronal death induced by perinatal hypoxic-ischemia injury (Aberg et al., 2006; Gustafson, Hagberg, Bengtsson, Brantsing, & Isgaard, 1999; Isgaard, Aberg, & Nilsson, 2007; Scheepens et al., 1999; Scheepens, Williams, Breier, Guan, & Gluckman, 2000; Scheepens et al., 2001) or following pilocarpine-induced epilepsy (Yu, Song, & An, 2001). Exogenous GH has also been shown to be neuroprotective in the spinal cord. Repeated topical application of GH before and after spinal cord injury was found to preserve the amplitude of spinal cord potentials and to reduce the formation of edema and the cellular changes that normally accompany spinal cord injury (Nyberg & Sharma, 2002; Winkler et  al., 2000). Exogenous GH similarly protects the spinal cord against radiationinduced damage (Isla et al., 2007). The neuroprotective mechanism of GH action is not completely known but it is probably mediated by the inhibition of caspase activity (Han et al., 2007; Tamatani, Ogawa, & Tohyama, 1998). For instance, exogenous GH is neuroprotective against opioid-induced apoptosis in hippocampal neurons by reducing the activity of caspase-3 and by reducing lactate dehydrogenase (LDH) release (Svensson et al., 2008). GH-induced neuroprotection has also been correlated with decreased levels of the apoptotic proteins bad and bax (Shin et al., 2004) and with reduced levels of inducible nitric oxide synthase (iNOS) and endothelial NOS (eNOS), which would reduce the neurotoxic effect of released nitric oxide (Shin et  al., 2004). The importance of GH in CNS neuroprotection is also indicated by the ubiquitous distribution of GH-inducible transmembrane protein (Ghitm) throughout the brain (Reimers, Choi, Bucan, & Vogt, 2007; Yoshida, Nagata, & Kataoka, 2006) This protein is a member of the bax inhibitor-1 (BI-1) family of cyto-protective proteins that inhibit the pro-apoptotic actions of bax and is dependent upon GH for its expression. The neuroprotective action of GH may also be mediated, in part, by its induction of IGF-1 expression, since IGF-1 is a well-established neuroprotective agent (D’Ercole et al., 1996; Guan & Gluckman 2009; Isgaard et al., 2007)

Proteomic Actions of Growth Hormone in the Nervous System

147

and is GH inducible. Frago et  al. (2002), for instance, demonstrated that IGF-1 mRNA levels were increased in the hypothalamus cerebellum and hippocampus of GH-treated rats, together with an increase in the intracellular Pl3 kinase pathways involved in cell survival, and an increase in the activity of the antiapoptotic protein Bcl-1. GH-induced neuroprotection against oxidative damage caused by ozone exposure has also been shown to be mediated by an increased activity of the enzyme cyclo-oxygenase-2 (COX-2) (Martinez-Canabal et al., 2008). GH-induced antioxidative defense in the brain is also accompanied by a suppression of glutathione (GSH) degradation (Brown-Borg et al., 2004) and by increased brain GSH levels (Brown-Borg & Rakoczy, 2003). The level of N-acetyl aspartate (NAA) (a marker of neuronal density and integrity), is also increased in the brains of GH-treated hypoxic-ischemic rats (Han et al., 2007). Cell death in the nervous system is also induced by expression of the HIV-1 protein in neural stem cells and differentiated neurons. Exogenous GH is, however, neuroprotective and prevents the neuronal injury caused by HIV-1, and, moreover, improves neurocognitive performance (Silva et al., 2003; Van Marle et al., 2005). GH-induced neuroprotection in response to HIV-1 infection is again likely to be mediated by an increase in IGF-1 expression and by an increase in the expression of nestin, a marker of neuronal health. GH-induced neuroprotection has also been documented in the neural retina of chick embryos, in which endogenous GH promotes the survival of RGCs in an autocrine/paracrine way (Harvey, Baudet, & Sanders, 2009). This antiapoptotic action involves a suppression of caspase-3 expression and the expression of AIF (apoptosis-inducing factor), which acts through caspase-independent death pathways (Harvey, Baudet, & Sanders, 2006). The physiological importance of this action is demonstrated after the immunoneutralization of endogenous GH in purified RGCs, which markedly increases RGC death in vitro and in vivo (Sanders et al., 2005), as a result of increased caspase-3 and caspase-9 activation and increased PARP-1 cleavage (Sanders et al., 2006, 2008). Calpain activation by GH also causes PARP-1 cleavage via a caspase-independent pathway and a specific calpain inhibitor abrogates the apoptotic activity of GH antiserum on RGC death (Sanders et al., 2006). The Akt signaling system also participates in GH-induced RGC neuroprotection, since GH treatment of immunopanned RGCs reduces Akt levels whilst concomitantly raising the level of phosphorylated Akt (Akt-phos) (Sanders et al., 2006). GH-induced neuroprotection also involves an activation of cytosolic tyrosine kinases (Trks) and extracellular signal regulated kinases (ERKs) and the activation of CREB (cAMP response element binding protein), which initiates the transcription of pro- or anti-apoptotic genes (Sanders et al., 2008). As observed in the rat brain, these neuroprotective actions of GH are likely to be mediated, in large part, through the actions of IGF-1, since GH induces IGF-1 expression in the retina (Baudet et al., 2003) and because the simultaneous immunoneutralization of GH and IGF-1 does not increase the level of cell death in RGC cultures above that achieved by immunoneutralization of GH alone (Sanders et al., 2009).

148

S. Harvey and M.-L. Baudet

6 GH and Neurotransmission It is well established that GH has neuromodulatory actions in discrete brain regions (Harvey, Minami, Sugihara, & Wakabayashi, 1993). For instance, in the hypothalamus, exogenous GH increases the affinity of b-adrenergic receptors (Popova, Ivanova, et al., 1991), promoting somatostatin (SRIF) release (Rettori et al., 1990), but it reduces the affinity of these receptors in the cerebral cortex. The number and affinity of serotonin (5HT) receptors in the rat hypothalamus, but not in the cerebral cortex, are also reduced by exogenous GH (Popova, Ivanova, et  al., 1991), suppressing SRIF release (Richardson, Hollander, Prasad, & Hirooka, 1981). GH treatment also promotes a decrease in the affinity of hypothalamic muscarinic receptors (Popova et al., 1990), which might promote SRIF release (Torsello et al., 1988). A role for GH in cholinergic transmission is also indicated by the deficiency of choline acetyltransferase (ChAT) activity in the hippocampus, olfactory tubules and striatum of GH-deficient Snell dwarf mice, especially as normal ChAT activity is restored by exogenous GH (Fuhrmann, Durkin, Thiriet, Kempf, & Ebel, 1985, Fuhrmann, Kempf, & Ebel, 1986). The excitatory amino acid aspartate is also increased in concentrations following GH administration in man (Burman et  al., 1996). Since the activation of NMDA receptors inhibits dopamine release, increased aspartate concentrations may provide a mechanism for GH inhibition of dopamine turnover. The induction of NMDA receptor subunits NR1, NR2A, NR2B in the hippocampus of GH-treated rodents (Le Greves et al., 2002, 2006; Svensson et al., 2008) also suggests actions of GH in NMDA transmission. This possibility is supported by the stimulatory action of exogenous GH on AMPA- and NMDA-receptormediated excitatory postsynaptic potentials (EPSPs) in hippocampal area CA1 of the rat brain (Mahmoud & Grover, 2006). It is also supported by the reduced cognitive function, reduced memory and attention in GH-deficient subjects, in which brain NAA concentrations are reduced (Van Dam et al., 2005). Opioid signaling may also be increased by GH action, since b-endorphin levels in cerebrospinal fluid are increased after GH treatment (Johansson et  al., 1995), which also increases the abundance of opioid receptor proteins (Persson et al., 2005).

7 GH and Neuroendocrine Function GH autoregulates its own secretion from the pituitary by actions at central sites. This short-loop negative feedback pathway involves a decrease in the hypothalamic synthesis and release of GH-releasing hormone (GHRH) (Bertherat et al., 1993; Burton et al., 1992, 1995; Kamegai, Unterman, Frohman, & Kineman, 1998; Pellegrini et al., 1996, 1997) and an increase in somatostatin (SRIF) synthesis and release (Aguila & McCann, 1993; Sato & Frohman, 1993; Kamegai et al., 1994; Lopez-Fernandez et al., 1996). The increased secretion of SRIF is largely mediated by actions of GH on neuropeptide Y (NPY) neurons and increased NPY gene expression (Kamegai, Minami, Sugihara.,

Proteomic Actions of Growth Hormone in the Nervous System

149

Higuchi, & Wakabayashi, 1996; Chan, Steiner, & Clifton, 1996; Minami, Kamegai, Sugihara, Suzuki, & Wakabayashi, 1998). This is a direct action of GH and is not mediated though the induction of IGF-1 (Minami et al., 1997). The autoregulatory action of GH on SRIF neurons occurs through an increased expression of STAT5b (Bennett et al., 2005). Actions of GH in hypothalamic neurons also include an increased expression of the genes for SOCS3 and CIS (cytokine inducible SH2 protein), which provides another mechanism to attenuate GH action (Kasagi et al., 2004).

8 GH and Behavior GH-induced protein synthesis in the brain has been associated with some behavioral responses. Sleep, for instance, is induced by GH and this promotes the synthesis of proteins that are critically involved in the regulation of both rapid eye movement (REM) sleep and non-REM sleep (Methippara et  al., 2008). Another behavioral response to GH is the stimulation of food intake, by promoting the hypothalamic expression of agouti-related protein and NPY (Bohlooly et  al., 2005). GH also improves cognitive function and memory by inducing the hippocampal expression of the NMDA receptor gene (Deijen, de Boer, & van der Veen, 1998; Le Greves et al., 2002, 2006; Ohsumi et al., 2008; Schneider-Rivas, Rivas-Arancibia, VazquezPereyra, Vazquez-Sandoval, & Borgonio-Perez, 1995; Schneider-Rivas et al., 2007) and the expression of PSD-95 (Le Greves et  al., 2006). GH also improves wellbeing and the quality of life (Hull & Harvey, 2003) reflecting the induction by GH of opioid receptor proteins in the brain (Persson et al., 2005). GH has also recently been found to be epileptogenic, inducing epilepsy progression in the hippocampus by activating STAT5 and increasing the expression of the SRE-regulated genes, Erg-1, fos and the Jun B oncogene (Kato et al., 2009). In fish, GH stimulates swimming behavior, probably by suppressing hypothalamic dopamine activity (Jonsson et al., 2003).

8.1 Summary The nervous system is a target site for GH action. These actions result in changes in the proteome that promote neurogenesis, neuroprotection, neurotransmission, neuroendocrine function and behavior. These actions are receptor-mediated and may be direct or indirectly mediated through other growth factors. Finally, these actions may be induced by pituitary GH after its uptake through the blood–brain barrier or reflect autocrine or paracrine actions of GH produced locally in the brain. Acknowledgment  Supported by the National Scientific Research Council of Canada.

150

S. Harvey and M.-L. Baudet

References Aberg, N. D., Brywe, K. G., & Isgaard, J. (2006). Aspects of growth hormone and insulin-like growth factor-I related to neuroprotection, regeneration, and functional plasticity in the adult brain. The Scientific World Journal, 6, 53–80. Aberg, N. D., Carlsson, B., Rosengren, L., Oscarsson, J., Isaksson, O. G. P., Ronnback, L., et al. (2000). Growth hormone increases connexin-43 expression in the cerebral cortex and hypothalamus. Endocrinology, 141, 3879–3886. Aguila, M. C., & McCann, M. (1993). Growth hormone increases somatostatin release and messenger ribonucleic acid levels in the rat hypothalamus. Brain Research, 623, 89–94. Ajo, R., Cacicedo, L., Navarro, C., & Sanchez-Franco, F. (2003). Growth hormone action on proliferation and differentiation of cerebral cortical cells from fetal rat. Endocrinology, 144, 1086–1097. Almazan, G., Honegger, P., Matthieu, J.-M., & Guentert-Lauber, B. (1985). Epidermal growth factor and bovine growth hormone stimulate differentiation and myelination of brain cell aggregates in culture. Developmental Brain Research, 21, 257–264. Alvarez, X. A., & Cacabelos, R. (1993). Influence of growth hormone GH and GH-releasing factor on locomotor activity in rats. Peptides, 14, 707–712. Andreadis, A. (2005). Tau gene alternative splicing: expression patterns, regulation and modulation of function in normal brain and neurodegenerative diseases. Biochimica et Biophysica Acta, 1739, 91–103. Arckens, L., Van der Gucht, E., Van den Bergh, G., Massie, A., et al. (2003). Differential display implicates cyclophilin A in adult cortical plasticity. The European Journal of Neuroscience, 18, 61–75. Arikan, M. C., Memmott, J., Broderick, J. A., Lafyatis, R., et  al. (2002). Modulation of the membrane-binding projection domain of tau protein: splicing regulation of exon 3. Brain Research. Molecular Brain Research, 101, 109–121. Attardo, D., & Harvey, S. (1990). Growth hormone-binding sites in chicken hypothalamus. Journal of Molecular Endocrinology, 4, 23–29. Baudet, M. L., Hassanali, Z., Sawicki, G., List, E. O., Kopchick, J. J., & Harvey, S. (2008). Growth hormone action in the developing neural retina: A proteomic analysis. Proteomics, 8, 389–401. Baudet, M. L., Rattray, D., Martin, B. T., & Harvey, S. (2009). Growth hormone promotes axon growth in the developing nervous system. Endocrinology, 150, 2758–2766. Baudet, M. L., Sanders, E. J., & Harvey, S. (2003). Retinal growth hormone in the chick embryo. Endocrinology, 144, 5459–5468. Bennett, P. A., Levy, A., Sophokleous, S., Robinson, I. C. A. F., & Lightman, S. L. (1995). Hypothalamic GH receptor gene expression in the rat: Effects of altered GH status. The Journal of Endocrinology, 147, 225–234. Bennett, E., McGuinness, L., Gevers, E. F., Thomas, G. B., Robinson, I. C., Davey, H. W., et al. (2005). Hypothalamic STAT proteins: regulation of somatostatin neurones by growth hormone via STAT5b. Journal of Neuroendocrinology, 17, 186–194. Bertherat, J., Timsit, J., Bluet-Pajot, M. T., Mercadier, J. J., Gourdji, D., Kordon, C., et al. (1993). Chronic growth hormone GH hypersecretion induces reciprocal and reversible changes in mRNA levels from hypothalamic GH-releasing hormone and somatostatin neurons in the rat. The Journal of Clinical Investigation, 91, 1783–1791. Berti-Mattera, L. N., Gomez, C. J., & Krawiec, L. (1983). Effects of bovine growth hormone on RNA synthesis in brain and liver of neonatal hypothyroid rats. Hormone and Metabolic Research, 15, 286–289. Biffo, S., Verhaagen, J., Schrama, L. H., Schotman, P., et al. (1990). B-50/GAP43 expression correlates with process outgrowth in the embryonic mouse nervous system. The European Journal of Neuroscience, 2, 487–499.

Proteomic Actions of Growth Hormone in the Nervous System

151

Bohlooly, M., Olsson, B., Bruder, C. E., Lindent, D., Sjogren, K., Bjursell, M., et  al. (2005). Growth hormone overexpression in the central nervous system results in hyperphagia-induced obesity associated with insulin resistance and dyslipidemia. Diabetes, 54, 51–62. Brown-Borg, H. M., & Rakoczy, S. G. (2003). Growth hormone administration to long-living dwarf mice alters multiple components of the antioxidative defense system. Mechanicsms of Ageing and Development, 124, 1013–1024. Brown-Borg, H. M., Rakoczy, S. G., & Uthus, E. O. (2004). Growth hormone alters components of the glutathione metabolic pathway in Ames dwarf mice. Annals of the New York Academy of Sciences, 1019, 317–320. Burman, P., Hetta, J., Wide, L., Mansson, J.-E., Ekman, R., & Karlsson, F. A. (1996). Growth hormone treatment affects brain neurotransmitters and thyroxine. Clinical Endocrinology, 44, 319–324. Burton, K. A., Kabigting, E. B., Clifton, D. K., & Steiner, R. A. (1992). Growth hormone receptor messenger ribonucleic acid distribution in the adult male rat brain and its colocalization in hypothalamic somatostatin neurons. Endocrinology, 130, 958–963. Burton, K. A., Kabigting, E. B., Steiner, R. A., & Clifton, D. K. (1995). Identification of target cells for growth hormone’s action in the arcuate nucleus. The American Journal of Physiology, 269, E716–E722. Cai, W., Rook, S. L., Jiang, Z. Y., Takahara, N., & Aiello, L. P. (2000). Mechanisms of hepatocyte growth factor-induced retinal endothelial cell migration and growth. Investigative Ophthalmology & Visual Science, 41, 1885–1893. Chan, Y. Y., Steiner, R. A., & Clifton, D. K. (1996). Regulation of hypothalamic neuropeptide-Y neurons by growth hormone in the rat. Endocrinology, 137, 1319–1325. Chen, L., Lund, P. K., Burgess, S. B., Rudisch, B. E., & McIlwain, D. L. (1996). Growth hormone, insulin-like growth factor I, and motoneuron size. Journal of Neurobiology, 32, 202–212. Coculescu, M. (1999). Blood-brain barrier for human growth hormone and insulin-like growth factor-I. Journal of Pediatric Endocrinology & Metabolism, 12, 113–124. Creyghton, W. M., van Dam, P. S., & Koppeschaar, H. P. (2004). The role of the somatotropic system in cognition and other cerebral function. Seminars in Vascular Medicine, 4, 167–172. Cui, T. X., Kwok, R., & Schwartz, J. (2008). Cooperative regulation of endogenous cAMPresponse element binding protein and CCATT/enhancer-binding protein beta in GH-stimulated c-fos expression. The Journal of Endocrinology, 196, 89–100. Cui, D. S., Zhang, K. L., Pei, F., Wei, S. Z., & Hu, L. Q. (2006). Growth hormone enhances regeneration of cavernous nerves after their transplantation in rats. Zhonghua Nan Ke Xue, 12(784–786), 790. D’Ercole, A. J., Ye, P., Calikoglu, A. S., & Gutierrez-Ospina, G. (1996). The role of the insulinlike growth factors in the central nervous system. Molecular Neurobiology, 13, 227–255. Deijen, J. B., de Boer, H., & van der Veen, E. A. (1998). Cognitive changes during growth hormone replacement in adult mem. Psychoneuroendocrinology, 23, 45–55. Donahue, C. P., Jensen, R. V., Ochiishi, T., Eisenstein, I., Zhao, M., Shors, T., et  al. (2002). Transcriptional profiling reveals regulated genes in the hippocampus during memory formation. Hippocampus, 12, 821–833. Donahue, C. P., Kosik, K. S., & Shors, T. (2006). Growth hormone is produced within the hippocampus where it responds to age, sex, and stress. Proceedings of the National Academy of Sciences of the United States of America, 103, 6031–6036. Dorrell, M. I., & Friedlander, M. (2006). Mechanisms of endothelial cell guidance and vascular patterning in the developing mouse retina. Progress in Retinal and Eye Research, 25, 277–295. Fiumelli, H., Kiraly, M., Ambrus, A., Magistretti, P. J., & Martin, J. L. (2000). Opposite regulation of calbindin and calretinin expression by brain-derviced neurotrophic factor in cortical neurons. Journal of Neurochemistry, 74, 1870–1877. Frago, L. M., Paneda, C., Dickson, S. L., Hewson, A. K., Argente, J., & Chowen, J. A. (2002). Growth hormone GH and GH-releasing peptide-6 increases brain insulin-like growth factor-I

152

S. Harvey and M.-L. Baudet

expression and activate intracellular signaling pathways involved in neuroprotection. Endocrinology, 143, 4113–4132. Fraser, R. A., Attardo, D., & Harvey. (1990). Growth hormone receptors in hypothalamic and extra-hypothalamic tissues. Journal of Molecular Endocrinology, 5, 231–238. Fuhrmann, G., Durkin, T., Thiriet, G., Kempf, E., & Ebel, A. (1985). Cholinergic neurotransmission in the central nervous system of the Snell dwarf mouse. Journal of Neuroscience Research, 13, 417–430. Fuhrmann, G., Kempf, E., & Ebel, A. (1986). Effects of hormone therapy on the central cholinergic neurotransmission of the Snell dwarf mouse. Journal of Neuroscience Research, 16, 527–539. Gossard, F., Dihl, F., Pelletier, G., Dubois, P. M., & Morel, G. (1987). In situ hybridization to rat brain and pituitary gland of growth hormone cDNA. Neuroscience Letters, 79, 251–256. Guan, J., & Gluckman, P. D. (2009). IGF-1 derived small neuropeptides and analogues: a novel strategy for the development of pharmaceuticals for neurological conditions. British Journal of Pharmacology, 157, 881–889. Gustafson, K., Hagberg, H., Bengtsson, B. A., Brantsing, C., & Isgaard, J. (1999). Possible protective role of growth hormone in hypoxia-ischemia in neonatal rats. Pediatric Research, 45, 318–323. Han, T. R., Chun, M. H., Jang, D. H., Kim, K.-S., Lim, K. H., & Cho, H. J. (2007). Neuroprotective effects of growth hormone against hypoxic-ischemic brain injury in neonatal rats: 1H magnetic resonance spectroscopic study. Journal of Korean Medical Science, 22, 122–126. Harvey, S., Baudet, M. L., & Sanders, E. J. (2006). Growth hormone and cell survival in the neural retina: caspase dependence and independence. NeuroReport, 17, 1715–1718. Harvey, S., Baudet, M. L., & Sanders, E. J. (2009). Growth hormone-induced neuroprotection in the neural retina during chick embryogenesis. Annals of the New York Academy of Sciences, 1163, 414–416. Harvey, S., & Hull, K. (2003). Neural growth hormone. Journal of Molecular Neuroscience, 20, 1–14. Harvey, S., Hull, K. L., & Fraser, R. A. (1993). Growth hormone: neurocrine and neuroendocrine perspectives. Growth Regulations, 3, 161–171. Harvey, S., Johnson, C. D., & Sanders, E. J. (2001). Growth hormone in neural tissues of the chick embryo. The Journal of Endocrinology, 169, 487–498. Harvey, S., Lavelin, I., & Pines, M. (2002). Growth hormone “GH” action in the brain: neural expression of a GH-response gene. Journal of Molecular Neuroscience, 18, 89–95. Hasegawa, O., Minami, S., Sugihara, H., & Wakabayashi, I. (1993). Developmental expression of the growth hormone receptor gene in the rat hypothalamus. Developmental Brain Research, 74, 287–290. Huang, X., Hu, L. Q., & Tian, B. Q. (2005). Growth hormone enhances regeneration of nNOScontaining nerve fibers in the aged rat penis. Zhonghua Nan Ke Xue, 11, 98–100. Hull, K. L., & Harvey, S. (1998a). Autoregulation of central and peripheral growth hormone receptor mRNA in domestic fowl. The Journal of Endocrinology, 156, 323–329. Hull, K. L., & Harvey, S. (1998b). Autoregulation of growth hormone receptor and growth hormone binding protein transcripts in brain and peripheral tissues of the rat. Growth Hormone & IGF Research, 8, 167–173. Hull, K. L., & Harvey, S. (2003). Growth hormone therapy and quality of life: possibilities, pitfalls and mechanisms. The Journal of Endocrinology, 179, 311–333. Isgaard, J., Aberg, D., & Nilsson, M. (2007). Protective and regenerative effects of the GH/IGF-I axis on the brain. Minerva Endocrinologica, 32, 103–113. Isla, A., Budke, M., Garcia-Grande, A., Gomez de la Riva, A., Morales, C., & Rey, J. (2007). Protective effects of the growth hormone GH on the irradiated spinal cord in rats. Neurocirugía, 18, 89–94. Ivanov, D., Dvoriantchikova, G., Nathanson, L., McKinnon, S., & Shestopalov, V. I. (2006). Microarray analysis of gene expression in adult retinal ganglion cells. FEBS Letters, 580, 331–335. Johansson, J. O., Larson, G., Andersson, M., Elmgren, A., Hynsjo, L., Lindahl, A., et al. (1995). Treatment of growth hormone-deficient adults with recombinant human growth hormone

Proteomic Actions of Growth Hormone in the Nervous System

153

increases the concentration of growth hormone in the cerebrospinal fluid and affects neurotransmitters. Neuroendocrinology, 61, 57–66. Johansson, V., Winberg, S., & Bjornsson, B. T. (2005). Growth hormone-induced stimulation of swimming and feeding behaviour of rainbow trout is abolished by D1 dopamine antagonist SCH23390. General and Comparative Endocrinology, 141, 58–65. Jonsson, E., Johansson, V., Bjornsson, B. T., & Winberg, S. (2003). Central nervous system actions of growth hormone on brain monoamine levels and behavior of juvenile rainbow trout. Hormones and Behavior, 43, 367–374. Kamegai, J., Minami, S., Sugihara, H., Higuchi, H., & Wakabayashi, I. (1994) Growth hormone induces expression of the c-fos gene on hypothalamic neuropeptide-Y and somatostatin neurons in hypophysectomized rats. Endocrinology, 135, 2765–2771. Kamegai, J., Minami, S., Sugihara, H., Higuchi, H., & Wakabayashi, I. (1996). Growth hormone receptor gene is expressed in neuropeptide Y neurons in hypothalamic arcuate nucleus of rats. Endocrinology, 137, 2109–2112. Kamegai, J., Unterman, T. G., Frohman, L. A., & Kineman, R. D. (1998). Hypothalamic/pituitaryaxis of the spontaneous dwarf rat: autofeedback regulation of growth hormone GH includes suppression of GH releasing-hormone receptor messenger ribonucleic acid. Endocrinology, 139, 3554–3560. Kanje, M., Skottner, A., & Lundborg, G. (1988). Effects of growth hormone treatment on the regeneration of rat sciatic nerve. Brain Research, 475, 254–258. Kasagi, Y., Tokita, R., Nakata, T., Imaki, T., & Minami, S. (2004). Human growth hormone induces SOCS3 and CIS mRNA increase in the hypothalamic neurons of hypophysectomized rats. Endocrine Journal, 51, 145–154. Kastrup, Y., Le Greves, M., Nyberg, F., & Blomqvist, A. (2005). Distribution of growth hormone receptor mRNA in the brain stem and spinal cord of the rat. Neuroscience, 130, 419–425. Kato, H. (2002). Molecular biology of protein metabolism. In J. Kakinuma (Ed.), Molecular nutrition (pp. 50–64). Tokyo: Koseikan. Kato, K., Suzuki, M., Kanno, H., Kekino, S., Kusakabe, K., Okada, T., et al. (2009). Distinct role of growth hormone on epilepsy progression in a model of temporal lobe epilepsy. Journal of Neurochemistry, 110, 509–519. Kim, S. H., Lessner, S. M., Sakurai, Y., & Galis, Z. S. (2004). Cyclophilin A as a novel biphasic mediator of endothelial activation and dysfunction. The American Journal of Pathology, 164, 1567–1574. Lai, Z. N., Emtner, M., Roos, P., & Nyberg, F. (1991). Characterization of putative growth hormone receptors in human choroid plexus. Brain Research, 546, 222–226. Lai, Z., Roos, P., Zhai, O., Olsson, Y., Fholenhag, K., Larsson, C., et al. (1993). Age-related reduction of human growth hormone-binding sites in the human brain. Brain Research, 621, 260–266. Laron, Z., & Galatzer, A. (1985). Growth hormone, somatomedin and prolactin–relationship to brain function. Brain & Development, 7, 559–567. Le Greves, M., Fholenhag, K., Zhou, Q., Berg, M., Myerson, B., & Nyberg, F. (2000). Growth hormone memory. Growth Hormone & IGF Research, 10, 130. Le Greves, M., Steensland, P., Le Greves, P., & Nyberb, F. (2002). Growth hormone induces agedependent alteration in the expression of hippocampal growth hormone receptor and N-methyl-D-aspartate receptor subunits gene transcripts in male rats. Proceedings of the National Academy of Sciences of the United States of America, 99, 7119–7123. Le Greves, M., Zhou, Q., Bergt, M., Le Greves, P., Fholenhag, K., Meyerson, B., et al. (2006). Growth hormone replacement in hypophysectomized rats affects spatial performance and hippocampal levels of NMDA receptor subunit and PSD-95 gene transcript levels. Experimental Brain Research, 173, 267–273. Lea, R. W., & Harvey, S. (1992). Growth hormone GH suppression of catecholamine turnover in the chicken hypothalamus: implications for GH autoregulation. The Journal of Endocrinology, 136, 245–251. Li, K., Arikan, M. C., & Andreadis, A. (2003). Modulation of the membrane-binding domain of tau protein: splicing regulation of exon 2. Brain Research. Molecular Brain Research, 116, 94–105.

154

S. Harvey and M.-L. Baudet

Lobie, P. E., Garcia-Aragon, J., Lincoln, D. T., Barnard, R., Wilcox, J. N., & Waters, M. J. (1993). Localization and ontogeny of growth hormone receptor gene expression in the central nervous system. Brain Research. Developmental Brain Research, 74, 225–233. Lobie, P. E., Zhu, T., Graichen, R., & Goh, E. L. (2000). Growth hormone insulin-like growth factor I and the CNS: localization, function and mechanism of action. Growth Hormone & IGF Research, 10, S51–S56. Lopez-Fernandez, J., Sanchez-Franco, F., Belasco, B., Tolon, R. M., Pazos, F., & Cacicedo, L. (1996). Growth hormone induces somatostatin and insulin-like growth factor I gene expression in the cerebral hemispheres of aging rats. Endocrinology, 137, 4384–4391. Lyuh, E., Kim, H.-J., Kim, M., Lee, J.-K., Park, K.-S., Yoo, K.-Y., et al. (2007). Dose-specific or dose-dependant effect of growth hormone treatment on the proliferation and differentiation of cultured neuronal cells. Growth Hormone & IGF Research, 17, 315–322. Mahmoud, G. S., & Grover, L. M. (2006). Growth hormone enhances excitatory synaptic transmission in area CA1 of rat hippocampus. Neurophysiology, 95, 2962–2974. Martinez-Canabal, A., Angoa-Perez, M., Rugerio-Vargas, C., Borgonio-Perez, G., & RivasArancibia, S. (2008). Effects of growth hormone on cyclo-oxygenase-2 expression in the hippocampus of rats chronically exposed to ozone. International Journal of Neuroscience, 118, 445–469. Martinoli, M. G., Ouellet, J., Rheaume, E., & Pelletier, G. (1991). Growth hormone and somatostatin gene expression in adult and aging rats as measured by quantitative in situ hybridization. Neuroendocrinology, 54, 607–615. McGauley, G. (1996). Growth hormone treatment, brain neurotransmitters and thyroxine. Clinical Endocrinology, 44, 325–326. McLenachan, S., Lum, M. G., Waters, M. J., & Turnley, A. M. (2009). Growth hormone promotes proliferation of adult neurosphere cultures. Growth Hormone & IGF Research, 19, 212–218. Methippara, M. M., Alam, M. N., Kumar, S., Bashir, T., Szymusia, R., & McGinty, D. (2008). Administration of the protein synthesis inhibitor, anisomycin, has distinct sleep-promoting effects in lateral preoptic and perifornical hypothalamic sites in rats. Neuroscience, 151, 1–11. Minami, S., Kamegai, J., Hasegawa, O., Sugihara, H., Okada, K., & Wakabayashi, I. (1993). Expression of growth hormone receptor gene in rat hypothalamus. Journal of Neuroendocrinology, 5, 691–696. Minami, S., Kamegai, J., Sugihara, H., Hasegawa, O., & Wakabayashi, I. (1992). Systemic administration of recombinant human growth hormone induces expression of the c-fos gene in the hypothalamic arcuate and periventricular nuclei in hypophysectomized rats. Endocrinology, 131, 247–253. Minami, S., Kamegai, J., Sugihara, H., Suzuki, N., & Wakabayashi, I. (1998). Growth hormone inhibits its own secretion by acting on the hypothalamus through its receptors on neuropeptide Y neurons in the arcuate nucleus and somatostatin neurons in the periventricular nucleus. Endocrine Journal, 45, S19–S26. Minami, S., Suzuki, N., Sugihara, H., Tamura, H., Emoto, N., & Wakabayashi, I. (1997). Microinjection of rat GH but not human IGF-I into a defined area of the hypothalamus inhibits endogenous GH secretion in rats. The Journal of Endocrinology, 153, 283–290. Mosevitsky, M. I. (2005). Nerve ending “signal” proteins GAP-43, MARCKS, and BASP1. International Review of Cytology, 245, 245–325. Muramatsu, T., & Miyauchi, T. (2003). Basigin (CD147): A multifunctional transmembrane protein involved in reproduction, neural function, inflammation and tumor invasion. Histology and Histopathology, 18, 981–987. Mustafa, A., Adem, A., Roos, P., & Nyberg, F. (1994). Sex differences in binding of human growth hormone to rat brain. Neuroscience Research, 19, 93–99. Mustafa, A., Nyberg, F., Bogdanovic, N., Islam, A., Roos, P., & Adem, A. (1994). Somatogenic and lactogenic binding sites in rat brain and liver: Quantitative autoradiographic localization. Neuroscience Research, 20, 257–263. Noguchi, T. (1996). Effects of growth hormone on cerebral development: Morphological studies. Hormone Research, 45, 5–17.

Proteomic Actions of Growth Hormone in the Nervous System

155

Noguchi, T., & Sugisaki, T. (1986). Stimulatory effects of growth hormone and thyroxine on the concentration of gangliosides in the Snell dwarf cerebrum. Journal of Neurochemistry, 47, 1785–1792. Noguchi, T., Sugisaki, T., & Tsukada, Y. (1982). Postnatal action of growth and thyroid hormones on the retarded cerebral myelinogenesis of Snell dwarf mice dw. Journal of Neurochemistry, 38, 257–263. Noguchi, T., Sugisaki, T., & Tsukada, Y. (1983). Stimulation of Snell mouse neuronal growth by GH and T4. Neurochemical Pathology, 2, 123–138. Noguchi, T., Sugisaki, T., Watanabe, M., Kohsaka, S., & Tsukada, Y. (1982). Effects of bovine growth hormone on the retarded cerebral development induced by neonatal hydrocortisone intoxication. Journal of Neurochemistry, 38, 246–256. Nyberg, F. (2000). Growth hormone in the brain: characteristics of specific brain targets for the hormone and their functional significance. Frontiers in Neuroendocrinology, 21, 330–348. Nyberg, F., & Burman, P. (1996). Growth hormone and its receptors in the central nervous systemlocation and functional significance. Hormone Research, 45, 18–22. Nyberg, D., & Sharma, H. S. (2002). Repeated topical application of growth hormone attenuates blood-spinal cord barrier permeability and edema formation following spinal cord injury: and experimental study in the rat using Evans blue, ([125])I-sodium and lanthanum tracers. Amino Acids, 23, 231–239. Obal, F., Jr., Bodosi, B., Szilagyi, A., Kacsoh, B., & Krueger, J. M. (1997). Antiserum of growth hormone decreases sleep in the rat. Neuroendocrinology, 66, 9–16. Ohsumi, M., Shi, X., Tuchiya, T., Tujioka, K., Lyou, S., Hayase, K., et  al. (2007). The role of growth hormone and amino acids on brain protein synthesis in aged rats given proteins of different quantity and quality. Amino Acids, 32, 247–253. Ohsumi, M., Tujioka, K., Hayase, K., Nagata, S., & Yokogoshi, H. (2008). The growth hormone affects the brain protein synthesis rate in hypophysectomized aged rats. Journal of Nutrional Science and Vitaminology, 54, 76–81. Pan, W., Yu, Y., Cain, C. M., Nyberg, F., Couraud, P. O., & Kastin, A. J. (2005). Permeation of growth hormone across the blood-brain barrier. Endocrinology, 146, 4898–4904. Pavel, M. E., Lohmann, T., Hahn, E. G., & Hoffmann, M. (2003). Impact of growth hormone on central nervous activity, vigilance, and tiredness after short-term therapy in growth hormonedeficient adults. Hormone and Metabolic Research, 35, 114–119. Pellegrini, E., Bluet-Pajot, M. T., Mounier, F., Bennett, P., Kordon, C., & Epelbaum, J. (1996). Central administration of a growth hormone (GH) receptor mRNA antisense increases GH pulsatility and decreases hypothalamic somatostatin expression in rats. The Journal of Neuroscience, 16, 8140–8448. Pellegrini, E., Carmignac, D. F., Bluet-Pajot, M. T., Mounier, F., Bennett, P., Epelbaum, J., et al. (1997). Intrahypothalamic growth hormone feedback: From dwarfism to acromegaly in the rat. Endocrinology, 138, 4543–4551. Persson, A. I., Thorlin, T., & Eriksson, P. S. (2005). Comparison of immunoblotted delta opioid receptor proteins expressed in the adult rat brain and their regulation of growth hormone. Neuroscience Research, 52, 1–9. Philp, N. J., Ochrietor, J. D., Rudoy, C., Muramatsu, T., & Linser, P. J. (2003). Loss of MCT1, MCT3, and MCT4 expression in the retinal pigment epithelium and neural retina of the 5A11/ basigin-null mouse. Investigative Ophthalmology & Visual Science, 44, 1305–1311. Politi, L. E., Insua, F., & Buzzi, E. (1998). Selective outgrowth and differential tropism of amacrine and photoreceptor axons to call targets during early development in vitro. Journal of Neuroscience Research, 52, 105–117. Popova, J., Ivanova, E., Tosheva, T., & Iavorska, N. (1991). Growth hormone and somatostatin treatment change 5-HT receptor activity. General Pharmacology, 22, 1143–1146. Popova, J., Robeva, A., Iavorska, N., & Zaharieva, J. (1991). Beta-adrenoreceptor activity change after prolonged treatment with growth hormone and somatostatin. Comparative Biochemistry and Physiology, 100, 543–546.

156

S. Harvey and M.-L. Baudet

Popova, J., Robeva, A., & Zaharieva, S. (1990). Muscarinic receptor activity change after prolonged treatment with growth hormone and somatostatin. Comparative Biochemistry and Physiology Part C, 96, 119–123. Ransome, M. I., Goldshmit, Y., Bartlett, P. F., Waters, M. J., & Turnley, A. M. (2004). Comparative analysis of CNS populations in knockout mice with altered growth hormone responsiveness. The European Journal of Neuroscience, 19, 2069–2079. Reimers, K., Choi, C. Y. U., Bucan, V., & Vogt, P. M. (2007). The growth-hormone inducible transmembrane protein Ghitm belong to the Bax inhibitory protein-like family. International Journal of Biological Sciences, 3, 471–476. Render, C. L., Hull, K. L., & Harvey, S. (1995). Neural expression of the pituitary GH gene. The Journal of Endocrinology, 147, 413–422. Rettori, V., Milenkovic, L., Aguila, M. C., & McCann, S. M. (1990). Physiologically significant effect of neuropeptide Y to suppress growth hormone release by stimulating somatostatin discharge. Endocrinology, 126, 2296–2301. Richardson, S. B., Hollander, C. S., Prasad, J. A., & Hirooka, Y. (1981). Somatostatin release from rat hypothalamus in vitro: Effects of melatonin and serotonin. Endocrinology, 109, 602–606. Sanders, E. J., Baudet, M. L., Parker, E., & Harvey, S. (2009). Signaling mechanisms mediating local GH action in the neural retina of the chick embryo. General and Comparative Endocrinology, 163, 63–69. Sanders, E. J., Parker, E., Aramburo, C., & Harvey, S. (2005). Retinal growth hormone is an anti apoptotic factor in embryonic retinal ganglion cell differentiation. Experimental Eye Research, 81, 551–560. Sanders, E. J., Parker, E., & Harvey, S. (2006). Retinal ganglion cell survival in development: Mechanisms of retinal growth hormone action. Experimental Eye Research, 83, 1205–1214. Sanders, E. J., Parker, E., & Harvey, S. (2008). Growth hormone-mediated survival of embryonic retinal ganglion cells: Signaling mechanisms. General and Comparative Endocrinology, 156, 613–621. Sarlieve, L. L., Bouchon, R., Koehl, C., & Neskovic, N. M. (1983). Cerebroside and sulfatide biosynthesis in the brain of Snell dwarf mouse: Effects of thyroxine and growth hormone in the early postnatal period. Journal of Neurochemistry, 40, 1058–1062. Sartorio, A., Conti, A., Molinari, E., Riva, G., Morabito, F., & Faglia, G. (1996). Growth, growth hormone and cognitive functions. Hormone Research, 45, 23–29. Sato, M., & Frohman, L. A. (1993). Differential effects of central and peripheral administration of growth hormone (GH) and insulin-like growth factor on hypothalamic GH-releasing hormone and somatostatin gene expression in GH-deficient dwarf rats. Endocrinology, 133, 793–7799. Scheepens, A., Moderscheim, T. A., & Gluckman, P. D. (2005). The role of growth hormone in neural development. Hormone Research, 64, 66–72. Scheepens, A., Sirimanne, E., Beilharz, E., Breier, B. H., Waters, M. J., Gluckman, P. D., et al. (1999). Alterations in the neural growth hormone axis following hypoxic-ischemic brain injury. Brain Research. Molecular Brain Research, 68, 88–100. Scheepens, A., Sirimanne, E. S., Breier, B. H., Clark, R. G., Gluckman, P. D., & Williams, C. E. (2001). Growth hormone as a neuronal rescue factor during recovery from CNS injury. Neuroscience, 104, 677–687. Scheepens, A., Williams, C. E., Breier, B. H., Guan, J., & Gluckman, P. D. (2000). A role for the somatotropic axis in neural development, injury and disease. Journal of Pediatric Endocrinology & Metabolism, 13, 1483–1491. Schneider, H. J., Pagotto, U., & Stalla, G. K. (2003). Central effects of the somatotropic system. European Journal of Endocrinology, 149, 377–392. Schneider-Rivas, S., Paredes-Carbajal, C., Mascher, D., Angoa-Perez, M., Jaramillo-Gonzalez, E., Borgonio-Perez, G., et al. (2007). Effects of testosterone and growth hormone on long-term retention and extinction of a passive avoidance response in young and aged rats. The International Journal of Neuroscience, 117, 1443–1456. Schneider-Rivas, S., Rivas-Arancibia, S., Vazquez-Pereyra, F., Vazquez-Sandoval, R., & Borgonio-Perez, G. (1995). Modulation of long-term memory and extinction responses induced by growth hormone (GH) and growth hormone releasing hormone (GHRH) in rats. Life Sciences, 56, 433–441.

Proteomic Actions of Growth Hormone in the Nervous System

157

Shin, D. H., Lee, E., Kim, J.-W., Kwon, B.-S., Jung, M. K., Jee, Y. H., et al. (2004). Protective effect of growth hormone on neuronal apoptosis after hypoxia-ischemia in the neonatal rat brain. Neuroscience Letters, 354, 64–68. Silva, C., Zhang, K., Tsutsui, S., Holden, J. K., Gill, M. J., & Power, C. (2003). Growth hormone prevents human immunodeficiency virus-induced neuronal p53 expression. Annals of Neurology, 54, 605–614. Sonntag, W. E., Lynch, C., Thornton, P., Khan, A., Bennett, S., & Ingram, R. (2000). The effects of growth hormone and IGF-1 deficiency on cerebrovascular and brain ageing. Journal of Anatomy, 4, 575–585. Sugisaki, T., Noguchi, T., & Tsukada, Y. (1985). Cerebral myelinogenesis in the Snell Dwarf mouse: stimulatory effects of GH and T4 restricted to the first 20 days of postnatal life. Neurochemical Research, 10, 767–778. Sun, L. Y., Al-Regaiey, K., Masternak, M. M., Wang, J., & Bartke, A. (2005). Local expression of GH, IGF-1 in the hippocampus of GH-deficient long-lived mice. Neurobiology of Aging, 26, 929–937. Sun, L. Y., Evans, M. S., Hsieh, J., Panici, J., & Bartke, A. (2005). Increased neurogenesis in dentate gyrus of long-lived Ames dwarf mice. Endocrinology, 146, 1138–1144. Svensson, A.-L., Bucht, N., Hallbert, M., & Nyberg, F. (2008). Reversal of opiate-induced apoptosis by human recombinant growth hormone in murine foetus primary hippocampal neuronal cell cultures. Proceedings of the National Academy of Sciences of the United States of America, 105, 7304–7308. Tamatani, M., Ogawa, S., & Tohyama, M. (1998). Roles of BCL-2 and caspases in hypoxiainduced neuronal cell death: a possible neuroprotective mechanism of peptide growth factors. Brain Research. Molecular Brain Research, 58, 27–39. Torsello, A., Panzeri, G., Cermenati, P., Caroleo, M. C., Ghigo, E., Camanni, F., et al. (1988). Involvement of the somatostatin and cholinergic systems in the mechanism of growth hormone autofeedback regulation in the rat. Journal of Endocrinology, 117, 273–281. Turnley, A. M. (2005). Growth hormone and SOCS2 regulation of neuronal differentiation: possible role in mental function. Pediatric Endocrinology Review, 2, 366–371. Turnley, A. M., Faux, C. H., Rietz, R. L., Coonan, J. R., & Bartlett, P. F. (2002). Suppression of cytokine signaling by inhibiting growth hormone signaling regulates neuronal differentiation. Nature Neuroscience, 5, 1155–1162. van Cauter, E., & Copinschi, G. (2000). Interrelationships between growth hormone and sleep. Growth Hormone & IGF Research, 10, S57–S62. van Cauter, E , Latta, F., Nedeltcheva, A., Spiegel, K., Leproult, R., Vandenbril, C., et al. (2004). Reciprocal interactions between the GH axis and sleep. Growth Hormone & IGF Research, 14, S10–S17. van Dam, S. P., de Winter, C. F., de Vries, R., van der Grond, J., Drent, M. L., Lijffitj, M., et al. (2005). Childhood-onset growth hormone deficiency, cognitive function and brain N-acetylaspartate. Psychoneuroendocrinology, 30, 357–363. van Marle, G., Antony, J. M., Silva, C., Sullivan, A., & Power, C. (2005). Aberrant cortical neurogenesis in a pediatric neuroAIDS model: neurotrophic effects of growth hormone. AIDS, 19, 1781–1791. Walsh, R. J., Slaby, F. J., & Posner, B. I. (1987). A receptor-mediated mechanism for the transport of prolactin from blood to cerebrospinal fluid. Endocrinology, 120, 1846–1850. Wang, X., Day, J. R., Zhou, Y., Beard, J. L., & Vasilatos-Younken, R. (2000). Evidence of a role for neuropeptide Y and monoamines in mediating the appetite-suppressive effect of GH. The Journal of Endocrinology, 66, 621–630. Winkler, T., Sharma, H. S., Stalberg, E., Badgaiyan, R. D., Westman, J., & Nyberg, F. (2000). Growth hormone attenuates alterations in spinal cord evoked potentials and cell injury following trauma to the rat spinal cord. An experimental study suing topical application of rat growth hormone. Amino Acids, 19, 363–371. Yang, J. W., Raizada, M. K., & Fellows, R. E. (1981). Effects of insulin on cultured rat brain cells: stimulation of ornithine decarboxylase activity. Journal of Neurochemistry, 36, 1050–1057. Yoshida, T., Nagata, S., & Kataoka, H. (2006). Ghitm is an ortholog of the Bombyx mori prothoracic gland-derived receptor (Pgdr) that is ubiquitously expressed in mammalian cells and requires an N-terminal signal sequence for expression. Biochemical and Biophysical Research Communications, 341, 13–18.

158

S. Harvey and M.-L. Baudet

Yoshizato, H., Fujikawa, T., Soya, H., Tanaka, M., & Nakashima, K. (1998). The growth hormone GH gene is expressed in the lateral hypothalamus: enhancement by GH-releasing hormone and repression by restraint stress. Endocrinology, 139, 2545–2551. Yu J.H., Song K.S., & An R.Z. (2001). The neuroprotective effect of growth hormone on neuronal injury of brain in pilocarpine induced epilepticus. Program and Abstract 83rd Annual Meeting at the Endocrine Society. Denver, Colorado (pp. 2–308). Zearfoss, N. R., Alarcon, J. M., Trifilieff, P., Kandel, E., & Richter, J. D. (2008). A molecular circuit composed of CPEB-1 and c-jun controls growth hormone-mediated synaptic plasticity in the mouse hippocampus. The Journal of Neuroscience, 28, 8502–8509.

Part II

Learning and Memory

Gene Expression and Signal Transduction Cascades Mediating Estrogen Effects on Memory Kristina K. Aenlle and Thomas C. Foster

Abstract  Estrogen treatment has been shown to influence memory function and protect against age-related cognitive decline. However, the mechanism through which estrogen acts to regulate these processes is not well understood. Research in this area has focused on the hippocampus and other limbic structures, due to their well-established role in memory processes. In particular, the impact of estrogen on Ca2+-dependent synaptic plasticity is of interest, since synaptic plasticity provides a potential memory mechanism. Estrogen acts through rapid Ca2+ signaling cascades to modify induction of synaptic plasticity and control translational mechanisms via the phosphorylation state of transcription factors, leading to structural modifications. Estrogen can also influence transcription through estrogen receptor (e.g., ERa and ERb) interactions with estrogen response elements located on DNA. Thus, in addition to rapid effects on synaptic function, transcriptional mechanisms lead to longer term trophic and neuroprotective benefits that may maintain hippocampal health in the face of aging. Keywords  Estrogen • Aging • Memory • Synaptic plasticity • Hippocampus • Signal transduction cascade • Estrogen receptor

1 Introduction Numerous studies have suggested a role for estrogen in nonreproductive behaviors. In particular, research has pointed to a role for estrogen in influencing memory and protection against age-related cognitive decline. As life expectancy has increased, the age at which women enter menopause has remained relatively constant; therefore, women are spending potentially one-third of their lifespan in an estrogen-depleted state and

T.C. Foster (*) Department of Neuroscience, McKnight Brain Institute, University of Florida, 100244, Gainesville, FL 32610-0244, USA e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_6, © Springer Science+Business Media, LLC 2011

161

162

K.K. Aenlle and T.C. Foster

s­ ubsequent increased likelihood of cognitive impairment. However, the mechanism by which estrogen can prevent age-related cognitive impairments is unknown and the data supporting estrogen’s positive influence on cognition are not without controversy. The divergence between copious basic science, epidemiological, and clinical studies that support the potential benefit of hormone therapy in preventing ageassociated cognitive impairment and recent studies suggesting adverse effects is emphasized by the results of the Women’s Health Initiative Memory Study (WHIMS), which indicate that estrogen may augment cognitive decline (Espeland et al., 2004). Following the conclusion of WHIMS, many scientists were immediately prepared to provide possible explanations of why the study failed to confirm previous findings of the beneficial effects of estrogen treatment (Foster, 2005; Sherwin, 2005; Wise, Dubal, Rau, Brown, & Suzuki, 2005; Zhao, O’Neill, & Diaz Brinton, 2005). The divergence in opinion emphasizes that our understanding of estrogen neurobiology is lacking. The current review focuses on the effects of estrogen on the brain processes that may influence memory function. Much of the research concerning estrogen effects on memory has focused on the limbic system, including the hippocampus and amygdala, due to the rich history detailing involvement of these systems in specific memory processes. Furthermore, injected radioactively labeled estrogen is concentrated in these limbic structures (Pfaff & Keiner, 1973), and estrogen treatment results in cell growth, observed as an increase in the number of morphological synapses within these structures (Gould, Woolley, Frankfurt, & McEwen, 1990; Nishizuka & Arai, 1982; Woolley, Gould, Frankfurt, & McEwen, 1990). This work suggests that estrogen has longterm, possibly genomic influences within the limbic system that outlasts the presence of estrogen in the plasma.

2 Synaptic Plasticity: A Memory Model It is generally believed that synaptic plasticity provides a model, if not the mechanism, for learning and memory. Several forms of synaptic plasticity have been described and, in many cases, the mechanisms are intrinsic to the synapse. The induction of synaptic modifications depends on neural activity in order to initiate a Ca2+ second messenger signaling cascade. The Ca2+ signaling cascade forms the foundation of an early phase involving posttranslational modification of proteins, including glutamate receptors. In addition, this signaling cascade sets in motion a late phase that comprises transcriptional and translational mechanisms for longer term structural modifications with the sprouting of dendritic branches and the creation of new synapses. This series of steps, from rapid activation of Ca2+ signaling to transcription, is subject to the influence of neuromodulators that adjust the induction or maintenance of synaptic plasticity. Indeed, estrogen can interact at each step, contributing to the induction of synaptic plasticity, priming the signaling pathway involved in transcription, and regulating the function of neuromodulators. Thus, estrogen acts directly on these signaling cascades and provides control

Gene Expression and Signal Transduction Cascades Mediating Estrogen Effects

163

mechanisms which adjust the induction of synaptic plasticity (Bi, Foy, Vouimba, Thompson, & Baudry, 2001; Carrer, Araque, & Buno, 2003).

2.1 Long-Term Potentiation Figure 1 provides a simplified model for memory formation and the induction of one form of synaptic plasticity, long-term potentiation (LTP). The signaling cascade involves two distinct processes; a mechanism for rapid induction and a slower mechanism involving gene regulation and structural changes. The rapid phase requires neural activity, which results in an intracellular rise in Ca2+ mainly from the opening of NMDA receptor channels. The rise in Ca2+ initiates the activation of protein kinases, including the cyclic AMP-dependent calmodulin kinase II (CamKII) (Nguyen & Woo, 2003; Selcher, Weeber, Varga, Sweatt, & Swank, 2002),

NMDAR

ER

2+ Ca PKA

ERK1/2

Calcineurin CaMKII Glutamate Receptor

ms s nis ism n e ha ec M d M i r p Ra lowe S

p

Protein Synthesis CREB p

a ch

CRE GENE

ERE GENE

Nucleus

RNA Synthesis ESTROGEN

Fig. 1  Estrogen induction of rapid and slower mechanism of synaptic plasticity. Estrogen induces the rapid rise of Ca2+ through NMDA receptor (NMDAR) channels, initiates the activation of cyclic AMP-dependent protein kinase (PKA) and calmodulin kinase II (CaMKII), and inhibits calcineurin. This shift in the balance of kinase and phosphatase activity leads to the phosphorylation of glutamate receptors. Within minutes of estrogen exposure, the rise in Ca2+ also leads to the activation of extracellular signal-regulated MAP kinase1/2 (ERK1/2). Through slower mechanisms, estrogen regulation of kinase and phosphatase activity influences the phosphorylation of the transcription factor cyclic AMP response element binding factor (CREB). The estrogen receptor (ER) contributes to both the rapid and slower mechanism of these signaling cascades. Finally, altered transcription/translation can feedback to influence the expression of ERs and NMDARs

164

K.K. Aenlle and T.C. Foster

as well as the inhibition of phosphatases such as calcineurin (Winder & Sweatt, 2001). In turn, the shift in the balance of kinases and phosphatases in synapses that have previously been active, results in phosphorylation of glutamate receptors and a rapid increase in postsynaptic responsiveness to neurotransmitters.

3 Genomic Processes For many systems, the cellular response to estrogen involves classic nuclear receptor-mediated transcriptional activation of genes involved in cell survival and proliferation. The two main estrogen receptors are estrogen receptor alpha and beta (ERa and ERb, respectively). Binding of estrogen to homodimers and/or heterodimers of ERa and ERb facilitates receptor–ligand complex binding to estrogen response elements (ERE) within the DNA. The form of the receptor–ligand complex (i.e., homodimer or heterodimer) determines, in part, which other components (activator protein 1, Sp1 transcription factor, CCAAT/enhancer binding protein, Fos) of the transcriptional machinery are recruited (Bjornstrom & Sjoberg, 2005; Paech et al., 1997; Vanacker, Pettersson, Gustafsson, & Laudet, 1999).

3.1 Estrogen Receptors Despite the fact that the two estrogen receptor subtypes have a similar structure and equivalent binding affinity for estrogen, the combination of subtypes can have different or even opposite effects on transcription-related cellular activities (Paech et al., 1997). More precisely, even though human ERa and ERb share 97% homology, their C-terminal extension of the DNA-binding domain is moderately divergent, suggesting that they share a set of primary EREs but differ in their ability to associate with other EREs, resulting in differences in transcriptional activation properties (Barkhem et al., 1998). Indeed, estrogen effects on the dendritic structure may depend on the receptor subtype expressed (Patrone et al., 2000). For instance, these studies found that ERa activation increased neurite length and number, whereas ERb activation only influenced neurite elongation. 3.1.1 Estrogen Receptor Localization ERa and ERb mRNA are present in the hippocampus and amygdala (Osterlund & Hurd, 2001; Shughrue, Lane, & Merchenthaler, 1997), but differ in expression level and the distribution of the translated protein. Early on, it was recognized that nuclear localization of ERa was found, for the most part, in interneurons of the hippocampus (Weiland, Orikasa, Hayashi, & McEwen, 1997). Similarly, ERa is predominately localized to the nucleus of the medial and lateral amygdale (Kalita, Szymczak, & Kaczmarek, 2005). In contrast, while ERb is localized to the nuclei of cells in the

Gene Expression and Signal Transduction Cascades Mediating Estrogen Effects

165

amygdala, there is considerable extranuclear expression of ERb in the hippocampus. (Kalita et al., 2005; Mitra et al., 2003; Nishio, Kuroki, & Watanabe, 2004). However, both estrogen receptors have been localized to subcellular structures including mitochondria (Chen, Delannoy, Cooke, & Yager, 2004; Milner et al., 2005). In addition, it should be noted that ERb and ERa staining has been observed in astroglia and select astrocytic processes in close association with dendritic spines (Azcoitia, Sierra, & Garcia-Segura, 1999; Milner et al., 2001), and gonadal steroids have been shown to regulate glial morphology, the expression of glial fibrillary acid protein (McQueen et al., 1992; Day et al., 1993; Del Cerro et al., 1995, 1996), and glial tyrosine kinase A receptor immunoreactivity (McCarthy, Barker-Gibb, Alves, & Milner, 2002). Estrogen can increase the transcription of glutamine synthetase within 2 h following treatment. Glutamine synthetase is restricted to glial cells and converts glutamate to glutamine. As such, estrogen could alter mechanisms for neurotransmission (Blutstein et al., 2006).

4 Rapid Signaling Cascades Estrogen receptors have been observed at the plasma membrane (Kuroki, Fukushima, Kanda, Mizuno, & Watanabe, 2000; Milner et  al., 2005; Ramirez & Zheng, 1996), and recent studies indicate that ERa and ERb are found on dendritic spines, axon terminals, and within axons of hippocampal CA1 pyramidal cells (Adams et al., 2002; Milner et al., 2001, 2005; Romeo, McCarthy, Wang, Milner, & McEwen, 2005). The results suggest that putative membrane receptors could mediate rapid nongenomic changes in physiology. Indeed, rapid effects are thought to be due to interactions between estrogen and membrane receptors such as G-protein coupled receptors. As such, estrogen can influence G-protein and Ca2+ signaling cascades (Boulware et al., 2005; Revankar, Cimino, Sklar, Arterburn, & Prossnitz, 2005), leading to change in cAMP levels, increasing the activity of protein kinases (Sawai et  al., 2002; Setalo, Singh, Guan, & Toran-Allerand, 2002; Shingo & Kito, 2002) and decreasing the activity of protein phosphatases (Sharrow, Kumar, & Foster, 2002). Thus, the rapid effects of estrogen appear to interact with the same signaling cascades involved in synaptic plasticity (Fig. 1).

4.1 Calcium Regulation Application of estrogen results in a biphasic shift in intracellular Ca2+. The rapid rise in Ca2+ may originate from NMDA receptors (Foy et al., 1999), voltage-dependent Ca2+ channels (VDCC) (Wu, Wang, Chen, & Brinton, 2005), or intracellular stores (Tanabe, Kimoto, & Kawato, 2006). This process is likely to be influenced by feedback; however, the timing of events and the exact mechanisms remain to be

166

K.K. Aenlle and T.C. Foster

determined. Evidence has been provided to suggest that the increased intracellular Ca2+ influx through VDCCs and intracellular stores is short-lived, such that intracellular levels of Ca2+ fall below baseline 5 min after application (Wu et al., 2005). This reversal may explain why other reports indicate that estrogen inhibits L type calcium channel activity (Kurata, Takebayashi, Kagaya, Morinobu, & Yamawaki, 2001; Mermelstein, Becker, & Surmeier, 1996). Furthermore, in the hippocampus, estrogencan influence Ca2+ -dependent processes in a manner directly opposite to that observed during aging; including, alterations in Ca2+ dependent processes leading to a reduction on Ca2+ dependent after hyperpolarization (AHP), facilitation of LTP, and blockade of long-term depression (LTD) (for review, see Foster, 2005).

4.2 Alteration in Kinase and Phosphatase Activity This brief rise in Ca2+ contributes to estrogen effects on subsequent events, including an increase in the activity of kinases, PKA (Shingo & Kito, 2005), and CaMKII (Sawai et al., 2002), as well as inhibition of the phosphatase calcineurin (Sharrow et al., 2002). In some cases, increased kinase activity is quickly reversed (Hayashi et al., 2005), and in others the activity is maintained for several hours (Lee et al., 2004). The initial shift in enzyme activity is thought to result in the phosphorylation of non-NMDA glutamate receptors and underlie an increase synaptic transmission which occurs rapidly after estrogen application to hippocampal slices (Foy et al., 1999; Fugger, Kumar, Lubahn, Korach, & Foster, 2001; Gu & Moss, 1996). In addition, the estrogen-mediated shift in synaptic plasticity, with facilitation of LTP induction (Sharrow et al., 2002), may result from phosphorylation of NMDA receptors (Bi et al., 2001). Extensive research demonstrates that estrogen-initiated kinase activity is part of a coordinated molecular cascade that impels synaptic growth and neuroprotection. Particularly important are tyrosine kinase and mitogen-activated protein (MAP) kinase pathways. The extracellular signal-regulated MAP kinases1/2 (ERK1/2 MAPK) are activated within minutes of estrogen exposure (Kuroki et  al., 2000; Singh, Setalo, Guan, Warren, & Toran-Allerand, 1999; Wade & Dorsa, 2003; Watters, Campbell, Cunningham, Krebs, & Dorsa, 1997; Wu et al., 2005). These kinases are important for regulation of transcription and may be involved in memory function (Blum, Moore, Adams, & Dash, 1999; Selcher, Atkins, Trzaskos, Paylor, & Sweatt, 1999), suggesting a potential link between cell-surface transmembrane receptors and transcription regulatory mechanisms involved in memory (Chang & Karin, 2001).

5 Interaction of Rapid Signaling Cascades and Transcription As note above, the rapid induction phase of synaptic plasticity involving a Ca2+dependent signaling cascade is followed by a slower consolidation phase. The shift in the balance of phosphatases and kinases results in the phosphorylation of tran-

Gene Expression and Signal Transduction Cascades Mediating Estrogen Effects

167

scription factors which initiates transcriptional activity, culminating in ­structural modifications. These transcriptional modifications include the induction of the immediate early gene c-fos, a transcription factor regulated by neural activity including activation of the NMDA receptor. In turn, c-fos may act to stabilize structural modifications (Guzowski, 2002). Similar to activity-dependent synaptic plasticity, estrogen regulation of kinase and phosphatase activity ultimately influences transcription through the phosphorylation state of transcription factors such as the cyclic AMP response element binding factor (CREB) (Aronica, Kraus, and Katzenellenbogen, 1994; Lee et  al., 2004). In turn, CREB influences the transcription of c-fos, and the cascade of events could contribute to estrogen’s long-term influence on structure or function. Estrogen treatment produces an initial and transient wave of c-fos expression in hippocampal pyramidal cells (Cattaneo & Maggil, 1990), which is followed 24 h later by a second wave (Rudick & Woolley, 2003). Interestingly, the rapid expression of c-fos is not blocked by traditional estrogen receptor antagonists, suggesting that this first wave is mediated by nongenomic rather than classical genomic mechanisms (Rudick & Woolley, 2003). In contrast, another group found that the rapid expression of Fos is not observed in the hippocampus of ERa/ERb double knockout mice (Dominguez-Salazar, Shetty, & Rissman, 2006), which suggests that classical ER mechanisms are required for the rapid induction of c-fos. One possible explanation of this discrepancy is that estrogen priming and thus the history of estrogen receptor activity, alters the system to enhance rapid responses to subsequent estrogen exposure (Foster, 2005). As noted above, rapid effects of estrogen include an increase in NMDA receptor function and cell excitability. Importantly, estrogen priming exerts transcriptional/translation effects that increase the expression of NMDA receptors (Adams, Fink, Janssen, Shah, & Morrison, 2004; Cyr, Thibault, Morissette, Landry, & Di Paolo, 2001; Weiland, 1992) and regulate proteins involved in cell excitability (Carrer et al., 2003; Foy, Chiaia, & Teyler, 1984; Wong & Moss, 1992). Moreover, the rapid effects of estrogen on synaptic transmission are reduced but not eliminated in ERa knockout mice (Fugger et al., 2001), consistent with the idea that the previous history of estrogen receptor activity primes the system for subsequent responses. The results suggest a feed forward interaction between genomic mechanisms and subsequent rapid effects on hippocampal function. Similarly, estrogen can rapidly modify G-protein signaling cascades and has longer term effects on the function of neuromodulators which act through G-protein coupled receptors. Within the limbic system, these changes include modulation of serotonin receptor expression and function (McEwen, 2002), catecholamine levels and responsiveness (Bowman, Ferguson, & Luine, 2002; Favit, Fiore, Nicoletti, & Canonico, 1991; Heikkinen, Puolivali, Liu, Rissanen, & Tanila, 2002; Hruska & Pitman, 1982), and the functional state of the cholinergic system (Gibbs & Aggarwal, 1998; Luine, 1985). Taken together, the results suggest that previous transcriptional activity associated with estrogen treatment may alter the responsiveness of nongenomic estrogen actions.

168

K.K. Aenlle and T.C. Foster

5.1 Trophic and Neuroprotective Influences Another important aspect of estrogen control of memory function includes trophic and neuroprotective influences that support cell vitality and defend against toxic pressures and the rigors of aging. A role for estrogen in structural plasticity in the young adult is well established. Estrogen promotes the proliferation of neurons (Galea, Spritzer, Barker, & Pawluski, 2006; Tanapat, Hastings, Reeves, & Gould, 1999) and the growth of new dendritic spines (Gould et al., 1990; Shors, Chua, & Falduto, 2001; Woolley et al., 1990). Recent work has extended these initial observations to better characterize the type of structural changes, and extended the findings to different species (Eberling, Wu, Tong-Turnbeaugh, & Jagust, 2004; Leranth, Shanabrough, & Redmond, 2002; Xu & Zhang, 2006). Similar to synaptic plasticity induced by neural activity, estrogen-mediated growth is initiated by activation of Ca2+ signaling cascades leading to the translocation of CREB (Murphy & Segal, 1997; Zhao, Chen, Ming Wang, & Brinton, 2005). Subsequent transcription through CREB activation includes expression of the dendritic spine marker spinophilin (Lee et  al., 2004), and the presynaptic markers growth associated protein (GAP-43) (Ferrini et al., 2002) and synaptophysin (Brake et al., 2001). The chain of events likely involves binding of estrogen to its receptor (McEwen, Tanapat, & Weiland, 1999); however, ERa and ERb may play different roles in the generation and elimination of dendritic spines (Segal & Murphy, 2001, Szymczak et al., 2006). In addition to a role for estrogen receptors, spine growth requires activation of NMDA receptors (Woolley & McEwen, 1994). The activation of NMDA receptors may determine the site of estrogen-induced growth similar to activity-dependent changes in synaptic growth (Lin et  al., 2005; Pang & Lu, 2004). It is likely that estrogen is interacting with and modulating other growth factors since the signaling cascades involved in estrogen-induced trophic effects are activated by several different growth factors (Improta-Brears et al., 1999; Scharfman & Maclusky, 2006). For example, estrogen increases the expression of insulin-like growth factor-1 (IGF-1) and can interact with IGF-1 to induce dendritic growth and synaptic plasticity (Cardona-Gomez, Trejo, Fernandez, & Garcia-Segura, 2000; Mendez, Wandosell, & Garcia-Segura, 2006). Furthermore, an ERE is located on the promoter region of brain-derived neurotrophic factor (BDNF), suggesting that enhanced BDNF following estrogen treatment involves classic transcriptional mechanisms (Cavus & Duman, 2003; Gibbs, 1998, 1999; Singh, Meyer, & Simpkins, 1995; Sohrabji, Miranda, & Toran-Allerand, 1995), as well as signaling cascades (Zhou, Zhang, Cohen, & Pandey, 2005). These same signaling cascades are likely involved in protecting the brain from apoptosis associated with Alzheimer’s disease (Mattson , Robinson, & Guo, 1997; Shah, Anderson, Rapoport, & Ferreira, 2003), altered Ca2+ regulation associated with normal aging (Brewer, Reichensperger, & Brinton, 2006), and disruption of mitochondrial function-associated neural stressors (Dykens, Simpkins, Wang, & Gordon, 2003). Furthermore, estrogen promotes an increase in transcription of

Gene Expression and Signal Transduction Cascades Mediating Estrogen Effects

169

genes which are neuroprotective (Benvenuti et  al., 2005; Stoltzner, Berchtold, Cotman, & Pike, 2001; Wu et al., 2005; Yalcin, 2005) and decreases transcription of genes associated with oxidative stress (Chan, Tammariello, Estus, & Mattson, 1999). The results suggest that estrogen may reverse age-related changes in transcription associated with stress (Blalock et al., 2003; Prolla, 2002); however, this idea remains to be tested.

5.2 Age-Related Changes in Estrogen Pathways It might be expected that the decline in estrogen level with advanced aged or due to oophorectomy should be associated with radical decline in memory function. The fact that memory changes are not extreme may related to local production of estrogen. The enzyme aromatase, which converts steroids into estrogen, is found in the brain (Prange-Kiel & Rune, 2006; Sasano, Takashashi, Satoh, Nagura, & Harada, 1998), and estrogen production may be shifted during times of stress when it would act to maintain neuroprotection and trophic influences (Carswell et  al., 2005; McCullough, Blizzard, Simpson, Oz, & Hurn, 2003; Veiga, Azcoitia, & Garcia-Segura, 2005). In contrast, genomic and nongenomic effects of estrogen appear to decline with advanced age, including neuroprotecive influences. The decreased effectiveness may result from a loss of estrogen receptors (Adams et al., 2002; Wilson et al., 2002) or changes in the expression of coactivators and corepressors (Jezierski & Sohrabji, 2001; Rodriguez-Calvo et al., 2006). In addition, the rapid signaling mechanisms may be altered with age including down-regulation of tropic factor and G-protein cascades and a shift in Ca2+ regulation and the activity of kinases and phosphatases (Foster, 2005). Indeed, the fact that beneficial effects of estrogen are mainly observed in middle-aged animals relative to senescent animals suggests that estrogen protects brain function against detrimental effects of aging rather than reversing the effects of age. Furthermore, age-related pressures may bring about an increased sensitivity to the loss of estrogen.

6 Conclusion A body of work indicates that estrogen has beneficial effects on cognitive function, particularly during aging. While the exact mechanisms for memory remain to be elucidated, evidence suggests that memory formation depends on activation of rapid Ca2+ signaling cascades. In addition, these cascades contribute to the consolidation of memory through the regulation of transcription in order to bring about structural changes. These same cascades are regulated by numerous neural modulators and hormones including estrogen. Thus, estrogen can promote rapid activity associated with synaptic activity and induce longer term trophic and neuroprotecive effect that may maintain the hippocampus.

170

K.K. Aenlle and T.C. Foster

References Adams, M. M., Fink, S. E., Janssen, W. G., Shah, R. A., & Morrison, J. H. (2004). Estrogen modulates synaptic N-methyl-D-aspartate receptor subunit distribution in the aged hippocampus. The Journal of Comparative Neurology, 474, 419–426. Adams, M. M., Fink, S. E., Shah, R. A., Janssen, W. G., Hayashi, S., Milner, T. A., et al. (2002). Estrogen and aging affect the subcellular distribution of estrogen receptor-alpha in the ­hippocampus of female rats. The Journal of Neuroscience, 22, 3608–3614. Aronica, S. M., Kraus, W. L., & Katzenellenbogen, B. S. (1994). Estrogen action via the cAMP signaling pathway: Stimulation of adenylate cyclase and cAMP-regulated gene transcription. Proceedings of the National Academy of Sciences of the United States of America, 91, 8517–8521. Azcoitia, I., Sierra, A., & Garcia-Segura, L. M. (1999). Localization of estrogen receptor betaimmunoreactivity in astrocytes of the adult rat brain. Glia, 26, 260–267. Barkhem, T., Carlsson, B., Nilsson, Y., Enmark, E., Gustafsson, J., & Nilsson, S. (1998). Differential response of estrogen receptor alpha and estrogen receptor beta to partial estrogen agonists/antagonists. Molecular Pharmacology, 54, 105–112. Benvenuti, S., Luciani, P., Vannelli, G. B., Gelmini, S., Franceschi, E., Serio, M., et al. (2005). Estrogen and selective estrogen receptor modulators exert neuroprotective effects and ­stimulate the expression of selective Alzheimer’s disease indicator-1, a recently discovered antiapoptotic gene, in human neuroblast long-term cell cultures. The Journal of Clinical Endocrinology and Metabolism, 90, 1775–1782. Bi, R., Foy, M. R., Vouimba, R. M., Thompson, R. F., & Baudry, M. (2001). Cyclic changes in estradiol regulate synaptic plasticity through the MAP kinase pathway. Proceedings of the National Academy of Sciences of the United States of America, 98, 13391–13395. Bjornstrom, L., & Sjoberg, M. (2005). Mechanisms of estrogen receptor signaling: Convergence of genomic and nongenomic actions on target genes. Molecular Endocrinology, 19, 833–842. Blalock, E. M., Chen, K. C., Sharrow, K., Herman, J. P., Porter, N. M., Foster, T. C., et al. (2003). Gene microarrays in hippocampal aging: Statistical profiling identifies novel processes ­correlated with cognitive impairment. The Journal of Neuroscience, 23, 3807–3819. Blum, S., Moore, A. N., Adams, F., & Dash, P. K. (1999). A mitogen-activated protein kinase cascade in the CA1/CA2 subfield of the dorsal hippocampus is essential for long-term spatial memory. The Journal of Neuroscience, 19, 3535–3544. Blutstein, T., Devidze, N., Choleris, E., Jasnow, A. M., Pfaff, D. W., & Mong, J. A. (2006). Oestradiol up-regulates glutamine synthetase mRNA and protein expression in the hypothalamus and hippocampus: implications for a role of hormonally responsive glia in amino acid neurotransmission. Journal of Neuroendocrinology, 18, 692–702. Boulware, M. I., Weick, J. P., Becklund, B. R., Kuo, S. P., Groth, R. D., & Mermelstein, P. G. (2005). Estradiol activates group I and II metabotropic glutamate receptor signaling, leading to opposing influences on cAMP response element-binding protein. The Journal of Neuroscience, 25, 5066–5078. Bowman, R. E., Ferguson, D., & Luine, V. N. (2002). Effects of chronic restraint stress and estradiol on open field activity, spatial memory, and monoaminergic neurotransmitters in ovariectomized rats. Neuroscience, 113, 401–410. Brake, W. G., Alves, S. E., Dunlop, J. C., Lee, S. J., Bulloch, K., Allen, P. B., et al. (2001). Novel target sites for estrogen action in the dorsal hippocampus: An examination of synaptic ­proteins. Endocrinology, 142, 1284–1289. Brewer, G. J., Reichensperger, J. D., & Brinton, R. D. (2006). Prevention of age-related dysregulation of calcium dynamics by estrogen in neurons. Neurobiology of Aging, 27, 306–317. Cardona-Gomez, G. P., Trejo, J. L., Fernandez, A. M., & Garcia-Segura, L. M. (2000). Estrogen receptors and insulin-like growth factor-I receptors mediate estrogen-dependent synaptic plasticity. Neuroreport, 11, 1735–1738.

Gene Expression and Signal Transduction Cascades Mediating Estrogen Effects

171

Carrer, H. F., Araque, A., & Buno, W. (2003). Estradiol regulates the slow Ca2+-activated K+ current in hippocampal pyramidal neurons. The Journal of Neuroscience, 23, 6338–6344. Carswell, H. V., Dominiczak, A. F., Garcia-Segura, L. M., Harada, N., Hutchison, J. B., & Macrae, I. M. (2005). Brain aromatase expression after experimental stroke: topography and time course. The Journal of Steroid Biochemistry and Molecular Biology, 96, 89–91. Cattaneo, E., & Maggi, A. (1990). c-fos Induction by estrogen in specific rat brain areas. European Journal of Pharmacology, 188, 153–159. Cavus, I., & Duman, R. S. (2003). Influence of estradiol, stress, and 5-HT2A agonist treatment on brain-derived neurotrophic factor expression in female rats. Biological Psychiatry, 54, 59–69. Chan, S. L., Tammariello, S. P., Estus, S., & Mattson, M. P. (1999). Prostate apoptosis response-4 mediates trophic factor withdrawal-induced apoptosis of hippocampal neurons: Actions prior to mitochondrial dysfunction and caspase activation. Journal of Neurochemistry, 73, 502–512. Chang, L., & Karin, M. (2001). Mammalian MAP kinase signalling cascades. Nature, 410, 37–40. Chen, J. Q., Delannoy, M., Cooke, C., & Yager, J. D. (2004). Mitochondrial localization of ERalpha and ERbeta in human MCF7 cells. American Journal of Physiology. Endocrinology and Metabolism, 286, E1011–E1022. Cyr, M., Thibault, C., Morissette, M., Landry, M., & Di Paolo, T. (2001). Estrogen-like activity of tamoxifen and raloxifene on NMDA receptor binding and expression of its subunits in rat brain. Neuropsychopharmacology, 25, 242–257. Day, J. R, Laping, N. J, Lampert-Etchells, M., Brown, S. A, O’Callaghan, J. P, McNeill, T.H., & Finch, C. E. (1993) Gonadal steroids regulate the expression of glial fibrillary acidic protein in the adult male rat hippocampus. Neuroscience, 55, 435–443. Del Cerro, S., Garcia-Estrada, J., & Garcia-Segura, L. M. (1995) Neuroactive steroids regulate astroglia morphology in hippocampal cultures from adult rats. Glia, 14, 65–71. Dominguez-Salazar, E., Shetty, S., & Rissman, E. F. (2006). Rapid neural Fos responses to oestradiol in oestrogen receptor alphabeta double knockout mice. Journal of Neuroendocrinology, 18, 195–202. Dykens, J. A., Simpkins, J. W., Wang, J., & Gordon, K. (2003). Polycyclic phenols, estrogens and neuroprotection: A proposed mitochondrial mechanism. Experimental Gerontology, 38, 101–107. Eberling, J. L., Wu, C., Tong-Turnbeaugh, R., & Jagust, W. J. (2004). Estrogen- and tamoxifenassociated effects on brain structure and function. Neuroimage, 21, 364–371. Espeland, M. A., Rapp, S. R., Shumaker, S. A., Brunner, R., Manson, J. E., Sherwin, B. B., et al. (2004). Conjugated equine estrogens and global cognitive function in postmenopausal women: Women’s Health Initiative Memory Study. JAMA, 291, 2959–2968. Favit, A., Fiore, L., Nicoletti, F., & Canonico, P. L. (1991). Estrogen modulates stimulation of inositol phospholipid hydrolysis by norepinephrine in rat brain slices. Brain Research, 555, 65–69. Ferrini, M., Bisagno, V., Piroli, G., Grillo, C., Deniselle, M. C., & De Nicola, A. F. (2002). Effects of estrogens on choline-acetyltransferase immunoreactivity and GAP-43 mRNA in the forebrain of young and aging male rats. Cellular and Molecular Neurobiology, 22, 289–301. Foster, T. C. (2005). Interaction of rapid signal transduction cascades and gene expression in mediating estrogen effects on memory over the life span. Frontiers in Neuroendocrinology, 26, 51–64. Foy, M. R., Chiaia, N. L., & Teyler, T. J. (1984). Reversal of hippocampal sexual dimorphism by gonadal steroid manipulation. Brain Research, 321, 311–314. Foy, M. R., Xu, J., Xie, X., Brinton, R. D., Thompson, R. F., & Berger, T. W. (1999). 17betaestradiol enhances NMDA receptor-mediated EPSPs and long-term potentiation. Journal of Neurophysiology, 81, 925–929. Fugger, H. N., Kumar, A., Lubahn, D. B., Korach, K. S., & Foster, T. C. (2001). Examination of estradiol effects on the rapid estradiol mediated increase in hippocampal synaptic transmission in estrogen receptor alpha knockout mice. Neuroscience Letters, 309, 207–209. Galea, L. A., Spritzer, M. D., Barker, J. M., & Pawluski, J. L. (2006). Gonadal hormone modulation of hippocampal neurogenesis in the adult. Hippocampus, 16, 225–232. Gibbs, R. B. (1998). Levels of trkA and BDNF mRNA, but not NGF mRNA, fluctuate across the estrous cycle and increase in response to acute hormone replacement. Brain Research, 787, 259–268.

172

K.K. Aenlle and T.C. Foster

Gibbs, R. B. (1999). Treatment with estrogen and progesterone affects relative levels of brainderived neurotrophic factor mRNA and protein in different regions of the adult rat brain. Brain Research, 844, 20–27. Gibbs, R. B., & Aggarwal, P. (1998). Estrogen and basal forebrain cholinergic neurons: ­implications for brain aging and Alzheimer’s disease-related cognitive decline. Hormones and Behavior, 34, 98–111. Gould, E., Woolley, C. S., Frankfurt, M., & McEwen, B. S. (1990). Gonadal steroids regulate dendritic spine density in hippocampal pyramidal cells in adulthood. The Journal of Neuroscience, 10, 1286–1291. Gu, Q., & Moss, R. L. (1996). 17 beta-Estradiol potentiates kainate-induced currents via activation of the cAMP cascade. The Journal of Neuroscience, 16, 3620–3629. Guzowski, J. F. (2002). Insights into immediate-early gene function in hippocampal memory consolidation using antisense oligonucleotide and fluorescent imaging approaches. Hippocampus, 12, 86–104. Hayashi, S., Ueyama, T., Kajimoto, T., Yagi, K., Kohmura, E., & Saito, N. (2005). Involvement of gamma protein kinase C in estrogen-induced neuroprotection against focal brain ischemia through G protein-coupled estrogen receptor. Journal of Neurochemistry, 93, 883–891. Heikkinen, T., Puolivali, J., Liu, L., Rissanen, A., & Tanila, H. (2002). Effects of ovariectomy and estrogen treatment on learning and hippocampal neurotransmitters in mice. Hormones and Behavior, 41, 22–32. Hruska, R. E., & Pitman, K. T. (1982). Distribution and localization of estrogen-sensitive dopamine receptors in the rat brain. Journal of Neurochemistry, 39, 1418–1423. Improta-Brears, T., Whorton, A. R., Codazzi, F., York, J. D., Meyer, T., & McDonnell, D. P. (1999). Estrogen-induced activation of mitogen-activated protein kinase requires mobilization of intracellular calcium. Proceedings of the National Academy of Sciences of the United States of America, 96, 4686–4691. Jezierski, M. K., & Sohrabji, F. (2001). Neurotrophin expression in the reproductively senescent forebrain is refractory to estrogen stimulation. Neurobiology of Aging, 22, 309–319. Kalita, K., Szymczak, S., & Kaczmarek, L. (2005). Non-nuclear estrogen receptor beta and alpha in the hippocampus of male and female rats. Hippocampus, 15, 404–412. Kurata, K., Takebayashi, M., Kagaya, A., Morinobu, S., & Yamawaki, S. (2001). Effect of betaestradiol on voltage-gated Ca(2+) channels in rat hippocampal neurons: A comparison with dehydroepiandrosterone. European Journal of Pharmacology, 416, 203–212. Kuroki, Y., Fukushima, K., Kanda, Y., Mizuno, K., & Watanabe, Y. (2000). Putative membranebound estrogen receptors possibly stimulate mitogen-activated protein kinase in the rat hippocampus. European Journal of Pharmacology, 400, 205–209. Lee, S. J., Campomanes, C. R., Sikat, P. T., Greenfield, A. T., Allen, P. B., & McEwen, B. S. (2004). Estrogen induces phosphorylation of cyclic AMP response element binding (pCREB) in primary hippocampal cells in a time-dependent manner. Neuroscience, 124, 549–560. Leranth, C., Shanabrough, M., & Redmond, D. E., Jr. (2002). Gonadal hormones are responsible for maintaining the integrity of spine synapses in the CA1 hippocampal subfield of female nonhuman primates. The Journal of Comparative Neurology, 447, 34–42. Lin, B., Kramar, E. A., Bi, X., Brucher, F. A., Gall, C. M., & Lynch, G. (2005). Theta stimulation polymerizes actin in dendritic spines of hippocampus. The Journal of Neuroscience, 25, 2062–2069. Luine, V. N. (1985). Estradiol increases choline acetyltransferase activity in specific basal forebrain nuclei and projection areas of female rats. Experimental Neurology, 89, 484–490. Mattson, M. P., Robinson, N., & Guo, Q. (1997). Estrogens stabilize mitochondrial function and protect neural cells against the pro-apoptotic action of mutant presenilin-1. Neuroreport, 8, 3817–3821. McCarthy, J. B., Barker-Gibb, A. L., Alves, S. E., & Milner, T. A. (2002). TrkA immunoreactive astrocytes in dendritic fields of the hippocampal formation across estrous. Glia, 38, 36–44. McCullough, L. D., Blizzard, K., Simpson, E. R., Oz, O. K., & Hurn, P. D. (2003). Aromatase cytochrome P450 and extragonadal estrogen play a role in ischemic neuroprotection. The Journal of Neuroscience, 23, 8701–8705.

Gene Expression and Signal Transduction Cascades Mediating Estrogen Effects

173

McEwen, B. (2002). Estrogen actions throughout the brain. Recent Progress in Hormone Research, 57, 357–384. McEwen, B. S., Tanapat, P., & Weiland, N. G. (1999). Inhibition of dendritic spine induction on hippocampal CA1 pyramidal neurons by a nonsteroidal estrogen antagonist in female rats. Endocrinology, 140, 1044–1047. McQueen, J. K., Dow, R. C., & Fink, G. (1992) Gonadal steroids regulate number of astrocytes immunostained for glial fibrillary acidic protein in mouse hippocampus. Molecular and Cellular Neurosciences, 3, 482–486. Mendez, P., Wandosell, F., & Garcia-Segura, L. M. (2006). Cross-talk between estrogen receptors and insulin-like growth factor-I receptor in the brain: Cellular and molecular mechanisms. Frontiers in Neuroendocrinology, 27, 391–403. Mermelstein, P. G., Becker, J. B., & Surmeier, D. J. (1996). Estradiol reduces calcium currents in rat neostriatal neurons via a membrane receptor. The Journal of Neuroscience, 16, 595–604. Milner, T. A., Ayoola, K., Drake, C. T., Herrick, S. P., Tabori, N. E., McEwen, B. S., et al. (2005). Ultrastructural localization of estrogen receptor beta immunoreactivity in the rat hippocampal formation. The Journal of Comparative Neurology, 491, 81–95. Milner, T. A., McEwen, B. S., Hayashi, S., Li, C. J., Reagan, L. P., & Alves, S. E. (2001). Ultrastructural evidence that hippocampal alpha estrogen receptors are located at extranuclear sites. The Journal of Comparative Neurology, 429, 355–371. Mitra, S. W., Hoskin, E., Yudkovitz, J., Pear, L., Wilkinson, H. A., Hayashi, S., et  al. (2003). Immunolocalization of estrogen receptor beta in the mouse brain: comparison with estrogen receptor alpha. Endocrinology, 144, 2055–2067. Murphy, D. D., & Segal, M. (1997). Morphological plasticity of dendritic spines in central neurons is mediated by activation of cAMP response element binding protein. Proceedings of the National Academy of Sciences of the United States of America, 94, 1482–1487. Nguyen, P. V., & Woo, N. H. (2003). Regulation of hippocampal synaptic plasticity by cyclic AMP-dependent protein kinases. Progress in Neurobiology, 71, 401–437. Nishio, M., Kuroki, Y., & Watanabe, Y. (2004). Subcellular localization of estrogen receptor beta in mouse hippocampus. Neuroscience Letters, 355, 109–112. Nishizuka, M., & Arai, Y. (1982). Synapse formation in response to estrogen in the medial amygdala developing in the eye. Proceedings of the National Academy of Sciences of the United States of America, 79, 7024–7026. Osterlund, M. K., & Hurd, Y. L. (2001). Estrogen receptors in the human forebrain and the relation to neuropsychiatric disorders. Progress in Neurobiology, 64, 251–267. Paech, K., Webb, P., Kuiper, G. G., Nilsson, S., Gustafsson, J., Kushner, P. J., et  al. (1997). Differential ligand activation of estrogen receptors ERalpha and ERbeta at AP1 sites. Science, 277, 1508–1510. Pang, P. T., & Lu, B. (2004). Regulation of late-phase LTP and long-term memory in normal and aging hippocampus: Role of secreted proteins tPA and BDNF. Ageing Research Reviews, 3, 407–430. Parducz, A., Hajszan, T., Maclusky, N. J., Hoyk, Z., Csakvari, E., Kurunczi, A., et al. (2006). Synaptic remodeling induced by gonadal hormones: Neuronal plasticity as a mediator of neuroendocrine and behavioral responses to steroids. Neuroscience, 138, 977–985. Patrone, C., Pollio, G., Vegeto, E., Enmark, E., de Curtis, I., Gustafsson, J. A., et  al. (2000). Estradiol induces differential neuronal phenotypes by activating estrogen receptor alpha or beta. Endocrinology, 141, 1839–1845. Pfaff, D., & Keiner, M. (1973). Atlas of estradiol-concentrating cells in the central nervous system of the female rat. The Journal of Comparative Neurology, 151, 121–158. Prange-Kiel, J., & Rune, G. M. (2006). Direct and indirect effects of estrogen on rat hippocampus. Neuroscience, 138, 765–772. Prolla, T. A. (2002). DNA microarray analysis of the aging brain. Chemical Senses, 27, 299–306. Ramirez, V. D., & Zheng, J. (1996). Membrane sex-steroid receptors in the brain. Frontiers in Neuroendocrinology, 17, 402–439.

174

K.K. Aenlle and T.C. Foster

Revankar, C. M., Cimino, D. F., Sklar, L. A., Arterburn, J. B., & Prossnitz, E. R. (2005). A transmembrane intracellular estrogen receptor mediates rapid cell signaling. Science, 307, 1625–1630. Rodriguez-Calvo, R., Jove, M., Coll, T., Camins, A., Sanchez, R. M., Alegret, M., et al. (2006). PGC-1beta down-regulation is associated with reduced ERRalpha activity and MCAD expression in skeletal muscle of senescence-accelerated mice. The Journals of Gerontology. Series A: Biological Sciences and Medical Sciences, 61, 773–780. Romeo, R. D., McCarthy, J. B., Wang, A., Milner, T. A., & McEwen, B. S. (2005). Sex differences in hippocampal estradiol-induced N-methyl-D-aspartic acid binding and ultrastructural localization of estrogen receptor-alpha. Neuroendocrinology, 81, 391–399. Rudick, C. N., & Woolley, C. S. (2003). Selective estrogen receptor modulators regulate ­phasic activation of hippocampal CA1 pyramidal cells by estrogen. Endocrinology, 144, 179–187. Sasano, H., Takashashi, K., Satoh, F., Nagura, H., & Harada, N. (1998). Aromatase in the human central nervous system. Clinical Endocrinology, 48, 325–329. Sawai, T., Bernier, F., Fukushima, T., Hashimoto, T., Ogura, H., & Nishizawa, Y. (2002). Estrogen induces a rapid increase of calcium-calmodulin-dependent protein kinase II activity in the hippocampus. Brain Research, 950, 308–311. Scharfman, H. E., & Maclusky, N. J. (2006). Estrogen and brain-derived neurotrophic factor (BDNF) in hippocampus: Complexity of steroid hormone-growth factor interactions in the adult CNS. Frontiers in Neuroendocrinology, 27, 415–435. Segal, M., & Murphy, D. (2001). Estradiol induces formation of dendritic spines in hippocampal neurons: Functional correlates. Hormones and Behavior, 40, 156–159. Selcher, J. C., Atkins, C. M., Trzaskos, J. M., Paylor, R., & Sweatt, J. D. (1999). A necessity for MAP kinase activation in mammalian spatial learning. Learning & Memory, 6, 478–490. Selcher, J. C., Weeber, E. J., Varga, A. W., Sweatt, J. D., & Swank, M. (2002). Protein kinase signal transduction cascades in mammalian associative conditioning. The Neuroscientist, 8, 122–131. Setalo, G., Jr., Singh, M., Guan, X., & Toran-Allerand, C. D. (2002). Estradiol-induced phosphorylation of ERK1/2 in explants of the mouse cerebral cortex: the roles of heat shock protein 90 (Hsp90) and MEK2. Journal of Neurobiology, 50, 1–12. Shah, R. D., Anderson, K. L., Rapoport, M., & Ferreira, A. (2003). Estrogen-induced changes in the microtubular system correlate with a decreased susceptibility of aging neurons to beta amyloid neurotoxicity. Molecular and Cellular Neurosciences, 24, 503–516. Sharrow, K. M., Kumar, A., & Foster, T. C. (2002). Calcineurin as a potential contributor in estradiol regulation of hippocampal synaptic function. Neuroscience, 113, 89–97. Sherwin, B. B. (2005). Estrogen and memory in women: How can we reconcile the findings? Hormones and Behavior, 47, 371–375. Shingo, A. S., & Kito, S. (2005). Estradiol induces PKA activation through the putative membrane receptor in the living hippocampal neuron. Journal of Neural Transmission, 112, 1469–1473. Shingo, A. S., & Kito, S. (2002). Estrogen induces elevation of cAMP-dependent protein kinase activity in immortalized hippocampal neurons: imaging in living cells. Journal of Neural Transmission, 109, 171–174. Shors, T. J., Chua, C., & Falduto, J. (2001). Sex differences and opposite effects of stress on dendritic spine density in the male versus female hippocampus. The Journal of Neuroscience, 21, 6292–6297. Shughrue, P. J., Lane, M. V., & Merchenthaler, I. (1997). Comparative distribution of estrogen receptor-alpha and -beta mRNA in the rat central nervous system. The Journal of Comparative Neurology, 388, 507–525. Singh, M., Meyer, E. M., & Simpkins, J. W. (1995). The effect of ovariectomy and estradiol replacement on brain-derived neurotrophic factor messenger ribonucleic acid expression in cortical and hippocampal brain regions of female Sprague-Dawley rats. Endocrinology, 136, 2320–2324. Singh, M., Setalo, G., Jr., Guan, X., Warren, M., & Toran-Allerand, C. D. (1999). Estrogen-induced activation of mitogen-activated protein kinase in cerebral cortical explants: convergence of estrogen and neurotrophin signaling pathways. The Journal of Neuroscience, 19, 1179–1188.

Gene Expression and Signal Transduction Cascades Mediating Estrogen Effects

175

Sohrabji, F., Miranda, R. C., & Toran-Allerand, C. D. (1995). Identification of a putative estrogen response element in the gene encoding brain-derived neurotrophic factor. Proceedings of the National Academy of Sciences of the United States of America, 92, 11110–11114. Stoltzner, S. E., Berchtold, N. C., Cotman, C. W., & Pike, C. J. (2001). Estrogen regulates bcl-x expression in rat hippocampus. Neuroreport, 12, 2797–2800. Szymczak, S., Kalita, K., Jaworski, J., Mioduszewska, B., Savonenko, A., Markowska, A., et al. (2006). Increased estrogen receptor beta expression correlates with decreased spine formation in the rat hippocampus. Hippocampus, 16, 453–463. Tanabe, N., Kimoto, T., & Kawato, S. (2006). Rapid Ca2+ signaling induced by Bisphenol A in cultured rat hippocampal neurons. Neuro Endocrinology Letters, 27, 97–104. Tanapat, P., Hastings, N. B., Reeves, A. J., & Gould, E. (1999). Estrogen stimulates a transient increase in the number of new neurons in the dentate gyrus of the adult female rat. The Journal of Neuroscience, 19, 5792–5801. Vanacker, J. M., Pettersson, K., Gustafsson, J. A., & Laudet, V. (1999). Transcriptional targets shared by estrogen receptor- related receptors (ERRs) and estrogen receptor (ER) alpha, but not by ERbeta. The EMBO Journal, 18, 4270–4279. Veiga, S., Azcoitia, I., & Garcia-Segura, L. M. (2005). Extragonadal synthesis of estradiol is protective against kainic acid excitotoxic damage to the hippocampus. Neuroreport, 16, 1599–1603. Wade, C. B., & Dorsa, D. M. (2003). Estrogen activation of cyclic adenosine 5¢-monophosphate response element-mediated transcription requires the extracellularly regulated kinase/mitogenactivated protein kinase pathway. Endocrinology, 144, 832–838. Watters, J. J., Campbell, J. S., Cunningham, M. J., Krebs, E. G., & Dorsa, D. M. (1997). Rapid membrane effects of steroids in neuroblastoma cells: effects of estrogen on mitogen activated protein kinase signalling cascade and c-fos immediate early gene transcription. Endocrinology, 138, 4030–4033. Weiland, N. G. (1992). Estradiol selectively regulates agonist binding sites on the N-methyl-Daspartate receptor complex in the CA1 region of the hippocampus. Endocrinology, 131, 662–668. Weiland, N. G., Orikasa, C., Hayashi, S., & McEwen, B. S. (1997). Distribution and hormone regulation of estrogen receptor immunoreactive cells in the hippocampus of male and female rats. The Journal of Comparative Neurology, 388, 603–612. Wilson, M. E., Rosewell, K. L., Kashon, M. L., Shughrue, P. J., Merchenthaler, I., & Wise, P. M. (2002). Age differentially influences estrogen receptor-alpha (ERalpha) and estrogen receptorbeta (ERbeta) gene expression in specific regions of the rat brain. Mechanisms of Ageing and Development, 123, 593–601. Winder, D. G., & Sweatt, J. D. (2001). Roles of serine/threonine phosphatases in hippocampal synaptic plasticity. Nature Reviews. Neuroscience, 2, 461–474. Wise, P. M., Dubal, D. B., Rau, S. W., Brown, C. M., & Suzuki, S. (2005). Are estrogens protective or risk factors in brain injury and neurodegeneration? Reevaluation after the Women’s health initiative. Endocrine Reviews, 26, 308–312. Wong, M., & Moss, R. L. (1992). Long-term and short-term electrophysiological effects of estrogen on the synaptic properties of hippocampal CA1 neurons. The Journal of Neuroscience, 12, 3217–3225. Woolley, C. S., Gould, E., Frankfurt, M., & McEwen, B. S. (1990). Naturally occurring fluctuation in dendritic spine density on adult hippocampal pyramidal neurons. The Journal of Neuroscience, 10, 4035–4039. Woolley, C. S., & McEwen, B. S. (1994). Estradiol regulates hippocampal dendritic spine density via an N-methyl-D-aspartate receptor-dependent mechanism. The Journal of Neuroscience, 14, 7680–7687. Wu, T. W., Wang, J. M., Chen, S., & Brinton, R. D. (2005). 17Beta-estradiol induced Ca2+ influx via L-type calcium channels activates the Src/ERK/cyclic-AMP response element binding protein signal pathway and BCL-2 expression in rat hippocampal neurons: a potential ­initiation mechanism for estrogen-induced neuroprotection. Neuroscience, 135, 59–72. Xu, X., & Zhang, Z. (2006). Effects of estradiol benzoate on learning-memory behavior and synaptic structure in ovariectomized mice. Life Sciences, 79, 1553–1560.

176

K.K. Aenlle and T.C. Foster

Yalcin, A., Kanit, L., Durmaz, G., Sargin, S., Terek, C. H., & Tanyolac, B. (2005) Altered level of apurinic/apyrimidinic endonuclease/redox factor-1 (APE/REF-1) mRNA in the hippocampus of ovariectomized rats treated by raloxifene against kainic acid. Cilinical and Experimental Pharmacology and Physiology, 32, 611–614. Zhao, L., Chen, S., Ming Wang, J., & Brinton, R. D. (2005). 17beta-estradiol induces Ca2+ influx, dendritic and nuclear Ca2+ rise and subsequent cyclic AMP response element-binding protein activation in hippocampal neurons: a potential initiation mechanism for estrogen neurotrophism. Neuroscience, 132, 299–311. Zhao, L., O’Neill, K., & Diaz Brinton, R. (2005). Selective estrogen receptor modulators (SERMs) for the brain: current status and remaining challenges for developing NeuroSERMs. Brain Research. Brain Research Reviews, 49, 472–493. Zhou, J., Zhang, H., Cohen, R. S., & Pandey, S. C. (2005). Effects of estrogen treatment on expression of brain-derived neurotrophic factor and cAMP response element-binding protein expression and phosphorylation in rat amygdaloid and hippocampal structures. Neuroendocrinology, 81, 294–310.

Diagnostic Genome Profiling in Mental Retardation David A. Koolen, Joris A. Veltman, and Bert B.A. de Vries

Abstract  Mental retardation occurs in 2–3% of the general population. Chromosomal aberrations are one of the major causes of mental retardation, but despite the significant progresses in the elucidation of mental retardation, the genetic causes of mental retardation remain largely unknown. Conventional karyotyping using light microscopy has been the primary tool for diagnosing chromosomal aberrations in mental retardation for more than 30 years. Several novel methods based on fluorescence in situ hybridization (FISH) and polymerase chain reaction (PCR)-based methods have been developed over recent years to increase the detection yield of copy-number changes (CNVs) at the submicroscopic level (<5–10 Mb) in individuals with mental retardation. In the last few years, genome-wide microarray technologies have resulted in significant increases in the resolution of chromosome analysis. Microarray ­technologies allow genomewide detection of multiple genomic submicroscopic CNVs. The implementation of these novel molecular-cytogenetic technologies not only showed that submicroscopic genomic aberrations are an important cause of mental retardation, resulting in newly recognized microdeletion/microduplication syndromes, but also allowed for the identification of novel genes causing mental retardation. Keywords  Copy-number variants • Genome profiling • Mental retardation • Microarray • Molecular karyotyping • Syndrome

1 Introduction Mental retardation can be defined as a significant limitation of intellectual ­functioning and of adaptive behavior originating before the age of 18 years (Luckasson et al., 2002). It is a complex disorder and often involves slow learning

B.B.A. de Vries (*) Department of Human Genetics, Nijmegen Centre for Molecular Life Sciences, Radboud University Nijmegen Medical Centre, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_7, © Springer Science+Business Media, LLC 2011

177

178

D.A. Koolen et al.

of basic motor and language skills during childhood, and a significantly below -normal global intellectual capacity as an adult. The prevalence of mental retardation is estimated to be in the order of 2–3% of the general population (Leonard et al., 2002). The prevalence, the lifelong severity and poor curability of mental retardation emphasize the profound impact on families and the high disease burden to society. The American Association on Mental Retardation subdivides the disorders associated with mental retardation into three general areas: prenatal causes, perinatal causes, and postnatal causes. Prenatal causes can be divided into two groups: ­exogenic factors that may cause brain damage, such as intra-uterine infections, excess maternal alcohol consumption during pregnancy and premature birth, and genetic factors. The latter group includes chromosome aberrations and single gene mutations, and are the most common cause of the severe mental retardation accounting for 20–30% of the cases (McLaren & Bryson, 1987). Chromosome aberrations can be divided into changes of overall copy number (aneuploidy), changes of copy number over a specific region (segmental ­aneuploidy), and chromosome rearrangements. These structural chromosome aberrations are often associated, in addition to mental retardation, with dysmorphisms and congenital malformations. Genes that influence cognitive function have predominantly been found on the X chromosome. This might be because of the gene richness of the X chromosome but also because of the ease of identifying families with a X-linked pattern of mental retardation (Ropers et al., 2005). More than 60 genes on the X chromosome are known to be involved in mental retardation (http://xlmr.interfree.it/home.htm) (de Brouwer et al., 2007), whereas the majority of autosomal mental retardation genes await identification. Despite the significant progresses in the elucidation of mental retardation, the genetic causes of mental retardation remain largely unknown (Chelly et al., 2006; Raymond & Tarpey, 2006; Ropers & Hamel, 2005). In clinical practice, even after extensive investigations, about half the individuals with mental retardation remain without a diagnosis (de Vries et  al., 1997; Moog, 2005; Rauch et al., 2006). Various molecular genetic testing using DNA-based methods can be used for diagnosing specific single-gene disorders, whereas conventional cytogenetic testing can only detect chromosome rearrangements spanning more than 5–10 million basepairs (bp) or 5–10 megabases (Mb). More recently developed molecular cytogenetic methods, such as genomic microarrays technologies (Pinkel et al., 1998; Solinas-Toldo et al., 1997), provide powerful tools to bridge the technical divide between molecular genetic testing and conventional cytogenetics. These genomic microarray technologies allow us to detect all DNA ­copy-number variants (CNVs) in the human genome in a single experiment. The implementation of genomic microarrays was a revolution in the mental retardation research. Genomic microarray technologies showed that de  novo CNVs that are too small to be detected by conventional cytogenetic techniques (so-called submicroscopic rearrangements or microdeletions/microduplications) form an important cause of mental retardation, ushering in a new era in diagnostic genome profiling in mental retardation.

Diagnostic Genome Profiling in Mental Retardation

179

2 Conventional Cytogenetics Conventional cytogenetics (conventional karyotyping) refers to the analysis of metaphase chromosomes using light microscopy. It includes routine analysis of Giemsabanded chromosomes (G-banding) and other cytogenetic banding techniques. Since the 1970s, G-banding has been the primary tool for diagnosing ­chromosomal aberrations in overall copy number (aneuploidy), changes in copy number over a region (segmental aneuploidy), and chromosome rearrangements (Mao & Pevsner, 2005). Conventional cytogenetics analysis in mental retardation started with the discovery of the correct number of chromosomes in man of 46 (23 pairs) (Ford & Hamerton, 1956a, 1956b; Tjio & Levan, 1956). This discovery enabled the identification of several numerical chromosome aberrations in individuals with mental retardation and/or congenital malformations, like trisomy 21 in Down syndrome (Lejeune & Strong, 1959), 45,X in Turner syndrome (Ford et al., 1959), 47,XXY in Klinefelter syndrome (Jacobs & Strong, 1959), trisomy 13 (Patau et  al., 1960), and trisomy 18 (Edwards et  al., 1960). The implementation of chromosome banding techniques and the improvements of cell culture methods from the 1970s resulted in a huge increase in the number of routine cytogenetic studies and in the detection of various structural aberrations causally related to mental retardation and malformation syndromes, such as translocations, inversions, deletions, and duplications (Smeets, 2004). Conventional cytogenetic banding is widely used in diagnostic genome profiling. However, the resolution of chromosome banding techniques is limited to approximately 500 bands per haploid genome, i.e., only CNVs larger than ~5–10 Mb can be detected (Raymond &Tarpey, 2006). In addition, conventional karyotyping is labor-intensive and difficult to automate, rendering these methods unsuitable for high throughput routine genome profiling. High-resolution banding (Yunis, 1976) allows the detection of cryptic CNVs usually not seen with conventional banding (so-called submicroscopic deletions or microdeletions). It involves the fixation of cells at an earlier stage of mitosis such as prometaphase when the chromosomes are not fully contracted and can achieve 1,000 chromosome bands in a haploid genome (3–5 Mb) (Shaffer & Bejjani, 2004). Using high-resolution chromosome banding, several well-known clinical syndromes associated with mental retardation are found to be caused by small chromosome deletions that cannot be detected using routine chromosome analyses, such as Prader–Willi and Angelman syndrome (15q11.2q13) (Ledbetter et al., 1981; Magenis et al., 1987, 1990; Malcolm et al., 1991), Miller–Dieker syndrome (17p13.3) (Dobyns et al., 1991), and DiGeorge/velocardiofacial syndrome (22q11.2) (Scambler et al., 1992). However, high-resolution banding is a laborintensive application and is not used routinely in the clinical cytogenetics laboratory.

3 Molecular Cytogenetics 3.1 Fluorescence In Situ Hybridization Fluorescence in situ hybridization (FISH) (Trask, 1991) is a cytogenetic technique that can be used to detect and localize the presence or absence of specific small

180

D.A. Koolen et al.

chromosome loci. The main application of FISH in the diagnostic process of individuals with mental retardation is the detection of subtelomere rearrangements and known microdeletion syndromes such as Prader–Willi and Angelman syndrome (15q11.2q13), Smith Magenis syndrome (17p11.2) (Smith et al., 1986), Miller–Dieker syndrome (17p13.3), and DiGeorge/velocardiofacial syndrome (22q11.2). For FISH analyses, large fragments of DNA, such as bacterial artificial chromosome (BAC) clones (~40,000–200,000 bp) are labeled with a fluorescent dye and hybridized to either metaphase or interphase chromosome spreads. Next, fluorescence microscopy can be used to assess if and where the fluorescent probe has bound to the chromosome. FISH using locus-specific probes is a powerful tool for identifying subtle ­submicroscopic aberrations, including microdeletions associated with clinically wellknown syndromes. FISH is predominately used to confirm a clinical diagnosis, because it can only screen a limited number of genomic targets in a single hybridization experiment. FISH-based techniques using multiple fluorescence dyes, such as Multiplex-FISH (M-FISH) (Speicher et al., 1996) and spectral karyotyping (SKY) (Lu et al., 2002) enable simultaneous visualization of all pairs of chromosomes in different colors allowing the identification of complex or submicroscopic abnormalities. Other multiprobe FISH applications can interrogate the telomeres in a routine procedure (Knight et al., 1999). Rearrangements involving telomeres are an important cause of human genetic diseases, for example, deletion 4pter (Wolf–Hirschhorn syndrome), deletion 5pter (Cri du Chat syndrome), and deletion 17pter (Miller–Dieker syndrome) (de Vries et al., 2003). Submicroscopic subtelomeric chromosome anomalies can be identified in 3–5% of individuals with unexplained mental retardation (de Vries et al., 2003; Flint et al., 1995; Flint & Knight, 2003; Ravnan et al., 2006). Several polymerase chain reaction (PCR)-based methods also allow routine diagnostic subtelomere screening, such as multiplex amplifiable probe hybridization (MAPH) (Sismani et al., 2001), multiplex ligation-dependent probe amplification (MLPA) (Koolen et  al., 2004; Rooms et al., 2004; Schouten et al., 2002), and quantitative PCR (Boehm et al., 2004). The introduction of these routine screening technologies of the subtelomeric regions led to the characterization of new mental retardation syndromes associated with submicroscopic chromosome deletions, such as the 1p36 microdeletion syndrome (Slavotinek et al., 1999). Moreover, subtelomeric deletions of 1q44 (de Vries et al., 2001), 2q37.3 (Casas et al., 2004), 3q29 (Willatt et al., 2005), 9q34 (Stewart et al., 2004), and 22q13 (Nesslinger et  al., 1994) were identified in numerous individuals with mental ­retardation and are now considered to cause recognizable syndromes.

3.2 Molecular Karyotyping The identification of microdeletions in clinical syndromes and the detection of cryptic subtelomeric aberrations using high-resolution chromosome banding and FISH analyses suggested that numerous submicroscopic CNVs are present at various sites within the human genome, in particular in patients suffering from mental retardation. The

Diagnostic Genome Profiling in Mental Retardation

181

detection of these cryptic chromosomal anomalies associated with mental retardation became feasible with the introduction of array-based comparative genomic hybridization (array CGH), also referred to as “molecular karyotyping.” Molecular karyotyping refers to all technologies that allow a genome-wide detection of multiple copy-number changes at the submicroscopic level (Vermeesch et  al., 2005; Vissers et  al., 2005). These microarray-based technologies can be divided in clone-based and oligonucleotide-based microarray systems, based on the microarray target used. 3.2.1 Clone-Based Genomic Microarrays Array CGH using clone-based genomic microarray platforms (Pinkel et al., 1998; Solinas-Toldo et al., 1997) is a molecular method to detect chromosome CNVs, at the submicroscopic level. Clone-based genomic microarray platforms contain genomic fragments from large-insert genomic clones, such as cosmids, P1, phage artificial chromosomes (PAC) clones (Solinas-Toldo et  al., 1997), BAC clones (Pinkel et al., 1998), and cDNA fragments (Pollack et al., 1999). The development of these clone-based microarrays was mediated by (1) the generation of genomewide clone resources integrated into the finished human genome sequence, (2) the development of high throughput microarray platforms, and (3) the optimization of comparative genomic hybridization (CGH) protocols and data analysis systems (Vissers et al., 2005). CGH was developed in the early 1990s (Kallioniemi et al., 1992). In this application, equal amounts of patient DNA and normal reference DNA are fluorescently labeled with different colors and hybridized to normal human metaphase chromosomes. The DNA copy number is proportional to the test/reference fluorescence ratio. However, the detection of small, cryptic aberrations is limited, because metaphase chromosomes are used as the hybridization target (Kallioniemi et  al., 1992). Clone-based genomic microarray platforms do not use metaphase chromosomes, but mapped DNA fragments which are immobilized on glass slides (Pinkel et al., 1998; Solinas-Toldo et al., 1997). Therefore, the resolution of these microarrays only depends on the size and the number of targets selected. The principle of genomic array CGH is outlined in Fig. 1. The resolution of microarrays containing collections of large-insert genomic clones is limited by the size of the targets used: aberrations below 100 kilobases (kb) cannot be detected, because of the size of the targets is between 100 and 200 kb. The production of microarrays containing more than 100,000 targets is not practically achievable. Moreover, the implementation of clone-based genomic microarray platforms requires a dedicated microarray facility hampering the widespread implementation in a routine diagnostic setting. 3.2.2 High-Density Oligonucleotide Microarrays High-density oligonucleotide arrays are microarray platforms encompassing large numbers of short DNA targets (oligonucleotides). Most of these microarrays

182

D.A. Koolen et al. Test DNA

Reference DNA

differential labeling

cohybridization with Cot1-DNA

Log2 Test/Reference

unique target sequence labeled DNA

microarray slide 1,0

gain

detection of labeled DNA

computer analyses

0,3 0 –0,3 loss –1,0 Clones ordered by Mb position on chromosome

Fig. 1  The principle of genomic array CGH. Equal amounts of isolated and fragmented genomic test DNA and reference DNA are differentially labeled using fluorescence dyes. Subsequently, test (green) and reference samples (red) are mixed with Cot-1 DNA, co-precipitated, and resuspended in a hybridization solution. After denaturation of probe and target DNA, the DNA mix is hybridized to the target DNA on the glass slide. After several washing steps, images of the fluorescent signals are captured and the ratio of patient over reference signals is quantified digitally for each target on the array. Copy number is related to the test/reference fluorescence ratio on the array targets

encompass 25- to 85-mer oligonucleotides targeting random genomic sequences (Barrett et  al., 2004; Selzer et  al., 2005; van den IJssel et  al., 2005) or single ­nucleotide polymorphisms (SNPs) (Friedman et  al., 2006; Huang et  al., 2004; Nannya et al., 2005; Peiffer et al., 2006; Slater et al., 2005). The advantages of these platforms are numerous: (1) they provide a higher genome coverage than most clone-based genomic microarrays, (2) they can be produced in large quantities according to industrial quality standards, (3) they are available to all researchers,

Diagnostic Genome Profiling in Mental Retardation

183

including those without dedicated microarray facilities, and (4) their widespread use generates large comparable datasets that facilitate data comparison and scientific cooperation among research groups. SNP oligonucleotide arrays (SNP arrays) allow the concurrent analysis of loss of heterozygosity (LOH) and copy-number variation (Zhou et al., 2004). The SNP signal intensities are analyzed to identify chromosomal copy-number changes. It is also important to be aware that homozygosity might reflect a hemizygous deletion. Both approaches complement and also cross-validate each other, resulting in a powerful molecular genetic tool for the high resolution screening of genomic alterations. Most SNP array platforms require the amplification of portions of the genome by PCR in order to reduce the complexity of the genome. Genomic DNA is digested with a restriction enzyme, and ligated to adapters. A generic primer recognizing the adapter sequence is used to selectively amplify PCR fragments within a certain size range. The amplified DNA is fragmented, labeled, and hybridized to the arrays targeting 104–105 SNPs. The different high-density oligonucleotide microarray platforms available require either a one-color hybridization of labeled genomic DNA to a microarray, or a two-color hybridization in which differentially labeled test and reference DNAs are co-hybridized. The interpretation of the microarrays requiring one-color hybridization involves an in silico comparison against control DNAs hybridized separately to identical microarrays, i.e., the genomic profiles obtained by these arrays are comparable to that of two-color microarrays.

4 High Throughput Screening Technologies in Mental Retardation 4.1 Identification of Genes in Known Mental Retardation Syndromes Many previously described clinical syndromes are associated with mental retardation, dysmorphisms and congenital anomalies, but for the majority of these syndromes, the underlying cause is still unknown. Screening for DNA CNV using array CGH is able to disclose an underlying gene defect in well-known syndromic and previously elusive mental retardation disorders, for example in the CHARGE syndrome. CHARGE is an acronym which stands for “Coloboma of the eye, Heart defects, Atresia of the choanae, Retardation of growth and/or development, Genital and/or urinary abnormalities, and Ear abnormalities and deafness.” Array CGH revealed two microdeletions on 8q12 in two CHARGE patients, and, based on these imbalances, mutations in the CHD7 gene were identified as the cause of this disorder (Vissers et al., 2004). Since then, CHD7 mutation analysis has shown that mutations in the gene account for the majority of the cases with CHARGE syndrome (Jongmans et  al., 2006). Other examples are the autosomal recessive Peters-Plus

184

D.A. Koolen et al.

syndrome, the autosomal dominant Pitt–Hopkins syndrome, and mutations in X-linked mental retardation. Peters-Plus syndrome is characterized by the association of Peters’ congenital glaucoma with dwarfism, mental retardation, facial ­dysmorphism including ear abnormalities, and cleft palate. Biallelic truncating mutations in the B3GALTL gene were identified in patients with Peters-Plus ­syndrome after detection of a microdeletion at 13q12.3q13.1 by array CGH (Lesnik Oberstein et al., 2006). In the same way, mutations in TCF4 (18q21.2) were identified in Pitt–Hopkins syndrome (mental retardation, facial dysmorphism, and intermittent hyperventilation) (Amiel et  al., 2007; Zweier et  al., 2007) and also in ZNF674 (Xp11.3), in individuals with X-linked mental retardation (Lugtenberg et al., 2006).

4.2 Targeted Microarray Applications Microarrays covering different parts of the human genome, such as the subtelomeric regions (Harada et  al., 2004; Veltman et  al., 2002), distal 1p36 (Yu et  al., 2003), 4pter (van Buggenhout et al., 2004), 6p25 (Ekong et al., 2004), 15q11q13 (Locke et al., 2004), proximal 17p (Shaw et al., 2004), and Xq26.1q27.3 (Solomon et  al., 2004), have been constructed for various research purposes. Additionally, arrays covering individual chromosomes have been developed, such as microarrays covering chromosome 18 (Veltman et al., 2003), 20 (Pinkel et al., 1998), 22 (Bruder et  al., 2001; Buckley et  al., 2002), and chromosome X (Bauters et  al., 2005; Veltman et  al., 2004). Other targeted genomic arrays are designed to interrogate regions that are flanked by segmental duplications or low-copy repeats (LCRs) which provide substrates for recombination and recurrent CNVs (Sharp et al., 2005, 2006). Finally, specially designed microarrays targeted to specific loci of known clinical significance can be used in the diagnostic screening for CNVs in mental retardation (Lu et  al., 2007; Shaffer et  al., 2006). These arrays target genomic regions of known clinical significance in order to facilitate the clinical interpretation of submicroscopic CNVs in a diagnostic setting (Lu et al., 2007; Shaffer et al., 2006). In most cases, parental samples are not requisite for the clinical interpretation of the microarray findings, and much is known about the clinical consequences of these submicroscopic CNVs, facilitating the genetic counseling of the families. However, in contrast to genome-wide microarrays, targeted approaches miss ­sporadic CNVs in mental retardation, as these are not spotted on the microarrays.

4.3 Genome-Wide Screening for DNA Copy-Number Variation Genome-wide genomic microarrays cover the human genome at various resolutions. The first arrays for measurement of CNVs across the human genome were assembled in 2001 (Snijders et al., 2001). These first arrays contained approximately 2,400

Diagnostic Genome Profiling in Mental Retardation

185

BAC clones covering the genome with an average of 1 clone per Mb, whereas the current tiling path resolution BAC arrays (Ishkanian et  al., 2004) and oligonucleotide microarrays have a resolving power of less than 200 kb. Several pilot studies showed the diagnostic value of the genome-wide 1-Mb resolution microarrays in mental retardation (Schoumans et al., 2005; Shaw-Smith et al., 2004; Vissers et al., 2003). In these studies, de novo submicroscopic genome imbalances were identified in approximately 10–15% of individuals with mental retardation (Schoumans et al., 2005; Shaw-Smith et al., 2004; Vissers et al., 2003). The clinical usefulness of genome profiling was underscored in larger cohorts of patients with unexplained mental retardation, using 0.5- to 1-Mb resolution BAC arrays (Engels et  al., 2007; Krepischi-Santos et  al., 2006; Menten et  al., 2006; Miyake et al., 2006; Rosenberg et al., 2006; Tyson et al., 2005), tiling-resolution BAC arrays (de Vries et al., 2005), and 100-k SNP arrays (Friedman et al., 2006). In the latter genome-wide studies, the diagnostic yield of de  novo CNVs among unselected individuals with unexplained mental retardation is approximately 10% (5–12%). Obviously, the diagnostic yield of genome-wide diagnostic approaches largely depends on the previous cytogenetic studies performed, patient selection, and the microarray used. The majority of de novo CNVs identified using genomewide array CGH technologies are single and unique cases. Therefore, the interpretation of the results is a skilled process, requiring a joint approach between cytogeneticists, molecular geneticists, and medical professionals. The interpretation of the diagnostic consequences of novel submicroscopic rearrangements is facilitated by international databases that capture (sub) microscopic chromosome rearrangements, such as DECIPHER (www.sanger.ac.uk/PostGenomics/decipher/) and ECARUCA (www.ECARUCA.net). 4.3.1 Identification of New Genomic Disorders Clinical syndromes which are caused by genomic rearrangements mediated by the flanking LCRs (also known as segmental duplications) are called genomic disorders (Lupski, 1998; Stankiewicz & Lupski, 2002). Several genomic disorders are associated with mental retardation, such as Williams–Beuren, Prader–Willi, Angelman, Smith– Magenis, Sotos, and DiGeorge/velocardiofacial syndrome (Stankiewicz & Lupski, 2002; Visser et al., 2005) (Table 1). Flanking LCRs predispose these regions to recurrent rearrangement by non-allelic homologous recombination (NAHR) (Stankiewicz & Lupski, 2002). During meiosis, mispairing occurs between non-allelic homologous LCRs and subsequent crossing over between strands results in duplication or deletion of the intervening sequence. Many genomic disorders, such as DiGeorge/velocardialfacial syndrome, Williams–Beuren syndrome, Prader–Willi syndrome, and Angelman syndrome, had been clinically recognized long before the underlying cytogenetic cause was elucidated. However, the widespread clinical implementation of high-resolution genome-wide screening techniques enables the delineation of new genomic disorders after screening of large heterogeneous cohorts of individuals with mental

186

D.A. Koolen et al.

Table 1  Genomic disorders associated with mental retardation Name Del/Dup Location Size (Mb) Sotos syndrome Deletion 5q35 0.7 Williams–Beuren syndrome Deletion 7q11.23 1.6 7q11.23 duplication syndrome Duplication 7q11.23 1.6 10q22q23 deleton syndrome Deletion 10q22q23 7.2 Angelman/Prader–Willi syndrome Deletion 15q11.2q13a 3.5 15q24 deleton syndrome Deletion 15q24 1.7–3.9 Smith–Magenis syndrome Deletion 17p11.2 3.7 17p11.2 duplication syndrome Duplication 17p11.2 4.0 Charcot-Marie-Tooth disease 1A Duplication 17p12 1.4 17q21.31 deletion syndrome Deletion 17q21.31 0.6 Velocardiolfacial syndrome Deletion 22q11.2 3.0/1.5 a Angelman syndrome, maternal deletion; Prader–Willi syndrome, paternal deletion b Mutations in UBE3A result in Angelman syndrome in a subset of cases

Gene NSD1 – – – UBE3Ab – RAI1 – PMP22 – –

retardation. As an example, recurrent overlapping de  novo microdeletions at 17q21.31 were identified in patients with mental retardation using array CGH (Koolen et al., 2006; Sharp et al., 2006; Shaw-Smith et al., 2006). Detailed clinical comparison of these patients revealed marked phenotypic similarities in these cases, showing that the deletion is pathogenic and represents a previously undescribed recurrent microdeletion syndrome (Koolen et al., 2006; Sharp et al., 2006; Shaw-Smith et al., 2006). This new syndrome is characterized by mental retardation, hypotonia, and characteristic face, including a long hypotonic face with ptosis, blepharophimosis, large and low-set ears, tubular pear-shaped nose with bulbous nasal tip, long columella with hypoplastic alae nasi, and a broad chin (Koolen et al., 2006; Sharp et al., 2006; Shaw-Smith et al., 2006). Other overlapping CNVs that represent new genomic disorders are deletions 10q22q23 (Balciuniene et al., 2007) and deletions 15q24q24 (Sharp et al., 2007). The reciprocal duplications are predicted to occur at the same frequency. The 17p11.2 microduplication syndrome (also referred to as “Potocki–Lupski syndrome”) was the first predicted reciprocal microduplication syndrome described, being the homologous recombination reciprocal of the Smith–Magenis syndrome microdeletion del(17)(p11.2p11.2) (Potocki et al., 2000). Similarly, reciprocal microduplications of the Williams syndrome have been reported (Kirchhoff et  al., 2007; Somerville et al., 2005).

5 Copy-Number Variants in the General Population The increased resolution of genome-wide microarray techniques revealed the presence of CNVs in the general population (Iafrate et  al., 2004; Sebat et  al., 2004). A firstgeneration map of CNVs in healthy individuals showed that up to approximately 12% of the human genome involve CNVs (Redon et  al., 2006). These CNVs of DNA sequences range from kilobases to megabases in size (Conrad et al., 2006; Hinds et al.,

Diagnostic Genome Profiling in Mental Retardation

187

2006; Iafrate et al., 2004; McCarroll et al., 2006; Sebat et al., 2004; Sharp et al., 2006; Tuzun et al., 2005). In genome-wide diagnosis of mental retardation, these CNVs challenge the medical professional interpretation of test results and the establishment of accurate translation to the clinical phenotype. Using parent–offspring trios, Redon et al. demonstrated that the vast majority of CNVs in healthy individuals are inherited and so far no de  novo occurrence of CNVs has been demonstrated (Redon et  al., 2006). Therefore, de novo occurrence of a CNV indicates a role in the development of human disease.

6 The Clinical Consequences of Submicroscopic Copy-Number Variants The interpretation of molecular findings obtained by high-resolution genome-wide screening studies performed in individuals with sporadic forms of mental retardation is a skilled process based on a combination of clinical and molecular variables. First, it is essential to rule out inherited CNVs from unaffected parents before drawing any conclusions about whether an aneusomic segment is causative for the mental retardation and/or congenital anomalies. It is of note that inherited genome imbalances might still be important in the pathogenesis of mental retardation, as the aneusomic segment can unmask a recessive mutation of a gene at the other allele. Moreover, great clinical variation and reduced penetrance of the clinical phenotype has to be considered, as can be observed for instance in the DiGeorge/velocardiofacial syndrome (Lindsay, 2001). Second, it is important to check whether an alteration has been described previously in a large cohort of unaffected individuals, since this will reduce the likelihood that it is causative. Third, the description of patients with similar chromosomal aberrations will aid in the clinical interpretation. Finally, the presence of protein coding genes within a chromosomal aberration may underscore the causative effect of a submicroscopic aberration. It is useful to know whether these proteins are expressed in a dosage-dependent fashion, whether a specific function and expression pattern of the gene makes sense in the light of the disorder, whether the gene has sequence or functional homology with genes associated with human disease, and finally, whether there are animal studies indicating a functional role for the protein. The difficulties in the interpretation of the clinical consequences of submicroscopic CNVs are illustrated in the following case. A severely mentally retarded girl was referred for genome-wide CNV screening using genomic array CGH (Fig. 2a). Previous routine chromosome analysis, subtelomeric analysis and routine metabolic screening had not shown any abnormalities. Array CGH revealed a 2.3-Mb gain at 12q24.21q24.23 (Fig. 2b) that was confirmed and proven to be de novo using MLPA (Schouten et al., 2002). MLPA is a PCRbased method that permits the detection of copy-number loss or copy-number gain of a specific genomic region of interest. Subsequently, FISH analysis showed that

188

D.A. Koolen et al.

b

2

Log2 Test /Reference

this copy-number gain represented an in-tandem microduplication within the 12q24.21q24.23 region. This small microduplication had never been reported before. However, several factors indicate that this submicroscopic CNV is indeed the cause of the pathology of this girl. Strongly supportive of a pathogenic effect of the microduplication is the de novo occurrence. Moreover, several clinical features, such as mental retardation, microcephaly, short stature, recurrent infections, hypotonia, hypertelorism, epicanthal folds, and a broad nasal bridge, were also described in patients with larger duplications overlapping the 12q24.21q24.23 region. Finally, the duplicated region contained several genes, such as THRAP2, RFC5 and NOS1, that are expressed in the brain and/or are involved in embryogenesis. Based on this information, the 12q24.21q24.23 microduplication identified in this girl was reported to be causative for the developmental delay/learning disability and multiple congenital anomalies (Ruiter et al., 2006).

1 0.3 0 −0.3 −1

−2

0

50 100 Clones ordered by Mb position on chromosome 12

Fig.  2  (a) The proband at 7.5 years of age. Note the epicanthal folds, hypertelorism, arched ­eyebrows, low set ears, short philtrum, open mouth appearance, full lips, and irregular position of the lower teeth. (b) Chromosome 12 plot as obtained by array comparative genomic hybridization. The 2.3-Mb gain of 12q24.21q24.23 is indicated by an arrow

Diagnostic Genome Profiling in Mental Retardation

189

7 Conclusions and Future Prospects Conventional karyotyping using light microscopy has been the primary tool for diagnosing chromosomal aberrations in mental retardation for more than 30 years. In the last few years, novel genome-wide microarray technologies have resulted in a significant increase in the resolution of chromosome analysis, and have revolutionized the genome-wide profiling in individuals with mental retardation. Microarray technologies allow a genome-wide detection of multiple CNVs at the submicroscopic level, and are leading to a new era of “molecular karyotyping.” Acknowledgements  This work was supported by grants from the Netherlands Organization for Health Research and Development (ZON-MW) (D.A.K., J.A.V., and B.B.A.d.V.), and the Hersenstichting Nederland (B.B.A.d.V.).

References Amiel, J., et al. (2007). Mutations in TCF4, encoding a class I basic helix-loop-helix transcription factor, are responsible for Pitt-Hopkins syndrome, a severe epileptic encephalopathy associated with autonomic dysfunction. American Journal of Human Genetics, 80(5), 988–993. Balciuniene, J., et al. (2007). Recurrent 10q22-q23 deletions: A genomic disorder on 10q associated with cognitive and behavioral abnormalities. American Journal of Human Genetics, 80(5), 938–947. Barrett, M. T., et al. (2004). Comparative genomic hybridization using oligonucleotide microarrays and total genomic DNA. Proceedings of the National Academy of Sciences of the United States of America, 101(51), 17765–17770. Bauters, M., et  al. (2005). X chromosome array-CGH for the identification of novel X-linked mental retardation genes. European Journal of Medical Genetics, 48(3), 263–275. Boehm, D., et al. (2004). Rapid detection of subtelomeric deletion/duplication by novel real-time quantitative PCR using SYBR-green dye. Human Mutation, 23(4), 368–378. Bruder, C. E., et al. (2001). High resolution deletion analysis of constitutional DNA from neurofibromatosis type 2 (NF2) patients using microarray-CGH. Human Molecular Genetics, 10(3), 271–282. Buckley, P. G., et  al. (2002). A full-coverage, high-resolution human chromosome 22 genomic microarray for clinical and research applications. Human Molecular Genetics, 11(25), 3221–3229. Casas, K. A., et al. (2004). Chromosome 2q terminal deletion: report of 6 new patients and review of phenotype-breakpoint correlations in 66 individuals. American Journal of Medical Genetics. Part A, 130(4), 331–339. Chelly, J., et al. (2006). Genetics and pathophysiology of mental retardation. European Journal of Human Genetics, 14(6), 701–713. Conrad, D. F., et  al. (2006). A high-resolution survey of deletion polymorphism in the human genome. Nature Genetics, 38(1), 75–81. de Brouwer, A. P., et  al. (2007). Mutation frequencies of X-linked mental retardation genes in families from the EuroMRX consortium. Human Mutation, 28(2), 207–208. de Vries, B. B., et al. (1997). Screening and diagnosis for the fragile X syndrome among the mentally retarded: an epidemiological and psychological survey. Collaborative Fragile X Study Group. American Journal of Human Genetics, 61(3), 660–667.

190

D.A. Koolen et al.

de Vries, B. B., et al. (2001). Submicroscopic subtelomeric 1qter deletions: a recognisable phenotype? Journal of Medical Genetics, 38(3), 175–178. de Vries, B. B., et al. (2003). Telomeres: A diagnosis at the end of the chromosomes. Journal of Medical Genetics, 40(6), 385–398. de Vries, B. B., et  al. (2005). Diagnostic genome profiling in mental retardation. American Journal of Human Genetics, 77(4), 606–616. Dobyns, W. B., et  al. (1991). Clinical and molecular diagnosis of Miller-Dieker syndrome. American Journal of Human Genetics, 48(3), 584–594. Edwards, J. H., et al. (1960). A new trisomic syndrome. Lancet, 1, 787–790. Ekong, R., et al. (2004). Chromosomal anomalies on 6p25 in iris hypoplasia and Axenfeld-Rieger syndrome patients defined on a purpose-built genomic microarray. Human Mutation, 24(1), 76–85. Engels, H., et al. (2007). DNA microarray analysis identifies candidate regions and genes in unexplained mental retardation. Neurology, 68(10), 743–750. Flint, J., et al. (1995). The detection of subtelomeric chromosomal rearrangements in idiopathic mental retardation. Nature Genetics, 9(2), 132–140. Flint, J., & Knight, S. (2003). The use of telomere probes to investigate submicroscopic rearrangements ­associated with mental retardation. Current Opinion in Genetics & Development, 13(3), 310–316. Ford, C. E., & Hamerton, J. L. (1956a). A colchicine, hypotonic citrate, squash sequence for mammalian ­chromosomes. Stain Technology, 31(6), 247–251. Ford, C. E., & Hamerton, J. L. (1956b). The chromosomes of man. Nature, 178(4541), 1020–1023. Ford, C. E., et al. (1959). A sex-chromosome anomaly in a case of gonadal dysgenesis (Turner’s syndrome). Lancet, 1(7075), 711–713. Friedman, J. M., et  al. (2006). Oligonucleotide microarray analysis of genomic imbalance in children with mental retardation. American Journal of Human Genetics, 79(3), 500–513. Harada, N., et al. (2004). Subtelomere specific microarray based comparative genomic hybridisation: A rapid detection system for cryptic rearrangements in idiopathic mental retardation. Journal of Medical Genetics, 41(2), 130–136. Hinds, D. A., et  al. (2006). Common deletions and SNPs are in linkage disequilibrium in the human genome. Nature Genetics, 38(1), 82–85. Huang, J., et  al. (2004). Whole genome DNA copy number changes identified by high density oligonucleotide arrays. Human Genomics, 1(4), 287–299. Iafrate, A. J., et  al. (2004). Detection of large-scale variation in the human genome. Nature Genetics, 36(9), 949–951. Ishkanian, A. S., et al. (2004). A tiling resolution DNA microarray with complete coverage of the human genome. Nature Genetics, 36(3), 299–303. Jacobs, P. A., & Strong, J. A. (1959). A case of human intersexuality having a possible XXY sexdetermining mechanism. Nature, 183(4657), 302–303. Jongmans, M. C., et al. (2006). CHARGE syndrome: The phenotypic spectrum of mutations in the CHD7 gene. Journal of Medical Genetics, 43(4), 306–314. Kallioniemi, A., et  al. (1992). Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science, 258(5083), 818–821. Kirchhoff, M., et  al. (2007). MLPA analysis for a panel of syndromes with mental retardation reveals imbalances in 5.8% of patients with mental retardation and dysmorphic features, including duplications of the Sotos syndrome and Williams-Beuren syndrome regions. European Journal of Medical Genetics, 50(1), 33–42. Knight, S. J., et  al. (1999). Subtle chromosomal rearrangements in children with unexplained mental retardation. Lancet, 354(9191), 1676–1681. Koolen, D. A., et  al. (2004). Screening for subtelomeric rearrangements in 210 patients with unexplained mental retardation using multiplex ligation dependent probe amplification (MLPA). Journal of Medical Genetics, 41(12), 892–899. Koolen, D. A., et al. (2006). A new chromosome 17q21.31 microdeletion syndrome associated with a common inversion polymorphism. Nature Genetics, 38(9), 999–1001.

Diagnostic Genome Profiling in Mental Retardation

191

Krepischi-Santos, A. C., et  al. (2006). Whole-genome array-CGH screening in undiagnosed syndromic patients: Old syndromes revisited and new alterations. Cytogenetic and Genome Research, 115(3–4), 254–261. Ledbetter, D. H., et  al. (1981). Deletions of chromosome 15 as a cause of the Prader-Willi ­syndrome. The New England Journal of Medicine, 304(6), 325–329. Lejeune, J., et al. (1959). Study of somatic chromosomes from 9 mongoloid children. Competes Rendus Hebdomadaires des Seances de l’Academis des Sciences, 248(11), 1721–1722. Leonard, H., et al. (2002). The epidemiology of mental retardation: challenges and opportunities in the new millennium. Mental Retardation and Developmental Disabilities Research Reviews, 8(3), 117–134. Lesnik Oberstein, S. A., et al. (2006). Peters Plus syndrome is caused by mutations in B3GALTL, a putative glycosyltransferase. American Journal of Human Genetics, 79(3), 562–566. Lindsay, E. A. (2001). Chromosomal microdeletions: dissecting del22q11 syndrome. Nature Reviews. Genetics, 2(11), 858–868. Locke, D. P., et al. (2004). BAC microarray analysis of 15q11-q13 rearrangements and the impact of segmental duplications. Journal of Medical Genetics, 41(3), 175–182. Lu, X. Y., et al. (2002). The utility of spectral karyotyping in the cytogenetic analysis of newly diagnosed pediatric acute lymphoblastic leukemia. Leukemia, 16(11), 2222–2227. Lu, X., et al. (2007). Clinical implementation of chromosomal microarray analysis: summary of 2513 postnatal cases. PLoS ONE, 2, e327. Luckasson, R., Borthwick-Duffy, S., Buntinx, W. H. E., Coulter, D. L., Snell, M. E., Spitalnik, D. M., et al. (2002). Mental retardation: Definition, classification, and systems of support (10th ed.). Washington: American Association on Mental Retardation. Lugtenberg, D., et al. (2006). ZNF674: A new kruppel-associated box-containing zinc-finger gene involved in nonsyndromic X-linked mental retardation. American Journal of Human Genetics, 78(2), 265–278. Lupski, J. R. (1998). Genomic disorders: structural features of the genome can lead to DNA rearrangements and human disease traits. Trends in Genetics, 14(10), 417–422. Magenis, R. E., et  al. (1987). Is Angelman syndrome an alternate result of del(15)(q11q13)? American Journal of Medical Genetics, 28(4), 829–838. Magenis, R. E., et  al. (1990). Comparison of the 15q deletions in Prader-Willi and Angelman syndromes: specific regions, extent of deletions, parental origin, and clinical consequences. American Journal of Medical Genetics, 35(3), 333–349. Malcolm, S., et  al. (1991). Uniparental paternal disomy in Angelman’s syndrome. Lancet, 337(8743), 694–697. Mao, R., & Pevsner, J. (2005). The use of genomic microarrays to study chromosomal abnormalities in mental retardation. Mental Retardation and Developmental Disabilities Research Reviews, 11(4), 279–285. McCarroll, S. A., et al. (2006). Common deletion polymorphisms in the human genome. Nature Genetics, 38(1), 86–92. McLaren, J., & Bryson, S. E. (1987). Review of recent epidemiological studies of mental retardation: prevalence, associated disorders, and etiology. American Journal of Mental Retardation, 92(3), 243–254. Menten, B., et al. (2006). Emerging patterns of cryptic chromosomal imbalance in patients with idiopathic mental retardation and multiple congenital anomalies: a new series of 140 patients and review of published reports. Journal of Medical Genetics, 43(8), 625–633. Miyake, N., et  al. (2006). BAC array CGH reveals genomic aberrations in idiopathic mental retardation. American Journal of Medical Genetics. Part A, 140(3), 205–211. Moog, U. (2005). The outcome of diagnostic studies on the etiology of mental retardation: Considerations on the classification of the causes. American Journal of Medical Genetics. Part A, 137(2), 228–231. Nannya, Y., et al. (2005). A robust algorithm for copy number detection using high-density oligonucleotide single nucleotide polymorphism genotyping arrays. Cancer Research, 65(14), 6071–6079.

192

D.A. Koolen et al.

Nesslinger, N. J., et al. (1994). Clinical, cytogenetic, and molecular characterization of seven patients with deletions of chromosome 22q13.3. American Journal of Human Genetics, 54(3), 464–472. Patau, K., et al. (1960). Multiple congenital anomaly caused by an extra autosome. Lancet, 1, 790–793. Peiffer, D. A., et al. (2006). High-resolution genomic profiling of chromosomal aberrations using Infinium whole-genome genotyping. Genome Research, 16(9), 1136–1148. Pinkel, D., et al. (1998). High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays. Nature Genetics, 20(2), 207–211. Pollack, J. R., et al. (1999). Genome-wide analysis of DNA copy-number changes using cDNA microarrays. Nature Genetics, 23(1), 41–46. Potocki, L., et al. (2000). Molecular mechanism for duplication 17p11.2- the homologous recombination reciprocal of the Smith-Magenis microdeletion. Nature Genetics, 24(1), 84–87. Rauch, A., et  al. (2006). Diagnostic yield of various genetic approaches in patients with unexplained developmental delay or mental retardation. American Journal of Medical Genetics. Part A, 140(19), 2063–2074. Ravnan, J. B., et al. (2006). Subtelomere FISH analysis of 11 688 cases: An evaluation of the frequency and pattern of subtelomere rearrangements in individuals with developmental disabilities. Journal of Medical Genetics, 43(6), 478–489. Raymond, F. L., & Tarpey, P. (2006). The genetics of mental retardation. Human Molecular Genetics, 15 Spec No 2, R110–R116. Redon, R., et  al. (2006). Global variation in copy number in the human genome. Nature, 444(7118), 444–454. Rooms, L., et al. (2004). Subtelomeric deletions detected in patients with idiopathic mental retardation using multiplex ligation-dependent probe amplification (MLPA). Human Mutation, 23(1), 17–21. Ropers, H. H., & Hamel, B. C. (2005). X-linked mental retardation. Nature Reviews. Genetics, 6(1), 46–57. Rosenberg, C., et al. (2006). Array-CGH detection of micro rearrangements in mentally retarded individuals: clinical significance of imbalances present both in affected children and normal parents. Journal of Medical Genetics, 43(2), 180–186. Ruiter, M., et  al. (2006). A novel 2.3 Mb microduplication of 12q24.21q24.23 detected by genome-wide tiling-path resolution array comparative genomic hybridization in a girl with syndromic mental retardation. Clinical Dysmorphology, 15(3), 133–137. Scambler, P. J., et al. (1992). Velo-cardio-facial syndrome associated with chromosome 22 deletions encompassing the DiGeorge locus. Lancet, 339(8802), 1138–1139. Schoumans, J., et al. (2005). Detection of chromosomal imbalances in children with idiopathic mental retardation by array based comparative genomic hybridisation (array-CGH). Journal of Medical Genetics, 42(9), 699–705. Schouten, J. P., et al. (2002). Relative quantification of 40 nucleic acid sequences by multiplex ligation-dependent probe amplification. Nucleic Acids Research, 30(12), e57. Sebat, J., et al. (2004). Large-scale copy number polymorphism in the human genome. Science, 305(5683), 525–528. Selzer, R. R., et al. (2005). Analysis of chromosome breakpoints in neuroblastoma at sub-kilobase resolution using fine-tiling oligonucleotide array CGH. Genes, Chromosomes & Cancer, 44(3), 305–319. Shaffer, L. G., & Bejjani, B. A. (2004). A cytogeneticist’s perspective on genomic microarrays. Human Reproduction Update, 10(3), 221–226. Shaffer, L. G., et al. (2006). Targeted genomic microarray analysis for identification of chromosome abnormalities in 1500 consecutive clinical cases. Jornal de Pediatria, 149(1), 98–102. Sharp, A. J., et  al. (2005). Segmental duplications and copy-number variation in the human genome. American Journal of Human Genetics, 77(1), 78–88. Sharp, A. J., et al. (2006). Discovery of previously unidentified genomic disorders from the duplication architecture of the human genome. Nature Genetics, 38(9), 1038–1042.

Diagnostic Genome Profiling in Mental Retardation

193

Sharp, A. J., et al. (2007). Characterization of a recurrent 15q24 microdeletion syndrome. Human Molecular Genetics, 16(5), 567–572. Shaw, C. J., et  al. (2004). Comparative genomic hybridisation using a proximal 17p BAC/PAC array detects rearrangements responsible for four genomic disorders. Journal of Medical Genetics, 41(2), 113–119. Shaw-Smith, C., et al. (2004). Microarray based comparative genomic hybridisation (array-CGH) detects submicroscopic chromosomal deletions and duplications in patients with learning ­disability/mental retardation and dysmorphic features. Journal of Medical Genetics, 41(4), 241–248. Shaw-Smith, C., et al. (2006). Microdeletion encompassing MAPT at chromosome 17q21.3 is associated with developmental delay and learning disability. Nature Genetics, 38(9), 1032–1037. Sismani, C., et al. (2001). Screening for subtelomeric chromosome abnormalities in children with idiopathic mental retardation using multiprobe telomeric FISH and the new MAPH telomeric assay. European Journal of Human Genetics, 9(7), 527–532. Slater, H. R., et  al. (2005). High-resolution identification of chromosomal abnormalities using oligonucleotide arrays containing 116, 204 SNPs. American Journal of Human Genetics, 77(5), 709–726. Slavotinek, A., et al. (1999). Monosomy 1p36. Journal of Medical Genetics, 36(9), 657–663. Smeets, D. F. (2004). Historical prospective of human cytogenetics: From microscope to microarray. Clinical Biochemistry, 37(6), 439–446. Smith, A. C., et  al. (1986). Interstitial deletion of (17)(p11.2p11.2) in nine patients. American Journal of Medical Genetics, 24(3), 393–414. Snijders, A. M., et al. (2001). Assembly of microarrays for genome-wide measurement of DNA copy number. Nature Genetics, 29(3), 263–264. Solinas-Toldo, S., et  al. (1997). Matrix-based comparative genomic hybridization: Biochips to screen for genomic imbalances. Genes, Chromosomes & Cancer, 20(4), 399–407. Solomon, N. M., et al. (2004). Array comparative genomic hybridisation analysis of boys with X linked hypopituitarism identifies a 3.9 Mb duplicated critical region at Xq27 containing SOX3. Journal of Medical Genetics, 41(9), 669–678. Somerville, M. J., et  al. (2005). Severe expressive-language delay related to duplication of the Williams-Beuren locus. The New England Journal of Medicine, 353(16), 1694–1701. Speicher, M. R., et  al. (1996). Karyotyping human chromosomes by combinatorial multi-fluor FISH. Nature Genetics, 12(4), 368–375. Stankiewicz, P., & Lupski, J. R. (2002). Molecular-evolutionary mechanisms for genomic disorders. Current Opinion in Genetics & Development, 12(3), 312–319. Stewart, D. R., et  al. (2004). Subtelomeric deletions of chromosome 9q: a novel microdeletion syndrome. American Journal of Medical Genetics. Part A, 128(4), 340–351. Tjio, J. H., & Levan, A. (1956). The chromosome number in man. Hereditas, 42, 1–6. Trask, B. J. (1991). Fluorescence in situ hybridization: applications in cytogenetics and gene mapping. Trends in Genetics, 7(5), 149–154. Tuzun, E., et al. (2005). Fine-scale structural variation of the human genome. Nature Genetics, 37(7), 727–732. Tyson, C., et al. (2005). Submicroscopic deletions and duplications in individuals with intellectual disability detected by array-CGH. American Journal of Medical Genetics. Part A, 139(3), 173–185. van Buggenhout, G., et al. (2004). Mild Wolf-Hirschhorn syndrome: Micro-array CGH analysis of atypical 4p16.3 deletions enables refinement of the genotype-phenotype map. Journal of Medical Genetics, 41(9), 691–698. van den IJssel, P., et  al. (2005). Human and mouse oligonucleotide-based array CGH. Nucleic Acids Research, 33(22), e192. Veltman, J. A., et al. (2002). High-throughput analysis of subtelomeric chromosome rearrangements by use of array-based comparative genomic hybridization. American Journal of Human Genetics, 70(5), 1269–1276.

194

D.A. Koolen et al.

Veltman, J. A., et al. (2003). Definition of a critical region on chromosome 18 for congenital aural atresia by arrayCGH. American Journal of Human Genetics, 72(6), 1578–1584. Veltman, J. A., et  al. (2004). High resolution profiling of X chromosomal aberrations by array comparative genomic hybridisation. Journal of Medical Genetics, 41(6), 425–432. Vermeesch, J. R., et al. (2005). Molecular karyotyping: Array CGH quality criteria for constitutional genetic diagnosis. The Journal of Histochemistry and Cytochemistry, 53(3), 413–422. Visser, R., et al. (2005). Identification of a 3.0-kb major recombination hotspot in patients with Sotos syndrome who carry a common 1.9-Mb microdeletion. American Journal of Human Genetics, 76(1), 52–67. Vissers, L. E., et al. (2003). Array-based comparative genomic hybridization for the genomewide detection of submicroscopic chromosomal abnormalities. American Journal of Human Genetics, 73(6), 1261–1270. Vissers, L. E., et al. (2004). Mutations in a new member of the chromodomain gene family cause CHARGE syndrome. Nature Genetics, 36(9), 955–957. Vissers, L. E., et al. (2005). Identification of disease genes by whole genome CGH arrays. Human Molecular Genetics, 14 Spec No 2, R215–R223. Willatt, L., et al. (2005). 3q29 microdeletion syndrome: clinical and molecular characterization of a new syndrome. American Journal of Human Genetics, 77(1), 154–160. Yu, W., et al. (2003). Development of a comparative genomic hybridization microarray and demonstration of its utility with 25 well-characterized 1p36 deletions. Human Molecular Genetics, 12(17), 2145–2152. Yunis, J. J. (1976). High resolution of human chromosomes. Science, 191(4233), 1268–1270. Zhou, X., et al. (2004). Concurrent analysis of loss of heterozygosity (LOH) and copy number abnormality (CNA) for oral premalignancy progression using the Affymetrix 10K SNP mapping array. Human Genetics, 115(4), 327–330. Zweier, C., et al. (2007). Haploinsufficiency of TCF4 causes syndromal mental retardation with intermittent hyperventilation (Pitt-Hopkins Syndrome). American Journal of Human Genetics, 80(5), 994–1001.

Genomic Imprinting and Sexual ExperienceDependent Learning in the Mouse William T. Swaney and Eric B. Keverne

Abstract  Sexual experience is a significant modulator of behaviour in male rodents, inducing mating behaviour changes, increased sexual motivation and olfactory learning. Sexually experienced males exhibit greater motivation to investigate females and show preferences for receptive oestrous females and their odours which are not seen in virgins. These behavioural effects of sexual experience are accompanied by neurobiological changes affecting forebrain sensitivity to steroid hormones, mesolimbic dopamine function and neural activity in the basal hypothalamus and olfactory pathways. These changes suggest that sexually experienced males are better able to detect receptive females, are more motivated to pursue them and are more proficient copulators. Furthermore, this response to sexual experience appears to be mediated by imprinted genes. Imprinted genes are a small class of mammalian autosomal genes that are expressed in parent-of-origin fashion and which are key regulators of placentation and development in mammals. Mice carrying a knockout of the paternally expressed gene Peg3 have deficits in maternal care and offspring development, but Peg3 mutant males also fail to show any sexual experience-dependent changes in behaviour or learned olfactory preferences. There are also no changes in female odour-evoked neural activity in the hypothalamus, vomeronasal system or main olfactory pathway of Peg3 mutants after sexual experience, suggesting a deficit in sexual experience-dependent forebrain plasticity. Peg3 appears to regulate male behavioural traits that would enhance its own transmission down the male line, suggesting that this imprinted gene has evolved to directly influence plasticity in male reproductive behaviour. Keywords  Sexual experience • Experience sexual • Sexual behaviour • Behaviour sexual • Olfactory learning • Learning olfactory • Pheromones • Odours • Genomic

W.T. Swaney (*) Behavioural Biology and Helmholtz Institute, Utrecht University, 3508 TB, Utrecht, The Netherlands e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_8, © Springer Science+Business Media, LLC 2011

195

196

W.T. Swaney and E.B. Keverne

imprinting • Imprinting genomic • Imprinted genes • Genes imprinted • Peg3 • Apoptosis • Dopamine • Mesolimbic dopamine system • Vomeronasal system • Main olfactory system • Mouse • Mice • Rodents • Mammals

1 Introduction Sexual experience has marked effects on male behaviours, modulating sexual behaviour, male–male aggression, anxiety-like behaviour and inducing learned preferences for female cues. These changes in behaviour appear to be mediated by interactions between sensory inputs, the basal forebrain nuclei that control sexual behaviour and dopaminergic pathways involved in sexual reward. The rewarding component of sexual behaviour appears to be crucial to many of the changes associated with sexual experience, suggesting that mechanisms known to be involved in general reward-mediated learning may also play a role in the learned behavioural changes induced by sexual experience in both male and female mammals.

1.1 Olfaction and Male Reproduction In rodents such as rats and mice, the dominant sensory modality is olfaction. Odour cues and pheromones sensed by the main and accessory olfactory system are necessary for the induction of sexual behaviour, and these systems appear to be strongly responsive to sexual experience. The two olfactory systems detect and process behaviourally relevant chemosignals, the main olfactory system primarily sensing volatile odours in the air, while the accessory olfactory system senses non-volatile chemicals detected through physical investigation of conspecifics and their odours (Zufall & Leinders-Zufall, 2007). These odours and pheromones regulate diverse aspects of reproduction in both genders, including female sexual receptivity, lordosis, mate preferences, inter-male and maternal aggression, and mate recognition (Brennan & Zufall, 2006). Male sexual behaviour is dependent on the detection of female odours and pheromones, without which they will not mate regardless of the availability of other sensory cues (Keller, Douhard, Baum, & Bakker, 2006a; Powers & Winans, 1975). This sensitivity to female odour cues is modulated by previous mating experience, and sexually experienced male rodents display increased interest in female odours as well as specific preferences for the odours of receptive females (Hayashi & Kimura, 1974; Stern, 1970). Sexual experience has been shown to have adaptive consequences for male reproductive success, increasing fecundity (Rastogi, Milone, & Chieffi, 1981) and attractiveness to potential

Genomic Imprinting and Sexual Experience-Dependent Learning in the Mouse

197

mates (Galef, Lim, & Gilbert, 2008), and the experience-led changes in olfactory responses to female odours may also have an impact on male sexual success, increasing the ability of males to seek out receptive females based on odour cues. Such responses to sexual experience are important factors when considering the evolution of reproduction in mammals, as individuals that display adaptive learned changes in behaviour after sexual experience are likely to acquire a competitive advantage over rival males when competing for mating opportunities.

1.2 Genomic Imprinting and Reproduction Our research group has been studying the role of imprinted genes in adult behaviour (Champagne, Curley, Swaney, Hasen, & Keverne, 2009; Curley, Barton, Surani, & Keverne, 2004; Swaney, Curley, Champagne, & Keverne, 2007, 2008) and have shown that male responses to sexual experience are among the behaviours that are regulated by the paternally expressed imprinted gene Peg3. Genomic imprinting is an unusual mechanism of gene regulation, first reported 25 years ago, which involves the parent-of-origin expression of a small subset of autosomal genes in mammals (Barton, Surani, & Norris, 1984; McGrath & Solter, 1984). Imprinted genes are not expressed biallelically, but in a monoallelic fashion according to the parent from which they are inherited. There are thus paternally expressed genes, with a functional paternal allele and a silent maternal allele, and maternally expressed genes with a functional maternal allele and a silent paternal allele. The existence of autosomal genes that are expressed in such haploid fashion has aroused great interest since their discovery, particularly as this mechanism of gene regulation dispenses with the protection from mutation which is conferred by diploidy. While there are relatively few imprinted genes – approximately 100 have been identified in humans and mice (Glaser, Ramsay, & Morison, 2006; Morison, Ramsay, & Spencer, 2005) – they are significant regulators of mammalian development and are thought to have played key roles in the evolution of placentation in mammals (Kaneko-Ishino, Kohda, & Ishino, 2003) and the expansion of the brain that has occurred over the course of primate evolution (Keverne, 2001). Imprinted genes have wide-ranging developmental roles and appear to be particularly important in brain development (Wilkinson, Davies, & Isles, 2007). The imprinted gene Peg3 is paternally expressed and is particularly strongly expressed in the hypothalamus, the primary regulatory area for reproductive behaviour in mammals. By studying the behaviour of mice carrying a knockout of Peg3, we have examined the role of this imprinted gene in adult behaviours and shown that Peg3 is important for plasticity in hypothalamically-mediated behaviours. There has been much research into the regulation of embryonic development and metabolism by imprinted genes, but we have attempted to examine the regulation of adult reproductive behaviours by imprinted genes to explore the importance of adult behaviour on the evolution of genomic imprinting.

198

W.T. Swaney and E.B. Keverne

2 The Effects of Sexual Experience on Male Behaviour Successful reproduction is a primary driver of animal behaviour, and the intensity of intra-sexual competition for mates means that even slight advantages in mating ability can have considerable consequences for male reproductive success. As such, it is unsurprising that sexual behaviour is not fixed to a rigid template and that there are significant experiential effects on both pre-copulatory and sexual behaviour in both sexes. Mating behaviour involves the processing of sensory input to detect potential mates; motivation to pursue potential mates; activation of the hypothalamic-pituitary-gonadal axis; initiation of appropriate patterns of motor output, and copulation-induced reward to reinforce these motivated beha­ viours. The modification of sexual behaviour by sexual experience thus potentially involves changes to all these disparate systems and the interactions between them. Perhaps unsurprisingly, sexual experience has thus been reported to trigger changes not just in sexual behaviour itself but also in associated behaviours which are also mediated to differing degrees by the same neuroendocrine and sensory pathways. Sexual experience affects diverse behavioural realms in males, modulating behaviours, including mating itself but also aggression, olfactory preferences and ­anxiety-like behaviours. While the effects of sexual experience on male sexual behaviour have been a topic of inquiry for many years (Beach, 1942), more recently the influence of sexual experience on other male behaviours has become clear. The specific mechanisms that mediate the behavioural consequences of sexual experience have also begun to be unpicked, revealing a complex interplay between gonadal steroid hormone responses, forebrain reward networks and sensory systems. In rodents, sexual experience has been shown to have important effects on olfactory acuity and sensitivity to conspecific odours, suggesting that sexual experience can modify sensory perception and so mediate changes in behaviours which are regulated by olfactory signals in these animals. Changes in dopaminergic signalling pathways and in steroid hormone-sensitive forebrain circuitry are also important consequences of sexual experience that affect other behaviours.

2.1 Behavioural Responses to Sexual Experience 2.1.1 Sexual Experience and Reproductive Behaviours Direct effects of sexual experience on male sexual behaviour occur in many different species and involve different aspects of male mating behaviour. Copulatory behaviour in male rodents is generally assayed by measuring mounting, intromission and ejaculation, and sexual experience has been shown to affect all of these behaviours. Sexually experienced rats have shorter latencies to mount and intromit, perform fewer intromissions before ejaculation, ejaculate more frequently and have shorter inter-ejaculatory periods (Dewsbury, 1969; Larsson, 1959, 1978). In guinea pigs,

Genomic Imprinting and Sexual Experience-Dependent Learning in the Mouse

199

sexual experience increases the frequency of both mounting and intromission, and sexually experienced guinea pigs are more likely than virgins to ejaculate during a 10-min sexual behaviour test (Valenstein & Goy, 1957; Valenstein, Riss, & Young, 1955). Sexual experience has similar effects in hamsters, increasing frequency of mounting and reducing latency to mount (Pfeiffer & Johnston, 1994). While these studies all show that sexual experience has direct effects on motor components of sexual behaviour, such as mounting and intromission, and that it can increase the frequency of ejaculation in mating tests, sexual experience has also been shown to affect mating success and fecundity. When housed with receptive females for a 12-h period, male mice with prior sexual experience sire more offspring than virgin males (Rastogi et al., 1981), a finding that can be explained by either increased copulatory efficiency or increased sperm fertility. Such experiential fecundity effects, as well as changes in ejaculatory frequency and inter-ejaculatory interval, suggest that sexually experienced males have enhanced reproductive potential and may indicate adaptive shaping of reproductive behaviour by sexual experience. As well as modulating copulatory behaviour, sexual experience also has significant effects on proceptive behaviour, increasing male motivation to investigate female cues. In goal box tests with a female rat located at the end of a runway, sexually experienced male rats run to investigate the female faster than virgin males (Lopez, Olster, & Ettenberg, 1999). This change in behaviour is apparent after a single ejaculation, suggesting that even one copulatory experience is enough to have significant effects on male motivated behaviour. In a demonstration of the importance of dopamine systems for behavioural reinforcement by sexual experience, administration of a dopamine antagonist prior to sexual experience does not disrupt copulation, but does prevent this reduction in approach time to the female (Lopez & Ettenberg, 2000) indicating that this increase in female-targeted motivation is dopamine-dependent. 2.1.2 Sexual Experience and Non-Reproductive Behaviours Among non-sexual behaviours that are affected by sexual experience, the most consistent effects have been reported in studies of male aggression. Inter-male aggression is common in rodents and is used to assert territorial and social dominance and to defend mates. In male rats, co-housing with intact, cycling females causes an increase in aggression in male rats that is not seen when housed with ovariectomised females or with males (Barr, 1981; Flannelly & Lore, 1977) or when simply exposed to intact cycling females without copulation (Flannelly, Blanchard, Muraoka, & Flannelly, 1982). A continuum of aggression is seen in male mice that correlates with extent of sexual experience: in tests of males with either no sexual experience, a few hours cohabitation with a receptive female or continuous housing with a female, virgin males were least likely to attack an intruder and extensively experienced males were most likely to do so, while males with only brief experience of females had intermediate aggression scores (Goyens & Noirot, 1975). Both virgin and sexually experienced male mice were used as intruders in this study and the sexual experience of the intruder male also influenced the likelihood of attack by the resident animal. Sexually

200

W.T. Swaney and E.B. Keverne

experienced intruders were both more likely than virgin intruders to attack the resident themselves but were also less likely to be attacked by the resident, suggesting not only that aggression is increased in sexually experienced animals but also that perception of status by conspecifics is modified by sexual experience. Sexual experience has also been shown to have effects on anxiety-like behaviour and stress responses in rodents, suggesting that mating experience does not just modify hormone or reward-mediated behaviours but has more global effects. Exposure to the odours of parasitised animals induces temporary analgesia in mice, an aversive response that is thought to indicate stress. However, the size of the analgesic response is modulated by sexual experience and naive males are more sensitive to nociceptive stimuli after exposure to parasitised female mice than sexually experienced males (Kavaliers, Colwell, & Choleris, 1998). This suggests either that sexually experienced males are more sensitive to the odours of parasitised females or that they have an increased stress response to aversive female odours. However, in rats, ejaculation has a short-term anxiolytic effect (Fernandez-Guasti, Roldan-Roldan, & Saldivar, 1989; Rodriguez-Manzo, Lopez-Rubalcava, & FernandezGuasti, 1999) and males with extensive sexual experience have been reported to exhibit reduced anxiety-like behaviour in behavioural tests using open field and elevated plus maze apparatus (Edinger & Frye, 2007). It may be that these differential effects of sexual experience are due to species differences or to the different experimental methodologies employed in these studies – the explicit use of olfactory stimuli in the study by Kavaliers et al. (1998) is notable as olfaction is an area which appears to be particularly sensitive to sexual experience.

2.2 Male Olfactory Learning and Sexual Experience Olfactory cues are essential for priming and inducing social and reproductive behaviours in rodents (Brennan & Kendrick, 2006). The importance of such cues for male sexual behaviour has been demonstrated in lesion and gene knockout experiments which have revealed important roles for both the main and accessory olfactory systems in rodents. These two olfactory systems have classically been thought of as having distinct and segregated functions, the main olfactory system sensing volatile odours from the general environment while the accessory olfactory system, or vomeronasal system, detects non-volatile pheromones secreted by conspecifics (Buck, 2000). Evidence for these specific roles comes from the structure of the two sensory epithelia: the main olfactory epithelium (MOE) is distributed across the nasal cavity and is exposed to the air while the vomeronasal organ (VNO) is a mucus-filled, blind-ended tube into which ligands are pumped during direct physical investigation of a stimulus. The MOE and VNO also have their own processing pathways and send segregated inputs to the main olfactory bulb (MOB) and accessory olfactory bulb (AOB), respectively. This separation of circuitry continues via separate MOB projections to the cortical amygdala (CoAmg) and AOB projections to the medial amygdala (MeAmg) before the two olfactory systems converge at the level of the basal hypothalamus which

Genomic Imprinting and Sexual Experience-Dependent Learning in the Mouse Vomeronasal System VNO

AOB

201

BNST NAcc MeAmg MPOA CoAmg

MOE Main Olfactory System

MOB piriform cortex

Fig.  1  Summary of the key nuclei involved in male responses to sexual experience. Female odours and pheromones are detected by the main olfactory epithelium (MOE) and vomeronasal organ (VNO) of the vomeronasal and main olfactory systems. The VNO projects via the accessory olfactory bulb (AOB) to the bed nucleus of stria terminalis (BNST) and medial amygdala (MeAmg), which also has links with the nucleus accumbens (NAcc) of the mesolimbic dopamine system. The MOE sends projections via the main olfactory bulb (MOB) to the piriform cortex and the cortical amygdala (CoAmg). Information processed in the two olfactory systems converges at the level of the amygdala and in projections to the medial pre-optic area (MPOA) of the hypothalamus, which is the principal regulatory area for male sexual behaviour

regulates male reproductive behaviour (Fig.  1). However, this clear demarcation of functional roles and connections is no longer seen as absolute, as evidence has accumulated that the main olfactory system has important functions in pheromonallyinduced behaviours (Zufall & Leinders-Zufall, 2007) and that there is significant cross-talk between the main and accessory olfactory systems, both within the amygdala and between the two olfactory bulbs (Martel & Baum, 2009; Yoon, Enquist, & Dulac, 2005). Nevertheless, while the precise roles of the two olfactory systems in male sexual behaviour are not yet fully resolved, olfactory cues are essential for eliciting male mating behaviour in rodents, and the functions of the main and accessory olfactory systems are sensitive to sexual experience. 2.2.1 Sexual Experience and the Accessory Olfactory System Classical work in the hamster demonstrated that removal of the VNO causes severe deficits in male sexual behaviour but that deafferentation of both olfactory systems is necessary to abolish sexual behaviour completely (Murphy & Schneider, 1970; Powers & Winans, 1975). However, the VNO is only critical for sexual behaviour in virgin male hamsters – removal of the VNO causes severe deficits in virgin hamsters but has little impact on sexual behaviour in males that have previously mated (Meredith, 1986). Thus, the VNO processes a pheromonal cue that is an important trigger of male sexual behaviour in naive animals but not in sexually experienced animals, in which sexual behaviour can be evoked by VNO-independent mechanisms.

202

W.T. Swaney and E.B. Keverne

This is possibly due to associations between sexual pheromones and other female cues which are reinforced during mating experience. Female hamster vaginal fluid (HVF) appears to have many of the properties of such a pheromone and elicits enormous interest and extensive investigation by males (Murphy, 1973). The ability of male hamsters to discriminate between females on the basis of HVF is dependent on a functional accessory olfactory system and not on an intact MOE (Steel & Keverne, 1985), suggesting that the VNO mediates the detection of HVF chemosignals. Exposure to HVF causes an increase in circulating testosterone in both virgins and sexually experienced hamsters that is abolished by VNO removal regardless of mating experience. However, if exposed to intact female hamsters rather than just HVF, males that have undergone bulbectomy will exhibit the increase in circulating testosterone only if they have had prior sexual experience. Thus, HVF produces an increase in testosterone in naive and experienced animals alike, but other female cues are only able to elicit this endocrine response if males have mated. These studies suggest that VNO-stimulating pheromones in HVF are necessary for sexual behaviour in naive male hamsters and that, through subsequent successful copulation, they learn an association between other female cues and HVF, which enables these other female stimuli to induce the same behavioural and endocrine responses as HVF itself. These effects of sexual experience on olfactory behaviour are echoed by demonstrated effects of sexual experience on neural activity in the accessory olfactory system. As well as disrupting sexual behaviour in virgin male hamsters, VNO removal blocks HVF-induced c-Fos expression in the medial pre-optic area (MPOA), the primary hypothalamic nucleus that regulates male sexual behaviour. c-Fos is an immediate early gene which is strongly expressed in recently activated neurons and as such is a useful histochemical marker of neuronal activity. It has been commonly used as a proxy indicator of neuronal activity in the olfactory systems and hypothalamus in studies of social and sexual behaviour (Baum & Everitt, 1992; Heeb & Yahr, 1996). Exposure to HVF increases c-Fos in intact experienced males relative to intact males; however, VNO-lesioned naive males show no MPOA c-Fos expression whatsoever while VNO-lesioned experienced males have similar c-Fos levels after HVF exposure to those of intact experienced males (Fewell & Meredith, 2002). MPOA c-Fos in hamsters is potentiated by sexual experience and neural activity in the MPOA of virgins appears to be dependent on HVF pheromones but can be evoked by cues sensed by the MOE in experienced males. Sexual experience increases the range of cues to which MPOA neurons will respond, indicating mating-induced plasticity in the olfactory pathways or pre-optic area. 2.2.2 The Main Olfactory System and Male Reproductive Behaviour While the VNO appears to mediate male sexual behaviour in the hamster, in the mouse the main olfactory system has a crucial role as well and studies involving combined genetic and behavioural approaches have helped to illustrate this. Studies of mice carrying knockouts of genes that are crucial to the function of each olfactory system have demonstrated that while both are involved in male sexual behaviour,

Genomic Imprinting and Sexual Experience-Dependent Learning in the Mouse

203

they fulfil different roles. Mice carrying a knockout of the VNO-specific Trp2 cation channel have very limited VNO function as indicated by greatly reduced VNO ligand-evoked potentials in VNO receptor neurons (Leypold et  al., 2002). The knockout of Trp2 does not disrupt the motor components of male copulatory behaviour; however males carrying the knockout are unable to discriminate between males and females and attempt to mount and mate with other mice regardless of gender (Stowers, Holy, Meister, Dulac, & Koentges, 2002). Conversely, mice carrying a knockout of the MOE-specific Cnga2 ion channel gene have no main olfactory function and, despite having a functional VNO, display no sexual behaviour whatsoever (Mandiyan, Coats, & Shah, 2005). This suggests that a functioning MOE is necessary for sexual behaviour in mice, but it is noteworthy that Cnga2 knockout animals do not engage in olfactory investigation of conspecifics at all, a significant effect of the knockout given the necessity of physical contact and investigation for VNO stimulation. It is possible that an absence of MOE stimulation in the Cnga2-knockout mice leads to a generalised reduction in motivation to investigate odour sources, leading to a failure to physically investigate other animals and a contingent reduction in VNO stimulation alongside the disruption to MOE signal detection. Further evidence for the importance of the main olfactory system in male sexual behaviour is provided by lesion work showing that irrigation of the MOE with zinc sulphate abolishes both MOE signalling and also all reproductive behaviours in male mice (Keller, Douhard, Baum, & Bakker, 2006b). MOE lesions disrupted sexual behaviour regardless of whether the subject males were sexually naive or experienced, indicating that these male mice were unable to initiate sexual behaviour in response to other, non-olfactory cues. This research points to a crucial role for the main olfactory system in male sexual behaviour in mice which is not seen in some other rodent species. 2.2.3 Sexual Experience and the Main Olfactory System Despite the similar effects on sexual behaviour of MOE lesions in experienced and naive males, sexual experience has been shown to play a role in olfactory responses to female odours in mice. Hayashi & Kimura (1974) demonstrated that virgin males are unable to discriminate between volatile urinary odours from oestrous and dioestrous females but that sexually experienced males have a preference for oestrous odours. A preference for the odours of such potential mates suggests that sexually experienced males are better able than virgins to locate oestrous females and so are less likely to expend resources pursuing unreceptive, dioestrous females. The acquisition of similar preferences for oestrous odours through sexual experience has also been demonstrated in rats (Carr, Loeb, & Dissinger, 1965; Stern, 1970), and such learned odour preferences may provide males with a competitive advantage over less experienced males. Sexual experience also influences male ultrasonic vocalisations, an important component of proceptive and courtship behaviours in male mice and one which is elicited by females or their odours (Nyby, Wysocki, Whitney, & Dizinno, 1977). While all male mice will vocalise to

204

W.T. Swaney and E.B. Keverne

fresh female urine, only sexually experienced males will vocalise in response to aged urine (Sipos, Kerchner, & Nyby, 1992). Thus, a short-lived pheromone in fresh urine elicits ultrasonic vocalisations from all males, but sexual experience is necessary for male mice to learn to associate other more stable chemosignals in female urine with this ephemeral pheromone. Nyby et al. (1977) noted that animals only vocalised after physical contact with female stimuli, suggesting that the VNO might mediate ultrasonic vocalisations; however, as in hamsters, this is dependent on whether the animal is sexually experienced. While VNO removal abolishes vocalising in virgins, it only causes a reduction in vocalisations in sexually experienced males which also occurs if the MOE is lesioned and the VNO is left intact (Sipos, Wysocki, Nyby, Wysocki, & Nemura, 1995). Lesioning both the VNO and MOE stops vocalisations in sexually experienced males indicating that, while mating allows MOE-sensed cues to evoke vocalisations, similar associations are not formed with non-olfactory cues during sexual experience. As in hamsters, sexual experience modifies the range of olfactory cues that can elicit reproductive behaviours suggesting that ­copulation induces changes in the male olfactory and forebrain circuitry.

2.3 Neuroendocrine Responses to Sexual Experience in the Male Brain Male sexual behaviour is regulated by an interplay between hypothalamic and preoptic areas that are sensitive to gonadal steroids, sensory systems such as the main and accessory olfactory pathways and the mesocorticolimbic dopamine system that mediates reward and motivated behaviour. Behavioural changes after sexual experience can involve changes to all these systems, as well to other neurochemical systems less explicitly involved in male sexual behaviour. 2.3.1 Sexual Experience and Androgens The critical importance of androgens for sexual behaviour is demonstrated by the dramatic decline in male sexual behaviour that occurs after castration. In several species, this effect is not uniform but is sensitive to sexual experience. While castrated virgin animals fail to exhibit any sexual behaviour, certain strains of hamsters and mice exhibit prolonged resistance to the effects of castration if they were sexually experienced prior to surgery (Lisk & Heimann, 1980; Manning & Thompson, 1976). Sexually experienced males are more likely to continue mounting and intromitting after castration, and these experienced, copulating castrates exhibit preferences for oestrous female odours that are not exhibited by virgin animals (Costantini et  al., 2007). While male sexual behaviour is largely dependent on androgens, sexual experience appears to allow for a certain emancipation of sexual behaviour from hormonal regulation, a phenomenon which coincides with learned olfactory

Genomic Imprinting and Sexual Experience-Dependent Learning in the Mouse

205

preferences for sexually relevant odour cues. The administration of androgens to non-copulating castrates rescues sexual behaviour; however, this restoration of reproductive behaviour by hormone treatment is also modulated by experience. One of the behaviours disrupted by castration in mice is ultrasonic vocalisation to female mice, and males that were sexually naive prior to castration will vocalise again after administration of testosterone or large doses of DHTP, but not if given estradiol benzoate. However, castrates that had previous sexual experience will vocalise again if treated with small doses of testosterone, DHTP or estradiol benzoate (Nunez & Tan, 1984). While male ultrasonic vocalisations remain hormonedependent, the range and concentration of steroid hormones necessary for male vocalising is changed by sexual experience indicating changes in downstream circuitry that mediates the activation of male behaviour by gonadal steroids.

2.3.2 Sexual Experience and Dopamine The learned responses to female odours suggest that during copulation male rodents form associations between unconditioned stimuli (whether female pheromones or the rewarding component of sexual behaviour) and previously unarousing female stimuli which can then elicit male behaviours themselves. Such conditioned learning in the context of reward is mediated by mesolimbic dopamine projections from the ventral tegmental area to the nucleus accumbens (NAcc) (Kelley, 2004) and this pathway appears to play an important role in sexual experience-dependent behavioural changes. If allowed to mate in the presence of volatile, neutral odours, male rats will subsequently ejaculate preferentially with females scented with the same odour (Kippin, Talianakis, Schattmann, Bartholomew, & Pfaus, 1998). Such learned associations involve activation of forebrain dopamine both during the mating-dependent conditioning and when responses to the conditioned stimuli are subsequently evoked. After pairing of volatile odours with mating, male rats will display increased interest in the mating-associated odour, but this is blocked if dopamine receptor antagonists are administered prior to the tests of olfactory interest (Lopez & Ettenberg, 2002b). Copulation also increases interest in female odours themselves, and this change in motivation can be blocked by administration of dopamine antagonists prior to mating (Lopez & Ettenberg, 2000). Dopamine is thus involved in both the acquisition of motivation changes during copulation and the subsequent display of dependent behavioural changes. Copulation activates the mesolimbic dopamine system in males (Balfour, Yu, & Coolen, 2004) and exposure to receptive female odours causes dopamine release at the NAcc in male rats which is further increased by copulation itself (Damsma, Pfaus, Wenkstern, Phillips, & Fibiger, 1992; Pfaus et  al., 1990). This dopamine efflux is not triggered by the odours of unreceptive females and the further copulatory increase in dopamine is mediated by intromission specifically and not by mounting (Wenkstern, Pfaus, & Fibiger, 1993). Moreover, NAcc activity in response to the odours of oestrous females is greater in sexually experienced males (Lopez & Ettenberg, 2002a),

206

W.T. Swaney and E.B. Keverne

showing that mesolimbic activity in response to mating-linked cues is potentiated by sexual experience. Dopamine is also released from periventricular neurons of the incertohypothalamic dopamine circuit after exposure to oestrous female odours and during copulation itself (Dominguez & Hull, 2005; Hull, Du, Lorrain, & Matuszewich, 1995) and this is an important circuit in male sexual behaviour. Dopamine agonists injected into the MPOA facilitate copulation (Hull et  al., 1986) while injection of dopamine antagonists disrupts male sexual behaviour (Warner et al., 1991). Infusion of dopamine antagonists into either the MPOA or the NAcc produces different effects which appear to indicate that MPOA dopamine receptors are involved in both anticipatory and consummatory components of male sexual behaviour while the NAcc only regulates pre-copulatory anticipation (Pfaus & Phillips, 1991). Examination of neural activity in these brain areas in response to sexual cues suggests that sexual experience modifies activity in both these circuits. Exposure to female odours causes an increase in c-Fos expression in the vomeronasal processing pathway of sexually experienced male rats and also in the MPOA and in the core and shell of the NAcc (Hosokawa & Chiba, 2005). Although female odours stimulate the VNO pathway in virgin males, they do not cause any increase in neural activity in the MPOA or in mesolimbic reward circuitry. Post-synaptic activity in the MPOA and NAcc is only increased once the males are sexually experienced, suggesting that dopamine release during sexual experience might be necessary to subsequently activate these circuits in response to female cues alone. An interaction between the vomeronasal processing pathway and MPOA dopamine has been demonstrated in the context of MeAmg regulation of male sexual behaviour. Lesions to the MeAmg cause severe deficits in consummatory aspects of male sexual behaviour (Kondo, 1992) which are reversed by infusion of dopamine agonists into the MPOA (Dominguez, Riolo, Xu, & Hull, 2001). MeAmg lesions also block femaleinduced release of dopamine in the MPOA, suggesting that the MeAmg acts as a gatekeeper for MPOA dopamine release in the context of mating. Indeed, chemical stimulation of the MeAmg causes an immediate increase in MPOA dopamine (Dominguez & Hull, 2001) indicating direct amygdala control of dopamine efflux in the MPOA. Dopamine release at the MPOA and NAcc during copulation may reinforce this signal transmission from the vomeronasal amygdala, strengthening subsequent activation of the MPOA and NAcc in response to female pheromones and chemosignals themselves. 2.3.3 Sexual Experience and Nitric Oxide An important regulator of dopamine release in the MPOA is nitric oxide (NO), the synthesis of which is modulated by sexual experience. Infusion of the NO precursor l-arginine into the MPOA causes an increase in dopamine in the MPOA which is blocked if an NO synthase inhibitor is administered simultaneously (Lorrain & Hull, 1993). The mating-induced increase in extracellular dopamine in the MPOA can be blocked by injection of an NO synthase inhibitor (Lorrain, Matuszewich, Howard, Du, & Hull, 1996), suggesting that the synthesis of NO is an upstream

Genomic Imprinting and Sexual Experience-Dependent Learning in the Mouse

207

signal for MPOA dopamine release. NO has a half-life of a few seconds and so is not stockpiled but is actively synthesised by NO synthase, thus levels of NO ­synthase are the rate-limiting step in NO signalling. Sexual experience has direct effects on NO signalling through NO synthase, which is more abundant in the MPOA of sexually experienced male rats than in virgins (Dominguez, Brann, Gil, & Hull, 2006). This up-regulation of NO synthase would potentially increase DA release in the MPOA of sexually experienced animals and so facilitate pre-copulatory and sexual behaviours. MPOA activation can also be evoked in sexually experienced animals in the absence of signalling via at least one dopamine signalling pathway. Mating increases MPOA c-Fos in male rats; however, treatment before copulation with a D1 receptor antagonist blocks this c-Fos increase in virgin males but not in sexually experienced males (Lumley & Hull, 1999). Sexually experienced males have higher levels of copulation-induced c-Fos in the MPOA than virgins and other non-D1 receptor mediated pathways may be mediating an increase in neural activity in the MPOA during copulation in sexually experienced males, masking the effects of D1 receptor blockade on MPOA c-Fos in these animals. These studies show that, while sexual experience potentiates dopamine signalling in the MPOA, other pathways are also activated in sexually experienced animals indicating mating-dependent synaptic plasticity involving multiple signalling mechanisms.

3 Genomic Imprinting Imprinted genes are mammalian genes, primarily autosomal, that are expressed in a haploid fashion according to the parent-of-origin of each allele. For paternally expressed genes, the allele located on the paternal chromosome is functional and the allele located on the maternal chromosome is silent, while the opposite is true for maternally expressed genes. The existence of genomic imprinting was revealed through studies of parthenogenesis in mammals which demonstrated that parthenogenetic mouse embryos fail at a very early stage. In some amniotes, notably certain lizard and fish species (Chapman et al., 2007; Cuellar, 1976; Watts et al., 2006), parthenogenesis is a viable method of reproduction. However, mouse embryos created using either two male (androgenetic) or two female pronuclei (parthenogenetic) fail at a very early stage, indicating that simple diploidy is insufficient for the normal development of mammalian embryos and that both the maternal and paternal genomes are required (Barton et  al., 1984; McGrath & Solter, 1984). This is due to the presence of imprinted genes which are selectively silenced during gametogenesis according to the gender of the individual and are inappropriately expressed in androgenetic (AG) and parthenogenetic (PG) embryos. In AG embryos, paternally expressed genes are overexpressed and maternally expressed genes are silent, while the reverse is true in PG embryos. The contribution of each parental genome to development was elucidated by creating chimeric embryos through the fusion of AG and PG blastocysts with wild-type blastocysts. This rescued the uniparental embryos and allowed the fate of the AG

208

W.T. Swaney and E.B. Keverne

and PG cells in these chimeras to be followed, revealing the role that the maternal and paternal genomes play in the development of different body systems and structures (Allen et al., 1995; Barton, Ferguson-Smith, Fundele, & Surani, 1991). AG chimeras are born significantly larger than wild-type pups while PG chimeras are smaller, suggesting that paternally and maternally derived genes regulate offspring growth in opposite directions. While AG cells contributed substantially to skeletal muscle and PG cells were largely absent from this tissue, PG cells were overrepresented in the brain where relatively few AG cells were found. These experiments demonstrated not only that the two parental genomes are non-equivalent but also showed that paternal and maternal imprinted genes regulate different developmental events in mammals. The studies of PG and AG mice showed that the differences between maternal and paternal genomes must be epigenetic rather than genetic (Surani, Barton, & Norris, 1984). In the years since the discovery of genomic imprinting, the mechanism by which parent-of-origin expression is established has been extensively studied. Imprinting status is not “hard-wired” into the genetic code, but involves an epigenetic mark that is applied during gametogenesis and which identifies parental origin (Surani, 1998). Imprinted genes generally occur in clusters close to “imprinting control regions” (ICRs) (Spahn & Barlow, 2003) which regulate the imprinted status of nearby imprinted genes (Williamson et al., 2006). These ICRs are characterised by CpG islands which are differentially methylated, and it is the methylation status of these ICR elements that appears to be the principal mechanism for regulating the expression of imprinted genes (Edwards & Ferguson-Smith, 2007; Sleutels, Barlow, & Lyle, 2000), although other epigenetic modifications also play a part in genomic imprinting (Keverne & Curley, 2008). Not only are imprinted genes topographically linked but they also share differentially methylated ICRs that regulate their expression, whether paternal or maternal. For example, deletion of the differentially methylated ICR linked to the Igf2 and H19 imprinted genes causes a loss of imprinted expression of both genes (Thorvaldsen, Duran, & Bartolomei, 1998), while disruption of the ICR of the Gnas imprinted gene cluster affects the expression of all imprinted genes in this area (Williamson et al., 2004). This suggests that the mechanism of imprinting evolved separately from the imprinted genes themselves and that it was applied to certain key regulatory genes over the course of mammalian evolution if haploid expression was adaptive.

3.1 Imprinted Genes in the Brain 3.1.1 Imprinted Genes and Brain Development While functional roles have not been identified for all of the imprinted genes, many encode developmental regulators such as transcription factors, growth factors, oncogenes, metabolic regulators and cell-cycle proteins (Morison et al., 2005). As shown in the AG and PG mouse experiments, imprinted genes have important roles in ­pre-natal

Genomic Imprinting and Sexual Experience-Dependent Learning in the Mouse

209

growth and resource supply (Tycko & Morison, 2002), but imprinted genes are also important regulators of brain development (Davies, Isles, Humby, & Wilkinson, 2008; Davies et al., 2005). Moreover, maternally and paternally imprinted genes are differentially expressed in specific regions of the brain. Tracking the fate of AG and PG cells in chimeric mouse embryos revealed that PG cells were overrepresented in the brain and that comparatively few AG cells were found. More detailed investigation of the contribution of AG and PG cells to the brain revealed a more nuanced picture: PG cells segregated primarily to the cortex and striatum (the “executive” brain) while AG cells were found exclusively in the hypothalamus, septum and preoptic area (the “emotional” brain) (Allen et al., 1995; Keverne, Fundele, Narasimha, Barton, & Surani, 1996). The two parental genomes appear thus to have evolved to regulate the development of separate parts of the mammalian brain. Such parent-oforigin regulation of the development of different parts of the brain has important functional ramifications, as these brain areas mediate different behaviours. While the hypothalamus and associated structures mediate hormonally regulated behaviour and physiology (reproduction, aggression, appetite), the cortex controls complex behaviours such as cognition, memory, attention and, in humans, language and thought. 3.1.2 Imprinted Genes and Brain Dysfunction Parent-of-origin regulation of the development of different brain areas leads to parent-of-origin inheritance of certain traits and diseases, many of which include deficits in learning and memory among their symptoms (Davies, Isles, & Wilkinson, 2001; Isles & Humby, 2006). Parent-of-origin effects are seen in many mental disorders including Alzheimer’s, autism, epilepsy and schizophrenia, suggesting that disruption of normal patterns of imprinting is involved to some degree in these illnesses (Morison et  al., 2005). However, disruption of normal expression of imprinted genes is also the direct cause of diseases which affect brain function. This is typically either caused by uniparental disomy, in which two copies of a chromosome or part thereof are inherited from one parent, by deletion of an imprinted domain or by mutations at an ICR (Davies et  al., 2001). Prader–Willi syndrome (PWS) and Angelman syndrome (AS) are caused by abnormal imprinting of chromosome 15q11-q14, an area which contains paternal and maternal imprinted genes (Jiang, Tsai, Bressler, & Beaudet, 1998). In PWS, an absence of expression of paternally expressed genes, specifically small nucleolar RNAs on chromosome 15 (Sahoo et al., 2008), has various effects on metabolism, including neonatal hypotonia and suckling deficits and then hyperphagia and obesity in later life, as well as mild mental retardation (Whittington et al., 2004). AS involves reciprocal deletions of chromosome 15q11-q14 and an absence of maternal expression at this site resulting in severe mental retardation with no speech development, motor disorders, hyperactivity, microcephaly and a happy disposition (Glenn, Driscoll, Yang, & Nicholls, 1997). The symptoms seen in these diseases involve systems regulated by areas of the brain where imprinted genes are developmentally important. PWS, in which paternal expression of imprinted genes is absent as described above,

210

W.T. Swaney and E.B. Keverne

has considerable effects on metabolism and appetite, which are regulated by the hypothalamus, and has relatively mild effects on cognition, a function of the cortex. Conversely, AS involves an absence of maternal expression of imprinted genes and results in severe mental retardation as well as cortical abnormalities (Dorries, Spohr, & Kunze, 1988) but does not involve disruption to hypothalamically-mediated behaviours. Further evidence for the differential role of maternally and paternally imprinted genes in brain function can be drawn from studies of patients suffering Turner’s syndrome. This disease occurs exclusively in women and is caused by the deletion of one of the two X chromosomes. However, different pathologies are seen depending on whether the deletion involves the maternal or paternal chromosome, indicating the presence of imprinted genes on the X chromosome. If the remaining X chromosome is maternal (and paternal gene expression from the X chromosome is absent), patients exhibit reductions in behavioural inhibition and in verbal skills, while patients with a single paternal X chromosome (and hence no maternal expression) have deficits in visuospatial memory. Given the effects of abnormal imprinting on brain function and the importance of imprinted genes for normal brain development, the role that imprinted genes play in behaviour and cognitive function is a significant one.

3.2 Evolution of Genomic Imprinting 3.2.1 Genomic Imprinting in Mammals Imprinted genes were shown to be involved in placental formation and development in the early experiments with AG and PG embryos which had enlarged and reduced placentas respectively (Surani, Barton, & Norris, 1986). Imprinted genes seem to directly regulate foetal growth in part by regulating placental development and the majority of imprinted genes identified so far are expressed in the placenta (Coan, Burton, & Ferguson-Smith, 2005). High levels of placental expression of imprinted genes, and the fact that genomic imprinting is only found in placental mammals, suggest links between placentation and imprinted genes. A further connection between genomic imprinting and placentation is also indicated by surveys of imprinting across mammalian classes, which show that the extent of genomic imprinting is not uniform but correlates with the existence and complexity of placental tissues. Of the approximately 100 imprinted genes that have been identified in eutherian species, only a subset of these is imprinted in the marsupials, which have short-lived and relatively unsophisticated placentas (Hore, Rapkins, & Graves, 2007; Suzuki et  al., 2005), while genomic imprinting appears to be absent in the egg-laying monotremes (Killian et  al., 2001; Warren et  al., 2008). The function, expression patterns and speciesspecificity of genomic imprinting are strongly suggestive of the co-evolution of imprinting and placentation in mammals (Kaneko-Ishino et  al., 2003; Pask et  al., 2009; Reik et al., 2003). This raises the question of why the evolution of placentation made haploid, parent-of-origin expression of certain genes adaptive when diploid

Genomic Imprinting and Sexual Experience-Dependent Learning in the Mouse

211

expression and the protection against mutation which it confers (Orr, 1995) is ubiquitous in other vertebrates. 3.2.2 Hypotheses for the Evolution of Genomic Imprinting Several theories have been advanced to explain why imprinting evolved, including the regulation of gene dosage in the context of internal development (Solter, 1988) and as a defence mechanism against invasive foreign DNA (Barlow, 1993; Pask et  al., 2009) or parthenogenesis (Varmuza & Mann, 1994). However, the most widely accepted theory explaining the evolution of genomic imprinting in placental mammals is the kinship, or conflict theory (Moore & Haig, 1991). It is based on two presumptions – firstly, that a female is equally related to all her offspring, but a particular male mate is related only to the offspring he has sired and is highly unlikely to be related to her other offspring, and secondly, that mothers invest more in offspring than fathers. This creates a conflict as females can maximise their fitness by investing equally in all their offspring, while males will maximise their fitness if they can manipulate maternal investment in favour of their offspring, to the potential detriment of the mother’s subsequent offspring. The placenta and foetus, as structures within the mother that are genetically half paternal and half maternal, give fathers the opportunity to influence maternal investment in offspring by manipulating maternal physiology and behaviour. The conflict theory predicts that paternally derived genes will favour increased maternal investment in a particular father’s offspring while maternally derived genes will favour balancing investment across all offspring. This would result in an evolutionary arms race with the placenta and embryo as the battle ground, paternally derived genes acting to enhance foetal growth while maternally derived genes restrict foetal growth resulting in stable equilibrium. The conflict theory was first proposed in the wake of experimental knockouts of the Igf2 and Igf2r imprinted genes (Haig & Graham, 1991) which showed that mutation of the paternally expressed Igf2 results in growth retardation (DeChiara, Robertson, & Efstratiadis, 1991) and that knockout of the maternally expressed Igf2r results in mice that are significantly larger than wild-type animals at birth (Lau et al., 1994). Since these early investigations, the growth effects seen in most imprinted gene knockouts have followed the pattern predicted by the conflict theory; however, a number of exceptions have been seen (Hurst & McVean, 1997) which do not support the predictions of the conflict theory and which have aroused considerable debate (Moore & Mills, 2008; Morison et al., 2005). 3.2.3 Co-adaptation and Genomic Imprinting Much of the work on mouse knockouts of imprinted genes has focused on metabolic and growth effects during the pre- and peri-natal phase, however our group has been examining the impact of imprinted gene knockouts on adult behaviour to try and address the question of how imprinted genes might have evolved to influence their

212

W.T. Swaney and E.B. Keverne

own transmission to the next generation from one or other parental genome. Our work has revealed that imprinted genes regulate reproductive behaviour in the adult and suggests that different reproductive behaviours in males, females and offspring have co-evolved to be regulated by the paternally expressed gene Peg3. As effects of imprinted genes on adult behaviour occur when conflict over resource allocation is not as significant, this suggests that conflict may be not be the sole driver of the evolution of genomic imprinting (Hurst & McVean, 1998) and that co-adaptation between individuals may also have played a role (Curley et al., 2004).

4 Paternally Expressed Gene 3 (Peg3) Peg3 is a paternally expressed gene located on mouse proximal chromosome 7 which encodes a Kruppel-type zinc finger protein that contains 12 widely spaced C2H2 motifs and two proline-rich repeat domains (Kuroiwa et  al., 1996; Relaix et al., 1996). In developing mice, Peg3 is strongly expressed in the hypothalamus, in mesodermal tissue and in the placenta. The human homologue of Peg3 is found at chromosome 19q13.4, an area that is syntenic with the Peg3 locus in the mouse (Kim, Ashworth, Branscomb, & Stubbs, 1997). Like murine Peg3, human PEG3 is ubiquitously paternally expressed and transcripts are found both developmentally and post-natally in the brain (Murphy, Wylie, & Jirtle, 2001). The Peg3 protein appears to have apoptotic functions involving the Siah1a (Relaix et al., 2000) and Bax proteins (Deng & Wu, 2000), components of the p53-mediated apoptotic pathway. In vitro work also indicates that Peg3 is involved in apoptosis as up-regulation of Peg3 occurs in cortical neurons that have suffered either hypoxia (Yamaguchi et al., 2002) or DNA damage (Johnson, Wu, Aithmitti, & Morrison, 2002).

4.1 Peg3, Maternal Behaviour and Offspring Development 4.1.1 Maternal Behaviour in Peg3 Mutant Mice A mouse knockout of Peg3 was first generated in 1999 and exhibited a complex phenotype which included deficits in offspring growth that were predicted by the conflict theory, as well as effects on maternal behaviour in adult Peg3 mutants that were not. One of the consequences of the unusual expression of paternally and maternally expressed imprinted genes is that heterozygote knockout animals exhibit the same phenotype as homozygote animals, so long as the mutant copy of the allele is inherited from the appropriate parent. Thus, if a heterozygote Peg3 mutant male is mated with a wild-type female, the resulting litter will be half wild-type and half mutant and the mutant animals will exhibit the same phenotype as homozygotes, due to the wild-type Peg3 allele being of maternal origin and therefore silenced. This enabled the effects of the Peg3 knockout to be studied independently in mutant pups born to a wild-type

Genomic Imprinting and Sexual Experience-Dependent Learning in the Mouse

213

mother and in mutant mothers carrying wild-type litters. This revealed that pups carrying the Peg3 mutation are growth retarded but also that wild-type pups born to Peg3 mutant females are smaller at birth and reach normal weight once they have been weaned (Li et al., 1999). The Peg3 mutant females were shown to have a lactational deficit which explained the failure of wild-type pups born to these mothers to thrive before weaning. Mammary glands in the mutant mothers were histologically normal; however, brain immunohistochemistry showed that Peg3 mutant females have significantly fewer oxytocin neurons in the paraventricular nucleus of the hypothalamus (PVN) and the MPOA where the release of oxytocin is essential for milk letdown. 4.1.2 Reciprocal Effects in Peg3 Mutant Mothers and Offspring Peg3 mutant females exhibit deficits in other aspects of maternal care: in retrieval tests with displaced pups, they are slower than wild-type females to retrieve the pups to the nest site, to rebuild the nest and to crouch over the pups. While such effects of the knockout of a paternally expressed gene on maternal behaviour were interesting, closer examination of the relationship between mother and pups in these mutant animals revealed a more complex role for Peg3 in mother–pup interactions. Peg3 mutant females spend less time licking, grooming and nursing pups (Champagne et al. 2009), important maternal behaviours which influence beha­ vioural and brain outcomes in offspring (Cameron et  al., 2005). Comparisons of mutant females rearing wild-type pups and mutant pups born to wild-type mothers revealed reciprocal effects of the mutation on behaviour in mother and offspring. While Peg3 mutant mothers suffer deficits in maternal food intake, lactation and nest building, mutant pups show deficits in reciprocal behaviours, namely pre-natal growth, suckling and thermoregulation (Curley et al., 2004). Each of these beha­ viours is vital for offspring to thrive but are also dependent on the appropriate response from the other partner in the mother–pup relationship. Peg3 is strongly expressed in both the developing and adult hypothalamus (Champagne et al., 2009); Curley et  al., 2004) which regulates these maternal and metabolic behaviours and appears to regulate reciprocal behaviours in mother and offspring, suggesting that co-adaptation of behaviour, and not just conflict, has been a component of the evolution of Peg3.

4.2 Peg3 and Male Behaviour While the maternal and developmental behaviours regulated by Peg3 clearly have adaptive consequences for successful reproduction, it is unclear why these traits are paternally transmitted. As Peg3 is only expressed when inherited down the patriline, offspring benefitting from Peg3-mediated maternal care would inherit a silent copy of the gene from their mother and the behavioural effects would skip a generation until transmitted by male offspring to female grand-offspring. However, if the gene

214

W.T. Swaney and E.B. Keverne

additionally regulates adaptive male behaviour, then Peg3 would also directly ­regulate male behaviour in every generation and so increase the likelihood of successful transmission of the gene across generations. 4.2.1 Peg3 and Sexual Experience-Dependent Behavioural Changes Peg3 mutant males do not show any gross copulatory deficits and breed successfully when paired with intact wild-type females. In sexual behaviour tests, virgin Peg3 mutants are indistinguishable from virgin wild-type male mice on all copulatory measures, including latency to mount, intromit or ejaculate, and frequency of mounting and intromission. Wild-type males that were allowed to mate repeatedly for 4 weeks were faster to mount females and mounted and intromitted more frequently than virgin animals; however, Peg3 mutant males showed no change in copulatory behaviour after sexual experience but exhibited similar sexual behaviour to that seen in virgin males (Swaney et al., 2007). As discussed above, sexual experience has important effects on olfactory abilities as well as sexual behaviour in rodents, and so we also examined the interest of Peg3 mutant animals in female odours cues and the acquisition of sexual experiencedependent preferences for female odours in these animals. When able to physically investigate volatile and involatile cues in female urine using both the MOE and VNO, Peg3 mutant males showed a preference for female urine over male urine akin to that of naive wild-type males, indicating that they do not suffer gross anosmia. However, while sexually naive and experienced Peg3 mutant males had similar preferences for female urine, wild-type males had several-fold larger preferences for female urine once sexually experienced (Swaney et al., 2007) (Table 1). Such effects on preferences for female non-volatiles were matched in tests in which males could only sniff volatile odours from oestrous and dioestrous female urine, which are primarily sensed by the MOE. Wild-type male mice have previously been shown to acquire a preference for oestrous volatiles though sexual experience (Hayashi & Kimura, 1974), a finding which we replicated. Peg3 mutant males displayed no preference for oestrous female urine regardless of sexual experience and indeed appeared unable to discriminate between the volatile odours of the two female urines (Table  1) in habituation– dishabituation tests in which one urine type is repeatedly presented, before being switched with the other urine type. Like Peg3 mutant males, wild-type virgins were unable to discriminate between oestrous and dioestrous urinary volatiles. However, sexually experienced wild-type males showed increased investigatory behaviour when the oestrous and dioestrous urines were switched, indicating that they were able to discriminate between the urine types (Swaney et al., 2008). The Peg3 mutation does not block olfactory ability, as mutant males were able to discriminate urinary volatiles from water and preferred female non-volatile cues to those from males. However, the Peg3 knockout seems to disrupt all sexual experience-dependent learning, whether it involves the MOE or the VNO. Peg3 does not appear to regulate behaviours specifically mediated by either olfactory system but rather is involved in behavioural plasticity in response to sexual experience. This appears to involve both olfactory systems as well as the basal hypothalamic circuit that regulates male sexual behaviour.

Genomic Imprinting and Sexual Experience-Dependent Learning in the Mouse

215

Table 1  A summary of the olfactory abilities of sexually naive and experienced wild-type and Peg3 mutant males Sexually experienced Olfactory Virgin wild- Sexually experienced Virgin Peg3 behaviour type males wild-type males mutant males Peg3 mutant males Discrimination     of volatile female & male odours     Preference for non-volatile female odours –  – – Discrimination of volatile oestrousdioestrous odours – – – Preference for  volatile oestrous odours One tick indicates that the group displays the olfactory ability in tests, two ticks indicates increased performance relative to other groups

4.3 Peg3 and Forebrain Plasticity in Response to Sexual Experience To examine which brain areas are affected by the Peg3 mutation, neural activity in response to oestrous female urine was investigated by c-Fos immunohistochemistry in key hypothalamic and olfactory nuclei in wild-type and mutant males that were either sexually naive or experienced. The hypothalamus, vomeronasal system and main olfactory system are all areas of high expression of Peg3, suggesting that the mutation might have direct effects on neural activity in these nuclei which are functionally important for male reproductive behaviour (Burns-Cusato, Scordalakes, & Rissman, 2004) and which have been implicated in behavioural changes caused by sexual experience. 4.3.1 Peg3 and the Olfactory Systems In the vomeronasal system, sexual experience increases female urine-induced c-Fos expression in both wild-type and Peg3 mutant males in the AOB, anterior and posterior MeAmg and BNST of the vomeronasal processing pathway (Swaney et al., 2007). However, the size of the increase in c-Fos activity is significantly greater in sexually experienced wild-type males than in mutant males (Table 2), indicating a differential effect of sexual experience on neural activity in the two genotypes. This mirrors the behavioural effects of sexual experience when males could physically investigate urinary stimuli: while all males preferred female urine to male urine, sexually experienced wild-type males showed a large increase in preference relative

216

W.T. Swaney and E.B. Keverne

to mutant males. In the main olfactory system, c-Fos expression was assayed in the MOB and in the piriform cortex, an important node in main olfactory processing which is responsive to acute odour exposure (Illig & Haberly, 2003) and has been implicated in olfactory learning (Brennan & Keverne, 1997). While no change was seen in MOB activation in any of the male groups, c-Fos expression in the piriform cortex was increased by female urine exposure in sexually experienced wild-type animals but not in virgin wild-type males or in either mutant group, irrespective of sexual experience (Table 2) (Swaney et al., 2008). This is particularly noteworthy as it was only the sexually experienced wild-type males that showed a preference for oestrous female volatile odours in behavioural tests, and these data suggest that this learned preference may involve potentiation of the main olfactory response to female odours at the level of the piriform cortex, a change that the Peg3 mutation disrupts. Sexual experience appears to modulate the sensitivity of both the main and accessory olfactory system to female chemosignals, changes which are blocked by the Peg3 mutation. Peg3 is not a key regulator of either olfactory system, however, it mediates the effects of sexual experience in both systems. Knocking out Peg3 blocks plasticity in these circuits, preventing experience-dependent olfactory learning about both volatile and non-volatile female cues. 4.3.2 Peg3 and the Basal Forebrain The MPOA is the primary regulatory nucleus in male sexual behaviour and there was a huge increase in MPOA c-Fos after female urine exposure in sexually experienced wild-type male mice while no change in c-Fos neuron number was seen in any other males (Table 2) (Swaney et al., 2007). Copulation triggers increased neural activity and neuropeptide release in the male MPOA, and the female odour-evoked activity in the MPOA of sexually experienced wild-type males may be a neural correlate of sexual anticipation induced by learned association between female odours and ­mating. Such an activity change may also reflect the facilitation of sexual behaviour by sexual experience in wild-type males. The absence of female odour-evoked Table 2  A summary of the effects of exposure to female urine on numbers of c-Fos neurons in vomeronasal, main olfactory, hypothalamic and mesolimbic nuclei of the male brain Sexually experienced Virgin wild- Sexually experienced Virgin Peg3 Brain area type males wild-type males mutant males Peg3 mutant males AOB ↑ ↑↑ – ↑ Anterior MeAmg – ↑↑ – ↑ Posterior MeAmg –   ↑ – ↑ BNST – ↑↑ – ↑ Piriform cortex – ↑↑ – – MPOA – ↑↑ – – NAcc core ↑   ↑ – – NAcc shell ↑   ↑ – – A single arrow represents a within-group increase in c-Fos due to exposure to female urine, two arrows represent an increase in c-Fos within a brain area relative to urine-exposed males in other groups

Genomic Imprinting and Sexual Experience-Dependent Learning in the Mouse

217

MPOA activity in Peg3 mutant males correlates with the absence of any behavioural effect of sexual experience in these animals. As previously discussed, the mesolimbic dopamine system plays a significant role in motivation and reward in the context of sexual behaviour and is thought to be important in the reinforcement of behaviour by sexual experience. c-Fos expression in the core and shell of the NAcc, the major target for mesolimbic dopamine, was increased by female odours in virgin and experienced wild-type males but was unaffected by exposure to female urine in all Peg3 mutants (Swaney et al., 2007). This suggests that knocking out Peg3 has a disruptive effect on the induction of mesolimbic reward by female pheromones. A decline in female stimulus-evoked reward would potentially reduce reinforcement of behaviour and of responses to sexual stimuli by mating. If the Peg3 knockout reduces activation of mesolimbic reward by arousing stimuli, it is unsurprising that these mutant males do not experience significant behavioural reinforcement by normally rewarding behaviour such as copulation. The importance of mesolimbic system deficits in the behavioural phenotype of the Peg3 mutant males would need to be confirmed by pharmacological manipulation; however, the c-Fos data offer intriguing indications that this may be a significant component of the failure to respond to sexual experience exhibited by these animals. 4.3.3 Peg3 and Hypothalamic Oxytocin The data from earlier experiments with Peg3 mutant females point to another possible mechanism by which the Peg3 mutation may be influencing sexual behaviour in male mice, namely hypothalamic oxytocin. Peg3 mutant females have fewer oxytocin neurons in the PVN and the MPOA, but this neuronal population has not been assessed in male Peg3 mutants. While generally not regarded as a key regulatory neuropeptide in male reproductive behaviour, the hypothalamic oxytocin circuit does appear to be responsive to sexual experience, although the data are somewhat unclear. Copulation has been reported to activate oxytocinergic neurons of the PVN (Witt & Insel, 1994), and either central or peripheral administration of oxytocin reduces the latency to ejaculate as well as the post-ejaculation refractory period in male rats (Arletti, Bazzani, Castelli, & Bertolini, 1985). Limited sexual experience has been shown to reduce numbers of oxytocin neurons in the PVN, while repeated sexual experience appears to restore them to pre-copulation levels (Jirikowski, Caldwell, Haussler, & Pedersen, 1991). Exposure to receptive females induces oxytocin release in virgin male rats but not in males that have mated between one and three times (Hillegaart, Alster, Uvnäs-Moberg, & Ahlenius, 1998). In sexually experienced male rats, either physical contact with an oestrous female or exposure to her volatile odours will increase the number of oxytocin cells that also express c-Fos. However, such a rise is only seen in virgins if they are able to physically contact the stimulus female and not if they can merely sniff volatile odours (Nishitani, Moriya, Kondo, Sakuma, & Shinohara, 2004). This suggests that, after sexual experience, female volatile odours are able to activate PVN oxytocin neurons in the same manner as involatile female pheromones, possibly due to

218

W.T. Swaney and E.B. Keverne

learned association between the two types of chemosignal. If oxytocin is significant in experience-dependent olfactory learning in males, the failure of Peg3 mutant males to learn through sexual experience may be due to an oxytocin deficit akin to that of the Peg3 mutant females. Knocking out the oxytocin gene has somewhat similar effects in male mice, as they are unable to recognise females that they have previously encountered, suggesting a specific deficit in formation of social olfactory memory as a result of a lack of oxytocin in these males (Ferguson et al., 2000). 4.3.4 Peg3, Brain Development and Plasticiy Peg3 is involved in p53-mediated apoptosis (Deng & Wu, 2000; Relaix et al., 2000) and disregulation of developmental apoptosis may be an important factor in the effects of the mutation on behaviour and forebrain function. Recent work by Broad, Curley, & Keverne (2009) indicates that apoptosis in 4- and 6-day-old Peg3 mutants is increased in the BNST, MeAmg, MPOA and NAcc, all areas in which a reduction of stimulus-evoked activity is seen in adult Peg3 mutant males. The altered patterns of apoptosis in brain areas that are essential for normal olfactory, vomeronasal and reproductive behaviour may lead to the formation of abnormal links between these areas. This in turn may lead not to gross disruption of behaviour but instead may prevent plasticity in these circuits in adulthood, blocking learning and behavioural adaptation in response to sexual experience.

5 Conclusion Imprinted genes are a unique and intriguing group of genes whose unusual pattern of expression seems to be intrinsically linked to several of the mammalian traits which distinguish this class of vertebrates including placentation and enlarged brains. While many imprinted genes have been shown to play key roles in resource allocation and metabolism during development, few have been shown to have effects on the adult. The behaviour of the Peg3 mutant animals has been extensively characterised, indicating that this gene regulates different aspects of reproduction in offspring, mother and father, suggesting co-evolution of many of these traits. While the roles of Peg3 in maternal care and offspring development enhance likely survival of the offspring, the effects of the knockout on male reproductive behaviour indicate that this gene also regulates behaviour that directly influences its own transmission down the patriline. While the female-specific effects of this paternally expressed gene will skip a generation when (silently) transmitted from a successful mother, the male behaviours that are regulated by Peg3 will be present in father and son. Sexual experience is an important modulator of male behaviour in rodents and the learned changes in behaviour induced by mating ­experience are likely to have adaptive consequences for male reproductive success. Sexually experienced males are more motivated to seek females and are able to discriminate receptive females

Genomic Imprinting and Sexual Experience-Dependent Learning in the Mouse

219

based on their odours alone. They are also more sexually vigorous, and all these traits are regulated by Peg3. The plasticity in the male sexual brain that underpins these learned responses appears to be dependent on normal, imprinted expression of Peg3, and the fine-tuning of this circuit by experience may have been significant for the successful transmission of this gene by males.

References Allen, N. D., Logan, K., Lally, G., Drage, D. J., Norris, M. L., & Keverne, E. B. (1995). Distribution of parthenogenetic cells in the mouse brain and their influence on brain development and behavior. Proceedings of the National Academy of Sciences of the United States of America, 92, 10782–10786. Arletti, R., Bazzani, C., Castelli, M., & Bertolini, A. (1985). Oxytocin improves male copulatory performance in rats. Hormones and Behavior, 19, 14–20. Balfour, M. E., Yu, L., & Coolen, L. M. (2004). Sexual behavior and sex-associated environmental cues activate the mesolimbic system in male rats. Neuropsychopharmacology, 29, 718–730. Barlow, D. (1993). Methylation and imprinting: From host defense to gene regulation? Science, 260, 309–310. Barr, G. A. (1981). Effects of different housing conditions on intraspecies fighting between male Long-Evans hooded rats. Physiology & Behavior, 27, 1041–1044. Barton, S. C., Ferguson-Smith, A. C., Fundele, R., & Surani, M. A. (1991). Influence of paternally imprinted genes on development. Development, 113, 679–687. Barton, S. C., Surani, M. A., & Norris, M. L. (1984). Role of paternal and maternal genomes in mouse development. Nature, 311, 374–376. Baum, M. J., & Everitt, B. J. (1992). Increased expression of c-Fos in the medial preoptic area after mating in male rats: Role of afferent inputs from the medial amygdala and midbrain central tegmental field. Neuroscience, 50, 627–646. Beach, F. A. (1942). Comparison of copulatory behavior of male rats raised in isolation, cohabitation, and segregation. The Pedagogical Seminary and Journal of Genetic Psychology, 60, 121–136. Brennan, P., & Kendrick, K. (2006). Mammalian social odours: Attraction and individual recognition. Philosophical Transactions of the Royal Society of London. Series B, 361, 2061–2078. Brennan, P. A., & Keverne, E. B. (1997). Neural mechanisms of mammalian olfactory learning. Progress in Neurobiology, 51, 457–481. Brennan, P. A., & Zufall, F. (2006). Pheromonal communication in vertebrates. Nature, 444, 308–315. Broad, K. D., Curley, J. P., & Keverne, E. B. (2009). Increased apoptosis during neonatal brain development underlies the adult behavioral deficits seen in mice lacking a functional paternally expressed gene 3 (Peg3). Developmental Neurobiology, 69, 314–325. Buck, L. B. (2000). The molecular architecture of odor and pheromone sensing in mammals. Cell, 100, 611–618. Burns-Cusato, M., Scordalakes, E. M., & Rissman, E. F. (2004). Of mice and missing data: What we know (and need to learn) about male sexual behavior. Physiology & Behavior, 83, 217–232. Cameron, N. M., Champagne, F. A., Parent, C., Fish, E. W., Ozaki-Kuroda, K., & Meaney, M. J. (2005). The programming of individual differences in defensive responses and reproductive strategies in the rat through variations in maternal care. Neuroscience and Biobehavioral Reviews, 29, 843–865. Carr, W. J., Loeb, L. S., & Dissinger, M. L. (1965). Responses of rats to sex odors. Journal of Comparative and Physiological Psychology, 59, 370–377.

220

W.T. Swaney and E.B. Keverne

Champagne, F. A., Curley, J. P., Swaney, W. T., Hasen, N. S., & Keverne, E. B. (2009). Paternal influence on female behavior: The role of Peg3 in exploration, olfaction and neuroendocrine regulation of maternal behavior of female mice. Behavioral Neuroscience, 123, 469–480. Chapman, D. D., Shivji, M. S., Louis, E., Sommer, J., Fletcher, H., & Prodohl, P. A. (2007). Virgin birth in a hammerhead shark. Biology Letters, 3, 425–427. Coan, P. M., Burton, G. J., & Ferguson-Smith, A. C. (2005). Imprinted genes in the placenta – a review. Placenta, 26(suppl A), S10–S20. Costantini, R. M., Park, J. H., Beery, A. K., Paul, M. J., Ko, J. J., & Zucker, I. (2007). Post-castration retention of reproductive behavior and olfactory preferences in male Siberian hamsters: Role of prior experience. Hormones and Behavior, 51, 149–155. Cuellar, O. (1976). Intraclonal histocompatibility in a parthenogenetic lizard: Evidence of genetic homogeneity. Science, 193, 150–153. Curley, J. P., Barton, S., Surani, A., & Keverne, E. B. (2004). Coadaptation in mother and infant regulated by a paternally expressed imprinted gene. Proceedings of the Royal Society of London. Series B, 271, 1303–1309. Damsma, G., Pfaus, J. G., Wenkstern, D., Phillips, A. G., & Fibiger, H. C. (1992). Sexual behavior increases dopamine transmission in the nucleus accumbens and striatum of male rats: Comparison with novelty and locomotion. Behavioral Neuroscience, 106, 181–191. Davies, W., Isles, A. R., Humby, T., & Wilkinson, L. S. (2008). What are imprinted genes doing in the brain? Advances in Experimental Medicine and Biology, 626, 62–70. Davies, W., Isles, A., Smith, R., Karunadasa, D., Burrmann, D., Humby, T., et al. (2005). Xlr3b is a new imprinted candidate for X-linked parent-of-origin effects on cognitive function in mice. Nature Genetics, 37, 625–629. Davies, W., Isles, A. R., & Wilkinson, L. S. (2001). Imprinted genes and mental dysfunction. Annali Medici, 33, 428–436. DeChiara, T. M., Robertson, E. J., & Efstratiadis, A. (1991). Parental imprinting of the mouse insulin-like growth factor II gene. Cell, 64, 849. Deng, Y. B., & Wu, X. W. (2000). Peg3/Pw1 promotes p53-mediated apoptosis by inducing Bax translocation from cytosol to mitochondria. Proceedings of the National Academy of Sciences of the United States of America, 97, 12050–12055. Dewsbury, D. A. (1969). Copulatory behaviour of rats (Rattus norvegicus) as a function of prior copulatory experience. Animal Behaviour, 17, 217–223. Dominguez, J. M., Brann, J. H., Gil, M., & Hull, E. M. (2006). Sexual experience increases nitric oxide synthase in the medial preoptic area of male rats. Behavioral Neuroscience, 120, 1389–1394. Dominguez, J. M., & Hull, E. M. (2001). Stimulation of the medial amygdala enhances medial preoptic dopamine release: Implications for male rat sexual behavior. Brain Research, 917, 225–229. Dominguez, J. M., & Hull, E. M. (2005). Dopamine, the medial preoptic area, and male sexual behavior. Physiology & Behavior, 86, 356–368. Dominguez, J., Riolo, J. V., Xu, Z., & Hull, E. M. (2001). Regulation by the medial amygdala of copulation and medial preoptic dopamine release. The Journal of Neuroscience, 21, 349–355. Dorries, A., Spohr, H. L., & Kunze, J. (1988). Angelman (“happy puppet”) syndrome – seven new cases documented by cerebral computed tomography: Review of the literature. European Journal of Pediatrics, 148, 270–273. Edinger, K. L., & Frye, C. A. (2007). Sexual experience of male rats influences anxiety-like behavior and androgen levels. Physiology & Behavior, 92, 443–453. Edwards, C. A., & Ferguson-Smith, A. C. (2007). Mechanisms regulating imprinted genes in clusters. Current Opinion in Cell Biology, 19, 281–289. Ferguson, J. N., Young, L. J., Hearn, E. F., Matzuk, M. M., Insel, T. R., & Winslow, J. T. (2000). Social amnesia in mice lacking the oxytocin gene. Nature Genetics, 25, 284–288. Fernandez-Guasti, A., Roldan-Roldan, G., & Saldivar, A. (1989). Reduction in anxiety after ejaculation in the rat. Behavioural Brain Research, 32, 23–29.

Genomic Imprinting and Sexual Experience-Dependent Learning in the Mouse

221

Fewell, G. D., & Meredith, M. (2002). Experience facilitates vomeronasal and olfactory influence on Fos expression in medial preoptic area during pheromone exposure or mating in male hamsters. Brain Research, 941, 91–106. Flannelly, K. J., Blanchard, R. J., Muraoka, M. Y., & Flannelly, L. (1982). Copulation increases offensive attack in male rats. Physiology & Behavior, 29, 381–385. Flannelly, K., & Lore, R. (1977). The influence of females upon aggression in domesticated male rats (Rattus norvegicus). Animal Behaviour, 25, 654–659. Galef, B. G., Jr., Lim, T. C. W., & Gilbert, G. S. (2008). Evidence of mate choice copying in Norway rats, Rattus norvegicus. Animal Behaviour, 75, 1117–1123. Glaser, R. L., Ramsay, J. P., & Morison, I. M. (2006). The imprinted gene and parent-of-origin effect database now includes parental origin of de novo mutations. Nucleic Acids Research, 34, D29–D31. Glenn, C. C., Driscoll, D. J., Yang, T. P., & Nicholls, R. D. (1997). Genomic imprinting: Potential function and mechanisms revealed by the Prader-Willi and Angelman syndromes. Molecular Human Reproduction, 3, 321–332. Goyens, J., & Noirot, E. (1975). Effects of cohabitation with females on aggressive behavior between male mice. Developmental Psychobiology, 8, 79–84. Haig, D., & Graham, C. (1991). Genomic imprinting and the strange case of the insulin-like growth factor II receptor. Cell, 64, 1045–1046. Hayashi, S., & Kimura, T. (1974). Sex-attractant emitted by female mice. Physiology & Behavior, 13, 563–567. Heeb, M. M., & Yahr, P. (1996). c-Fos immunoreactivity in the sexually dimorphic area of the hypothalamus and related brain regions of male gerbils after exposure to sex-related stimuli or performance of specific sexual behaviors. Neuroscience, 72, 1049–1071. Hillegaart, V., Alster, P., Uvnäs-Moberg, K., & Ahlenius, S. (1998). Sexual motivation promotes oxytocin secretion in male rats. Peptides, 19, 39–45. Hore, T. A., Rapkins, R. W., & Graves, J. A. M. (2007). Construction and evolution of imprinted loci in mammals. Trends in Genetics, 23, 440–448. Hosokawa, N., & Chiba, A. (2005). Effects of sexual experience on conspecific odor preference and estrous odor-induced activation of the vomeronasal projection pathway and the nucleus accumbens in male rats. Brain Research, 1066, 101–108. Hull, E. M., Bitran, D., Pehek, E. A., Warner, R. K., Band, L. C., & Holmes, G. M. (1986). Dopaminergic control of male sex behavior in rats: Effects of an intracerebrally-infused ­agonist. Brain Research, 370, 73–81. Hull, E. M., Du, J., Lorrain, D. S., & Matuszewich, L. (1995). Extracellular dopamine in the medial preoptic area: Implications for sexual motivation and hormonal control of copulation. The Journal of Neuroscience, 15, 7465–7471. Hurst, L. D., & McVean, G. T. (1997). Growth effects of uniparental disomies and the conflict theory of genomic imprinting. Trends in Genetics, 13, 436. Hurst, L. D., & McVean, G. T. (1998). Do we understand the evolution of genomic imprinting? Current Opinion in Genetics & Development, 8, 701–708. Illig, K. R., & Haberly, L. B. (2003). Odor-evoked activity is spatially distributed in piriform cortex. The Journal of Comparative Neurology, 457, 361–373. Isles, A. R., & Humby, T. (2006). Modes of imprinted gene action in learning disability. Journal of Intellectual Disability Research, 50, 318–325. Jiang, Y.-H., Tsai, T.-F., Bressler, J., & Beaudet, A. L. (1998). Imprinting in Angelman and Prader-Willi syndromes. Current Opinion in Genetics & Development, 8, 334. Jirikowski, G. F., Caldwell, J. D., Haussler, H. U., & Pedersen, C. A. (1991). Mating alters topography and content of oxytocin immunoreactivity in male mouse brain. Cell and Tissue Research, 266, 399–403. Johnson, M. D., Wu, X. W., Aithmitti, N., & Morrison, R. S. (2002). Peg3 / Pw1 is a mediator between p53 and Bax in DNA damage- induced neuronal death. The Journal of Biological Chemistry, 277, 23000–23007.

222

W.T. Swaney and E.B. Keverne

Kaneko-Ishino, T., Kohda, T., & Ishino, F. (2003). The regulation and biological significance of genomic imprinting in mammals. Journal of Biochemistry, 133, 699–711. Kavaliers, M., Colwell, D. D., & Choleris, E. (1998). Analgesic responses of male mice exposed to the odors of parasitized females: Effects of male sexual experience and infection status. Behavioral Neuroscience, 112, 1001–1011. Keller, M., Douhard, Q., Baum, M. J., & Bakker, J. (2006a). Destruction of the main olfactory epithelium reduces female sexual behavior and olfactory investigation in female mice. Chemical Senses, 31, 315–323. Keller, M., Douhard, Q., Baum, M. J., & Bakker, J. (2006b). Sexual experience does not compensate for the disruptive effects of zinc sulfate–lesioning of the main olfactory epithelium on sexual behavior in male mice. Chemical Senses, 31, 753–762. Kelley, A. E. (2004). Ventral striatal control of appetitive motivation: Role in ingestive behavior and reward-related learning. Neuroscience and Biobehavioral Reviews, 27, 765. Keverne, E. B. (2001). Genomic imprinting, maternal care, and brain evolution. Hormones and Behavior, 40, 146–155. Keverne, E. B., & Curley, J. P. (2008). Epigenetics, brain evolution and behaviour. Frontiers in Neuroendocrinology, 29, 398–412. Keverne, E. B., Fundele, R., Narasimha, S., Barton, S. C., & Surani, M. A. (1996). Genomic imprinting and the differential roles of parental genomes in brain development. Developmental Brain Research, 92, 91–100. Killian, J. K., Nolan, C. M., Stewart, N., Munday, B. L., Andersen, N. A., Nicol, S., et al. (2001). Monotreme IGF2 expression and ancestral origin of genomic imprinting. The Journal of Experimental Zoology, 291, 205–212. Kim, J., Ashworth, L., Branscomb, E., & Stubbs, L. (1997). The human homolog of a mouseimprinted gene, Peg3, maps to a zinc finger gene-rich region of human chromosome 19q13.4. Genome Research, 7, 532–540. Kippin, T. E., Talianakis, S., Schattmann, L., Bartholomew, S., & Pfaus, J. G. (1998). Olfactory conditioning of sexual behavior in the male rat (Rattus norvegicus). Journal of Comparative Psychology, 112, 389–399. Kondo, Y. (1992). Lesions of the medial amygdala produce severe impairment of copulatory behavior in sexually inexperienced male rats. Physiology & Behavior, 51, 939–943. Kuroiwa, Y., Kaneko-Ishino, T., Kagitani, F., Kohda, T., Li, L. L., Tada, M., et al. (1996). Peg3 imprinted gene on proximal chromosome 7 encodes for a zinc finger protein. Nature Genetics, 12, 186–190. Larsson, K. (1959). Experience and maturation in the development of sexual behaviour in male puberty rat. Behaviour, 14, 101–107. Larsson, K. (1978). Experiential factors in the development of sexual behaviour. In J. B. Hutchison (Ed.), Biological determinants of sexual behaviour (pp. 55–86). Chichester: Wiley. Lau, M. M., Stewart, C. E., Liu, Z., Bhatt, H., Rotwein, P., & Stewart, C. L. (1994). Loss of the imprinted IGF2/cation-independent mannose 6-phosphate receptor results in fetal overgrowth and perinatal lethality. Genes & Development, 8, 2953–2963. Leypold, B. G., Yu, C. R., Leinders-Zufall, T., Kim, M. M., Zufall, F., & Axel, R. (2002). Altered sexual and social behaviors in trp2 mutant mice. Proceedings of the National Academy of Sciences of the United States of America, 99, 6376–6381. Li, L., Keverne, E. B., Aparicio, S. A., Ishino, F., Barton, S. C., & Surani, M. A. (1999). Regulation of maternal behavior and offspring growth by paternally expressed Peg3. Science, 284, 330–333. Lisk, R. D., & Heimann, J. (1980). The effects of sexual experience and frequency of testing on retention of copulatory behavior following castration in the male hamster. Behavioral and Neural Biology, 28, 156–171. Lopez, H. H., & Ettenberg, A. (2000). Haloperidol challenge during copulation prevents subsequent increase in male sexual motivation. Pharmacology, Biochemistry and Behavior, 67, 387–393.

Genomic Imprinting and Sexual Experience-Dependent Learning in the Mouse

223

Lopez, H. H., & Ettenberg, A. (2002a). Exposure to female rats produces differences in c-fos induction between sexually-naive and experienced male rats. Brain Research, 947, 57–66. Lopez, H. H., & Ettenberg, A. (2002b). Sexually conditioned incentives: Attenuation of motivational impact during dopamine receptor antagonism. Pharmacology, Biochemistry and Behavior, 72, 65–72. Lopez, H. H., Olster, D. H., & Ettenberg, A. (1999). Sexual motivation in the male rat: The role of primary incentives and copulatory experience. Hormones and Behavior, 36, 176–185. Lorrain, D. S., & Hull, E. M. (1993). Nitric oxide increases dopamine and serotonin release in the medial preoptic area. NeuroReport, 5, 87–89. Lorrain, D. S., Matuszewich, L., Howard, R. V., Du, J., & Hull, E. M. (1996). Nitric oxide promotes medial preoptic dopamine release during male rat copulation. NeuroReport, 8, 31–34. Lumley, L. A., & Hull, E. M. (1999). Effects of a D1 antagonist and of sexual experience on copulation-induced Fos-like immunoreactivity in the medial preoptic nucleus. Brain Research, 829, 55–68. Mandiyan, V. S., Coats, J. K., & Shah, N. M. (2005). Deficits in sexual and aggressive behaviors in Cnga2 mutant mice. Nature Neuroscience, 8, 1660–1662. Manning, A., & Thompson, M. L. (1976). Postcastration retention of sexual behaviour in the male BDF1 mouse: The role of experience. Animal Behaviour, 24, 523–533. Martel, K. L., & Baum, M. J. (2009). A centrifugal pathway to the mouse accessory olfactory bulb from the medial amygdala conveys gender-specific volatile pheromonal signals. The European Journal of Neuroscience, 29, 368–376. McGrath, J., & Solter, D. (1984). Completion of mouse embryogenesis requires both the maternal and paternal genomes. Cell, 37, 179–183. Meredith, M. (1986). Vomeronasal organ removal before sexual experience impairs male hamster mating behavior. Physiology & Behavior, 36, 737–743. Moore, T., & Haig, D. (1991). Genomic imprinting in mammalian development – a parental ­tug-of-war. Trends in Genetics, 7, 45–49. Moore, T., & Mills, W. (2008). Evolutionary theories of imprinting – enough already! Advances in Experimental Medicine and Biology, 626, 116–122. Morison, I. M., Ramsay, J. P., & Spencer, H. G. (2005). A census of mammalian imprinting. Trends in Genetics, 21, 457–465. Murphy, M. R. (1973). Effects of female hamster vaginal discharge on the behavior of male hamsters. Behavioral Biology, 9, 367–375. Murphy, M. R., & Schneider, G. E. (1970). Olfactory bulb removal eliminates mating behavior in the male golden hamster. Science, 167, 302–304. Murphy, S. K., Wylie, A. A., & Jirtle, R. L. (2001). Imprinting of PEG3, the human homologue of a mouse gene involved in nurturing behavior. Genomics, 71, 110–117. Nishitani, S., Moriya, T., Kondo, Y., Sakuma, Y., & Shinohara, K. (2004). Induction of Fos immunoreactivity in oxytocin neurons in the paraventricular nucleus after female odor exposure in male rats: Effects of sexual experience. Cellular and Molecular Neurobiology, 24, 283–291. Nunez, A. A., & Tan, D. T. (1984). Courtship ultrasonic vocalizations in male Swiss-Webster mice: Effects of hormones and sexual experience. Physiology & Behavior, 32, 717–721. Nyby, J., Wysocki, C. J., Whitney, G., & Dizinno, G. (1977). Pheromonal regulation of male mouse ultrasonic courtship (Mus musculus). Animal Behaviour, 25, 333–341. Orr, H. A. (1995). Somatic mutation favors the evolution of diploidy. Genetics, 139, 1441–1447. Pask, A. J., Papenfuss, A. T., Ager, E. I., McColl, K. A., Speed, T. P., & Renfree, M. B. (2009). Analysis of the platypus genome suggests a transposon origin for mammalian imprinting. Genome Biology, 10, R1. Pfaus, J. G., Damsma, G., Nomikos, G. G., Wenkstern, D. G., Blaha, C. D., Phillips, A. G., et al. (1990). Sexual behavior enhances central dopamine transmission in the male rat. Brain Research, 530, 345–348.

224

W.T. Swaney and E.B. Keverne

Pfaus, J. G., & Phillips, A. G. (1991). Role of dopamine in anticipatory and consummatory aspects of sexual behavior in the male rat. Behavioral Neuroscience, 105, 727–743. Pfeiffer, C. A., & Johnston, R. E. (1994). Hormonal and behavioral responses of male hamsters to females and female odors: Roles of olfaction, the vomeronasal system, and sexual experience. Physiology & Behavior, 55, 129–138. Powers, J. B., & Winans, S. S. (1975). Vomeronasal organ: Critical role in mediating sexual behavior of the male hamster. Science, 187, 961–963. Rastogi, R. K., Milone, M., & Chieffi, G. (1981). Impact of socio-sexual conditions on the epididymis and fertility in the male mouse. Journal of Reproduction and Fertility, 63, 331–334. Reik, W., Constancia, M., Fowden, A., Anderson, N., Dean, W., Ferguson-Smith, A., et al. (2003). Regulation of supply and demand for maternal nutrients in mammals by imprinted genes. The Journal of Physiology, 547, 35–44. Relaix, F., Wei, X. J., Li, W., Pan, J. J., Lin, Y. H., Bowtell, D. D., et al. (2000). Pw1 / Peg3 is a potential cell death mediator and cooperates with Siah1a in p53-mediated apoptosis. Proceedings of the National Academy of Sciences of the United States of America, 97, 2105–2110. Relaix, F., Weng, X., Marazzi, G., Yang, E., Copeland, N., Jenkins, N., et al. (1996). Pw1, a novel zinc finger gene implicated in the myogenic and neuronal lineages. Developmental Biology, 177, 383–396. Rodriguez-Manzo, G., Lopez-Rubalcava, C., & Fernandez-Guasti, A. (1999). Anxiolytic-like effect of ejaculation under various sexual behavior conditions in the male rat. Physiology & Behavior, 67, 651–657. Sahoo, T., del Gaudio, D., German, J. R., Shinawi, M., Peters, S. U., Person, R. E., et al. (2008). Prader-Willi phenotype caused by paternal deficiency for the HBII-85 C/D box small nucleolar RNA cluster. Nature Genetics, 40, 719–721. Sipos, M. L., Kerchner, M., & Nyby, J. G. (1992). An ephemeral sex pheromone in the urine of female house mice (Mus domesticus). Behavioral and Neural Biology, 58, 138–143. Sipos, M. L., Wysocki, C. J., Nyby, J. G., Wysocki, L., & Nemura, T. A. (1995). An ephemeral pheromone of female house mice: Perception via the main and accessory olfactory systems. Physiology & Behavior, 58, 529–534. Sleutels, F., Barlow, D. P., & Lyle, R. (2000). The uniqueness of the imprinting mechanism. Current Opinion in Genetics & Development, 10, 229–233. Solter, D. (1988). Differential imprinting and expression of maternal and paternal genomes. Annual Review of Genetics, 22, 127–146. Spahn, L., & Barlow, D. P. (2003). An ICE pattern crystallizes. Nature Genetics, 35, 11–12. Steel, E., & Keverne, E. B. (1985). Effect of female odour on male hamsters mediated by the vomeronasal organ. Physiology & Behavior, 35, 195–200. Stern, J. J. (1970). Responses of male rats to sex odors. Physiology & Behavior, 5, 519–524. Stowers, L., Holy, T. E., Meister, M., Dulac, C., & Koentges, G. (2002). Loss of sex discrimination and male-male aggression in mice deficient for TRP2. Science, 295, 1493–1500. Surani, M. A. (1998). Imprinting and the initiation of gene silencing in the germ line. Cell, 93, 309. Surani, M. A., Barton, S. C., & Norris, M. L. (1984). Development of reconstituted mouse eggs suggests imprinting of the genome during gametogenesis. Nature, 308, 548–550. Surani, M. A. H., Barton, S. C., & Norris, M. L. (1986). Nuclear transplantation in the mouse: Heritable differences between parental genomes after activation of the embryonic genome. Cell, 45, 127–136. Suzuki, S., Renfree, M. B., Pask, A. J., Shaw, G., Kobayashi, S., Kohda, T., et al. (2005). Genomic imprinting of IGF2, p57KIP2 and PEG1/MEST in a marsupial, the tammar wallaby. Mechanisms of Development, 122, 213–222. Swaney, W. T., Curley, J. P., Champagne, F. A., & Keverne, E. B. (2007). Genomic imprinting mediates sexual experience-dependent olfactory learning in male mice. Proceedings of the National Academy of Sciences of the United States of America, 104, 6084–6089.

Genomic Imprinting and Sexual Experience-Dependent Learning in the Mouse

225

Swaney, W. T., Curley, J. P., Champagne, F. A., & Keverne, E. B. (2008). The paternally expressed gene Peg3 regulates sexual experience-dependent preferences for estrous odors. Behavioral Neuroscience, 122, 963–973. Thorvaldsen, J. L., Duran, K. L., & Bartolomei, M. S. (1998). Deletion of the H19 differentially methylated domain results in loss of imprinted expression of H19 and Igf2. Genes & Development, 12, 3693–3702. Tycko, B., & Morison, I. M. (2002). Physiological functions of imprinted genes. Journal of Cellular Physiology, 192, 245–258. Valenstein, E. S., & Goy, R. W. (1957). Further studies of the organization and display of sexual behavior in male guinea pigs. Journal of Comparative and Physiological Psychology, 50, 115–119. Valenstein, E. S., Riss, W., & Young, W. C. (1955). Experiential and genetic factors in the organization of sexual behavior in male guinea pigs. Journal of Comparative and Physiological Psychology, 48, 397–403. Varmuza, S., & Mann, M. (1994). Genomic imprinting – defusing the ovarian time bomb. Trends in Genetics, 10, 118–123. Warner, R. K., Thompson, J. T., Markowski, V. P., Loucks, J. A., Bazzett, T. J., Eaton, R. C., et al. (1991). Microinjection of the dopamine antagonist cis-flupenthixol into the MPOA impairs copulation, penile reflexes and sexual motivation in male rats. Brain Research, 540, 177–182. Warren, W. C., Hillier, L. W., Marshall Graves, J. A., Birney, E., Ponting, C. P., Grutzner, F., et al. (2008). Genome analysis of the platypus reveals unique signatures of evolution. Nature, 453, 175–183. Watts, P. C., Buley, K. R., Sanderson, S., Boardman, W., Ciofi, C., & Gibson, R. (2006). Parthenogenesis in Komodo dragons. Nature, 444, 1021–1022. Wenkstern, D., Pfaus, J. G., & Fibiger, H. C. (1993). Dopamine transmission increases in the nucleus accumbens of male rats during their first exposure to sexually receptive female rats. Brain Research, 618, 41–46. Whittington, J., Holland, A., Webb, T., Butler, J., Clarke, D., & Boer, H. (2004). Cognitive abilities and genotype in a population-based sample of people with Prader-Willi syndrome. Journal of Intellectual Disability Research, 48, 172–187. Wilkinson, L. S., Davies, W., & Isles, A. R. (2007). Genomic imprinting effects on brain development and function. Nature Reviews. Neuroscience, 8, 832–843. Williamson, C. M., Ball, S. T., Nottingham, W. T., Skinner, J. A., Plagge, A., Turner, M. D., et al. (2004). A cis-acting control region is required exclusively for the tissue-specific imprinting of Gnas. Nature Genetics, 36, 894. Williamson, C. M., Turner, M. D., Ball, S. T., Nottingham, W. T., Glenister, P., Fray, M., et al. (2006). Identification of an imprinting control region affecting the expression of all transcripts in the Gnas cluster. Nature Genetics, 38, 350–355. Witt, D. M., & Insel, T. R. (1994). Increased Fos expression in oxytocin neurons following ­masculine sexual behavior. Journal of Neuroendocrinology, 6, 13–18. Yamaguchi, A., Taniguchi, M., Hori, O., Ogawa, S., Tojo, N., Matsuoka, N., et al. (2002). Peg3/ Pw1 is involved in p53-mediated cell death pathway in brain ischemia/hypoxia. The Journal of Biological Chemistry, 277, 623–629. Yoon, H., Enquist, L. W., & Dulac, C. (2005). Olfactory inputs to hypothalamic neurons controlling reproduction and fertility. Cell, 123, 669–682. Zufall, F., & Leinders-Zufall, T. (2007). Mammalian pheromone sensing. Current Opinion in Neurobiology, 17, 483–489.

Proteomic Analysis of the Postsynaptic Density Ayse Dosemeci

Abstract  The postsynaptic density (PSD) is a large protein complex lining the postsynaptic membrane with an average mass around 1 million KDa. The structure appears to be an organized array containing neurotransmitter receptors and signal transduction molecules held together by specialized scaffold proteins. PSDs can be isolated by subcellular fractionation, the most widely used protocols consisting of detergent treatment of synaptosomal fractions followed by further density gradient centrifugation steps. Several groups used mass spectrometric approaches to identify proteins in these PSD fractions. An inherent problem in the determination of PSD composition through analysis of isolated fractions has been the presence of contaminants. A method for further purification of PSD fractions by an orthogonal affinity-based strategy and comparison of the parent and purified preparations to trace the enrichment of proteins allows the compilation of a more restricted list of putative PSD components. Ultimately, however, the presence of the identified proteins at the PSD has to be confirmed by immuno-electron microscopy. Proteomic techniques for relative quantification of proteins are applied for the detection of those proteins whose abundance in the PSD change following synaptic activity, as well as in response to various pathological conditions such as ischemia. Keywords  Postsynaptic density • PSD • Mass spectrometry • Proteomics • Electron microscopy

1 Introduction The postsynaptic density (PSD) is a very large neuronal protein complex lining the postsynaptic membrane at the active zone. A recent study estimates the average mass of the PSD as ~109 Da (Chen et al., 2005), equivalent to10,000 proteins with A. Dosemeci (*) Laboratory of Neurobiology, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD 20892, USA e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_9, © Springer Science+Business Media, LLC 2011

227

228

A. Dosemeci

an average molecular weight of 100 KDa each. By electron microscopy (EM) the PSD appears as an-electron dense structure on the cytoplasmic side of the postsynaptic membrane (Fig. 1). Although virtually all synapses have some sort of postsynaptic specialization at the active zone, the term PSD is most commonly used to refer to those more prominent specializations at excitatory glutamatergic synapses. The present review also uses the narrow definition of the PSD as a specialization of excitatory synapses. The position of the PSD at the active zone suggests a fundamental role in defining the receptivity of the postsynaptic neuron to incoming stimuli. Indeed, receptors are strategically concentrated there, facing the sites of presynaptic neurotransmitter release. The whole protein complex appears to be an array that organizes the ­postsynaptic reception by promoting specific positioning of receptors and signal transduction elements. A crucial task in the clarification of the mechanisms of postsynaptic reception and activity-induced changes in synaptic strength is identification of those constituent proteins of the PSD, determination of their stoichiometries and characterization of activity-induced changes in their composition. In the present chapter, I will ­discuss the use of proteomic techniques in the elucidation of PSD composition and of its activity-induced modification.

Fig. 1  Typical excitatory synapse. The electron-dense material lining the postsynaptic membrane (arrow) is the postsynaptic density (PSD). EM: courtesy of J-H. Tao Cheng

Proteomic Analysis of the Postsynaptic Density

229

2 Composition of the PSD The composition of the PSD has been largely deduced from the analysis of biochemically­ isolated PSD preparations. The introduction of powerful mass spectrometric techniques allowed for the global analysis of PSD fractions without a need for prior identification of candidate proteins or a priori hypotheses. In this section, a critical­review of ­published proteomic studies directed to the elucidation of PSD composition will be presented. I will argue that a major problem in the interpretation of these studies is the presence of contaminants – a problem inherent to organelle proteomics in general, but enhanced in the special case of the PSD, an organelle without well-­delineated boundaries­and whose components may also be found in other cellular compartments. Finally, strategies for improving the purity of PSD fractions and ­methods for the validation of the results obtained in proteomic analyses will be discussed.

2.1 Isolation of PSDs In the 1970s, several laboratories observed that PSDs are resistant to detergent treatments that cause solubilization of cellular membranes (Cohen, Blomberg, Berzins, & Siekevitz, 1977; Cotman, Banker, Churchill, & Taylor, 1974; Matus & Taff-Jones, 1978) and took advantage of this property in devising protocols for their isolation. To limit the complexity of the detergent insoluble material, virtually all protocols devised for the isolation of PSDs first involve preparation of a synaptosomal fraction. Treatment of the synaptosomal fraction with detergents results in the solubilization of membranes, including pre- and postsynaptic membranes at the cleft, and release membrane-free PSDs. The most widely applied method for the preparation of PSD fraction is the one developed in Siekevitz’s laboratory (Carlin, Grab, Cohen, & Siekevitz, 1980; Cohen et  al., 1977) which uses the neutral detergent TritonX-100. The protocol involves extraction of synaptosomal membranes with 0.5% Triton X-100; fractionation of the Triton-insoluble pellet on a sucrose gradient and a second extraction of the ­PSD-enriched fraction with Triton. EM examination of the Triton X-100 derived PSD fractions (Fig. 2) reveals particles of similar size and morphology to the in situ PSD, indicating that the bulk of the PSD survives TritonX-100 treatment.

2.2 Identification of Proteins in PSD Fractions Prior to mass spectrometric methods, western immunoblotting and, to a lesser extent, peptide sequencing had been applied to identify components of the PSD fraction. Most studies were confined to confirming the presence of a single protein or of a family of proteins, while a few aimed for a global analysis (Walsh & Kuruc, 1992). In the 2000s, with the advent of mass spectrometric methods, a flurry of proteomic studies of the PSD were published (Table 1).

230

A. Dosemeci

Fig. 2  TritonX-100 derived conventional PSD fraction. Thin section EM shows structures with a similar morphology to in situ PSDs. Arrow shows a typical isolated PSD. Bar 500  nm. EM: ­courtesy of J-H. Tao Cheng Table 1  Mass spectrometric analyses of PSD fractions Preparation strategy after References synaptosome/SPM stage Proteomic strategy Walikonis et al. 2× TritonX-100 extraction 1D-E, MS (2000) Li et al. (2004) 2× TritonX-100 extraction, 2D-E, MS/MS ICAT, LC-MS/MS 2D-LC-MS/MS Yoshimura et al. TritonX-100 extraction, (2004) density gradient centrifugation Peng et al. (2004) 2× TritonX-100 extraction 1D-E, LC-MS/MS 1D-E, LC-MS/MS Jordan et al. 2× TritonX-100 extraction, (2004) density gradient centrifugation 1D-E, LC-MS/MS Collins et al. TritonX-100 extraction, (2006) density gradient centrifugation Trinidad et al. N-octyl-b-d glucopyranoside 2D-LC-MS/MS (2006) extraction, density gradient centrifugation

Number of proteins identified 31 Not reported 492

374 452

620

1,264

SPM synaptic plasma membrane fraction; 1D-E, 2D-E one and two-dimensional electrophoresis; LC liquid chromatography; 2D-LC two-dimensional liquid chromatography; MS mass spectrometry; MS/MS tandem mass spectrometry

Proteomic analysis of a protein mixture such as the PSD typically involves digestion with a protease, usually trypsin, mass spectrometric analysis of resulting peptides, database searching for the identification of peptides and inference of intact proteins from identified peptides. Various strategies have been used for

Proteomic Analysis of the Postsynaptic Density

231

reducing the complexity of the peptide mixture that is analyzed by mass spectrometry­at any one time. The advantages and disadvantages of these strategies, which include pre-fractionation of proteins by 1D and 2D electrophoresis, fractionation of peptides by liquid chromatography and selection for cysteine-containing peptides, are discussed below. Fractionation of proteins: Prior fractionation by electrophoresis has the advantage of reducing the complexity of the mixture that is analyzed at any one time by mass spectrometry. On the other hand, electrophoretic separation can miss proteins in the very high and very low molecular weight ranges that may not enter or run off the gel respectively. In addition, 2D electrophoresis can miss out proteins in the extreme ranges of pI as well as those of high hydrophobicity due to their poor resolution in the first (isoelectric focusing) dimension. 2D electrophoresis is especially problematic for the analysis of PSD proteins, many of which are poorly solubilized in the absence of the strong detergent SDS. Indeed, comparing 1D and 2D separation profiles, Walsh & Kuruc (1992) concluded that many PSD proteins do not enter the first dimension in 2D electrophoresis in a quantitative fashion. Fractionation of peptides: While most proteomic studies of the PSD use reverse phase HPLC for fractionation of tryptic peptides, two studies, Yoshimura et  al. (2004) and Trinidad, Specht, Thalhammer, Schoepfer, & Burlingame (2006), added another dimension, strong cation exchange (SCX) chromatography. This additional liquid chromatography step reduces the complexity of the sample and allows elimination of the prior electrophoresis step for the fractionation of proteins. Trinidad et al. (2006) noted that their approach resulted in the identification of a far greater number of proteins in the PSD fraction. Another strategy used for reducing the complexity of the sample is selection of cysteine-containing peptides by isotope-coded affinity tagging (ICAT). Li et  al. (2004) used ICAT selection followed by LC and mass spectrometry as a complementary approach to 2D electrophoresis followed by mass spectrometry. The authors show that certain PSD proteins that were not detected by their 2D electrophoresis approach could be detected by the alternative ICAT approach. While fractionation at the peptide stage has the advantage of avoiding some of the pitfalls of electrophoresis mentioned above, it should be noted that elimination of the electrophoresis step is at the expense of the loss of an important piece of information for the evaluation of the data. Indeed, the ability to associate identified peptides with specific gel slices or spots, with defined mobilities (1D electrophoresis­) or defined mobilities and pIs (2D electrophoresis), places important constraints for the identification of intact proteins, especially in cases where the same peptides map to multiple proteins or protein isoforms. Proteins identified: Altogether, proteomic studies based on mass spectrometry listed in Table 1 identified more than 1,000 proteins in the PSD fraction. The proteins identified include scaffold proteins, most notably PSD-95, PSD-93 (Chapsyn-110), SAP97, SAP102, SAPAPs (GKAPs), Shanks and Homer; glutamate receptors; enzymes, including CaMKII which is the most abundant protein in PSD fractions, SynGAP and other G-protein regulators; cell adhesion molecules; cytoskeletal ­elements, including spectrins, neurofilaments, tubulins and actin; and proteins

232

A. Dosemeci

described to be localized in ribosomes, mitochondria and the presynaptic compartment­. Collins et al. (2006) combined data from their study and six previous studies, as well as the PSD literature, to compile a list of a total of 1,124 proteins detected in PSD fractions. A majority (58%) of these proteins, however, were detected in only one study. Because detection in multiple studies adds confidence to the status of a protein as a PSD constituent, they designated 466 proteins detected in two or more studies as “Consensus PSD.” The Consensus PSD list constitutes a good basis for the assessment of the PSD proteome. However, it should be emphasized that, in the absence of further verification, as discussed in Sects. 2.5 and 2.6, the proteins in the list should be considered only as candidate PSD constituents. Possible sources of error include inherent difficulties in identification of proteins from data generated by tandem mass spectrometric analysis of protein mixtures (see review by Nesvizhskii, Vitek, & Aebersold, 2007) and presence of contaminants in PSD preparations as discussed below.

2.3 Problem of Contaminants On replicas of the conventional PSD fraction (Fig. 3), PSDs are clearly identified by their specific morphology and exclusive labeling for PSD-95. In addition to PSDs, however, other particulate material is observed, including spectrin filaments, neurofilaments and spherical structures known as CaMKII clusters. All three of these proteins are major components of the PSD fraction, and indeed are among a handful of proteins detected in all the studies listed in Table  1, underlining the prominence of contaminants in the preparation. The problem of contamination is not unique to the analysis of PSD fractions. Indeed, subcellular fractionation never yields 100% pure preparations, and

Fig.  3  Particulate contaminants in the TritonX-100 derived PSD fraction. PSD fraction was adhered to glass, immunogold labeled and rotary shadowed to obtain replicas. A selected area of the grid illustrates presence of neurofilaments, spectrin and CaMKII clusters in addition to PSDs that typically label for PSD-95. Note that the area was chosen for the variety of particulate material and does not reflect the actual proportions of contaminants to PSD in the fraction. Bar 100 nm. From Vinade et al., (2003); printed with the permission of Wiley-Blackwell

Proteomic Analysis of the Postsynaptic Density

233

contamination­ is a general problem in organelle proteomics. On the other hand, the absence of precise membraneous boundaries for the PSD as an organelle and the presence of proteins that may not be uniquely located at the PSD make the distinction between genuine PSD components and contaminants even more problematic. Weeding out proteins that are known to be exclusively localized in other cells or organelles is a first step in dealing with the contamination issue. Some of the reports on the proteomic analysis of the PSD recognize such contaminant groups – glial, presynaptic, mitochondrial, nuclear, etc. (e.g., Yoshimura et al., 2004). On the other hand, the status of proteins for which no immuno-cytochemical localization has been described, or those known to be located in more than one cellular compartment, ­cannot be resolved without additional information. This problem is well illustrated by the case of CaMKII, the major protein in PSD fractions, which is shown by immuno-cytochemistry to be widely distributed throughout the neuron (Tao-Cheng, Dosemeci, Winters, & Reese, 2006).

2.4 PSD Fractions of Higher Purity Most proteomic studies focusing on the PSD used a Triton-X100-derived PSD fraction, with the exception of Trinidad et al. (2006), who adopted an alternative preparation technique with the detergent N-octyl-b-d glucopyranoside. Another variation in preparation protocol is the omission, in some studies, of the density gradient ­centrifugation step after detergent extraction of synaptosomal membranes (Table 1). At this stage it is hard to tell precisely how particular variations in preparation protocols affected the list of proteins identified, due to confounding factors such as  variations in proteomic analyses. In general, however, it is expected that the omission of the density gradient fractionation of the detergent-insoluble material would result in a greater degree of contamination. Indeed, Suzuki et  al. (2001) described the separation of detergent-insoluble, less dense lipid rafts from the PSDs by the density gradient. While the differences in the proteins identified in studies listed in Table 1 can be due to variations in the preparative techniques and strategies for data analysis, detection of the same contaminants in different studies is indicative of a systemic problem. Thus, a review of the general preparation strategies is useful. Conventional subcellular fractionation techniques are based on differential and density gradient centrifugation and separate organelles with respect to density. Two types of ­contaminants are likely to remain with the PSDs through centrifugation-based fractionation. One type of contaminants are proteins from other cellular compartments, such as synapsins, which attach to the PSD in low ionic strength buffers used during cell disruption and subcellular fractionation. Contaminants in this category may be removed by treatment with high ionic strength solutions. The second type of ­contaminants are proteins making up particulate material that co-fractionates with the PSDs. PSDs which contain very little or no lipids are denser than certain other detergent-insoluble material derived from synaptosomal fractions such as lipid rafts, and can be separated from

234

A. Dosemeci

them by density gradient centrifugation. However, other particulate material such as cytoskeletal elements and CaMKII clusters are also essentially all protein and thus have densities similar to PSDs. Separation of lipid-free particulate contaminants, therefore, cannot be achieved by the additional centrifugation steps. 2.4.1 N-Lauroyl-Sarcosinate-Derived PSD Fraction One strategy used for the removal of contaminants is treatment of the PSD fraction with the relatively strong, ionic detergent N-lauroyl sarcosinate (NLS Cho, Hunt, & Kennedy, 1992). This fraction, occasionally called PSD III (PSD I and PSD II referring to the fractions obtained after the first and second TritonX-100 treatment, respectively), contains fewer proteins than the Triton-derived fraction. NLS treatment presumably removes a number of contaminants but also certain PSD components that are weakly associated with the structure, leaving what has been described as the “PSD core” (Kennedy, 1997). On the other hand, further observations by EM (Dosemeci, Reese, Petersen, & Tao-Cheng, 2000) show that certain particulate contaminants in the Triton-derived PSD fraction, called CaMKII clusters, survive NLS treatment. This would explain the extremely high levels of CaMKII observed in the NLS-derived preparation. 2.4.2 Affinity Purification of PSD Fraction Using Magnetic Beads A challenge in further purification of PSDs is separating out particulate material that is of similar density to PSDs and therefore co-fractionates during centrifugation. For this purpose, affinity based separation can be used as an orthogonal approach to density based centrifugal fractionation. A method we devised uses magnetic beads coated with an antibody against PSD-95 (Vinade et  al., 2003). ImmunoEM studies indicate that PSD-95 is a good marker for PSDs, and studies with isolated PSD fractions demonstrate that the antibody can access PSD-95 from either surface of the PSD (Petersen et al., 2003). The affinity purification method consists of further purifying the conventional PSD fraction by taking advantage of selective binding of PSDs to magnetic beads. Other particulate material in the PSD fraction, including cytoskeletal elements and CaMKII clusters, do not bind to the beads, and are separated when beads are ­collected with a magnet and washed repeatedly. Repeated washes in relatively high ionic strength buffers (Tris-buffered saline) presumably also remove contaminants such as synapsin that may have associated with the PSD after cell disruption. EM examination of the affinity purified sample shows particles with the characteristic morphology of PSDs and none of the known particulate contaminants (Fig.  4). Control beads lacking the PSD-95 antibody coating do not exhibit any attached material, indicating an absence of nonspecific binding. Proteins in the affinity purified fraction as well as the parent PSD fraction were separated by 1D electrophoresis and analyzed by LC/MS/MS. Altogether, 288

Proteomic Analysis of the Postsynaptic Density

235

Fig.  4  Affinity-purification with magnetic beads coated with an antibody against PSD-95. Structures that have similar morphology to in situ PSDs associate with the beads, while other particulate material are washed away. Bar 100 nm Table 2  Major proteins in conventional PSD fraction that are depleted/greatly reduced upon further purification with PSD-95 antibody coated beads

Protein (family) name Protein origin/category Bassoon Presynaptic ERC protein 2 Presynaptic Piccolo Presynaptic RIM Presynaptic SNIP Presynaptic Synapsin 1&2 Presynaptic a-internexin Cytoskeletal (IF) Neurofilament-L & -M Cytoskeletal (IF) Plectin Cytoskeletal a- & b-spectrin Cytoskeletal acidic ribosomal proteins Ribosomal VDAC1&2 Mitochondrial Myelin basic protein Glial a-actinin Actin-binding synaptopodin Actin-binding b-catenin Actin-binding/cell adhesion N-cadherin Cell adhesion Relative abundance estimates based on summed ion current intensities for each protein in the ­parent and purified PSD fractions were used to rank proteins and evaluate co-purification (Dosemeci et al., 2007). The proteins listed are among the 50 top ranking in the parent fraction but are either undetected or greatly reduced in the purified fraction

p­ roteins were detected in the affinity-purified preparation, out of the 525 in the parent­ PSD preparation (Dosemeci et  al., 2007). Comparison of the purified and parent fractions indicates a drastic loss of likely contaminants including presynaptic, cytoskeletal, mitochondrial, ribosomal and glial elements (Dosemeci et  al., 2007; Table  2). Less clear is the status of a few proteins, a-actinin, synaptopodin, b-catenin/N-cadherin, which have been observed at the contact zone by immunoEM (Mundel et  al., 1997; Petralia, Sans, Wang, & Wenthold, 2005; Rubio, Curcio,

236

A. Dosemeci

Chauvet, & Bruses, 2005; Wyszynski et al., 1998), but whose levels decrease upon affinity purification. One possibility is that these actin-binding ­proteins dissociate from the PSD upon partial disassembly of actin during purification. In conclusion, the affinity-purified preparation contains drastically reduced levels of contaminants but may have lost few peripheral components – thus the designation “synaptic PSD95 complex” rather than PSD may be appropriate. On the other hand, EM observations suggest that the bulk of the PSD survives the purification protocol (Fig. 4).

2.5 Validation of PSD Constituents by Biochemical Techniques As discussed in the previous section, improvement of fractionation techniques to obtain a PSD preparation of high purity helps reduce the number of false positives in proteomic analyses. However, because it is virtually impossible to obtain a ­biochemical preparation of 100% purity, it is also important to develop criteria that would validate the findings. It has been standard procedure in biochemical fractionation protocols to follow co-purification of proteins with particular organelles to establish evidence of the localization of proteins in specific subcellular sites. The most common method for the comparison of consecutive subcellular fractions is western immunoblotting. Using this strategy, Jordan et  al. (2004) demonstrated enrichment in the TritonX-100-derived PSD preparation of more than a dozen proteins and Collins et al. (2006) demonstrated enrichment of 21 proteins out of 48 tested. Andersen et  al. (2003) introduced a proteomic approach for establishing ­co-purification of proteins with organelles. This method, called protein correlation profiling, uses relative quantification based on ion current peaks to measure the co-purification of proteins through successive fractionation steps during the isolation of an organelle. The group applied this approach to establish components of the centrosome (Andersen et al., 2003) and ten subcellular locations in liver cells (Foster et al., 2006). Li et al. (2005) adopted the correlation profiling strategy and used the relative quantification method of ICAT (described in the next section) to follow the enrichment of proteins in a PSD fraction as compared to the parent synaptic membrane preparation. A total of 60 proteins were found to have an ICAT ratio >1, indicating enrichment. The study also showed that most mitochondrial proteins, as well as a number of proteins in other categories including energy production and transporters­, are greatly reduced in the PSD fraction. On the other hand, certain other ­contaminants such as plectin and glial fibrillary acidic protein (GFAP) were also found to be enriched. As mentioned in the above section, the Triton-derived conventional PSD ­fraction was further affinity purified using magnetic beads coated with a PSD-95 antibody (Dosemeci et  al., 2007; Vinade et  al., 2003). Co-purification of a protein was assessed comparing the parent and purified fractions using a relative abundance index based on cumulative ion current intensities of corresponding peptides (see Dosemeci et al., 2007 for details). Some of the major proteins that were found to co-purify according to the set criteria are listed on Table 3. These include ­specialized

Proteomic Analysis of the Postsynaptic Density Table 3  Major proteins that co-purify upon isolation of PSDs PSD-95 antibody Enrichment from Protein family/name synaptosome to PSD Scaffolds PSD-95 family/PSD-95, Cho et al. (1992); Li et al. PSD-93 (2005)a; Collins et al. (2006) SAPAPs (GKAPs)/SAPAP1, Kim et al. (1997); SAPAP2, SAPAP4 Li et al. (2005)a Shanks/Shank2, Shank3 Naisbitt et al. (1999) Homer/homer 1 Sun et al. (1998); Li et al. (2005)a AIDA-1/Q8BZM2) Jordan, Fernholz, Khatri, & Ziff (2007)

237 with magnetic beads coated with

EM validation Aoki et al. (2001)

Naisbitt et al. (1997) Naisbitt et al. (1999) Tu et al. (1999)

Glutamate receptors and related proteins AMPA receptor subunits/ GluR1, GluR2, GluR3, GluR4

Collins et al. (2006)

NMDA receptor subunits/ NR1, NR2A, NR2B

Collins et al. (2006); Li et al. (2005)a

TARP/g8 (AMPA receptor anchoring) Syne-1/CPG2 protein (receptor internalization) G-protein regulators RasGAP/SynGAP

Tomita et al. (2003); Li et al. (2005)a Li et al. (2005)a

ArfGEFs/BRAG1, BRAG2b

Chen, Rojas-Soto, Oguni, & Kennedy (1998); Li et al. (2005)a Murphy, Jensen, & Walikonis (2006); Li et al. (2005)a

Other Neuroligin (cell adhesion) Cylindromatosis (ubiquitin thiolesterase)

Baude, Nusser, Molnar, McIlhinney, & Somogyi (1995); Kharazia & Weinberg (1997) Petralia et al. (1994); Petralia et al. (2005); Kharazia & Weinberg (1997) Tomita et al. (2003); Fukaya et al. (2006) Cottrell et al. (2004)b

Petralia et al. (2005)

Sakagami et al. (2008)

Song, Ichtchenko, Sudhof, & Brose (1999) Li et al. (2005)a

Conventional PSD fraction was further purified with antibody-coated magnetic beads. Relative abundance estimates based on summed ion current intensities for each protein in the parent and purified PSD fractions were used to rank proteins and evaluate co-purification. Proteins listed are among the 50 top ranking in the purified fraction (Dosemeci et al., 2007). It should be noted that this is not a comprehensive list of all major proteins present in the purified fraction. For example, the most abundant protein CaMKII is not listed because it does not show co-purification according to the chosen criteria. Other proteins (LRTM1, Protein FAM81A, Transmembrane protein 132B) that meet the criteria for co-purification are not listed in the absence of additional biochemical and EM validation a Li et  al. (2005) used a proteomic strategy (ICAT) to compare to levels of proteins in synaptic membrane and PSD fractions b CPG2 immunogold labeling is described to be “lateral and underneath the PSD,” indicating that the protein is not part of the main PSD matrix

238

A. Dosemeci

PSD scaffolds PSD-95, PSD-93, GKAPs, Shanks and Homer, glutamate receptors of AMPA and NMDA types, G-protein regulators, and neuroligin, a cell adhesion molecule. The list also contains a protein of the AIDA-1 family with two protein– protein interaction (SAM) domains that may function as another PSD scaffold. It should be noted that Table  3 does not list all the major proteins in the affinitypurified fraction. For example, the most abundant protein CaMKII does not meet the set criteria for co-purification because it is present both in PSDs and in contaminating particulate material. Co-purification in consecutive fractionation steps, especially if they are based on independent properties (such as density and affinity), is a more robust indicator that a protein is a bona fide element of a particular organelle. It can be observed from Table 3 (second column) that several of the proteins that co-purify on beads coated with PSD-95 antibody were also found to be enriched during the previous fractionation step (treatment of synaptosomal preparations with Triton X-100 and centrifugation), strengthening their status as genuine PSD components. On the other hand, a number of proteins that are substantially reduced upon affinity purification were found to be enriched during the previous purification step. These include the actin binding protein b-catenin and cytoskeletal elements spectrin, a-internexin and plectin (Collins et  al., 2006; Li et  al., 2005). Such conflicting results during two consecutive fractionation steps underlie the importance of additional validation strategies. As discussed for particular examples, and more generally below, immuno-EM can provide the most direct evidence for the presence of a protein at the PSD.

2.6 Validation of PSD Constituents by Immuno-Electron Microscopy Figure  5 shows the synaptic localization by immunoEM of two proteins that are prominent in conventional PSD fractions. Both PSD-95 and synapsin 1 were detected in all the proteomic analyses of PSD fractions listed in Table 1. Label for PSD-95 is on the PSD, confirming this protein as a bona fide PSD constituent, whereas label for the other, synapsin 1, is exclusively at the presynaptic compartment indicating that it is a contaminant. Immuno-EM studies have confirmed the localization at the PSD of several of major proteins identified in the affinity-purified fraction, including many specialized scaffolds and glutamate receptors (Table 3, third column). On the other hand, immunolabel for one of the proteins listed, CPG2, was observed to be not actually on the PSD but around it (Cottrell, Borok, Horvath, & Nedivi, 2004), suggesting that this protein may be associated as an appendage to the PSD. The status of certain other proteins identified remains to be clarified. As mentioned before, EM examination of the PSD fraction also gives important insights into the status of identified proteins. In this respect, the example of CaMKII, the major protein in PSD fractions, is worth examining. Immuno-EM

Proteomic Analysis of the Postsynaptic Density

239

Fig.  5  Labeling of the synapse with antibodies against PSD-95 and Synapsin 1, two prominent proteins in conventional PSD fractions. ImmunoEM observations confirm the status of PSD-95 as a genuine PSD constituent but indicate that Synapsin1 is a contaminant in the PSD fraction originating from the presynaptic compartment. EM: courtesy of J-H. Tao Cheng

studies of the conventional PSD fraction show that CaMKII is clearly located on PSDs, but also on spherical particulate material in the fraction that does not label for PSD-95 (Dosemeci et  al., 2000). These structures, called CaMKII clusters, presumably form during the postmortem handling of brain tissue to contaminate the PSD fraction. Thus, as discussed in Sect. 2.7.1, quantitative data based on conventional PSD fractions would give erroneously high estimates for CaMKII. These considerations indicate that, when quantification of proteins is the goal, immunoEM verification of the PSD fraction may be necessary.

2.7 Quantification 2.7.1 Absolute Quantification Once constituent proteins of the PSD are identified, an important second step in the elucidation of the molecular organization of the organelle is determination of the stoichiometry of these proteins. Mass spectrometric strategies for absolute quantification of PSD proteins involved use of isotope-labeled internal standards. These techniques are based on spiking the protein mixture with known quantities of either labeled recombinant proteins or labeled synthetic peptides corresponding to the proteins of interest before mass spectrometric analysis. The amounts of the endogenous proteins are then estimated by comparing the ion current intensities for the standards and the endogenous peptides. Because the digestion efficiency of proteins may not be 100%, the use of recombinant proteins that are expected to be proteolyzed with the same efficiency as

240

A. Dosemeci

e­ ndogenous proteins would be expected to yield more accurate estimates compared to peptides. On the other hand, labeled synthetic peptides are convenient for the quantification of multiple proteins. The use of labeled recombinant artificial protein standards containing signature peptides of multiple proteins with their natural flanking sequences is a promising new technique to mimic proteolysis conditions of endogenous proteins endogenous proteins for accurate estimates of stoichiometry (Kito, Ota, Fujita, & Ito, 2007; Nanavati, Gucek, Milne, Subramaniam, & Markey, 2008). Peng et al. (2004) used labeled recombinant proteins for the quantification of six selected PSD components and, in 2006, the same group used labeled synthetic peptides for 32 selected components (Cheng et al., 2006). The general agreement between the two sets of results for the same proteins argues for an efficient PSD protein digestion under the conditions employed. The study by Cheng et al. (2006) reports molar quantities of glutamate receptors and scaffolding molecules as well as enzymes such as CaMKII and SynGAP per unit weight of PSD fraction. While the reported relative quantities of scaffolding molecules and receptors fall within expected ranges, the quantities of CaMKII appear to be far greater than those of other PSD constituents – altogether, the molar amounts of CaMKII a and b ­subunits is about three times that of the other 30 core PSD proteins combined. The high levels of CaMKII, as the authors point out, could be partially due to the presence of contaminating CaMKII clusters in the PSD preparation. While the presence of contaminants in PSD fractions affects the accuracy of absolute quantification, molar ratios of those proteins that are exclusively in the PSD and not in contaminating particles should not be affected. For estimates for ubiquitous proteins such as CaMKII that are present both in the PSD and in contaminating particles, the use of purer fractions is necessary. Using scanning transmission EM, the molecular mass of an average PSD (360 nm diameter) was estimated to be 1.1 GDa (Chen et al., 2005). With this ­number, the average number of copies per PSD of a particular protein can be ­calculated as long as the mass ratio of that protein in the PSD is known. Using estimates from highly purified preparations and correction factors together with quantitative immunoblotting, Chen et al. (2005) estimated ~300 copies of PSD-95 per PSD. This number is in agreement with the estimate obtained by Sugiyama et al. (2005) using quantitative fluorescence imaging. Taking the estimated number of copies of PSD-95 per PSD (Chen et al., 2005) and the molar ratios of proteins to that of PSD-95 (Cheng et al., 2006). Sheng & Hoogenraad (2007) calculated the number of copies per PSD of other core proteins as: ~150 for SAPAPs (GKAPs),  ~150 for Shanks, ~60 for Homer, ~20 for NMDA receptors channels, and ~15 for AMPA receptor channels. These numbers are, for the most part, consistent with estimates from studies using alternative approaches. The reader is referred to the excellent review by Sheng & Hoogenraad (2007) for further details on this issue. 2.7.2 Relative Quantification PSD composition is not identical in different brain regions and also changes during development. Several methodologies have been devised for the comparison of

Proteomic Analysis of the Postsynaptic Density

241

p­ rotein levels in two or more protein mixtures. One of these approaches is based on isotope incorporation by chemical derivatization. The original method uses ICAT reagents, labeled with either deuterium or hydrogen, to introduce an affinity tag on cysteine residues. Samples to be compared are tagged with different isotopes then mixed and analyzed by LC-MS/MS. The ratio of the peaks within pairs (hydrogenand deuterium-labeled, respectively) reflects the ratio of the corresponding proteins in control and treated samples. Using the ICAT methodology, Cheng et al. (2006) compared levels of proteins in PSD preparations from forebrain and cerebellum. This study identified 43 proteins that showed significant variation in abundance between the two brain regions.

3 Catching the Dynamics of the PSD PSD morphology changes with activity. Among the types of morphological changes described are increases in surface area, perforations and thickening (Dosemeci et al., 2001; Harris, Fiala, & Ostroff, 2003; Toni et al., 2001). These morphological changes are assumed to involve addition or subtraction of particular proteins to and from the PSD complex. One well-described example is CaMKII, which accumulates on the PSD under a variety of excitatory conditions, including glutamate application, long-term potentiation (LTP) and ischemia (Dosemeci et al., 2001; Hu et  al., 1998; Otmakhov et  al., 2004). Changes in PSD composition following ­specific patterns of activity, pharmacological treatments or pathological conditions are thought to underlie related changes in synaptic function.

3.1 Studies Monitoring Changes in the PSD Composition There are few studies tracking changes in PSD composition on a global basis ­following specific treatments. An early study (Hu et al., 1998) compared 1D electrophoretic profiles of PSD fractions from control animals and from animals ­subjected to transient cerebral ischemia. Bands that showed a change were identified by internal peptide microsequencing and/or western immunoblotting. Ten proteins including TrkB, NSF, heat shock cognate protein-70, PSD-95 and CaMKII were listed as showing either an increase or a decrease in PSD fractions from treated animals. A later study by Satoh et al. (2002) compared 2D electrophoretic profiles of PSD fractions from control and kainate-treated animals and identified some of the modified spots, including one containing SynGAP. Ehlers (2003) studied the effects of activity on PSD composition by treating ­cultured cortical neurons with tetradoxin (blocks action potential-dependent activity) or bicuculline (blocks inhibitory transmission). After the treatments, PSD fractions were prepared and analyzed by quantitative immunoblotting to track changes in ~30 proteins. The levels of a number of proteins, including CaMKII, PSD-95,

242

A. Dosemeci

Homer and NR2A, were found to be increased with activity while that of others including SAPAP (GKAP), Shank, NR1 and NR2B were found to be decreased. In recent years, the development of proteomic techniques for relative quantification by differential labeling of samples opened new opportunities for the global examination of changes in PSD composition. One of these is the ICAT labeling as described in Sect. 2.7.2. Using the ICAT technique and LC/MS/MS, a recent study by Moron et al. (2007) examined changes in the composition of PSD fraction as a result of morphine treatment. The study identified ten proteins whose levels were consistently changed by morphine treatment. Most notable was an increase of clathrin heavy chain in PSD fractions from morphine-treated animals.

3.2 Problem of Postmortem Modifications When Using Whole Animals The studies discussed above, with the exception of Ehlers (2003) who used cultured neurons, examine changes in PSD fractions prepared from brain tissue following pharmacological or surgical treatment of whole animals. An important issue to be considered in such studies has been first raised by Suzuki, Okumura-Noji, Tanaka, & Tada (1994) who observed an increase in CaMKII levels in PSD fractions within minutes when the brains were left on ice prior to homogenization. Postmortem translocation of CaMKII and other proteins to the PSD is likely to occur due to ischemia-like conditions that prevail during dissection of brain tissue. These ­considerations imply that isolated PSDs reflect, both in terms of protein profile and morphology, the response of neurons to a certain degree of ischemia due to the ­cessation of the blood supply after decapitation. While it is impossible to completely avoid the ischemia-like interval between decapitation and cell disruption, making it as short as possible and/or standardizing the dissection time, so that all samples are modified to the same extent, can enhance the signal to noise ratio. The unavoidable modification of PSDs during dissection of brain tissue introduces a confounding factor, which makes detection of subtle induced modifications difficult, if not impossible. Alternative experimental systems, such as brain slices, slice cultures and dispersed cell cultures that allow instantaneous cell disruption, can eliminate this problem.

3.3 Another Experimental Model: PSD fraction from Hippocampal Slices Hippocampal slices as well as slices from other brain regions provide excellent experimental systems where the physiological and morphological consequences of specific treatments can be studied in parallel. Several treatment protocols using brain slices have

Proteomic Analysis of the Postsynaptic Density

243

been established for eliciting specific functional changes such as LTP and ­long-term depression (LTD), and rapid fixation of slices for EM observation is possible. PSDs in slices are thick immediately after dissection, implying exposure to ischemia-like conditions. However, this change is reversible, as they get substantially thinner after 2 h of recovery in oxygenated medium (Tao-Cheng, Vinade, Pozzo-Miller, Reese, & Dosemeci, 2002). Following recovery, excitatory conditions can, once again, elicit thickening of the PSD (Dosemeci et  al., 2001). Morphological modification of the PSD in response to various conditions, including LTP-inducing stimuli, can be observed by EM examination of slices (Harris et al., 2003). While detection of morphological changes at the PSD in response to treatments that produce defined physiological states is possible, parallel proteomic analyses of the PSD had been problematic due to the small quantity of starting material. Protocols for the preparation of PSD fractions typically use brain tissue in gram quantities while a hippocampal slice weighs only a few milligrams. Because of the threefold magnitude difference in the amounts of starting material, simple downsizing­is not likely to work. We recently developed a protocol for the preparation of PSD fractions from ­hippocampal slices (Dosemeci, Tao-Cheng, Vinade, & Jaffe, 2006). The protocol involves homogenization of slices in isotonic sucrose, separation of nuclei by a low speed centrifugation step, and a second, higher speed centrifugation step to obtain a mitochondria- and synaptosome-enriched pellet. The method of Hajos (1975) is then applied for the fractionation of the pellet to obtain synaptosomes. This method replaces the high speed equilibrium density gradient centrifugation step used in conventional protocols, with a relatively low speed differential centrifugal separation on a one step (0.32/0.8 M) sucrose gradient. The synaptosomal fraction is then extracted twice with 1% TritonX-100, 75 mM KCl to obtain the micro-scale crude PSD fraction. The protocol can be applied to as few as five hippocampal slices. When 5–6 slices were used per fractionation unit (centrifuge tube) the yield was about 4 mg per slice. The electrophoretic protein staining pattern of the micro-scale PSD fraction (mPSD) is similar to that of the standard PSD fraction (sPSD) obtained by conventional protocols from gram quantities of material (Fig. 6a). The micro-scale PSD fraction shows a significant enrichment in the marker protein PSD-95, while a ­presynaptic protein synaptophysin is virtually eliminated following Triton extraction. On the other hand, western immunoblots also show some enrichment in GFAP, a common contaminant in standard PSD fractions (Fig. 6b). Thin section electron microscopy shows PSDs similar to those observed in situ, although contaminating particulate material, most notably membraneous vesicles and myelin, are also present­ (Fig. 6c). In general, EM observation reveals a fraction enriched in PSDs but more contaminated than the standard PSD fraction derived from gram quantities of tissue. The greater level of contamination is most probably due to the omission/ modification of certain fractionation steps of the conventional protocol for the sake of obtaining a reasonable yield.

244

A. Dosemeci

Fig. 6  Micro-scale PSD preparation from hippocampal slices. A shortened protocol was devised for the preparation of a PSD fraction from milligram quantities of wet tissue (Dosemeci et al., 2006). Protein staining patterns (a) and western immunoblots (b) corresponding to homogenate (H), synaptosome (Syn) and micro-scale PSD (mPSD) preparations. mPSD has a similar protein profile to the standard PSD (sPSD) preparation obtained using gram quantities of tissue. Western immunoblots show significant enrichment in the PSD marker PSD-95, but also some enrichment in GFAP, a contaminant, while the presynaptic protein synaptohysin is depleted. (c) Thin section EM of the micro-scale PSD fraction shows structures of similar morphology to in situ PSDs but also contaminants as indicated. From Dosemeci et al. (2006); printed with the permission from Elsevier

The composition of the PSD fraction from hippocampal slices was analyzed by two-dimensional liquid chromatography and tandem mass spectrometry (2D-LC/ MS/MS). This approach, which utilizes as little as 10 mg total protein, allowed the identification of more than 100 proteins. These include specialized PSD scaffolds such as PSD-95, PSD-93 and Shanks, the protein kinases, CaMKII and PKC, a G-protein regulator SynGAP and several cytoskeletal elements. In analogy to ­conventional PSD fractions, the majority of cytoskeletal elements detected are likely to be contaminants. Other contaminants found in the micro-scale PSD fraction included presynaptic, mitochondrial and glial proteins. When compared with the results of four previous proteomic studies of PSD fractions (Jordan et al., 2004; Li et al., 2004; Peng et al., 2004; Yoshimura et al., 2004), it turns out that most of the proteins detected in the micro-scale PSD fraction were also identified in standard PSD fractions (Dosemeci et al., 2006). In conclusion, the micro-scale protocol devised allows for the preparation of PSD fraction of reasonable purity from hippocampal slices. The yield is sufficient for global analysis by mass spectrometric methods. Although the fraction contains contaminants, this should not be a problem for monitoring changes in the levels of

Proteomic Analysis of the Postsynaptic Density

245

proteins known to be located exclusively on the PSD, using comparative proteomic strategies such as ICAT. For those proteins that are also present in other material contaminating the fraction, affinity purification is possible, although at this stage the yield from this strategy may be insufficient for proteomic analyses (Dosemeci et al., 2006). The method developed for the isolation of PSD fractions from acute slices is also applicable to slice cultures, and opens the way for the parallel identification of compositional, posttranscriptional, and structural changes in the PSD following defined physiological or pharmacological treatments.

4 Conclusions and Perspective In recent years, proteomics based on mass spectrometry resulted in the identification of more than 1,000 proteins in PSD fractions. A major task now is distinguishing bona fide constituents of the PSD from contaminants. Strategies for further purification of PSDs and assessment of enrichment during consecutive ­purification steps are important tools that have already led to the selection of certain proteins as potential major constituents and the elimination of others as contaminants. While proteomic approaches narrow down the list of candidate PSD elements, further validation by immuno – EM is necessary. A number of proteins, including several specialized scaffolds, glutamate receptor subunits and their accessories, as well as certain enzymes, have been confirmed by the combination of biochemical and immunoEM strategies to be constituents of the PSD. The status of many others, however, remains to be clarified. Another crucial task for the elucidation of the molecular architecture of the PSD is determination of the average number of copies per PSD of constituent proteins. Evaluation of these copy numbers requires the use of either very pure PSD preparations or precise correction factors for contaminants. Combining data from different laboratories resulted in the estimation of copy numbers for several scaffold proteins and glutamate receptors. Finally, comparative proteomic strategies are now in place for the assessment of changes in the composition of the PSD following activity and pharmacological treatments. A particular problem for the case of the PSD is its extreme susceptibility to ischemia-like conditions that prevail during tissue dissection, which tends to blunt experimental modifications. For this reason, experimental models such as brain slices, where immediate cell disruption is possible, are more suitable than live animals for the study of PSD dynamics. A new method for the preparation of PSD fractions from hippocampal slices provides material of sufficient purity and yield that would allow the application of proteomic strategies for the clarification of activity-induced PSD modification. Acknowledgments  I would like to thank Drs Stanford Markey and Thomas Reese for a critical reading of the manuscript. Supported by the Intramural Research Program of the NIH, NINDS.

246

A. Dosemeci

References Andersen, J. S., Wilkinson, C. J., Mayor, T., Mortensen, P., Nigg, E. A., & Mann, M. (2003). Proteomic characterization of the human centrosome by protein correlation profiling. Nature, 426, 570–574. Aoki, C., Miko, I., Oviedo, H., Mikeladze-Dvali, T., Alexandre, L., Sweeney, N., et al. (2001). Electron microscopic immunocytochemical detection of PSD-95, PSD-93, SAP-102, and SAP-97 at postsynaptic, presynaptic, and nonsynaptic sites of adult and neonatal rat visual cortex. Synapse, 40, 239–257. Baude, A., Nusser, Z., Molnar, E., McIlhinney, R. A., & Somogyi, P. (1995). High-resolution immunogold localization of AMPA type glutamate receptor subunits at synaptic and non-synaptic sites in rat hippocampus. Neuroscience, 69, 1031–1055. Carlin, R. K., Grab, D. J., Cohen, R. S., & Siekevitz, P. (1980). Isolation and characterization of postsynaptic densities from various brain regions: Enrichment of different types of postsynaptic densities. The Journal of Cell Biology, 86, 831–845. Chen, H. J., Rojas-Soto, M., Oguni, A., & Kennedy, M. B. (1998). A synaptic Ras-GTPase activating protein (p135 SynGAP) inhibited by CaM kinase II. Neuron, 20, 895–904. Chen, X., Vinade, L., Leapman, R. D., Petersen, J. D., Nakagawa, T., Phillips, T. M., et al. (2005). Mass of the postsynaptic density and enumeration of three key molecules. Proceedings of the National Academy of Sciences of the United States of America, 102, 11551–11556. Cheng, D., Hoogenraad, C. C., Rush, J., Ramm, E., Schlager, M. A., Duong, D. M., et al. (2006). Relative and absolute quantification of postsynaptic density proteome isolated from rat forebrain and cerebellum. Molecular & Cellular Proteomics, 5, 1158–1170. Cho, K. O., Hunt, C. A., & Kennedy, M. B. (1992). The rat brain postsynaptic density fraction ­contains a homolog of the Drosophila discs-large tumor suppressor protein. Neuron, 9, 929–942. Cohen, R. S., Blomberg, F., Berzins, K., & Siekevitz, P. (1977). The structure of postsynaptic densities isolated from dog cerebral cortex. I. Overall morphology and protein composition. The Journal of Cell Biology, 74, 181–203. Collins, M. O., Husi, H., Yu, L., Brandon, J. M., Anderson, C. N., Blackstock, W. P., et al. (2006). Molecular characterization and comparison of the components and multiprotein complexes in the postsynaptic proteome. Journal of Neurochemistry, 97(Suppl 1), 16–23. Cotman, C. W., Banker, G., Churchill, L., & Taylor, D. (1974). Isolation of postsynaptic densities from rat brain. The Journal of Cell Biology, 63, 441–455. Cottrell, J. R., Borok, E., Horvath, T. L., & Nedivi, E. (2004). CPG2: A brain- and synapse-specific protein that regulates the endocytosis of glutamate receptors. Neuron, 44, 677–690. Dosemeci, A., Makusky, A. J., Jankowska-Stephens, E., Yang, X., Slotta, D. J., & Markey, S. P. (2007). Composition of the synaptic PSD-95 complex. Molecular & Cellular Proteomics, 6, 1749–1760. Dosemeci, A., Reese, T. S., Petersen, J., & Tao-Cheng, J. H. (2000). A novel particulate form of CaMKII-dependent protein kinase II in neurons. The Journal of Neuroscience, 20, 3076–3084. Dosemeci, A., Tao-Cheng, J. H., Vinade, L., & Jaffe, H. (2006). Preparation of postsynaptic density fraction from hippocampal slices and proteomic analysis. Biochemical and Biophysical Research Communications, 339, 687–694. Dosemeci, A., Tao-Cheng, J. H., Vinade, L., Winters, C. A., Pozzo-Miller, L., & Reese, T. S. (2001). Glutamate-induced transient modification of the postsynaptic density. Proceedings of the National Academy of Sciences of the United States of America, 98, 10428–10432. Ehlers, M. D. (2003). Activity level controls postsynaptic composition and signaling via the ubiquitin-proteasome system. Nature Neuroscience, 6, 231–242. Foster, L. J., de Hoog, C. L., Zhang, Y., Xie, X., Mootha, V. K., & Mann, M. (2006). A mammalian organelle map by protein correlation profiling. Cell, 125, 187–199. Fukaya, M., Tsujita, M., Yamazaki, M., Kushiya, E., Abe, M., Akashi, K., et al. (2006). Abundant distribution of TARP gamma-8 in synaptic and extrasynaptic surface of hippocampal neurons and its major role in AMPA receptor expression on spines and dendrites. The European Journal of Neuroscience, 24, 2177–2190.

Proteomic Analysis of the Postsynaptic Density

247

Hajos, F. (1975). An improved method for the preparation of synaptosomal fractions in high purity. Brain Research, 93, 485–489. Harris, K. M., Fiala, J. C., & Ostroff, L. (2003). Structural changes at dendritic spine synapses during long-term potentiation. Philosophical Transactions of The Royal Society of London. Series, B, 358, 745–748. Hu, B. R., Park, M., Martone, M. E., Fischer, W. H., Ellisman, M. H., & Zivin, J. A. (1998). Assembly of proteins to postsynaptic densities after transient cerebral ischemia. The Journal of Neuroscience, 18, 625–633. Jordan, B. A., Fernholz, B. D., Boussac, M., Xu, C., Grigorean, G., Ziff, E. B., et  al. (2004). Identification and verification of novel rodent postsynaptic density proteins. Molecular & Cellular Proteomics, 3, 857–871. Jordan, B. A., Fernholz, B. D., Khatri, L., & Ziff, E. B. (2007). Activity-dependent AIDA-1 nuclear signaling regulates nucleolar numbers and protein synthesis in neurons. Nature Neuroscience, 10, 427–435. Kennedy, M. B. (1997). The postsynaptic density at glutamatergic synapses. Trends in Neurosciences, 20, 264–268. Kharazia, V. N., & Weinberg, R. J. (1997). Tangential synaptic distribution of NMDA and AMPA receptors in rat neocortex. Neuroscience Letters, 238, 41–44. Kim, E., Naisbitt, S., Hsueh, Y. P., Rao, A., Rothschild, A., Craig, A. M., et al. (1997). GKAP, a novel synaptic protein that interacts with the guanylate kinase-like domain of the PSD-95/ SAP90 family of channel clustering molecules. The Journal of Cell Biology, 136, 669–678. Kito, K., Ota, K., Fujita, T., & Ito, T. (2007). A synthetic protein approach toward accurate mass spectrometric quantification of component stoichiometry of multiprotein complexes. Journal of Proteome Research, 6, 792–800. Li, K. W., Hornshaw, M. P., Van Der Schors, R. C., Watson, R., Tate, S., Casetta, B., et al. (2004). Proteomics analysis of rat brain postsynaptic density. Implications of the diverse protein functional groups for the integration of synaptic physiology. The Journal of Biological Chemistry, 279, 987–1002. Li, K., Hornshaw, M. P., van Minnen, J., Smalla, K. H., Gundelfinger, E. D., & Smit, A. B. (2005). Organelle proteomics of rat synaptic proteins: Correlation-profiling by isotope-coded affinity tagging in conjunction with liquid chromatography-tandem mass spectrometry to reveal postsynaptic density specific proteins. Journal of Proteome Research, 4, 725–733. Matus, A. I., & Taff-Jones, D. H. (1978). Morphology and molecular composition of isolated postsynaptic junctional structures. Proceedings of the Royal Society of London. Series B, 203, 135–151. Moron, J. A., Abul-Husn, N. S., Rozenfeld, R., Dolios, G., Wang, R., & Devi, L. A. (2007). Morphine administration alters the profile of hippocampal postsynaptic density-associated proteins: A proteomics study focusing on endocytic proteins. Molecular & Cellular Proteomics, 6, 29–42. Mundel, P., Heid, H. W., Mundel, T. M., Kruger, M., Reiser, J., & Kriz, W. (1997). Synaptopodin: An actin-associated protein in telencephalic dendrites and renal podocytes. The Journal of Cell Biology, 139, 193–204. Murphy, J. A., Jensen, O. N., & Walikonis, R. S. (2006). BRAG1, a Sec7 domain-containing protein, is a component of the postsynaptic density of excitatory synapses. Brain Research, 1120, 35–45. Naisbitt, S., Kim, E., Tu, J. C., Xiao, B., Sala, C., Valtschanoff, J., et al. (1999). Shank, a novel family of postsynaptic density proteins that binds to the NMDA receptor/PSD-95/GKAP complex and cortactin. Neuron, 23, 569–582. Naisbitt, S., Kim, E., Weinberg, R. J., Rao, A., Yang, F. C., Craig, A. M., et  al. (1997). Characterization of guanylate kinase-associated protein, a postsynaptic density protein at excitatory synapses that interacts directly with postsynaptic density-95/synapse-associated protein 90. The Journal of Neuroscience, 17, 5687–5696. Nanavati, D., Gucek, M., Milne, J. L., Subramaniam, S., & Markey, S. P. (2008). Stoichiometry and absolute quantification of proteins with mass spectrometry using fluorescent and isotope labeled concatenated peptide standards. Molecular Cell Proteomics, 7, 442–447.

248

A. Dosemeci

Nesvizhskii, A. I., Vitek, O., & Aebersold, R. (2007). Analysis and validation of proteomic data generated by tandem mass spectrometry. Nature Methods, 4, 787–797. Otmakhov, N., Tao-Cheng, J. H., Carpenter, S., Asrican, B., Dosemeci, A., Reese, T. S., et  al. (2004). Persistent accumulation of calcium/calmodulin-dependent protein kinase II in dendritic spines after induction of NMDA receptor-dependent chemical long-term potentiation. The Journal of Neuroscience, 24, 9324–9331. Peng, J., Kim, M. J., Cheng, D., Duong, D. M., Gygi, S. P., & Sheng, M. (2004). Semiquantitative proteomic analysis of rat forebrain postsynaptic density fractions by mass spectrometry. The Journal of Biological Chemistry, 279, 21003–21011. Petersen, J. D., Chen, X., Vinade, L., Dosemeci, A., Lisman, J. E., & Reese, T. S. (2003). Distribution of postsynaptic density (PSD)-95 and Ca2+/calmodulin-dependent protein kinase II at the PSD. The Journal of Neuroscience, 23, 11270–11278. Petralia, R. S., Sans, N., Wang, Y. X., & Wenthold, R. J. (2005). Ontogeny of postsynaptic density proteins at glutamatergic synapses. Molecular and Cellular Neurosciences, 29, 436–452. Petralia, R. S., Wang, Y. X., & Wenthold, R. J. (1994). The NMDA receptor subunits NR2A and NR2B show histological and ultrastructural localization patterns similar to those of NR1. Journal of Neuroscience, 14, 6102–6120. PubMed ID: 7931566. Rubio, M. E., Curcio, C., Chauvet, N., & Bruses, J. L. (2005). Assembly of the N-cadherin complex during synapse formation involves uncoupling of p120-catenin and association with presenilin 1. Molecular and Cellular Neurosciences, 30, 611–623. Sakagami, H., Sanda, M., Fukaya, M., Miyazaki, T., Sukegawa, J., Yanagisawa, T., et al. (2008). IQ-ArfGEF/BRAG1 is a guanine nucleotide exchange factor for Arf6 that interacts with PSD-95 at postsynaptic density of excitatory synapses. Neuroscience Research, 60, 199–212. Satoh, K., Takeuchi, M., Oda, Y., Deguchi-Tawarada, M., Sakamoto, Y., Matsubara, K., et  al. (2002). Identification of activity-regulated proteins in the postsynaptic density fraction. Genes to Cells, 7, 187–197. Sheng, M., & Hoogenraad, C. C. (2007). The postsynaptic architecture of excitatory synapses: A more quantitative view. Annual Review of Biochemistry, 76, 823–847. Song, J. Y., Ichtchenko, K., Sudhof, T. C., & Brose, N. (1999). Neuroligin 1 is a postsynaptic cell-adhesion molecule of excitatory synapses. Proceedings of the National Academy of Sciences of the United States of America, 96, 1100–1105. Sun, J., Tadokoro, S., Imanaka, T., Murakami, S. D., Nakamura, M., Kashiwada, K., et al. (1998). Isolation of PSD-Zip45, a novel Homer/vesl family protein containing leucine zipper motifs, from rat brain. FEBS Letters, 437, 304–308. Suzuki, T., Ito, J., Takagi, H., Saitoh, F., Nawa, H., & Shimizu, H. (2001). Biochemical evidence for localization of AMPA-type glutamate receptor subunits in the dendritic raft. Brain Research. Molecular Brain Research, 89, 20–28. Suzuki, T., Okumura-Noji, K., Tanaka, R., & Tada, T. (1994). Rapid translocation of cytosolic Ca2+/calmodulin-dependent protein kinase II into postsynaptic density after decapitation. Journal of Neurochemistry, 63, 1529–1537. Tao-Cheng, J. H., Dosemeci, A., Winters, C. A., & Reese, T. S. (2006). Changes in the distribution of calcium calmodulin-dependent protein kinase II at the presynaptic bouton after depolarization. Brain Cell Biology, 35, 117–124. Tao-Cheng, J. H., Vinade, L., Pozzo-Miller, L. D., Reese, T. S., & Dosemeci, A. (2002). Calcium/ calmodulin-dependent protein kinase II clusters in adult rat hippocampal slices. Neuroscience, 115, 435–440. Tomita, S., Chen, L., Kawasaki, Y., Petralia, R. S., Wenthold, R. J., Nicoll, R. A., et  al. (2003). Functional studies and distribution define a family of transmembrane AMPA receptor regulatory proteins. The Journal of Cell Biology, 161, 805–816. Toni, N., Buchs, P. A., Nikonenko, I., Povilaitite, P., Parisi, L., & Muller, D. (2001). Remodeling of synaptic membranes after induction of long-term potentiation. The Journal of Neuroscience, 21, 6245–6251. Trinidad, J. C., Specht, C. G., Thalhammer, A., Schoepfer, R., & Burlingame, A. L. (2006). Comprehensive identification of phosphorylation sites in postsynaptic density preparations. Molecular & Cellular Proteomics, 5, 914–922.

Proteomic Analysis of the Postsynaptic Density

249

Tu, J. C., Xiao, B., Naisbitt, S., Yuan, J. P., Petralia, R. S., Brakeman, P., et al. (1999). Coupling of mGluR/Homer and PSD-95 complexes by the Shank family of postsynaptic density proteins. Neuron, 23, 583–592. Valtschanoff, J. G., & Weinberg, R. J. (2001). Laminar organization of the NMDA receptor complex within the postsynaptic density. The Journal of Neuroscience, 21, 1211–1217. Vinade, L., Chang, M., Schlief, M. L., Petersen, J. D., Reese, T. S., Tao-Cheng, J. H., et al. (2003). Affinity purification of PSD-95-containing postsynaptic complexes. Journal of Neurochemistry, 87, 1255–1261. Walikonis, R. S., Jensen, O. N., Mann, M., Provance, D. W., Jr., Mercer, J. A., & Kennedy, M. B. (2000). Identification of proteins in the postsynaptic density fraction by mass spectrometry. Journal of Neuroscience, 20, 4069–4080. Walsh, M. J., & Kuruc, N. (1992). The postsynaptic density: Constituent and associated proteins characterized by electrophoresis, immunoblotting, and peptide sequencing. Journal of Neurochemistry, 59, 667–678. Wyszynski, M., Kharazia, V., Shanghvi, R., Rao, A., Beggs, A. H., Craig, A. M., et  al. (1998). Differential regional expression and ultrastructural localization of alpha-actinin-2, a putative NMDA receptor-anchoring protein, in rat brain. The Journal of Neuroscience, 18, 1383–1392. Yoshimura, Y., Yamauchi, Y., Shinkawa, T., Taoka, M., Donai, H., Takahashi, N., et al. (2004). Molecular constituents of the postsynaptic density fraction revealed by proteomic analysis using multidimensional liquid chromatography-tandem mass spectrometry. Journal of Neurochemistry, 88, 759–768.

Part III

Behavior

Functional Genomic Dissection of Speech and Language Disorders Sonja C. Vernes and Simon E. Fisher

Abstract  Mutations of the human FOXP2 gene have been shown to cause severe difficulties in learning to make coordinated sequences of articulatory gestures that underlie speech (developmental verbal dyspraxia or DVD). Affected individuals are impaired in multiple aspects of expressive and receptive linguistic processing and ­display abnormal grey matter volume and functional activation patterns in cortical and subcortical brain regions. The protein encoded by FOXP2 belongs to a divergent subgroup of forkhead-box transcription factors, with a distinctive DNA-binding domain and motifs that mediate hetero- and homodimerization. This chapter describes the successful use of FOXP2 as a unique molecular window into neurogenetic pathways that are important for speech and language development, adopting several complementary strategies. These include direct functional investigations of FOXP2 splice variants and the effects of etiological mutations. FOXP2’s role as a transcription factor also enabled the development of functional genomic routes for dissecting neurogenetic mechanisms that may be relevant for speech and language. By identifying downstream target genes regulated by FOXP2, it was possible to identify common regulatory themes in modulating synaptic plasticity, neurodevelopment, and axon guidance. These targets represent novel entrypoints into in  vivo pathways that may be disturbed in speech and language disorders. The identification of FOXP2 target genes has also led to the discovery of a shared neurogenetic pathway between clinically distinct language disorders; the rare Mendelian form of DVD and a complex and more common form of language ­disorder known as Specific Language Impairment. Keywords  FOXP2 • Speech and language • Chromatin immunoprecipitation • Developmental verbal dyspraxia • Specific language impairment • CNTNAP2 • Transcription factor S.C. Vernes (*) Wellcome Trust Centre for Human Genetics, University of Oxford, Roosevelt Drive, Oxford OX3 7BN, UK e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_10, © Springer Science+Business Media, LLC 2011

253

254

S.C. Vernes and S.E. Fisher

Abbreviations ASD CFDE ChIP DVD EMSA FOX NLS SLI

Autism spectrum disorder Cortical dysplasia–focal epilepsy Chromatin-immunoprecipitation Developmental verbal dyspraxia Electrophoretic mobility shift assay Forkhead-box Nuclear localisation signal Specific language impairment

1 Human Speech and Language Language typically develops in all children given adequate environmental input but without the need for formal instruction, and involves the use of thousands of words assembled into a potentially infinite number of meaningful sentences by following a set of grammatical rules (Doupe & Kuhl, 1999). Neural networks subserving language receive auditory, visual or tactile input followed by the activation of memory systems (phonological and semantic) and the integration of sensory/motor output systems. Speech is a motor output of language that involves the rapid sequencing of articulatory gestures to produce structured vocalisations. This complex process requires the integration of orofacial, laryngeal and respiratory motor control with sensorimotor and auditory feedback (Doupe & Kuhl, 1999; Price, 2000). Many of the neurological components necessary for speech and language are thought to reside in cortical areas of the brain, including Broca’s area, Wernicke’s area, the supramarginal gyrus and angular gyrus, the motor cortex, supplementary motor area and anterior cingulate (Ojemann, 1991; Price, 2000; Aboitiz & Garcia, 1997). Subcortical areas are also implicated in vocal learning and speech production, such as the basal ganglia, thalamus and cerebellum (Crosson, 1985). Thus, the current model of language implicates interaction between cortical and subcortical brain regions with distributed networks of circuits subserving different aspects of speech and language.

2 Disorders of Speech and Language A range of neurodevelopmental disorders affect language, such as autism spectrum disorders (ASD; MIM:209850) and specific language impairment (SLI; MIM:606711). ASD is characterised by atypical social behaviour, impaired communication (verbal and non-verbal) and repetitive and stereotyped behaviours (Geschwind & Levitt, 2007). General cognitive impairments are seen in many but not all children with ASD, and about 50% display significant delay in onset of

Functional Genomic Dissection of Speech and Language Disorders

255

language (age at first word) (Alarcon et  al., 2002; Geschwind & Levitt, 2007). Approximately 7% of school-age children display SLI, a failure to develop normal speech and language skills despite adequate environmental influences and in the absence of other medical/genetic causes, such as hearing loss, neurological disorders (e.g. mental retardation, cerebral palsy, autism, childhood schizophrenia) or motor impairments (Tomblin et al., 1997). Affected individuals commonly display limited vocabularies, use only simple sentence construction and struggle to learn to read and write (Bartlett et al., 2004). There is substantial evidence that developmental disorders of speech and language are highly heritable, including the clustering of these disorders in families (Neils & Aram, 1986; Lewis et al., 1989; Tallal et al., 1989; Tomblin, 1989; Lahey & Edwards, 1995), and the higher rates of concordance for speech and language deficits in monozygotic (near 100% concordance) vs. dizygotic twins (~50–70%) (Lewis & Thompson, 1992; Bishop et  al., 1995; Tomblin & Buckwalter, 1998). However, efforts to uncover the genetic risk factors associated with language disorders via classical mapping or association studies are hampered by the complex nature of the phenotype and the potentially large number of contributing (and potentially interacting) genes.

2.1 Developmental Verbal Dyspraxia and the KE Family In 1990, a pedigree was reported that was to prove pivotal to our understanding of the molecular basis of human speech and language development (Hurst et  al., 1990). The KE family is a large three-generation pedigree (of ~30 members) in which approximately half the family members display a complex disorder of speech and language primarily characterised by developmental verbal dyspraxia (DVD; MIM: 602081) (Fig.  1). The core feature of DVD is a severe deficit in ­controlling complex sequences of orofacial movements, impairing speech (VarghaKhadem et  al., 1995; Watkins, Dronkers, et  al., 2002; This was accompanied by impaired linguistic processing in both expressive and receptive domains (VarghaKhadem et  al., 1998). Previously described pedigrees of developmental language disorders had shown complex ­patterns of inheritance suggestive of multifactorial causes. By contrast, the pattern of inheritance for DVD in the KE family was compatible with an autosomal dominant mode of transmission, suggestive of a single gene disorder (Hurst et al., 1990). Tests of orofacial praxis revealed a significant impairment in affected KE family members, particularly in tests involving combinations or sequences of movements (Vargha-Khadem et al., 1995, 1998; Watkins, Dronkers, et al., 2002; It was tests of word repetition, non-word repetition and sequential orofacial movements that could most reliably distinguish affected from unaffected family members (Vargha-Khadem et  al., 1995, 1998; Watkins, Dronkers, et al., 2002; Watkins, Vargha-Khadem, et al., 2002). Intensive phenotypic evaluation of the affected and unaffected family members also demonstrated that the disorder affected language processing in multiple domains. Affected family members were impaired on verbal

256

S.C. Vernes and S.E. Fisher 1

I

2

1

3

4

2

5

6

7

8

9

10 11

II

10

11

12

13

14

15

16

17

18

19

20

23

24

III 1

2

3

4

5

6

7

8

9

21

22

Fig. 1  Pedigree of the KE family. The pattern of inheritance seen for verbal dyspraxia in this three generation (I–III) pedigree suggested that the disorder could be attributed to a single autosomal gene. Individuals displaying developmental verbal dyspraxia are represented by filled symbols. Squares represent males and circles represent females. A line through a symbol indicates a deceased individual. Individuals unavailable for genetic testing are indicated by an asterisk. Reproduced from (Lai et al., 2001)

and non-verbal language-based tests including tests of receptive and expressive grammar, word and non-word repetition, lexical decision, phoneme addition and deletion, non-word reading/spelling, and rhyme production (Gopnik & Crago, 1991; Vargha-Khadem et al., 1995; Watkins, Dronkers, et al., 2002; Neuroimaging studies in the KE family showed no overt structural abnormalities, but subtle bilateral defects in regions of the brain involved in speech and language processing (Vargha-Khadem et al., 1998; Watkins, Dronkers, et al., 2002; Belton et al., 2003; Liegeois et al., 2003). These included altered grey matter density and patterns of over- and underactivation during langage tasks in cortical and subcortical structures (Liegeois et al., 2003).

2.2 Identifying a Gene Underlying Verbal Dyspraxia The monogenic form of transmission observed in the KE family made it possible to undertake classical mapping studies in order to pinpoint the gene underlying their disorder. These studies found that the disorder segregated with an interval on chromosome 7q31, designated SPCH1 (Fisher et  al., 1998). An unrelated verbal dyspraxia case (named CS) was also identified, who carried a balanced translocation that interrupted the coding region of a novel gene within the SPCH1 region; the FOXP2 gene (Lai et al., 2000, 2001). Screening of the FOXP2 coding region in the

Functional Genomic Dissection of Speech and Language Disorders

257

KE family identified a heterozygous point mutation within exon 14 of all affected family members resulting in an arginine-histidine (R553H) amino acid substitution (Fig.  2) (Lai et  al., 2001). This mutation was not found in any of the unaffected family members or in large numbers of controls (Lai et  al., 2001). The protein encoded by FOXP2 was identified as belonging to the FOX superfamily of transcription factors, and the R553H residue, mutated in the KE family, is located within the highly conserved DNA binding domain. This R553 residue is invariant across all known FOX family members (Lai et al., 2001). Moreover, mutations of the paralogous position in other FOX genes (e.g. FOXC1.R127H) are known to severely affect the properties of the transcription factor and have been implicated in disease states (Lehmann et al., 2003; Saleem et al., 2003). Thus, it was concluded that the heterozygous R553H mutation in FOXP2 was likely to be a functional change responsible for the autosomal dominant monogenic transmission of verbal dyspraxia in the KE family (Lai et al., 2001).

3 FOX Transcription Factors In order to understand why mutations of FOXP2 cause disorders of speech and language, it is necessary to understand the properties and function of this gene. FOX transcription factors are named after their characteristic feature; the forkhead-box or FOX DNA binding domain. Accepted nomenclature for this gene family uses uppercase for human genes (FOXA), lowercase for mouse genes (Foxa) and mixed upper/lower case for all other species (FoxA) (Kaestner et  al., 2000). Subfamilies are named alphabetically, in order of their discovery, and to date 17 subfamilies have been

Fig. 2  Summary of base changes identified in verbal dyspraxia patients. Schematic diagram of the FOXP2 genomic locus. The gene spans >600 kb and 26 exons (including alternatively spliced exons). Black shading represents coding exons, atg indicates position of start codon in isoform I and tga the position of stop codon. Positions of coding changes (above locus) and non-coding variants (below locus) identified in a screen of 49 DVD probands are given, with frequency and amino acid changes where relevant. Asterisk indicate changes that correspond to those identified in Newbury et al. (2002). Reproduced from (MacDermot et al., 2005)

258

S.C. Vernes and S.E. Fisher

categorized (FOXA–FOXQ). Forkhead transcription factors regulate eukaryotic gene expression and are involved in a range of diverse processes including morphogenesis, cell specification and tumourigenesis. Mutation of either Foxg1 or Foxd1 in the mouse results in pathfinding errors of retinal neuron projections on the tectum (required for the relay of visual information to the brain) (Yuasa et al., 1996). In humans, many FOX genes have been implicated in developmental disorders; for example, they are associated with a range of ocular disorders including AxenfeldRieger syndrome (MIM: 602482) or glaucoma (MIM:602402) (FOXC1), Lymphoedema-Distichiasis (MIM:153400) (FOXC2), cataracts (MIM: 601094) (FOXE3) and Blepharophimosis-Ptosis-Epicanthus Inversus (BPES; MIM: 110100) (FOXL2) (Lehmann et al., 2003). The FOXP subfamily of transcription factors includes FOXP1, FOXP2, FOXP3 and FOXP4 (and their orthologues in other mammalian species) (Bennett et  al., 2001; Shu et al., 2001; Teufel et al., 2003; Li et al., 2004). FOXP is one of the most recently characterised and divergent of the subgroups that fall within the FOX umbrella (Shu et al., 2001; Katoh & Katoh, 2004). Although very different to other FOX superfamily proteins, the four FOXP proteins are highly similar to one another; FOXP1/2/4 display ~92% similarity across the three FOX domains (Lu et al., 2002). FOXP proteins feature a number of unusual characteristics not found in other FOX superfamily members, including truncation of the C-terminal region of the DNA binding domain, a glutamine-rich region, zinc-finger/leucinezipper domains and a C-terminal acidic tail region. Most FOX proteins have DNA binding domains of ~110 amino acids, while FOXP domains consist of only ~84 amino acids. As a result, they lack wing domains in this region which are though to stabilise DNA binding, and this may underlie the atypical DNA binding properties of FOXP proteins (Li et al., 2004). Unlike other FOX proteins, FOXP subfamily members must homo- or heterodimerise to bind target DNA and cannot bind efficiently as monomers (Li et al., 2004). It has been suggested that FOXP dimerisation, mediated via the highly conserved leucine-zipper/zinc-finger domain region, is a necessary prerequisite for DNA binding and transcriptional activity of the fulllength transcription factors (Shu et al., 2001; Li et al., 2004). Lastly, FOXP proteins are characterised by an N-terminal glutamine rich region, and some subfamily members contain polyglutamine tracts which may be important for protein–protein interactions. FOXP2 contains two such polyglutamine tracts (of 10 and 40 glutamines, respectively) within the glutamine rich region. FOXP2 is expressed in a range of tissues including the lung, spleen, small intestine, skeletal muscle, kidney and brain (Shu et al., 2001). In the brain, the gene follows a highly restricted pattern of expression in distinct structures throughout development, including the deep ­layers of the cortex, the striatum, the thalamus, the hypothalamus and Purkinje cells of the cerebellum (Ferland et al., 2003; Lai et al., 2003). FOXP2 displays striking conservation between species and is in the top 5% of the most highly conserved proteins between human and mouse (Enard et al., 2002). Ignoring slight differences in poly-glutamine tract length, there are only three amino acid substitutions between mouse and human orthologs (Enard et al., 2002), and human and mouse embryonic brains have shown highly overlapping patterns of expression; as yet no

Functional Genomic Dissection of Speech and Language Disorders

259

region of FOXP2 expression has been found exclusively in human samples (Lai et al., 2003).

4 Functional Properties of FOXP2 FOXP2 is located on human chromosome 7q31 and spans a genomic locus of at least 600 kb. Northern blots suggest expression of an approximately 6.5 kb transcript in human tissue and to date 26 exons have been identified; exons 1–7, plus exons 1b, 2a, 2b and 3a (alternatively spliced, predicted to be non-coding), exons 3b, 4a (coding exons, alternatively spliced) and exons s1-s3 (5’-UTR) (see Fig. 2) (Bruce & Margolis, 2002; MacDermot et al., 2005; Schroeder & Myers, 2008). Although the exon/intron structure of mouse and human FOXP2/Foxp2 appear to be conserved, transcripts of 9.0, 3.5 and 2.0 kb have been detected in mouse ­t issue, with the 9.0  kb transcript being predominant in the mouse brain (Shu et al., 2001).

4.1 Alternative Splicing of FOXP2 A number of alternative transcripts and protein isoforms have been predicted based on the genomic sequence of FOXP2 (Fig. 3). The major protein isoform (isoform I), is encoded by a transcript including exons 1–17, and is predicted to produce a 715 amino acid protein product (Lai et al., 2001). In this isoform, a start codon in exon 2 and a stop codon in exon 17 are thought to be utilised (Fig. 3). Across all known protein isoforms, the main functional domains remain the same (excluding FOXP2.10+, see below); the poly-glutamine tracts spanning exons 5 and 6, the zinc-finger/leucine-zipper region spanning exons 8–10 and the FOX domain encompassing exons 12–14 (Figs.  2 and 3) (Lai et  al., 2001; MacDermot et  al., 2005). Isoform II is the same as isoform I except that insertion of the alternatively spliced exon 3b between exons 3 and 4 into the transcript yields a larger predicted product of 740 amino acids (Lai et al., 2001). Isoform III differs from the others as it is translated from an alternative start codon in exon 4 (resulting from insertion of exon 3a or 3a–3b into the transcript), giving a 623-amino acid product, truncating the N-terminal of the predicted protein with a shorter glutamine rich region (Lai et al., 2001). A C-terminally truncated isoform of FOXP2 has been identified that is unusual because it lacks the FOX DNA-binding domain (Bruce & Margolis, 2002). In this truncated splice variant (known as FOXP2.10+ or FOXP2-S), exon 10 is replaced by exon 10+ (an elongated version of exon 10), which introduces ten extra amino acids and an early stop codon (Bruce & Margolis, 2002). This transcript carries a polyadenylation signal (78 bases downstream of exon 10+) and a poly-A-tail (a further 27 bases downstream) (Bruce & Margolis, 2002). This alternative splicing event is

260

S.C. Vernes and S.E. Fisher

a Transcript:

1

2 3 4

5

6

7

8 9 10 11 1213141516

5

6

7

8 9 10 11 1213141516

17

I atg 1

tga

2 33b 4

17

II tga

atg 1

2 33a 4

5

6

7

8 9 10 11 1213141516

5

6

7

8 9 10 11 1213141516

2,145 bp 715 amino acids (Protein Isoform I)

2,220 bp 740 amino acids (Protein Isoform Il)

17

III tga

atg 1

2 3 3a3b4

17

1,869 bp 623 amino acids (Protein Isoform IIl)

IV atg

tga

b

Fig. 3  Predicted isoforms of FOXP2. (a) Four alternative transcripts of FOXP2 (I–IV) were initially predicted from the genomic sequence. Alternative splicing of exons 3a and 3b (indicated by asterisk) results in products of altered length. Reproduced from Lai et al., 2001. (b) Schematic representation of the full length FOXP2 isoform I containing the Glutamine rich region (Q-rich), Zinc Finger (ZnF), Leucine Zipper (LeuZ) and forkhead-box (FOX) domains and the C-terminal acidic tail region (acidic). The predicted protein product of the coding changes identified in verbal dyspraxia patients (R553H, Q17L, R328X) and two forms of FOXP2 resulting from alternative splicing (FOXP2.10+, isoform III) are also given. Reproduced from Vernes et al., 2006

predicted to produce a 432-amino acid protein (Fig.  3) that contains the poly-­ glutamine regions, zinc finger and leucine zipper domains, but that lacks the FOX domain and C-terminal acidic tail (Bruce & Margolis, 2002). Of note, a similar isoform of Foxp1, lacking the FOX domain, has been identified in the mouse (Shu et al., 2001), suggesting that alternative splicing of this kind may be a conserved feature of the FOXP subfamily.

Functional Genomic Dissection of Speech and Language Disorders

261

Given that differential splicing of a gene can substantially alter protein function, affecting levels of expression, subcellular localisation, and/or DNA binding site recognition/affinity (Nakahata & Kawamoto, 2005), the properties of three isoforms of human FOXP2 were investigated: isoform I (major isoform), FOXP2.10+ (C-terminally truncated isoform) and isoform III (N-terminally truncated isoform) via functional analysis (Vernes et  al., 2006). Crucially, while lacking either N- or C-terminal ends found in isoform I, both the alternative isoforms include the known dimerisation domains of FOXP2, suggesting that they might be able to interact with isoform I and/or each other in regions of the brain where they are co-expressed. FOXP2 Isoform I was found to be predominantly localized to the nucleus (Fig. 4), a feature of the protein that is crucial to its activity as a transcription factor (Vernes et al., 2006). Putative nuclear localisation signals (NLSs) were identified within the FOX domain. These were located at the N-terminus and C-terminus of the FOX domain and both were shown to contribute to the nuclear targeting of the protein (Vernes et al., 2006; Mizutani et al., 2007). Furthermore, this protein could efficiently bind to a consensus sequence previously identified for the FOXP family and repress transcription from a promoter containing consensus binding sites

Fig.  4  Intracellular localisation of mutant FOXP2 proteins. Western blot analysis of recombinant proteins transiently expressed in HEK293T cells. Cells were fractionated into nuclear and cytoplasmic compartments prior to SDS PAGE and western blotting. Proteins were detected using an N-terminal FOXP2 Antibody (Santa Cruz Biotechnology) and equivalent loading was confirmed using the b-Tubulin internal loading control. The R328X protein could only be detected following extended exposure (overexposed panel). Reproduced from Vernes et al., 2006

262

S.C. Vernes and S.E. Fisher

(Vernes et al., 2006). Although FOXP2 isoform III showed no deviation from the properties of full length FOXP2, the FOXP.10+ protein displayed increased ­cytoplasmic location, most likely due to the lack of a FOX domain (Fig.  4). Interestingly, FOXP2.10+ was sequestered in cytoplasmic protein aggregates, comparable to those observed in a subsequent study in which a construct producing a protein similar to FOXP2.10+ was utilised. Given that FOXP2.10+ protein clusters were also found to co-localise with classical aggresome markers (such as ubiquitin and g-tubulin), these bodies are thought to represent aggresomes (Vernes et  al., 2006). Although the functions of aggresomes are not fully understood, recent evidence suggests that the sequestration of proteins into these intracellular bodies could play a regulatory role by controlling the quantity of active protein available in the nucleus (Kopito, 2000; Taylor et al., 2003, 2004; Webb et al., 2004). These data hint that the FOXP2.10+ isoform might play a post-translational role in ­regulating the availability/functionality of longer FOXP2 isoforms, via interactions involving the dimerisation domains. Consistent with this, in vitro luciferase assays suggest that transfected FOXP2.10+ may be able to influence expression from a reporter construct despite lacking a DNA-binding domain, perhaps by modulating the behaviour of endogenous FOXP2 present in these cells (Vernes et al., 2006).

4.2 Functional Consequences of FOXP2 Mutations Studies have investigated FOXP2 as a genetic risk factor in language-related neurodevelopmental disorders. Several cases of verbal dyspraxia patients with gross chromosomal abnormalities affecting the FOXP2 locus have been identified, including deletions and translocations (Feuk et  al., 2006; Lennon et  al., 2007). MacDermot and colleagues carried out mutation screening of all known exons of FOXP2 in 49 probands with verbal dyspraxia, and identified two novel heterozygous coding changes; Q17L and R328X (Fig.  2) (MacDermot et  al., 2005). The functional significance of the heterozygous substitution in exon 2 (Q17L) was unclear as it was found in an affected proband but not an affected sibling (MacDermot et  al., 2005). Nonetheless, screening of a large number of control chromosomes did not reveal the presence of this change in unaffected individuals. The second variant was predicted to introduce an early stop codon into exon 7 (R328X), such that the product would be a highly truncated form of the protein – lacking most of the defined functional domains (Fig. 2) (MacDermot et al., 2005). This mutation was found in a proband, and his affected sibling, and not observed in any control chromosomes (MacDermot et  al., 2005). The proband’s mother, who suffered from expressive and receptive language problems and had demonstrated speech delay in childhood, also carried the R328X mutation, whereas the phenotypically normal father did not (MacDermot et al., 2005). Thus, for the first time since the KE family, an independent pedigree was established in which a FOXP2 point mutation was segregating with a verbal dyspraxia phenotype.

Functional Genomic Dissection of Speech and Language Disorders

263

Heterozygous mutations that alter the dosage of a FOX transcription factor can have severe effects on protein function. Studies on the FOXPA orthologue (pha-4) in C.elegans showed that pha-4 expression is temporally regulated such that expression increases as development progresses (Gaudet & Mango, 2002). Early in development, when pha-4 expression is low, only high affinity pha-4 promoters are activated. Promoters for which pha-4 has a low affinity are not expressed until later in development when pha-4 levels are high. This example illustratates the importance of precise spatio-temporal regulation of FOX gene dosage, since a heterozygous mutant might retain the ability to regulate high affinity targets during particular developmental timepoints, but never achieve the highest levels of expression needed to regulate low affinity targets. The effects of etiological point mutations of FOXP2 were assessed using a range of functional genetic techniques (Vernes et  al., 2006), an approach that has also proven informative when studying disease-causing mutations in other forkhead genes (Saleem et al., 2001, 2003; Berry et al., 2002, 2005). In their study, Vernes et al. (2006) focussed on three point mutations; the heterozygous missense mutation inherited by the 15 affected members of the KE family (R553H), the heterozygous nonsense mutation found in affected members of a small unrelated pedigree (R328X) and the substitution present in a single affected proband (Q17L). While the Q17L change did not substantially affect any aspect of protein function tested, the R553H and R328X mutations each substantially interfered with the functional properties of the FOXP2 protein, as described below.

4.3 R553H Affects Multiple Aspects of FOXP2 Function A formal molecular analysis of the properties of the R553H form of the FOXP2 ­protein has shed light on why this simple substitution may have such profound consequences for brain development and function, with the mutation affecting the DNA binding, regulatory capacity and localisation of FOXP2 (Vernes et  al., 2006). The R553H mutation was predicted to affect DNA-binding properties, given its location in the recognition helix of the DNA binding domain and evidence suggesting the formation of protein–DNA and intra-protein interactions by this residue during target recognition (Stroud et al., 2006). EMSA (electrophoretic mobility shift assay) studies using purified proteins and nuclear lysates clearly demonstrated that introduction of the R553H change abolishes binding of FOXP2 to a FOXP consensus sequence (Vernes et al., 2006). The R553H change also interfered with the protein’s ability to regulate target gene expression. A luciferase reporter assay demonstrated that introduction of the R553H change resulted in loss of repression from a promoter containing consensus binding sites and actually caused increased expression from the target promoter (Vernes et al., 2006). This effect is likely to relate to FOXP dimerisation; as the overexpressed mutant form may interfere with the function of endogenous FOXP proteins. Of note, the R553H protein may still retain some capacity to bind to target DNA, however the substitution could result in altered binding-site specificity.

264

S.C. Vernes and S.E. Fisher

In addition, the R553H change impacts on the intracellular localisation of the encoded FOXP2 protein (Fig. 4). Although the substitution occurs at some distance from the conserved NLS in the FOX domain, it has been suggested that it could alter the electrostatic charge distribution of the domain, disrupting interaction with nuclear transport machinery (Saleem et al., 2003; Berry et al., 2005). The effects of the R553H mutation on FOXP2 localisation were much more subtle than those reported for the paralogous R-to-H changes in the FOXC subfamily (Saleem et al., 2003; Berry et al., 2005), and varied in different cell-types, with stronger cytoplasmic signals in cells where no endogenous (i.e. wild-type) FOXP2 protein could be detected (Vernes et al., 2006). Moreover, co-transfection of R553H with wild-type FOXP2 has been shown to result in increased nuclear localisation of R553H with concomitant increases in cytoplasmic localisation of wild-type FOXP2 (Mizutani et al., 2007). These findings underscore the unusual nature of the FOXP subfamily of forkhead proteins. Unlike FOXC proteins which are thought to act as monomers, the distantly related FOXP proteins have dimerisation domains which appear crucial for function (Wang et al., 2003). This suggests that, when FOXP2.R553H is expressed in the presence of endogenous or cotransfected FOXP2, dimerisation can occur between mutant and wild-type proteins, and that such heterodimers may perhaps show more effective nuclear localisation than dimers comprised only of mutant protein (i.e. a partial rescue of nuclear targeting). Such observations could be relevant for understanding aetiological pathways in affected members of the KE family, who are heterozygous for the R553H mutation and may thus express both wild-type and mutant protein. It remains to be determined whether the R553H mutation in heterozygous KE members leads to speech/language disorder solely through a loss of function (i.e. reduced dosage of functional protein), or if the presence of the mutant protein also interferes with the activity of wild-type FOXP2 (an additional dominant negative effect).

4.4 R328X Yields an Unstable Non-Functional Product The R328X mutation is currently the only nonsense mutation identified for a human speech/language disorder. In vitro and in vivo studies have shown reduced levels of expression of FOXP2 when R328X or similar early stop codons are introduced, likely due to nonsense-mediated decay of the transcript and/or an unstable protein product (Vernes et  al., 2006; Groszer et  al., 2008). R328X ­protein was also predominantly localized to the cytoplasm (Fig.  4), showed no DNA binding capacity and demonstrated no transactivation capacity (Vernes et  al., 2006), suggesting that the severe truncation of the protein produces a largely non-functional transcription factor. Taking these findings together, it is likely that the problems observed in affected people carrying the R328X mutation result from reduced dosage of FOXP2 protein, rather than a dominant negative effect of the truncated product. Given the limited available data describing the  behavioural/cognitive phenotype associated with this nonsense mutation in

Functional Genomic Dissection of Speech and Language Disorders

265

humans, it will be important in future to make a detailed phenotypic comparison between individuals carrying the R553H and R328X changes, and correlate these with the functional genetic findings.

5 Regulatory Networks Associated with Human Speech and Language FOXP2’s role as a transcription factor, modulating the expression of target genes, offers elegant functional genomic routes for dissecting the associated neurogenetic pathways. By isolating the targets of FOXP2, it is possible to identify downstream molecular pathways that could be important for language development, and furthermore discover new candidate genes that may be more widely associated with developmental language disorders. The chromatin-immunoprecipitation (ChIP) technique is a powerful method for identifying transcription factor target genes under physiological conditions (Kim & Ren, 2006), as it allows the characterisation of genomic sites bound by a protein of interest in native chromatin of living cells. Briefly, ChIP involves the cross-linking of DNA binding proteins to their target DNA in living cells before isolation of the native protein/DNA complex, using antibodies directed to a protein of interest (in this case FOXP2). Once isolated, cross-links can be reversed and the DNA purified, yielding a sample enriched for those sites throughout the genome that were bound by the protein of interest (Fig.  5; for a detailed review of protocol, see Lee et al., 2006). Identifying the chromatin that has been pulled down (via interactions with the transcription factor) requires the coupling of ChIP with further techniques. These include subcloning and/or library construction from precipitated chromatin (Impey et al., 2004; Sun et al., 2005; Wei et al., 2006), or more recently, the use of microarrays or high throughput sequencing to screen ChIP isolated DNA (Weinmann et  al., 2002; Oberley et  al., 2004; Kim & Ren, 2006; Johnson et al., 2007).

5.1 High Throughput Analysis of FOXP2 Target Genes The first high throughput identification of neural targets regulated by FOXP2 was recently carried out by coupling the ChIP technique with promoter microarrays (Spiteri et al., 2007; Vernes et al., 2007). These concurrent ChIP-chip studies utilised two different model systems: in vitro human neuronal-like cells (Vernes et al., 2007) and in vivo human foetal brain tissue (Spiteri et al., 2007). For both studies, the DNA obtained via FOXP2-ChIP was applied to human promoter microarrays representing ~6,000 genomic regions. Approximately 30% of the regions bound by FOXP2 in each of the in vivo datasets were also bound in the in vitro study, a level of overlap that was highly statistically significant. The identified targets displayed

266

S.C. Vernes and S.E. Fisher

Fig.  5  Overview of FOXP2 chromatin immunoprecipitation. Living cells are cross-linked by treatment with formaldehyde to preserve interactions between transcription factors and target DNA. Following cell-lysis, genomic DNA is sheared into fragments of ~0.5–1 kb. At this stage, an aliquot is removed to be used as “input” reference DNA. Chromatin fragments are immunoprecipitated using an antibody directed to the N-terminus of FOXP2. Cross-links are heat-reversed in both the immunoprecipitated and input samples. After DNA purification and amplification samples may be subjected to unbiased location analysis (low throughput shotgun cloning) or high throughput location analysis via labelling with fluorescent dyes and application to microarrays containing fragments from thousands of promoters. Genes identified via either technique can then be investigated using in silico analysis. Validation via qPCR to confirm enrichment, quantitative RT-PCR expression analysis and DNA binding assays is then carried out on a gene-by-gene basis. Reproduced from Vernes et al., 2007

diverse roles in modulating synaptic plasticity, neurodevelopment, neurotransmission and axon guidance, and represent novel entrypoints into in vivo pathways that may be disturbed in speech and language disorders (Fig. 6).

Response to nutrient

Response to chemical stimulus

Kinase activity

Nucleotide kinase activity

Peptidase activity

Serine-type peptidase activity

Protein binding

Binding

Kinase binding

Kinesin binding

Enzyme binding Cytoskeletal protein binding

Protein C terminus binding

Cellular Metabolism

Carrier activity

Notch binding

Anion:anion antiporter activity

Porter activity

Homeostasis

Anion exchanger activity

Phosphate transporter activity

Inorganic anion transporter activity

Anion transporter activity

lon transporter activity

Transporter activity

Cell homeostasis

Chloride transporter activity Electrochemical potential-driven transporter activity

Growth factor activity

Receptor binding

RNA processing

rRNA processing

rRNA metabolism

RNA metabolism

Signal transducer activity

Interleukin-1 receptor binding

Cytokine activity

Primary Metabolism Nucleobase, nucleoside & nucleic acid metabolism

Metabolism

Physiological process

Oxygen and reactive oxygen species metabolism

Cell migration

Cell motility

Locomotion

Molecular Function

Axon guidance

Axonogenesis

Neurite development

Neuron development

Neuron differentiation

Cellular process

Fig. 6  Gene ontology clustering of target genes enriched during in vitro FOXP2 high-throughput location analysis. GOTrees demonstrating gene ontology classifications that were over-represented in the target list of 303 genes and relationships between categories with regard to molecular function and biological process. Significantly over-represented categories (p < 0.05) are highlighted in black. Reproduced from Vernes et al., 2007

Subtilase activity

Serine-type endopeptidase activity

Transferase

Chemotaxis

Taxis

Locomotory behaviour

Behaviour

Hydrolase activity

Catalytic activity

Response to nutrient levels

Response to extracellular stimulus

Response to external stimulus

Response to stimulus

Biological Process

Functional Genomic Dissection of Speech and Language Disorders 267

268

S.C. Vernes and S.E. Fisher

Analysis of a subset of targets uncovered via ChIP-chip demonstrated that FOXP2 binds in a specific manner to consensus sites within relevant promoter regions leading to significant changes in target gene expression (Spiteri et al., 2007; Vernes et al., 2007). The majority of direct targets showed significant reductions in mRNA expression as a consequence of increased FOXP2 levels; however, a minority of targets were consistently upregulated in response to FOXP2 expression. This suggested that, while this regulatory factor usually represses transcription, it is also able to switch from repressor to activator under certain circumstances (Spiteri et  al., 2007; Vernes et al., 2007), consistent with findings from a number of FOX family members, including the closely related FOXP3 protein (Schubert et  al., 2001; Marson et al., 2007; Zheng et al., 2007). The switch from repressor to activator may be dependent upon binding site affinity (Bird et al., 2004), interaction with cofactors (Washburn & Esposito, 2001), or post-translational modification status. Interaction with cofactors is particularly pertinent when considering FOXP2 regulation of gene expression, given that it requires dimerisation for efficient DNA binding. Thus, the expression levels of known binding partners, such as FOXP family members (Wang et al., 2003; Li et al., 2004), CTBP1(Li et al., 2004), NFAT (Wu et al., 2006), or other as yet unidentified interactors, could affect not only the affinity for the target DNA but also the switch from repressor to activator. In addition, a number of targets identified by ChIP were further assessed using a mouse model carrying a homozygous nonsense mutation of Foxp2 (S321X). This mouse mutant closely models the nonsense mutation identified in the small human pedigree (R328X) and results in reduced transcript levels and no detectable Foxp2 protein expression in the homozygous state. Despite a lack of Foxp2 protein, homozygous mutants show no gross anomalies in anatomy or brain development during embryogenesis. Postnatally they display developmental delays and reduced cerebellar growth, dying ~3–4 weeks after birth for as yet unknown reasons (Groszer et  al., 2008). Given the 100% conservation of the FOXP2/Foxp2 DNA binding domain, it was hypothesised that these proteins would share some common target genes (Vernes et al., 2007). Chromatin immunoprecipitation studies in wild-type and homozygous mutant (S321X) littermates demonstrated promoter occupancy of Slc17a3, a target of FOXP2 identified in vitro and in vivo, accompanied by significant quantitative differences in the expression of this gene in the embryonic brains of mutant mice (Fig. 7) (Vernes et al., 2007). Thus, SLC17A3, a target identified in human model systems, displayed conserved binding and regulation by Foxp2 in the developing mouse brain.

5.2 Unbiased Identification of FOXP2 Binding FOXP2 chromatin immunoprecipitation using an in  vitro human neuron-like cell model was also coupled to shotgun cloning and sequencing (shotgun-ChIP) (Vernes et al., 2008). This approach identified a FOXP2-bound fragment within intron 1 of the CNTNAP2 gene (Fig.  8) (Vernes et  al., 2008). Although low throughput by

Functional Genomic Dissection of Speech and Language Disorders

269

Fig.  7  Foxp2 binds to and regulates target gene expression in  vivo in embryonic mouse brain. (a) Analysis of in vivo promoter occupancy. FOXP2 ChIP was performed using whole brains at E16 from either wild-type mice or homozygous mutant littermates that lack Foxp2 protein as a consequence of an exon-7 nonsense mutation. DNA was PCR amplified using primers directed towards Slc17a3, Cer1, Psen2 or the b-actin control promoter regions. The promoters of Slc17a3 and Cer1 were specifically enriched in ChIP samples isolated from wild-type brains compared to those from mutant brains. (b) Expression of Slc17a3 shows an inverse relationship with Foxp2 expression in E16 mouse brains. RT-PCR was performed with cDNA generated from 5 wild-type vs. 5 mutant whole brain samples. Note that reduced expression of Foxp2 in mutant mice is the result of nonsense-mediated RNA decay; in addition, the potential truncated product is highly unstable, leading to a lack of detectable Foxp2 protein (Groszer et al., 2008). Expression changes are represented as mean log2 expression ratios of comparisons between wild-type and mutant mice normalized for equal expression of the internal control, Gapdh. Statistical significance (**p < 0.01, ****p < 0.0001) was calculated using a two-tailed unpaired t test. Adapted from Vernes et al., 2007

comparison to array-based methods, shotgun-ChIP afforded an unbiased interrogation of binding sites identifying sites bound by FOXP2 not only in classical promoter regions but also intronic and 3’ regions. Because the FOXP2 binding site associated with CNTNAP2 was located within an intron and not within a classical promoter region, it would not have been detected via ChIP-chip with promoter or CpG island arrays. ChIP-chip limits detection to those regions that are represented as features on a given array (e.g. promoter arrays, CpG island arrays). Given the increasing body of evidence to suggest that transcription factor binding regularly occurs in regions not defined as classical promoters, (Cawley et  al., 2004; Johnson et  al., 2007; Marson et al., 2007; Zheng et al., 2007; Kimura et al., 2006; Birney et al., 2007; Trinklein et al., 2007), further unbiased screening will be necessary to characterise the FOXP2 binding pattern and this may shed further light on the molecular mechanisms by which FOXP2 exerts transcriptional control on its target genes. ChIP has also recently been coupled to ultra-high throughput sequencing (involving two to ten million sequence reads per experiment) in a technique termed ­ChIP-Seq (Johnson et al., 2007; Mikkelsen et al., 2007). This method boasts the

270

S.C. Vernes and S.E. Fisher

Fig. 8  An unbiased screen identifies CNTNAP2 as a direct neural target bound by human FOXP2. A ~300-bp clone was identified via FOXP2 shotgun-ChIP and localized to intron 1 of the human CNTNAP2 gene in 7q35. Semi-quantitative PCR indicated consistent enrichment of this region in multiple independent ChIP experiments in a neuron-like cell-line immunoprecipitated with an N-terminal FOXP2 antibody (lane 2) compared to no-antibody control (lane 3) and “input DNA” samples (lane 1). Lane 4 shows the water control. Two FOXP2 consensus binding sites were identified (CAAATT). Adapted from Vernes et al., 2008

high throughput screening associated with array-based platforms, but with the ­unbiased nature of sequence-based experiments and requires far less DNA as starting material, reducing the need for multiple rounds of amplification (which itself has the potential to introduce bias) (Jacobs et al., 2003; Johnson et al., 2007). The emergence of ChIP-Seq seems to provide an accessible method for unbiased whole genome interrogations of transcription factor binding and will likely help elucidate the frequency and pattern of non-promoter binding.

5.3 CNTNAP2 is a Direct Target of FOXP2 As noted above, a site within intron 1 of CNTNAP2 was shown to be bound by FOXP2 using shotgun-ChIP and confirmed via PCR and EMSA studies. PCR was

Functional Genomic Dissection of Speech and Language Disorders

271

used to amplify sequences spanning the FOXP2-bound fragment of CNTNAP2 in independent FOXP2 ChIP replicates as well as in control samples where no antibody control ChIP’s were performed (Vernes et  al., 2008). Enrichment was only observed when FOXP2-specific antibodies had been used to isolate the chromatin (Fig. 8). FOXP2 is thought to bind chromatin as a dimer, and in silico analyses of the FOXP2-ChIP-enriched fragment identified two adjacent sites, separated by 48 bases, matching a known consensus sequence for FOXP2 binding (CAAATT) (Fig. 8). EMSA analyses indicated that FOXP2 was able to bind both sites, and at each site binding could be disrupted by the mutation of three core nucleotides of the recognition sequences (i.e. CAAATT CGGGTT). Furthermore, CNTNAP2 expression could be seen to be repressed by FOXP2 in human neuronal model systems (Vernes et al., 2008). The CNTNAP2 gene encodes CASPR2, a member of the neurexin superfamily of transmembrane proteins involved in neuronal recognition and adhesion (Rasband, 2004). CASPR2 forms multi-protein complexes at juxtaparanodal regions of nodes of Ranvier necessary for the correct localisation and/or maintenance of Shaker-type voltage-activated potassium channels (Poliak et al., 2003; Traka et al., 2003). These potassium channels stabilise neural conduction and help to maintain internodal resting potential (Inda et al., 2006). In addition, CASPR2 has been implicated in human cortical histogenesis and is thought to mediate intercellular interactions ­during neuroblast migration and laminar organisation (Strauss et al., 2006). A recent search for genes involved in human cortical patterning identified CNTNAP2 as a differentially expressed gene in perisylvian regions of the developing human brain. CNTNAP2 gene expression was enriched in the frontal cortex compared to the superior temporal gyrus (Abrahams et  al., 2007). In mid-gestation human foetal brain slices, CNTNAP2 expression mirrored that of cortico-striato-thalamic ­circuitry involved in higher cognitive function, showing expression in frontal and anterior cortex, dorsal thalamus, caudate, putamen and amygdala (Abrahams et al., 2007). Given the striking and highly restricted pattern of expression in the mid-gestational brain and the fact that myelination does not occur until late gestational stages, Abrahams et  al. (2007) hypothesised that CNTNAP2 may also play a role in early patterning in cortico-striato-thalamic regions. Mutations in CNTNAP2 have also been associated with a number of neurological disorders. Chromosomal rearrangements involving CNTNAP2 have been identified in patients with Gilles de la Tourette syndrome (complex chromosomal insertion/translocation) (Verkerk et al., 2003), schizophrenia, and epilepsy (hemizygous deletions of between 1.5 and 11 Mb encompassing CNTNAP2) (Friedman et al., 2008). Strauss et al. (2006) reported the detection of CNTNAP2 mutations in patients diagnosed with cortical dysplasia–focal epilepsy (CDFE) syndrome (MIM: 610042). Patients displayed mild gross motor delay, seizures, language regression, and autistic characteristics. While this study suggested a role for CNTNAP2 during brain development, it also highlighted the potential for functional differences between mouse and human. Homozygous truncation of CASPR2 in these patients produced a severe phenotype including widespread tissue abnormalities throughout the hippocampus, amygdala, neocortex and

272

S.C. Vernes and S.E. Fisher

s­ ubcortex as well as abnormal neuronal migration, organisation and morphology (Strauss et  al., 2006). By comparison, Cntnap2 null mice do not display any gross histological or neurological abnormalities within the first 20 months of life, although K+ channel subunits were found to be mislocalised at nodes of Ranvier, similar to what was seen in human tissue (Poliak & Peles, 2003; Poliak et al., 2003).

5.4 FOXP2 and Common Forms of Language Disorder Identifying transcription factor targets presents an alternative methodology for elucidating candidate disease genes involved in the same molecular pathway. This technique is particularly applicable to complement positional cloning studies in complex disorders such as SLI and autism where the multiple loci contributing to the disorder make classical mapping techniques difficult (Fisher et  al., 2003; Abrahams & Geschwind, 2008). Although whole genome scans for SLIsusceptibility loci have been performed, none of these studies demonstrated significant linkage to FOXP2 or the SPCH1 region (7q31) (Bartlett et  al., 2002; SLIConsortium, 2002) Furthermore, mutation screening of the coding region of the FOXP2 gene in 43 SLI and 48 autistic probands did not identify any functional variants (Newbury et al., 2002). In order to investigate the hypothesis that targets of FOXP2 (i.e. genes functioning in the same molecular pathway) might be involved in more common forms of language disorder, CNTNAP2 was tested for association with quantitative markers of language impairment in SLI. A cluster of SNPs located within the exon 13–15 region of the CNTNAP2 gene demonstrated significant quantitative association with non-word repetition (a heritable marker of language impairment) in a large sample of children with typical forms of SLI (Vernes et al., 2008). These findings are of particular interest given independent data suggesting association of CNTNAP2 with autism (Alarcon et  al., 2008). A study of CNTNAP2 in children diagnosed with autism established an association between polymorphisms in the exon 13–15 region and age-at-first-word suggesting that common genetic variation in CNTNAP2 may be linked to delayed language onset in this disorder (Alarcon et  al., 2002, 2008; Geschwind & Levitt, 2007). The discovery that CNTNAP2 is a neural target of FOXP2, and that similar variants of this target are associated with deficits in common forms of language impairment (both typical SLI and delayed language in autism), illustrates how knowledge of the genetic cause of a rare single-gene disorder provides entrypoints into etiology of more complex phenotypes. By integrating functional genomics and quantitative trait analyses, a shared neurogenetic pathway was identified, one which is disturbed in distinct forms of language impairment. Further investigations of regulatory networks such as the FOXP2-CNTNAP2 pathway may lead to a better understanding of neurogenetic mechanisms involved in typical language disorders.

Functional Genomic Dissection of Speech and Language Disorders

273

6 Conclusions/Future Perspectives Although mutations in FOXP2 have only been implicated in a small proportion of individuals displaying speech/language disorder, the use of this gene as a molecular entrypoint into the neural pathways subserving language should not be underestimated. The fact that spoken language is unique to the human species, along with the complex nature of this trait, have made it difficult to study the critical neuromolecular networks. However, as has been illustrated via CNTNAP2, new candidates for involvement in common language disorders can be predicted by identifying genes that act in the same pathways as FOXP2. CNTNAP2 is, at the time of writing, the only FOXP2 target that has been tested for association with language impairment and future screens of other candidate genes identified in this way will reveal to what overall extent these FOXP2 associated pathways are implicated in language phenotypes. Furthermore, complementary approaches using animal models are expected to shed light on ancestral mechanisms that have been recruited towards speech/language development and processing (Fisher & Scharff, 2009). For example, the remarkable degree of conservation between mouse and human FoxP2 suggest high levels of shared neural function (Fisher & Marcus, 2006). Existing mutants include the S321X null mouse as well as R552H mutants – the model the missense mutation found in the KE family (Groszer et  al., 2008). Investigations into heterozygous mice carrying the R552H change have already revealed defects in motor-skill learning and abnormal synaptic plasticity in striatal and cerebellar neural circuitry (Groszer et  al., 2008). Songbirds are vocal learners and provide another in  vivo model system in which to study FoxP2 function given the strong concordance for the neurological systems used in birdsong and human speech and language (Doupe & Kuhl, 1999). Studies using a lentiviral knockdown system in juvenile and adult zebra finches have recently suggested an important role for FoxP2 in the postnatal songbird brain and may indicate that this gene is necessary for auditory guided motor learning during song development (Haesler et al., 2007). Studies in model systems such as these allow investigation of the role of FoxP2 in the context of the developing as well as the adult brain and complement data obtained from in vitro and human investigations. Only by integrating information gained from multiple approaches and model systems can we understand the complexities of the neuromolecular mechanisms underlying human speech and language.

References Aboitiz, F., & Garcia, V. R. (1997). The evolutionary origin of the language areas in the human brain. A neuroanatomical perspective. Brain Research Brain Research Reviews, 25(3), 381–396. Abrahams, B. S., & Geschwind, D. H. (2008). Advances in autism genetics: On the threshold of a new neurobiology. Nature Reviews Genetics, 9(5), 341–355.

274

S.C. Vernes and S.E. Fisher

Abrahams, B. S., Tentler, D., Perederiy, J. V., Oldham, M. C., Coppola G., & Geschwind D. H. (2007). Genome-wide analyses of human perisylvian cerebral cortical patterning. Proc Natl Acad Sci U S A, 104(45), 17849–17854. Alarcon, M., Abrahams, B. S., Stone, J. L., Duvall, J. A., Perederiy, J. V., Bomar, J. M., et al. (2008). Linkage, association, and gene-expression analyses identify CNTNAP2 as an autismsusceptibility gene. Am J Hum Genet, 82(1), 150–159. Alarcon, M., Cantor, R. M., Liu, J., Gilliam T. C., & Geschwind D. H. (2002). Evidence for a language quantitative trait locus on chromosome 7q in multiplex autism families. Am J Hum Genet, 70(1), 60–71. Bartlett, C. W., Flax, J. F., Logue, M. W., Smith, B. J., Vieland, V. J., Tallal, P., et al. (2004). Examination of potential overlap in autism and language loci on chromosomes 2, 7, and 13 in two independent samples ascertained for specific language impairment. Hum Hered, 57(1), 10–20. Bartlett, C. W., Flax, J. F., Logue, M. W., Vieland, V. J., Bassett, A. S., Tallal, P., et al. (2002). A major susceptibility locus for specific language impairment is located on 13q21. Am J Hum Genet, 71(1), 45–55. Belton, E., Salmond, C. H., Watkins, K. E., Vargha-Khadem, F., & Gadian, D. G. (2003). Bilateral brain abnormalities associated with dominantly inherited verbal and orofacial dyspraxia. Hum Brain Mapp, 18(3), 194–200. Bennett, C. L., Christie, J., Ramsdell, F., Brunkow, M. E., Ferguson, P. J., Whitesell, L., et al. (2001). The immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome (IPEX) is caused by mutations of FOXP3. Nat Genet 27(1), 20–21. Berry, F. B., Saleem, R. A., & Walter, M. A. (2002). FOXC1 transcriptional regulation is mediated by N- and C-terminal activation domains and contains a phosphorylated transcriptional inhibitory domain. J Biol Chem, 277(12), 10292–10297. Berry, F. B., Tamimi, Y., Carle, M. V., Lehmann, O. J., & Walter, M. A. (2005). The establishment of a predictive mutational model of the forkhead domain through the analyses of FOXC2 missense mutations identified in patients with hereditary lymphedema with distichiasis. Hum Mol Genet, 14(18), 2619–2627. Bird, A. J., Blankman, E., Stillman, D. J., Eide, D. J., & Winge, D. R. (2004). The Zap1 transcriptional activator also acts as a repressor by binding downstream of the TATA box in ZRT2. Embo J, 23(5), 1123–1132. Birney, E., Stamatoyannopoulos, J. A., Dutta, A., Guigo, R., Gingeras, T. R., Margulies, E. H., et al. (2007). Identification and analysis of functional elements in 1% of the human genome by the ENCODE pilot project. Nature, 447(7146), 799–816. Bishop, D. V., North, T., & Donlan, C. (1995). Genetic basis of specific language impairment: evidence from a twin study. Dev Med Child Neurol, 37(1), 56–71. Bruce, H. A., & Margolis, R. L. (2002). FOXP2: novel exons, splice variants, and CAG repeat length stability. Hum Genet, 111(2), 136–144. Cawley, S., Bekiranov, S., Ng, H. H., Kapranov, P., Sekinger, E. A., Kampa, D., et al. (2004). Unbiased mapping of transcription factor binding sites along human chromosomes 21 and 22 points to widespread regulation of noncoding RNAs. Cell, 116(4), 499–509. Crosson, B. (1985). Subcortical functions in language: a working model. Brain Lang, 25(2), 257–292. Doupe, A. J., & Kuhl, P. K. (1999). Birdsong and human speech: common themes and mechanisms. Annu Rev Neurosci, 22, 567–631. Enard, W., Przeworski, M., Fisher, S. E., Lai, C. S., Wiebe, V., Kitano, T., et al. (2002). Molecular evolution of FOXP2, a gene involved in speech and language. Nature, 418(6900), 869–872. Ferland, R. J., Cherry, T. J., Preware, P. O., Morrisey, E. E., & Walsh, C. A. (2003). Characterization of Foxp2 and Foxp1 mRNA and protein in the developing and mature brain. J Comp Neurol, 460(2), 266–279. Feuk, L., Kalervo, A., Lipsanen-Nyman, M., Skaug, J., Nakabayashi, K., Finucane, B., et al. (2006). Absence of a Paternally Inherited FOXP2 Gene in Developmental Verbal Dyspraxia. Am J Hum Genet, 79(5), 965–972.

Functional Genomic Dissection of Speech and Language Disorders

275

Fisher, S. E., & Marcus, G. F. (2006). The eloquent ape: genes, brains and the evolution of language. Nature Reviews Genetics, 7(1), 9–20. Fisher, S. E., & Scharff, C. (2009). FOXP2 as a molecular window into speech and language. Trends in Genetics, 25(4), 166–177. Fisher, S. E., Lai C. S., & Monaco, A. P. (2003). Deciphering the genetic basis of speech and language disorders. Annu Rev Neurosci, 26, 57–80. Fisher, S. E., Vargha-Khadem, F., Watkins, K. E , Monaco, A. P., & Pembrey, M. E. (1998). Localisation of a gene implicated in a severe speech and language disorder. Nat Genet, 18(2), 168–170. Friedman, J. I., Vrijenhoek, T., et al. (2008). CNTNAP2 gene dosage variation is associated with schizophrenia and epilepsy. Mol Psychiatry, 13(3), 261–266. Gaudet, J., & Mango, S. E. (2002). Regulation of organogenesis by the Caenorhabditis elegans FoxA protein PHA-4. Science 295(5556), 821–825. Geschwind, D. H., & Levitt, P. (2007). Autism spectrum disorders: developmental disconnection syndromes. Curr Opin Neurobiol, 17(1), 103–111. Gopnik, M., & Crago, M. B. (1991). Familial aggregation of a developmental language disorder. Cognition, 39(1), 1–50. Groszer, M., Keays, D. A., Deacon, R. M., de Bono, J. P., Prasad-Mulcare, S., Gaub, S., et al. (2008). Impaired synaptic plasticity and motor learning in mice with a point mutation implicated in human speech deficits. Curr Biol, 18(5), 354–362. Haesler, S., Rochefort, C., Georgi, B., Licznerski, P., Osten, P., & Scharff, C. (2007). Incomplete and inaccurate vocal imitation after knockdown of FoxP2 in songbird basal ganglia nucleus Area X. PLoS Biol, 5(12), e321. Hurst, J. A., Baraitser, M., Auger, E., Graham, F., & Norell, S. (1990). An extended family with a dominantly inherited speech disorder. Dev Med Child Neurol, 32(4), 352–355. Impey, S., McCorkle, S. R., Cha-Molstad, H., Dwyer, J. M., Yochum, G. S., Boss, J. M.. et al. (2004). Defining the CREB regulon: a genome-wide analysis of transcription factor regulatory regions. Cell, 119(7), 1041–1054. Inda, M. C., DeFelipe, J., & Munoz, A. (2006). Voltage-gated ion channels in the axon initial segment of human cortical pyramidal cells and their relationship with chandelier cells. Proc Natl Acad Sci U S A, 103(8), 2920–2925. Jacobs, F. M., van der Heide, L. P., Wijchers, P. J., Burbach, J. P., Hoekman, M. F., & Smidt, M. P. (2003). FoxO6, a novel member of the FoxO class of transcription factors with distinct shuttling dynamics. J Biol Chem, 278(38), 35959–35967. Johnson, D. S., Mortazavi, A., Myers, R. M., & Wold, B. (2007). Genome-wide mapping of in vivo protein-DNA interactions. Science, 316(5830), 1497–1502. Kaestner, K. H., Knochel, W., & Martinez, D. E. (2000). Unified nomenclature for the winged helix/forkhead transcription factors. Genes Dev, 14(2), 142–146. Katoh, M., & Katoh, M. (2004). Human FOX gene family (review). Int J Oncol, 25(5), 1495–1500. Kim, T. H., & Ren, B. (2006). Genome-wide analysis of protein-DNA interactions. Annu Rev Genomics Hum Genet, 7, 81–102. Kimura, K., Wakamatsu, A., Suzuki, Y., Ota, T., Nishikawa, T., Yamashita, R., et al. (2006). Diversification of transcriptional modulation: large-scale identification and characterization of putative alternative promoters of human genes. Genome Res, 16(1), 55–65. Kopito, R. R. (2000). Aggresomes, inclusion bodies and protein aggregation. Trends Cell Biol, 10(12), 524–530. Lahey, M., & Edwards, J. (1995). Specific language impairment: Preliminary investigation of factors associated with family history and with patterns of language performance. Journal of Speech Hear Res, 38(3), 643–657. Lai, C. S., Fisher, S. E., Hurst, J. A., Levy, E. R., Hodgson, S., Fox, M., et al. (2000). The SPCH1 region on human 7q31: genomic characterization of the critical interval and localization of translocations associated with speech and language disorder. Am J Hum Genet, 67(2), 357–368. Lai, C. S., Fisher, S. E., Hurst, J. A., Vargha-Khadem, F., & Monaco, A. P. (2001). A forkheaddomain gene is mutated in a severe speech and language disorder. Nature, 413(6855), 519–523.

276

S.C. Vernes and S.E. Fisher

Lai, C. S., Gerrelli, D., Monaco, A. P., Fisher, S. E., & Copp, A. J. (2003). FOXP2 expression during brain development coincides with adult sites of pathology in a severe speech and language disorder. Brain, 126(Pt 11), 2455–2462. Lee, T. I., Johnstone, S. E. & Young, R. A. (2006). Chromatin immunoprecipitation and microarray-based analysis of protein location. Nat Protoc, 1(2): 729–748. Lehmann, O. J., Sowden, J. C., Carlsson, P. Jordan, T., & Bhattacharya, S. S. (2003). Fox’s in development and disease. Trends Genet, 19(6), 339–344. Lennon, P. A., Cooper, M. L., Peiffer, D. A., Gunderson, K. L., Patel, A., Peters, S., et al. (2007). Deletion of 7q31.1 supports involvement of FOXP2 in language impairment: clinical report and review. Am J Med Genet A, 143A(8), 791–798. Lewis, B. A., Ekelman, B. L., & Aram, D. M., (1989). A familial study of severe phonological disorders. J Speech Hear Res, 32(4), 713–724. Lewis, B. A., & Thompson, L. A. (1992). A study of developmental speech and language disorders in twins. J Speech Hear Res, 35(5), 1086–1094. Li, S., Weidenfeld, J., & Morrisey, E. E. (2004). Transcriptional and DNA binding activity of the Foxp1/2/4 family is modulated by heterotypic and homotypic protein interactions. Mol Cell Biol, 24(2), 809–822. Liegeois, F., Baldeweg, T., Connelly, A., Gadian, D. G., Mishkin, M., & Vargha-Khadem, F. (2003). Language fMRI abnormalities associated with FOXP2 gene mutation. Nat Neurosci, 6(11), 1230–1237. Lu, M. M., Li, S., Yang, H., & Morrisey, E. E. (2002). Foxp4: a novel member of the Foxp subfamily of winged-helix genes co-expressed with Foxp1 and Foxp2 in pulmonary and gut tissues. Gene Expr Patterns, 2(3–4), 223–228. MacDermot, K. D., Bonora, E., Sykes, N., Coupe, A. M., Lai, C. S., Vernes, S. C., et al. (2005). Identification of FOXP2 truncation as a novel cause of developmental speech and language deficits. Am J Hum Genet, 76(6), 1074–1080. Marson, A., Kretschmer, K., Frampton, G. M., Jacobsen, E. S. Polansky, J. K., Macisaac, K. D., et al. (2007). Foxp3 occupancy and regulation of key target genes during T-cell stimulation. Nature, 445(7130), 931–935. Mikkelsen, T. S., Ku, M., Jaffe, D. B., Issac, B., Lieberman, E., Giannoukos, G., et al. (2007). Genome-wide maps of chromatin state in pluripotent and lineage-committed cells. Nature, 448(7153), 553–560. Mizutani, A., Matsuzaki, A., Momoi, M. Y., Fujita, E., Tanabe, Y., & Momoi, T. (2007). Intracellular distribution of a speech/language disorder associated FOXP2 mutant. Biochem Biophys Res Commun, 353(4), 869–874. Nakahata, S., & Kawamoto, S. (2005). Tissue-dependent isoforms of mammalian Fox-1 homologs are associated with tissue-specific splicing activities. Nucleic Acids Research, 33(7), 2078–2089. Neils, J., & Aram, D. M. (1986). Family history of children with developmental language disorders. Perceptual and Motor Skills, 63(2 Pt 1), 655–658. Newbury, D. F., Bonora, E., Lamb, J. A., Fisher, S. E., Lai, C. S., Baird, G., et al. (2002). FOXP2 is not a major susceptibility gene for autism or specific language impairment. Am J Hum Genet, 70(5), 1318–13127. Oberley, M. J., Tsao, J., Yau, P., & Farnham, P. J. (2004). High-throughput screening of chromatin immunoprecipitates using CpG-island microarrays. Methods Enzymol, 376, 315–334. Ojemann, G. A. (1991). Cortical organization of language. J Neurosci, 11(8), 2281–2287. Poliak, S., & Peles, E. (2003). The local differentiation of myelinated axons at nodes of Ranvier. Nat Rev Neurosci, 4(12), 968–980. Poliak, S., Salomon, D., Elhanany, H., Sabanay, H., Kiernan, B., Pevny, L., et al. (2003). Juxtaparanodal clustering of Shaker-like K+ channels in myelinated axons depends on Caspr2 and TAG-1. J Cell Biol, 162(6), 1149–1160. Price, C. J. (2000). The anatomy of language: Contributions from functional neuroimaging. J Anat, 197(Pt 3), 335–359.

Functional Genomic Dissection of Speech and Language Disorders

277

Rasband, M. N. (2004). It’s “juxta” potassium channel! J Neurosci Res, 76(6), 749–757. Saleem, R. A., Banerjee-Basu, S., Berry, F. B., Baxevanis, A. D., & Walter, M. A., (2001). Analyses of the effects that disease-causing missense mutations have on the structure and function of the winged-helix protein FOXC1. Am J Hum Genet, 68(3), 627–641. Saleem, R. A., Banerjee-Basu, S., Berry, F. B., Baxevanis, A. D., & Walter, M. A. (2003). Structural and functional analyses of disease-causing missense mutations in the forkhead domain of FOXC1. Hum Mol Genet, 12(22), 2993–3005. Schroeder, D. I., & Myers, R. M. (2008). Multiple transcription start sites for FOXP2 with varying cellular specificities. Gene, 413(1–2), 42–48. Schubert, L. A., Jeffery, E., Zhang, Y., Ramsdell, F., & Ziegler, S. F. (2001). Scurfin (FOXP3) acts as a repressor of transcription and regulates T cell activation. J Biol Chem, 276(40), 37672–37679. Shu, W., Yang, H., Zhang, L., Lu, M. M., & Morrisey, E. E. (2001). Characterization of a new subfamily of winged-helix/forkhead (Fox) genes that are expressed in the lung and act as transcriptional repressors. J Biol Chem, 276(29), 27488–27497. SLIConsortium. (2002). A genomewide scan identifies two novel loci involved in specific language impairment. Am J Hum Genet, 70(2), 384–398. Spiteri, E., Konopka, G., Coppola, G., Bomar, J., Oldham, M., Ou, J., et al. (2007). Identification of the Transcriptional Targets of FOXP2, a Gene Linked to Speech and Language, in Developing Human Brain. Am J Hum Genet, 81(6), 1144–1157. Strauss, K. A., Puffenberger, E. G., Huentelman, M. J., Gottlieb, S., Dobrin, S. E., Parod, J. M., et al. (2006). Recessive symptomatic focal epilepsy and mutant contactin-associated proteinlike 2. N Engl J Med, 354(13), 1370–1377. Stroud, J. C., Wu, Y., Bates, D. L., Han, A., Nowick, K., Paabo, S., et al. (2006). Structure of the Forkhead Domain of FOXP2 Bound to DNA. Structure, 14(1), 159–166. Sun, Y. M., Greenway, D. J., Johnson, R., Street, M., Belyaev, N. D., Deuchars, J., et al. (2005). Distinct profiles of REST interactions with its target genes at different stages of neuronal development. Mol Biol Cell, 16(12), 5630–5638. Tallal, P., Ross, R., & Curtiss, S. (1989). Familial aggregation in specific language impairment. J Speech Hear Disord, 54(2), 167–173. Tanaka, M., Kim, Y. M., Lee, G., Junn, E., Iwatsubo, T., & Mouradian, M. M. (2004). Aggresomes formed by alpha-synuclein and synphilin-1 are cytoprotective. J Biol Chem, 279(6), 4625–4631. Taylor, J. P., Tanaka, F., Robitschek, J., Sandoval, C. M., Taye, A., Markovic-Plese, S., et al. (2003). Aggresomes protect cells by enhancing the degradation of toxic polyglutamine-containing protein. Hum Mol Genet, 12(7), 749–757. Teufel, A., Wong, E. A., Mukhopadhyay, M., Malik, N., & Westphal, H. (2003). FoxP4, a novel forkhead transcription factor. Biochim Biophys Acta, 1627(2–3), 147–152. Tomblin, J. B. (1989). Familial concentration of developmental language impairment. J Speech Hear Disord, 54(2), 287–295. Tomblin, J. B., & Buckwalter, P. R. (1998). Heritability of poor language achievement among twins. J Speech, Lang Hear Res, 41(1), 188–199. Tomblin, J. B., Records, N. L., Buckwalter, P., Zhang, X., Smith, E., & O’Brien, M. (1997). Prevalence of specific language impairment in kindergarten children. J Speech Lang Hear Res, 40(6), 1245–1260. Traka, M., Goutebroze, L., Denisenko, N., Bessa, M., Nifli, A., Havaki, S., et al. (2003). Association of TAG-1 with Caspr2 is essential for the molecular organization of juxtaparanodal regions of myelinated fibers. J Cell Biol, 162(6), 1161–1172. Trinklein, N. D., Karaoz, U., Wu, J., Halees, A., Force Aldred, S., Collins, P. J., et al. (2007). Integrated analysis of experimental data sets reveals many novel promoters in 1% of the human genome. Genome Res, 17(6), 720–731. Vargha-Khadem, F., Watkins, K., Alcock, K., Fletcher, P., & Passingham, R. (1995). Praxic and nonverbal cognitive deficits in a large family with a genetically transmitted speech and language disorder. Proc Natl Acad Sci U S A, 92(3), 930–933.

278

S.C. Vernes and S.E. Fisher

Vargha-Khadem, F., Watkins, K. E., Price, C. J., Ashburner, J., Alcock, K. J., Connelly, A., et al. (1998). Neural basis of an inherited speech and language disorder. Proc Natl Acad Sci U S A, 95(21), 12695–12700. Verkerk, A. J., Mathews, C. A., Joosse, M., Eussen, B. H., Heutink, P., & Oostra, B. A. (2003). CNTNAP2 is disrupted in a family with Gilles de la Tourette syndrome and obsessive compulsive disorder. Genomics, 82(1), 1–9. Vernes, S. C., Newbury, D. F., Abrahams, B.S., Winchester, L., Nicod, J., Groszer, M., et al. (2008). A functional genetic link between distinct developmental language disorders. N Engl J Med, 359(22), 2337–2345. Vernes, S. C., Nicod, J., Elahi, F. M., Coventry, J. A., Kenny, N., Coupe, A. M., et al. (2006). Functional genetic analysis of mutations implicated in a human speech and language disorder. Hum Mol Genet, 15(21), 3154–3167. Vernes, S. C., Spiteri, E., Nicod, J., Groszer, M., Taylor, J. M., Davies, K. E., et al. (2007). HighThroughput Analysis of Promoter Occupancy Reveals Direct Neural Targets of FOXP2, a Gene Mutated in Speech and Language Disorders. Am J Hum Genet, 81(6), 1232–1250. Wang, B., Lin, D., Li, C., & Tucker, P. (2003). Multiple domains define the expression and regulatory properties of Foxp1 forkhead transcriptional repressors. J Biol Chem, 278(27), 24259–24268. Washburn, B. K., & Esposito, R. E. (2001). Identification of the Sin3-binding site in Ume6 defines a two-step process for conversion of Ume6 from a transcriptional repressor to an activator in yeast. Mol Cell Biol, 21(6), 2057–2069. Watkins, K. E., Dronkers, N. F., & Vargha-Khadem, F. (2002a). Behavioural analysis of an inherited speech and language disorder: comparison with acquired aphasia. Brain 125(Pt3), 452–464. Watkins, K. E., Vargha-Khadem, F., Ashburner, J., Passingham, R. E., Connelly, A., Friston, K. J., et al. (2002b). MRI analysis of an inherited speech and language disorder: structural brain abnormalities. Brain 125(Pt 3), 465–478. Webb, J. L., Ravikumar, B., & Rubinsztein, D. C. (2004). Microtubule disruption inhibits autophagosome-lysosome fusion: implications for studying the roles of aggresomes in polyglutamine diseases. Int J Biochem Cell Biol, 36(12), 2541–2550. Wei, C. L., Wu, Q., Vega, V. B., Chiu, K. P., Ng, P., Zhang, T., et al. (2006). A global map of p53 transcription-factor binding sites in the human genome. Cell, 124(1), 207–219. Weinmann, A. S., Yan, P. S., Oberley, M. J., Huang, T. H., & Farnham, P. J. (2002). Isolating human transcription factor targets by coupling chromatin immunoprecipitation and CpG island microarray analysis. Genes Dev, 16(2), 235–244. Wu, Y., Borde, M., Heissmeyer, V., Feuerer, M., Lapan, A. D., Stroud, J. C., et al. (2006). FOXP3 controls regulatory T cell function through cooperation with NFAT. Cell 126(2), 375–387. Yuasa, J., Hirano, S., Yamagata, M., & Noda, M. (1996). Visual projection map specified by topographic expression of transcription factors in the retina. Nature, 382(6592), 632–635. Zheng, Y., Josefowicz, S. Z., Kas, A., Chu, T. T., Gavin, M. A., & Rudensky, A. Y. (2007). Genomewide analysis of Foxp3 target genes in developing and mature regulatory T cells. Nature, 445 (7130), 936–940.

Studying Human Circadian Behaviour Using Peripheral Cells Lucia Pagani, Anne Eckert, and Steven A. Brown

Abstract  Complex behaviours are the product of intercellular signalling events, but their intracellular effectors are present in most cell types. The best-studied example of such architecture is the circadian clock, which directs all diurnal behaviour and physiology, and whose central mechanism is present in most body cells. We present below a method to look at its properties via transcriptional reporters virally delivered to primary cells. By studying primary fibroblasts cultivated from skin biopsies in different human subjects, we have been able to analyse the molecular underpinnings of variance in human daily behaviour. Similar methodologies could be applied to other signalling pathways. Keywords  Circadian • Fibroblast • Luciferase • Reporter • Biopsy • Human ­behaviour • Primary cells

1 Introduction Human behaviour is influenced by many genetic and environmental factors, and is the product of complex inter-neuron interactions; it is therefore often difficult to study by reductionist approaches. Nevertheless, it is increasingly clear that fundamental intracellular pathways are critical to behaviour, and that these same pathways are conserved in non-neuronal cell types. The premise of this chapter is that inter-individual genetic differences that create changes in behaviour do so at least in part by modifying conserved intracellular signalling pathways, and that these changes are paralleled by changes in non-neuronal cell types. Hence, they can be studied in peripheral cells obtained from human subjects.

S.A. Brown (*) Chronobiology and Sleep Research Group, Institut für Pharmakologie und Toxikologie, Medizinische Fakultät, Universität Zürich, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_11, © Springer Science+Business Media, LLC 2011

279

280

L. Pagani et al.

The most clearly illustrated example of this approach can be found in recent studies of the circadian clock, which is the molecular basis for human diurnal rhythms. This clock, which has a period of about 24 h – “circa diem” – is present in most cells of the body, and plays a key physiological role in the adaptation of living organisms to the alternation of day and night. In mammals, circadian rhythms influence nearly all aspects of physiology and behaviour, including sleep-wake cycles, cardiovascular activity, endocrinology, body temperature, renal activity, physiology of the gastro-intestinal tract, and hepatic metabolism (Gachon, Nagoshi, Brown, Ripperger, & Schibler, 2004). As we discuss in detail, although specific mutations in circadian clock components have profound effects upon behaviour, these effects are paralleled by changes in circadian gene expression in easily accessible tissues such as skin fibroblasts, and hence these cells can be used as both diagnostic and analytical tools in the study of circadian behaviour. This chapter is divided into three parts. In the first part, a general introduction to circadian clocks and their influence upon human behaviour is provided. The second part contains the protocols used by our laboratory to study circadian clock function in ­primary human cells. The chapter then closes with a discussion of future applications.

1.1 Basic Mechanisms of the Circadian Clock The idea that genes control diurnal behaviour in higher organisms was first established in the fruit fly by the demonstration that certain mutations cause altered ­circadian rhythms (Konopka and Benzer, 1971). The responsible gene, period (per), was subsequently isolated (Bargiello, Jackson, & Young, 1984) and was found to be expressed in a circadian pattern (Hardin, Hall, & Rosbash, 1990). This finding, and similar data from bacteria, fungi, and plants, has led to the general idea that periodically expressed genes constitute the physiological basis of circadian clocks in all living organisms (Dunlap, 1996; Hall and Rosbash, 1993; Takahashi, 1995). Moreover, because clocks exist in virtually all light-sensitive organisms, both multicellular and unicellular, it is not surprising that they are cell-autonomous. Their mechanism is based upon feedback loops of transcription, translation, and phosphorylation. In the simplest organism known to possess circadian rhythms, the cyanobacterium Synechococcus crassa, three genes – kaiA, kaiB, and kaiC – play essential roles in a central phosphorylation feedback loop, coupled to a transcriptional one (Ishiura et al., 1998; Nakajima et al., 2005). In Neurospora crassa, the hierarchy is reversed, with a transcription/translation feedback loop of Frequency and White Collar proteins (Froehlich, Loros, & Dunlap, 2003) probably playing a central role that is fine-tuned by phosphorylation by casein kinases 1 and 2 (He et al., 2003). Starting with metazoans, the proteins of the clock are highly conserved in sequence and in function. The same protein families that control the circadian clock in flies control the circadian clock in mammals, and polymorphisms in the genes that encode these proteins affect daily behaviour in human beings. Many reviews have been written about the molecular workings of the mammalian circadian oscillator (Ko and Takahashi, 2006), and we outline these mechanisms only

Studying Human Circadian Behaviour Using Peripheral Cells

281

briefly here. As in simpler organisms, the key mechanism is interconnected feedback loops of transcription and translation. It is conventional to divide clock components into a positive limb (consisting of transcriptional activators of clock genes) and a negative limb (inhibitors of these transcription factors). In mammals, the negative key components of the genetic circuitry are the four genes encoding cryptochrome 1 (Cry1) and cryptochrome 2 (Cry2) (Fig. 1, green rhombus), period 1 (Per1) and period 2 (Per2) (Fig. 1, violet ellipse). These genes are regulated by the two PAS domain basic helix-loop-helix transcription factors BMAL1 (Fig. 1, red octagon) and CLOCK (or in some tissues NPAS2) (Fig. 1, blue hexagon), the key positive components of the circadian oscillator. CLOCK and BMAL1 form a heterodimer that binds specific sequences present in Per and Cry promoters, the E-boxes, and activates the transcription of Per and Cry genes. Their gene products, PER and CRY proteins, in turn form heteropolymeric complexes of unknown stoichiometry that interact with the CLOCKBMAL1 heterodimer, leading to a block of Per and Cry expression. As a consequence, Cry and Per mRNAs and proteins decrease in concentration, and once the nuclear levels of the CRY–PER complexes are insufficient for auto-repression, a new cycle of Per and Cry transcription can start (Albrecht and Eichele, 2003; Reppert and Weaver, 2002). Additional interlocked feedback loops contribute to the robustness of this molecular clockwork circuitry. For example, the orphan nuclear receptor and repressor REVERBa (Fig. 1, yellow ellipse) interconnects circadian transcription of the positive and negative “limbs” of the oscillator. Rev-Erba transcription is activated by the CLOCKBMAL1 complex through the binding to E-box sequences present in its promoter, resulting in a circadian accumulation of REV-ERBa. REV-ERBa leads to periodic repression of Bmal1 transcription. In turn, this leads to a rhythmic expression of Bmal1 mRNA that is antiphasic to Rev-Erba expression (Preitner et al., 2002). Thus, indirectly, Bmal1 is positively regulated by PER and CRY proteins. Post-translational mechanisms such as protein phosphorylation also play important roles. For example, casein kinase 1e (CK1e) (Fig. 1, orange ellipse), initially identified as an essential Drosophila clock component (Price et al., 1998), phosphorylates PER, CRY, and BMAL1 proteins (Eide, Vielhaber, Hinz, & Virshup, 2002; Eide and Virshup, 2001; Lee, Weaver, & Reppert, 2004). CK1d (Fig.  1, orange ellipse), a close paralog of CK1e, has also been found to be associated with PER– CRY complexes and may therefore perform a similar function as CK1e (Lee et al., 2001). While some phosphorylations of PER proteins stabilize them, others destabilise them, leading to a complex balance of post-transcriptional regulation that helps to determine period length (Vanselow et al., 2006).

1.2 Central and Peripheral Oscillators Ablation and transplantation studies have firmly established that the suprachiasmatic nuclei (SCN) of the brain anterior hypothalamus are the site of the master circadian clock in mammals (Ralph, Foster, Davis, & Menaker, 1990). Surprisingly, recording of spontaneous action potentials from individual SCN neurons showed that the

282

L. Pagani et al. Dex shock

Period length

e.g. Rev-Erba:luc Time after pulse (hrs)

0

8

20

Increased expression of immediate early genes (e.g. per1, per2, c-fos)

Fig. 1  Schematic diagram of fibroblast synchronisation and period measurement. In each cell on a plate of cultured cells, a separate circadian clock ticks. The heterodimer of CLOCK (C, blue hexagon) and BMAL1 (B, red octagon) activates the transcription of period genes Per1-2 (P, ­violet ellipse), cryptochrome genes Cry1-2 (C, green rhombus) and Rev-Erba (R, yellow ellipse) through E-box enhancers. As the translated PER1-2 protein levels rise in the cytoplasm they ­multimerise with CRY1-2 protein and casein kinase e/d (CKe/d) (e/d, orange ellipse) and are phosphorylated (p, pink circle). The complex re-enters in the nucleus where it binds the dimer CLOCK-BMAL1, inhibiting its transcription activity. Meanwhile, REV-ERBa binds the ­Rev-Erb/ ROR responsive elements present in the Bmal1 promoter, repressing Bmal1 transcription. Since each cellular mechanism is unsynchronised with that of its neighbour, the population average of clock gene expression is constant (dashed line). The addition of dexamethasone results in ­synchronous activation of period genes (a), whose protein products multimerise with the more abundant cryptochrome proteins CRY1-2 (b) and are translocated to the nucleus (c). These ­complexes then inhibit the activation of circadian genes by CLOCK and BMAL1 heterodimers (d). The resultant repression of clock genes “resets” the clockwork and synchronises the oscillators of independent cells, allowing precise measurement of circadian period from population cultures (Adapted from Balsalobre et al., 1998.)

o­ scillatory machinery of the SCN is cell-autonomous as in simpler organisms (Welsh, Yoo, Liu, Takahashi, & Kay, 1995), and many laboratories therefore began looking for clock genes and other clock tissues.

Studying Human Circadian Behaviour Using Peripheral Cells

283

It was soon noted that these “clock genes” are expressed rhythmically outside the SCN (Sun et al., 1997), and even outside the brain (Zylka, Shearman, Weaver, & Reppert, 1998) in many peripheral tissues. The use of transgenic Drosophila that express green fluorescent protein under the control of the promoter of the clock gene period permitted researchers to follow in real-time the oscillation of the circadian rhythm in dissected peripheral tissues such as head, thorax, and abdominal tissues, demonstrating that autonomous circadian oscillators are present throughout the body of a fly (Plautz, Kaneko, Hall, & Kay, 1997). The same result was rapidly shown to be true for zebrafish and mammals, and even for immortalised fibroblasts (Balsalobre, Damiola, & Schibler, 1998; Whitmore, Foulkes, Strahle, & SassoneCorsi, 1998; Yamazaki et al., 2000).

1.3 Communication Between the SCN and Periphery The SCN uses humoral signals, neural efferents, and indirect physiological cues (­neural control of body temperature and feeding time) to convey timing information to other parts of the brain and the periphery. On an anatomical level, many hypothalamic nuclei receive SCN projections that mediate the circadian control of ­multiple ­systemic cues. For example, both direct and indirect projections from the SCN make synaptic contact with CRH (corticotropin-releasing hormone) neurons of the paraventricular nucleus (PVN). These neurons in turn impose rhythmic ACTH (adrenocorticotropic hormone) release from the pituitary and subsequent circadian corticosterone secretion from the adrenal glands. The sleep/arousal, reproductive, and endocrine systems are also regulated by the SCN in part through neuroanatomic connections. In addition, there exist SCN connections to all major organs via both sympathetic and parasympathetic pathways, and recent evidence suggests a critical role of the ANS (autonomic nervous system) in synchronising peripheral physiology (Buijs, van Eden, Goncharuk, & Kalsbeek, 2003). Interestingly, though, a transplanted SCN encapsulated in porous plastic is still able to direct circadian locomotor activity (Silver, LeSauter, Tresco, & Lehman, 1996), so at least this fundamental output of the circadian clock does not depend upon synaptic transmission directly from the SCN. Hormonal signals probably also play an important role in communication between the SCN and peripheral clocks. Glucocorticoid agonists can effectively shift peripheral clock gene expression in mice (Balsalobre, Brown, et  al., 2000). The precise role of other cyclically secreted hormones such as GH (growth hormone) and PRL (prolactin) in peripheral phase entrainment is yet to be examined. The SCN can also indirectly entrain peripheral oscillators by controlling daily activity/rest cycles, and consequently, feeding time (Damiola et al., 2000; Stokkan, Yamazaki, Tei, Sakaki, & Menaker, 2001). Hormones and metabolites related to feeding, such as insulin and glucose, might be involved in this phase setting (Hirota et al., 2002). Rhythmic body temperature – controlled by the SCN via the preoptic anterior hypothalamus – is also sufficient to entrain peripheral circadian oscillators in vivo and in vitro (Brown, Zumbrunn, Fleury-Olela, Preitner, & Schibler, 2002). Finally, activity rhythms as well as circadian outputs from other tissues can feed

284

L. Pagani et al.

back on the SCN and on other peripheral oscillators. For example, melatonin released from the pineal gland, which is itself under SCN control through autonomic pathways, functions as a circadian input signal for both the SCN and the anterior pituitary (Stehle, von Gall, & Korf, 2003).

1.4 Mutations in Clock Genes Affect Human Circadian Behaviour In organisms from mice to cyanobacteria, mutations in clock genes have furnished an excellent tool for studying circadian clocks. In human beings, polymorphisms located in known clock genes result in circadian pathologies. One of the most studied syndromes is familial advanced sleep-phase syndrome (FASPS). Individuals with this syndrome can wake up and go to sleep hours earlier than normal people. This phase change is believed to be related to a change in the endogenous freerunning period of the human circadian oscillator. Normally around 24 h, it has been measured to be only 20  h in an individual from an extensively studied FASPS ­lineage (Jones et  al., 1999). The source of the difference has been mapped to a change from serine to glycine at residue 662 of the Per2 gene. This mutation leads to a reduction in kinase activity of CK1e, resulting in a shortening of period length of the circadian rhythm (Toh et al., 2001). An unlinked mutation in casein kinase 1d also results in FASPS (Xu et al., 2005). Other polymorphisms in clock genes have been shown to be associated with a delayed sleep-phase syndrome (DSPS) (Archer et  al., 2003; Ebisawa et  al., 2001; Katzenberg et al., 1998; Pereira et al., 2005), though specific molecular mechanisms have not yet been identified. On a more general note, basic heritability studies of twins suggest that human daytime preference (“chronotype”) is 50% heritable (Koskenvuo, Hublin, Partinen, Heikkila, & Kaprio, 2007). Studies relating chronotype to circadian clock period length suggest that morning chronotype can result from a shorter period of the endogenous circadian oscillator, and evening chronotype from a longer clock period (Duffy, Rimmer, & Czeisler, 2001). Altogether, it appears that clock workings at a cellular and molecular level in different individuals can provide mechanistic explanations for differences observed at a behavioural level.

1.5 Interaction Between the Circadian Clock and Mood Disorders Many other diseases are also linked to sleep and circadian disturbances, For example, one striking feature of circadian rhythm sleep disorders is that they are often associated with other mood disorders. A part of this association is by definition: established clinical symptoms of diseases like major depressive disorder (MDD) and bipolar disorder (BD) are abnormal sleep/wake, appetite, and social rhythms (Boivin, 2000; Bunney and Bunney, 2000), which are also hallmarks of circadian rhythm disorders.

Studying Human Circadian Behaviour Using Peripheral Cells

285

Nevertheless, an increasing body of evidence suggests that there exists an interesting genetic basis for this correlation. Several circadian markers show abnormalities in depression (Avery, Wildschiodtz, & Rafaelsen, 1982; Duncan, 1996; Souetre et al., 1988, 1989; Szuba, Guze, & Baxter, 1997). One hypothesis for the involvement of circadian rhythm disturbances and development of depression is that, perhaps, depression may involve a weaker coupling process between internal pacemakers and involve abnormal sensitivity to environmental cues such as light (Souetre et  al., 1989). This could be a result of mutant clock genes or allelic variations leading to abnormal clock cycles or altered photosensitivity. In BD, a single nucleotide polymorphism in the 3¢ flanking region of the Clock-gene associates with a higher recurrence rate of bipolar episodes (Benedetti et al., 2003). This mutation is specific to bipolar depression: a similar association is not found in MDD (or unipolar depression) (Bailer et al., 2005). Another mutation, this time linked to the onset of illness in BD, has been localised to the glycogen synthase kinase 3b promoter (Benedetti et  al., 2005). This enzyme is the target of lithium, a common treatment for BD, and can phosphorylate the clock component REV ERBa (Yin, Wang, Klein, & Lazar, 2006). Seasonal affective disorder (winter depression, or SAD), is a syndrome characterised by recurrent depressions that occur at the same time every year (Rosenthal et al., 1984). Depressive phases are associated with hypersomnia, overeating, and carbohydrate craving. Abnormalities in circadian rhythms in SAD include sleep disturbances (Rosenthal et  al., 1984), and alteration of many circadian markers (Avery et  al., 1997; Dahl et  al., 1993; Lewy et  al., 1998; Lewy, Sack, Miller, & Hoban, 1987; Schwartz et al., 1997). Light therapy, sleep deprivation and phaseadvance treatment have been used either separately or in combination to treat SAD and depressive illness (Lewy et  al., 1998). Here, too, an understanding of the molecular workings of the circadian clock in affected individuals could furnish valuable insight into the causes of these disorders.

2 Measurement of Human Circadian Clocks Many physiological and biochemical parameters such as body temperature, melatonin, cortisol, heart rate, and blood pressure exhibit circadian rhythms. In principle, measurement of circadian parameters such as period length (the time of one complete clock cycle) or phase (the time of an organism’s clock within a given cycle) could be accomplished by measurement of any of these properties. However, endogenous circadian clocks in humans are masked by environmental factors. To dissect endogenous and exogenous components of a circadian rhythm, several protocols have been developed. These protocols include “constant routine,” in which subjects remain supine under constant environmental conditions, with small isocaloric meals and short naps at regular intervals (Duffy and Dijk, 2002), and “forced desynchrony,” in which subjects are in a dark–light environment so short or so long (e.g. 20  h period length or 28 h period length) that their endogenous clocks cannot adjust and

286

L. Pagani et al.

instead run freely (Blatter and Cajochen, 2007). Both these methods employ extensive subject observation under controlled laboratory conditions; hence, they are very expensive and labour-intensive. In addition, the significant time commitment required by subjects makes their recruitment problematic.

2.1 Reporter Assays in Peripheral Cells To circumvent the difficulties of direct measurement of circadian behaviour in humans, our laboratory has tried instead ex vivo analysis of clocks in tissues that can be obtained easily – skin, blood, or hair root cells. Evidence from rodent models suggests that clocks in peripheral tissues utilise similar or identical complements of basic clock proteins, and that mutations in clock genes that affect SCN function and locomotor behaviour also affect molecular oscillations in peripheral cells (Pando, Morse, Cermakian, & Sassone-Corsi, 2002; Yagita, Tamanini, van Der Horst, & Okamura, 2001). To measure clock function in peripheral human cells in high-throughput fashion, we have employed lentivirally-delivered circadian luciferase reporter vectors. By virtue of their easy preparation, wide host tropism, and stable integration into host DNA, lentiviral vectors are an excellent system to deliver reporters to hard-to-transfect primary cells (Salmon and Trono, 2007). Firefly luciferase is commonly used for the analysis of promoter function in eukaryotic systems (Alam and Cook, 1990). The enzyme catalyses the oxidative decarboxylation of beetle luciferin using O2, Mg2 and ATP as substrates. A photon is released at 560 nm in 90% of catalytic cycles: this light emission can be quantified in a luminometer. In mammalian cell cultures, the half-life of luciferase activity is approximately 3 h (Thompson, Hayes, & Lloyd, 1991), making it a convenient choice as a circadian reporter. Studies with transgenic mice containing the luciferase gene under control of clock promoters have shown that the molecular oscillations of clock gene expression reported in this fashion accurately recapitulate circadian transcription measured by conventional means in these tissues. Hence, clock gene: luciferase reporter systems have been widely used to measure expression of clock genes in prokaryotes, plants, Drosophila, and mammals (Brandes et al., 1996; Kondo et al., 1993; Millar, Carre, Strayer, Chua, & Kay, 1995; Millar, Short, Chua, & Kay, 1992).

2.2 Cell Type Considerations In principle, circadian reporter constructs could be introduced into any cell type. Practically, though, a limited range of primary tissues are available from human hosts. These include hair root cells, blood cells, and skin cells. Hair root cells include melanocytes and keratinocytes. Melanocytes are deep cells that seldom cling to the end of plucked hair. Since keratinocytes are present in the hair sheath,

Studying Human Circadian Behaviour Using Peripheral Cells

287

they are more easily obtained, but do not proliferate to high levels. Highly proliferative pre-keratinocytes are also present in the sheath, but are more difficult to cultivate. Blood monocytes are easily cultivated but have little amplificatory potential; and B- and T-cells can be amplified but have very weak endogenous clocks and are difficult to infect with lentiviral vectors. Because of their ease of manipulation, cultivation, and storage, our laboratory has mostly used adult dermal fibroblasts. Individual fibroblast cells are capable of functioning as independently phased circadian oscillators that are self-sustained for many days in vitro. They are also sensitive to a wide range of phase-shifting stimuli, including serum shock, glucocorticoids, and even medium changes (Balsalobre, Marcacci, & Schibler 2000). From these cells, a variety of properties can be measured: Period length:  By synchronising cultivated cells chemically (e.g. with dexamethasone) and then measuring their period under constant conditions, one can obtain a measure of the free-running period length of the circadian oscillator (t) (Fig. 1). A strong correlation has been found between in  vivo free-running behaviour from different mouse strains and period length of primary adult dermal fibroblasts from mouse tail (Brown et al., 2005). Hence, it appears possible to use peripheral oscillators as proxies for the clocks of the SCN. Recently, similar studies have been performed in humans, and fibroblasts from humans of extreme early chronotype show period lengths on average shorter than those from subjects of extreme late chronotype (Brown et al., 2008). Phase response:  To entrain to the daily light/dark cycle and respond optimally to changes in it, the circadian oscillator responds differently to light at different phases of its cycle. (Indeed, at some times of day – the so-called “dead zone” – the oscillator is completely insensitive to light.) This differential effect is most easily visualised as a phase-response curve (PRC), which plots the change in phase of the circadian clock as a function of the time that a stimulus is given. Under normal circumstances, fibroblasts do not respond to light. Nevertheless, the same signalling pathways that light uses to communicate to the clock in other cells – cAMP and calcium/cAMP-responsive element-binding protein (CREB) – are present. Thus, one can measure a PRC in response to an activator of cAMP (forskolin, which increases adenyl cyclase activity). Interestingly, the shape of the curve so obtained resembles the PRC of the human light response in vivo (Brown et al., 2008; Khalsa, Jewett, Cajochen, & Czeisler, 2003) (Fig. 2). By reintroducing a photopigment into fibroblast cells, one can even measure the sensitivity of the fibroblast clock directly to light (Pulivarthy et al., 2007). Phase:  After a few days of cultivation, fibroblast cells lose synchrony and phase is randomly distributed (Welsh et al., 2004). Since the methods described here involve amplification of cells for several days, timing information is therefore lost. For all experiments described here, a prior resynchronisation, usually with dexamethasone, is employed. Absolute phase information – i.e. the time at which the biopsy was taken, according to the biological clock – is thereby lost, but one can still ­measure the phase angle of normal entrainment for fibroblasts from any ­subject. In this case,

288

L. Pagani et al. f1

ldentical plates of cells, dex-synchronized

f2

Phase shifts

Addition of phase-shifting agent to individual plates at different times

Human PRC to bright light pulses

Phase shift (hours)

Phase shift (hours)

Human fibroblast PRC 2 1 0 −1 −2 −3 0

3

6

9 12 15 18 21

Time of light pulse

15 5 −5 30

40

50 hours

−15 Time of forskolin addition (after dex sync.)

Fig. 2  Diagram of a phase-shifting experiment. Identical plates of cells are synchronised with dexamethasone. After 1 day, a phase-shifting agent (in this case, forskolin) is added to each plate at a different time of day. The phase of the next circadian cycle is measured with respect to a control plate to which vehicle alone has been added. After all plates have been measured, the data can be plotted as a PRC (phase response curve) in which the magnitude of phase shift is shown for each addition time. The shape of this curve is very similar to that of a PRC of human physiology in response to bright light pulses (inset) (Adapted from Brown et al., 2008; Khalsa et al., 2003.)

fibroblasts are incubated in a shallow circadian temperature gradient (usually 12 h at 34°C and 12 h at 37°C) to mimic the daily entraining signals of the SCN. Cells will now adopt a period of 24 h, but the peak of reporter gene expression within this 24-h cycle will vary according to the period and amplitude of the underlying circadian oscillator (Brown et al., 2008). For example, the phase of fibroblasts carrying the FASPS mutation in Per2 have a phase of bioluminescence 6 h earlier than otherwise isogenic wild-type fibroblasts (Vanselow et al., 2006) in this protocol. Amplitude:  Because the magnitude of the reporter signals measured here is dependent upon the titer of virus used, amplitude is a property that cannot be directly measured at this time. It is possible, however, to cultivate identical plates of fibroblasts from the same subject, synchronise circadian oscillations with dexamethasone, and then harvest plates sequentially over a period of 24 h. From these cells, RNA or protein can be prepared to give a direct measure of the circadian amplitude of any gene or protein.

Studying Human Circadian Behaviour Using Peripheral Cells

289

3 Protocols 3.1 Overview of Requirements These protocols can be conducted by any laboratory comfortable with mammalian cell culture. Some special permissions, as well as the collaboration of a medical doctor, are usually required. Moreover, a real-time luminometry device capable of measuring bioluminescence in constant-temperature cell culture conditions is ­necessary to visualise results. Several commercial models exist (e.g. from Actimetrics, Hamamatsu, or Perkin-Elmer). Home-made devices are also feasible, and work equally well. 3.1.1 Ethical Considerations Collection of human fibroblasts by dermal punch biopsy is considered an invasive procedure in most countries, and therefore requires a medical doctor and prior approval of the Ethical Committee or Institute Review Board for the hospital or university in question. Typically, this permission is easily obtained for adults, but not necessarily for children. 3.1.2 Use of Viruses The self-inactivating lentiviral vectors used here are considered to be Biohazard Level II, which typically requires institutional notification and a separate cell ­culture facility. This classification applies only to the viral supernatants. Cell lines, once infected with virus, are considered to be only Biohazard Level I and may be cultivated normally. 3.1.3 Storage of Samples Fibroblast cells may be stored cryogenically as for any other immortalised cell line. We find that a storage medium consisting of 50% foetal bovine serum (FBS)/ Dulbecco’s Modified Eagle’s Medium supplemented with 4,500  mg/l glucose (DMEM)/7% DMSO works optimally. Viral supernatant at neutral pH may be stored indefinitely at −70°C, but has a half life of only a few hours at 37°C or room temperature, and a week at 4°C or on ice. Tissue biopsies may be stored in DMEM/50% FBS on ice for up to 5 days prior to fibroblast cultivation.

3.2 Producing the Reporter Virus The circadian measurement system described here relies upon lentiviral delivery of a bioluminescent reporter construct. We have equally used adenoviruses and

290

L. Pagani et al.

a­ deno-associated vectors, but favour the lentivirus for its stable integration into the host genome, allowing permanent storage of infected cells as stable lines. Since lentiviruses are amphotropic vectors (i.e. they infect most mammalian cell types including human ones), all production systems employ some strategy to prevent the generation of actively replicating virus and thereby minimise danger to the researcher. Our vectors are a so-called “second generation” system obtained from the laboratory of Didier Trono. (http://tronolab.epfl.ch/) One plasmid containing the vector itself is the only genetic material transferred to the target cells. It typically comprises the transgene cassette flanked by cis-acting elements necessary for its transcription, encapsidation, and integration. A second plasmid supplies all the genes necessary for packaging the virus, and a third supplies the coat protein, in our case the VSV-G coat protein for maximum host spectrum. All these plasmids are cotransfected into human embryonic kidney (HEK) cells immortalised by large T antigen (293T cells). Note that all HEK cells are not 293T cells, but can be immortalised by other means. The viral production plasmids contain an SV40 origin of replication, and will therefore amplify to high levels in 293T cells (but not in other HEK cells such as 293E) after transfection. 3.2.1 Calcium Phosphate Transfection in 293T Cells Typically, the transient transfection for viral production is done using calcium phosphate. In these cells, it is as efficient as lipophilic reagents, and certainly much cheaper. The procedure of virus production is 5 days long. The first day, amplified 293T cells must be at about 30–50% confluence in 10-cm Petri plates with 8 ml of medium. This medium is DMEM, penicillin-streptomycin-glutamine solution as recommended by the manufacturer, and 10% FBS. For each 10-cm dish, the following quantities of plasmids are transfected: 15 mg of viral reporter plasmid (e.g. pBluFpuro), 10 mg of packaging plasmid (e.g. psPAX2), and 6 mg of coat plasmid (e.g. pMD2G). This DNA is assembled in a total volume of 400 ml weakly buffered water (2 mM HEPES pH 7.05). Then, 100 ml of 2.5 M CaCl2, tissue culture grade, filter-sterilised, is added, and the mixture is allowed to equilibrate to room temperature. A second tube containing 500 ml of a solution of 2× HeBS (NaCl 0.283 M, HEPES 0.023 M, Na2HPO4 1.5 mM, pH 7.05) is also equilibrated to room temperature. The pH of HeBS is fundamental for high transfection efficiency: pH 7.05 is ideal, and pH outside the range 7.0–7.1 will result in significantly reduced efficiency. The plasmid and CaCl2 mix is dropwise added to the 2× HeBS under agitation. The combined solution is vortexed for 5 s, incubated for 30 min at room temperature, and added dropwise to 293T cells. The medium is mixed by rocking the plate back and forth, and cells are incubated overnight at 37°C, 5–7% CO2. The day after the transfection, 293T cells are washed with 5 ml of PBS 1× and then 10 ml of fresh medium enriched with 20 mM Hepes 7.8 is added to the cells to stabilise the pHlabile virus. Over the next 2 days, virus will be released into the culture medium. The following day, the virus-containing medium is harvested and stored on ice, and replaced with 10  ml of fresh medium enriched as before. The following day

Studying Human Circadian Behaviour Using Peripheral Cells

291

medium is harvested again from the cells. 293T plates are now discarded. ­Virus-containing medium is pooled and spun at 2G for 5 min to pellet debris, and then filtered through 0.4-mm filters to remove residual 293T cells. Virus can be concentrated through ultracentrifugation at 80,000g at 4°C for 90 min, ideally in conical-bottom tubes. The resulting very small, compact, whitishtranslucent virus pellet is resuspended overnight by medium-speed vortexing at 4°C. The virus is stored at −80°C. A concentration of virus between 7 and 10 times is ideal for a good cell infection. Higher titers can be toxic for primary cell cultures.

3.2.2 Optimising Virus Production Typically, we perform two tests to quantify virus production. In the first, we routinely transfect a plate of GFP-expressing vector in a duplicate plate in parallel with our larger preparations of virus. After 2 days, the efficiency of transient transfection can be analysed directly by fluorescence microscopy. About 50% is expected. The same plate can be left to produce virus, which is concentrated as usual. One aliquot of concentrated virus is then rediluted with medium to its original volume. Both of these aliquots are tested by infection of 293T cells (see below); 100% infection is expected in both cases. If desired, exact titers can be determined by serial dilution.

3.3 Processing of Skin Biopsies The punch biopsy is a fast, easy, inexpensive method to produce a cylinder of tissue from the skin surface to the underlying subcutaneous fat. A 2-mm round dermal punch is sufficient, and disposable needles for this purpose are commercially available. Typically, the punch biopsy is taken from the buttocks or upper arm of a subject. The area chosen for the sampling is an area that normally is not exposed to sunlight. The reasons for this choice are a reduced risk of mutagenesis caused by light rays to healing wounds and reduced aesthetical problems caused by the healing lesion for a few weeks after the sampling. No permanent scarring results. 3.3.1 Preparation of the Skin The skin should be prepared with chlorhexidine (Hibiclens), or povidone iodine (Betadine, Isodine), or alcohol. This is performed as a clean procedure, and full sterile technique with sterile drapes is unnecessary. Local anaesthesia with lidocaine can be used at patient request. One percent lidocaine with epinephrine is injected via a 30-gauge needle and Luer-Lok syringe, about 50 ml on each side of the planned biopsy site if desired.

292

L. Pagani et al.

3.3.2 Punch With the skin held taut at right angles to skin tension lines, the cutting edge of the biopsy needle is pressed down firmly, and simultaneously twisted one way and then the other in quick succession. As the punch traverses the dermis, resistance is lost, and there is a slight “give.” Typically, the punch will remain in the needle as it is removed. Sterile tweezers or a 26-gauge needle should be available to remove it. Occasionally, the biopsy may remain in the wound, and the same removal procedure applies. Bleeding is resolved with pressure, and no sutures are necessary for a 2-mm biopsy punch. A bandage is applied for a few days after the procedure. For every biopsy, a collection tube filled with medium must be prepared and kept on ice. The medium used to collect the biopsy is DMEM containing 50% FBS, plus a suitable concentration of penicillin, streptomycin, and fungicide such as Amphotericin B as recommended by the manufacturer. Tubes should be almost filled to facilitate immersion of biopsy, which adheres easily to dry surfaces. Biopsies in this ice cold solution can be kept for up to 5 days. Shipping of the biopsy can occur at this stage. 3.3.3 Biopsy Processing Biopsies are transferred in a 3.5-cm Petri dish for processing. Pour 2 ml of DMEM, 20% FBS, Amphotericin B, 200 ml of liberase in the dish with the biopsy. Liberase (Roche) is an enzyme blend of collagenase isoforms I and II from clostridium ­histoliticum and thermolysin from bacillus thermoproteolyticus; its action is to destroy partially the extracellular matrix, to allow the growth of fibroblasts outside the biopsy. Incubate the biopsy with the liberase-containing medium at 37°C, 5–7% CO2 for 4–10 h. At the end of digestion, multiple clumps containing tens to hundreds of cells are present, though the original biopsy has not entirely dissolved. The biopsy is transferred to 10 ml of PBS 1× and centrifuged for 5 min at 200g. The biopsy pellet is resuspended in 200 ml of DMEM, 20% FBS, penicillin-streptomycin+Amphotericin B and transferred back in the middle of the surface of the 3.5-cm petri dish. To maintain the biopsy attached to the bottom of the plate, a home-modified Millicell CM membrane disc (Millipore) from which spacer feet have been removed is placed on the biopsy. The membrane disc is covered with 1.5–1.8 ml of the same medium and incubated overnight. The following day the medium is changed. After a few days, the first fibroblasts start to grow around the biopsy. After 1 week of cultivation of the biopsy, it is not necessary to add amphotericin B to the medium. Seven to ten days after processing if the biopsy is attached to the bottom of the plate and the membrane disc can be carefully removed. The biopsy requires changing of medium every 3–4 days. Length of culture before the first harvest of fibroblasts will vary with the number of viable fibroblast foci. Cells are ready to be harvested when the combined volume of their foci is about half the volume of the plate to which they will be amplified. Trypsinise and replate cells normally. It should be noted that biopsy cultivation can lead initially to the growth of more than one kind of cells. Keratinocytes can grow from the biopsy; however, they will be overrun by the faster-growing fibroblasts.

Studying Human Circadian Behaviour Using Peripheral Cells

293

3.3.4 Splitting Procedure To split the cells, medium is removed and cells are washed once with PBS 1×. About 0.4 ml of trypsin is enough to detach fibroblasts that grow in a 3.5-cm Petri plate. The detachment of the cells can be followed under the microscope. The trypsin reaction should not be prolonged because human primary fibroblasts can be damaged. Normally, trypsinisation of fibroblasts takes about 1 min. After this time, the plate is vigorously shaken to promote the detachment of the cells. Then, 2 ml of serum-containing medium are added to block the reaction of trypsin. Cells are usually split 1:2 during each amplification passage. 3.3.5 Freezing Procedure Trypsinise fibroblasts as described in the splitting procedure. For every plate of fibroblasts a 15-ml polypropylene tube filled with 10  ml of PBS 1× is prepared. Cells are detached from the bottom of the plate and transferred in the tube. Centrifugation for 5 min at 200g is necessary to pellet fibroblasts. The supernatant is removed and cells are suspended in 1.5 ml of DMEM, 50% FBS, 7% DMSO. Cells are transferred into an ice-cold cryovial. The vial is left at least 20 min at −20°C and then placed at −80°C overnight. The following day, cells can be definitively stored in liquid nitrogen. Human primary fibroblasts can be stored in liquid nitrogen for years. 3.3.6 Thawing Process For every vial of cells to be thawed, a 15-ml polypropylene tube with 10 ml of 37°C warmed medium must be prepared. Unfreeze fibroblasts quickly, placing 0.5 ml of 37°C warmed medium in the cryovial and transferring the medium into the prepared tube. Repeat this operation until all the cells have been thawed and moved into the tube. Centrifuge the tube for 5 min at 200g to pellet the cells. Remove the ­supernatant and resuspend the cells in 2 ml growth medium (as above) and plate into the same size of plate as originally used for freezing.

3.4 Measurement of Circadian Oscillations from Skin Biopsies 3.4.1 Fibroblast Infection Even though lentivirus can infect non-replicating cells, the infection efficiency is increased when the cells are in mitosis. On the day of the infection, fibroblasts should be about 30–50% confluent. For a 3.5-cm Petri plate, it is enough to use about 750 ml of warmed concentrated virus prepared as above. To increase the

294

L. Pagani et al.

transfection efficiency, 8 mg/ml protamine sulphate is added to the virus aliquot as a polycation to block non-specific virus binding to the cell surface. Fibroblasts can be incubated overnight, though the infection is complete in about 6 h. The following day, the virus-containing medium is removed from the cells. Fibroblasts are rinsed with PBS 1× and standard DMEM high-glucose medium containing 20% FBS and penicillin-streptomycin is added to the cells. Three days after the viral transfection antibiotic selection of transfected cells can be started, if desired.

3.4.2 Measurement of Circadian Period Length When transfected fibroblasts have reached 90–100% confluence, the measurement of the circadian rhythm can be started. It is necessary to synchronise fibroblast rhythms. Commonly used methods are serum shock or a pulse of the glucocorticoid agonist dexamethasone. To induce synchronisation with serum shock, cells are incubated with DMEM and 50% FBS for 1  h. In the case of synchronisation through dexamethasone, fibroblasts are treated for 15 min with DMEM, 20% FBS, 100 nM dexamethasone. Cells are washed with PBS 1× to remove the presence of synchronisation agents and medium is replaced. The medium that is used to ­measure the circadian rhythm is DMEM without phenol red, 10% FBS, penicillinstreptomycin, and 0.1  mM luciferin. If evaporation is a problem, plates can be sealed with parafilm prior to placement in the measurement machine. In this case, medium should be buffered with 20  mM HEPES pH 7.6. Normally, circadian ­oscillations can be observed for 5–6 days. (The induced circadian oscillations are shown in Fig. 1.)

3.4.3 Phase Response Experiments For these experiments, duplicate plates – one for each timepoint desired in the phase response curve – are synchronised with dexamethasone as above. Plates are measured together for 1 day, and then the phase shifting agent is added at the desired time to one plate (experiment) and vehicle to the other (control). It is important that all reagents be prewarmed and that the procedure be done quickly, because temperature is also a phase shifting agent (Fig. 2). 3.4.4 Phase Entrainment Experiments For these experiments, it is necessary to change incubator temperature between 34 and 37°C every 12 h for 6 days. (One can purchase cell culture incubators that will do this automatically.) Fibroblast cells are placed into this regime, and on the seventh day, temperature is kept constant at 37°C and bioluminescence is measured.

Studying Human Circadian Behaviour Using Peripheral Cells

295

Because the electrical noise and quantum efficiency of the photomultiplier tube is itself temperature-sensitive, the temperature of the incubator should not be varied during the day of measurement unless a special insulating measurement device is constructed. 3.4.5 Data Analyses To analyse the circadian rhythm of human primary fibroblasts, either commercial software (e.g. Lumicycle Analysis, from Actimetrics), or Matlab routines (available from Hanspeter Herzel, Humboldt University, Berlin) can be used. Both programmes begin by normalising data to a 24-h running average. To determine period, the programmes then find the period and time constant that best fit the data to a sine wave multiplied by an exponential decay C = A sin(2π t / P − t0 )* e − t /τ . Empirically, we generally discard the first 18  h of data because noncircadian fluctuations associated with the synchronisation protocol make automatic sine fitting more problematic. To determine phase shifts during a PRC experiment, the timing of the peak of reporter expression after substance addition in experiment and control samples are manually determined and compared to each other. The phase response curve contains the magnitude of these differences on one axis, and the time of ­substance addition (relative to dexamethasone synchronisation) on the other axis. For phase entrainment experiments, the peak of reporter gene expression is reported relative to the onset of the 34:37 temperature cycle.

3.5 Troubleshooting Magnitude of signal:  Here, problems are usually related to viral titer. We recommend routinely testing transfection efficiency and viral titer at the time of viral production by using a GFP-expressing control as described above. Low viral titer:  Viral plasmids are large, and occasionally rearrange during bacterial expression. Also, calcium phosphate transfection is highly sensitive to pH, and this parameter should be regularly verified. Finally, 293T cells occasionally lose their ability to amplify viral plasmids, and should not be continuously cultivated. Unexplained cell death, slow growth, or senescence: Primary cell cultures are an excellent source of mycoplasma. New cultures should be tested prior to storage. Unusual period length or period length variance:  Since the circadian clock is temperature-overcompensated, increasing incubator temperature results in longer period lengths. (Even 0.5°C causes a period change of 30–60 min.) Hence, it is important that the incubator temperature be maintained rigidly constant for all except phase

296

L. Pagani et al.

entrainment experiments. Abnormally high or low concentrations of serum also alter period length, but this effect is relatively minor within standard ranges (5–20%).

4 Discussion The protocols used above measure circadian gene expression in adult dermal fibroblasts. In general, in both humans and mice, we have observed a correlation between the period length of this gene expression and behavioural phase (in humans) and the period of free-running locomotor activity (in mice). An important question to discuss, though, is how good this correlation is (or isn’t). At a cellular level, period length in both the SCN and peripheral tissues is ­stochastic: individual SCN neurons in culture and fibroblasts in culture express circadian rhythms of widely different phases and different period length (Nagoshi et  al., 2004; Welsh et  al., 1995). Luciferase-based measurement of fibroblasts as above yields a population-based measurement of the added luminescence of all cells in the plate. These differences are therefore not apparent in the overall average. At a population level, in most cases the mean of the circadian period length of SCN neurons or cultured fibroblasts is very similar to the mean of the period length of behavioural rhythms in the same strain of mouse (Liu, Weaver, Strogatz, & Reppert, 1997; Yagita et al., 2001). However, for individual cells, cell-to-cell variance is much greater in fibroblasts and dissociated SCN neurons than it is in intact SCN slices. A possible explanation of this difference is that coupling occurs in SCN tissue that is lacking in dissociated neurons or in fibroblasts in culture. In this way, aberrances in individual cells are “corrected” by input from neighbours. Mathematical modelling of simple oscillator systems suggests that such mechanisms can have major stabilising effects. As a result, with many clock gene mutations in mice, period length in fibroblasts or in dissociated SCN neurons is often more perturbed than in either intact brain slices or in animal behaviour. For periods within “normal” ranges, however, very good correlations are usually observed (Pagani et al., submitted). Coupling of SCN neurons occurs in multiple ways: conventional synapses, likely GABAergic (Wagner, Castel, Gainer, & Yarom, 1997); electrical synapses (i.e. gap junctions) (Colwell, 2000); and neuropeptidergic coupling via vasoactive intestinal peptide (VIP) and its receptor, VPAC2 (Colwell et al., 2003; Cutler et al., 2003). Recent evidence suggests that neuropeptidergic signalling plays a dominant role, since mutations that impair it result in severely decreased precision similar to what is observed in dissociated SCN neurons, and amplify the severity of point mutations in clock genes (Liu et al., 2007; Maywood et al., 2006). As we have presented it, the primary utility of peripheral circadian clocks is to screen more easily for circadian differences between sets of human subjects. Hence, practically speaking, the differences mentioned above could even help: the effects of any given clock variations would be expected to be larger as observed in peripheral cells than as observed in behaviour.

Studying Human Circadian Behaviour Using Peripheral Cells

297

4.1 Relationship Between the Circadian Oscillator and Other Pathologies A cellular method such as the one presented above, in which fundamental clock properties could be determined easily and precisely, could potentially permit an unbiased dissection of the relationship of circadian clock defects to unrelated pathologies such as depressive and affective disorders. As discussed in the Introduction, these maladies are frequently accompanied by severe alterations in the timing of sleep episodes. Whether these are directly linked to changes in the circadian clock is difficult to establish by conventional means. By comparing cellular clock properties, it would be possible to find out if genetic alterations in the circadian clock are risk factors for these diseases.

4.2 Ex Vivo Models Another big advantage of cellular circadian models is the possibility to manipulate the human genome. It is possible in fact to re-create the genetic mutation of a disease to investigate the molecular mechanisms that determine certain behaviour. Vanselow and colleagues introduced a mutation in Per2, believed to be responsible for human FASPS, into fibroblasts, and were able to recapitulate the phase advance in the behaviour of FASPS patients as an advanced phase of clockgene transcription in synchronised FASPS fibroblasts. Subsequent molecular analyses allowed them to show effects of this mutation upon phosphorylation at multiple sites in the PER2 protein, and to further demonstrate that these modifications affected both PER2 protein stability and nuclear localisation (Vanselow et al., 2006). Similarly, although fibroblasts are not themselves photosensitive, introduction of photopigments renders their circadian clocks sensitive to light. Exogenously supplied melanopsin acts as a sensory photopigment in fibroblasts and other cell types, where it probably signals via a native G-protein signalling cascade to activate the CREB (Melyan, Tarttelin, Bellingham, Lucas, & Hankins, 2005; Panda et al., 2005; Peirson and Foster, 2006; Qiu et al., 2005). Transcriptionally active phospho-CREB then binds to Per1 and Per2 promoter CRE sites and activates transcription, thereby altering the phase of the molecular oscillator (Lee et al., 2001; Meijer and Schwartz, 2003).

4.3 Future Potential Both for the circadian clock and for other systems, it is increasingly apparent that specific molecular defects cause a wide range of alterations in human

298

L. Pagani et al.

behaviour and physiology. Genetic studies have linked specific diseases to a wide range of cellular signalling pathways, and many of these are – like the circadian oscillator – conserved at a cellular level. Thus, in principle, they could be accessible to the same types of cellular analyses discussed above. Given the difficulties that revolve around the collection of human tissues, ex vivo analyses such as these could play an important future role in determining how molecular differences at an individual level contribute to variations in many aspects of physiology and behaviour.

References Alam, J., & Cook, J. L. (1990). Reporter genes: Application to the study of mammalian gene transcription. Analytical Biochemistry, 188, 245–254. Albrecht, U., & Eichele, G. (2003). The mammalian circadian clock. Current Opinion in Genetics & Development, 13, 271–277. Archer, S. N., Robilliard, D. L., Skene, D. J., Smits, M., Williams, A., Arendt, J., et  al. (2003). A  length polymorphism in the circadian clock gene Per3 is linked to delayed sleep phase ­syndrome and extreme diurnal preference. Sleep, 26, 413–415. Avery, D. H., Dahl, K., Savage, M. V., Brengelmann, G. L., Larsen, L. H., Kenny, M. A., et al. (1997). Circadian temperature and cortisol rhythms during a constant routine are phase-delayed in hypersomnic winter depression. Biological Psychiatry, 41, 1109–1123. Avery, D. H., Wildschiodtz, G., & Rafaelsen, O. J. (1982). Nocturnal temperature in affective disorder. Journal of Affective Disorders, 4, 61–71. Bailer, U., Wiesegger, G., Leisch, F., Fuchs, K., Leitner, I., Letmaier, M., et al. (2005). No association of clock gene T3111C polymorphism and affective disorders. European Neuropsycho­ pharmacology, 15, 51–55. Balsalobre, A., Brown, S. A , Marcacci, L., Tronche, F., Kellendonk, C., Reichardt, H. M., et al. (2000). Resetting of circadian time in peripheral tissues by glucocorticoid signaling. Science, 289, 2344–2347. Balsalobre, A., Damiola, F., & Schibler, U. (1998). A serum shock induces circadian gene expression in mammalian tissue culture cells. Cell, 93, 929–937. Balsalobre, A., Marcacci, L., & Schibler, U. (2000). Multiple signaling pathways elicit circadian gene expression in cultured Rat-1 fibroblasts. Current Biology, 10, 1291–1294. Bargiello, T. A., Jackson, F. R., & Young, M. W. (1984). Restoration of circadian behavioural rhythms by gene transfer in Drosophila. Nature, 312, 752–754. Benedetti, F., Serretti, A., Colombo, C., Barbini, B., Lorenzi, C., Campori, E., et  al. (2003). Influence of CLOCK gene polymorphism on circadian mood fluctuation and illness recurrence in bipolar depression. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 123B, 23–26. Benedetti, F., Serretti, A., Pontiggia, A., Bernasconi, A., Lorenzi, C., Colombo, C., et al. (2005). Long-term response to lithium salts in bipolar illness is influenced by the glycogen synthase kinase 3-beta -50 T/C SNP. Neuroscience Letters, 376, 51–55. Blatter, K., & Cajochen, C. (2007). Circadian rhythms in cognitive performance: Methodological constraints, protocols, theoretical underpinnings. Physiology & Behavior, 90, 196–208. Boivin, D. B. (2000). Influence of sleep-wake and circadian rhythm disturbances in psychiatric disorders. Journal of Psychiatry & Neuroscience, 25, 446–458. Brandes, C., Plautz, J. D., Stanewsky, R., Jamison, C. F., Straume, M., Wood, K. V., et al. (1996). Novel features of drosophila period Transcription revealed by real-time luciferase reporting. Neuron, 16, 687–692. Brown, S. A., Fleury-Olela, F., Nagoshi, E., Hauser, C., Juge, C., Meier, C. A., et al. (2005). The period length of fibroblast circadian gene expression varies widely among human individuals. PLoS Biology, 3, e338.

Studying Human Circadian Behaviour Using Peripheral Cells

299

Brown, S. A., Kunz, D., Dumas, A., Westermark, P. O., Vanselow, K., Tilmann-Wahnschaffe, A., et  al. (2008). Molecular insights into human daily behavior. Proceedings of the National Academy of Sciences of the United States of America, 105, 1602–1607. Brown, S. A., Zumbrunn, G., Fleury-Olela, F , Preitner, N., & Schibler, U. (2002). Rhythms of mammalian body temperature can sustain peripheral circadian clocks. Current Biology, 12, 1574–1583. Buijs, R. M., van Eden, C. G., Goncharuk, V. D., & Kalsbeek, A. (2003). The biological clock tunes the organs of the body: Timing by hormones and the autonomic nervous system. The Journal of Endocrinology, 177, 17–26. Bunney, W. E., & Bunney, B. G. (2000). Molecular clock genes in man and lower animals: Possible implications for circadian abnormalities in depression. Neuropsychopharmacology, 22, 335–345. Colwell, C. S. (2000). Rhythmic coupling among cells in the suprachiasmatic nucleus. Journal of Neurobiology, 43, 379–388. Colwell, C. S., Michel, S., Itri, J., Rodriguez, W., Tam, J., Lelievre, V., et al. (2003). Disrupted circadian rhythms in VIP- and PHI-deficient mice. American Journal of Physiology: Regulatory, Integrative and Comparative Physiology, 285, R939–R949. Cutler, D. J., Haraura, M., Reed, H. E., Shen, S., Sheward, W. J., Morrison, C. F., et al. (2003). The mouse VPAC2 receptor confers suprachiasmatic nuclei cellular rhythmicity and responsiveness to vasoactive intestinal polypeptide in vitro. The European Journal of Neuroscience, 17, 197–204. Dahl, K., Avery, D. H., Lewy, A. J., Savage, M. V., Brengelmann, G. L., Larsen, L. H., et  al. (1993). Dim light melatonin onset and circadian temperature during a constant routine in hypersomnic winter depression. Acta Psychiatrica Scandinavica, 88, 60–66. Damiola, F., Le Minh, N., Preitner, N., Kornmann, B., Fleury-Olela, F., & Schibler, U. (2000). Restricted feeding uncouples circadian oscillators in peripheral tissues from the central pacemaker in the suprachiasmatic nucleus. Genes & Development, 14, 2950–2961. Duffy, J. F., & Dijk, D. J. (2002). Getting through to circadian oscillators: Why use constant routines? Journal of Biological Rhythms, 17, 4–13. Duffy, J. F., Rimmer, D. W., & Czeisler, C. A. (2001). Association of intrinsic circadian period with morningness-eveningness, usual wake time, and circadian phase. Behavioral Neuroscience, 115, 895–899. Duncan, W. C., Jr. (1996). Circadian rhythms and the pharmacology of affective illness. Pharmacology & Therapeutics, 71, 253–312. Dunlap, J. C. (1996). Genetics and molecular analysis of circadian rhythms. Annual Review of Genetics, 30, 579–601. Ebisawa, T., Uchiyama, M., Kajimura, N., Mishima, K., Kamei, Y., Katoh, M., et  al. (2001). Association of structural polymorphisms in the human period3 gene with delayed sleep phase syndrome. EMBO Reports, 2, 342–346. Eide, E. J., Vielhaber, E. L., Hinz, W. A., & Virshup, D. M. (2002). The circadian regulatory proteins BMAL1 and cryptochromes are substrates of casein kinase Iepsilon. The Journal of Biological Chemistry, 277, 17248–17254. Eide, E. J., & Virshup, D. M. (2001). Casein kinase I: Another cog in the circadian clockworks. Chronobiology International, 18, 389–398. Froehlich, A. C., Loros, J. J., & Dunlap, J. C. (2003). Rhythmic binding of a WHITE COLLARcontaining complex to the frequency promoter is inhibited by FREQUENCY. Proceedings of the National Academy of Sciences of the United States of America, 100, 5914–5919. Gachon, F., Nagoshi, E., Brown, S. A., Ripperger, J., & Schibler, U. (2004). The mammalian circadian timing system: From gene expression to physiology. Chromosoma, 113, 103–112. Hall, J. C., & Rosbash, M. (1993). Oscillating molecules and how they move circadian clocks across evolutionary boundaries. Proceedings of the National Academy of Sciences of the United States of America, 90, 5382–5383. Hardin, P. E., Hall, J. C., & Rosbash, M. (1990). Feedback of the Drosophila period gene product on circadian cycling of its messenger RNA levels. Nature, 343, 536–540. He, Q., Cheng, P., Yang, Y., He, Q., Yu, H., & Liu, Y. (2003). FWD1-mediated degradation of FREQUENCY in Neurospora establishes a conserved mechanism for circadian clock regulation. The EMBO Journal, 22, 4421–4430.

300

L. Pagani et al.

Hirota, T., Okano, T., Kokame, K., Shirotani-Ikejima, H., Miyata, T., & Fukada, Y. (2002). Glucose down-regulates Per1 and Per2 mRNA levels and induces circadian gene expression in cultured Rat-1 fibroblasts. The Journal of Biological Chemistry, 277, 44244–44251. Ishiura, M., Kutsuna, S., Aoki, S., Iwasaki, H., Andersson, C. R., Tanabe, A., et  al. (1998). Expression of a gene cluster kaiABC as a circadian feedback process in cyanobacteria. Science, 281, 1519–1523. Jones, C. R., Campbell, S. S., Zone, S. E., Cooper, F., DeSano, A., Murphy, P. J., et al. (1999). Familial advanced sleep-phase syndrome: A short-period circadian rhythm variant in humans. Natural Medicines, 5, 1062–1065. Katzenberg, D., Young, T., Finn, L., Lin, L., King, D. P., Takahashi, J. S., et al. (1998). A CLOCK polymorphism associated with human diurnal preference. Sleep, 21, 569–576. Khalsa, S. B., Jewett, M. E., Cajochen, C., & Czeisler, C. A. (2003). A phase response curve to single bright light pulses in human subjects. Journal of Physiology, 549, 945–952. Ko, C. H., & Takahashi, J. S. (2006). Molecular components of the mammalian circadian clock. Human Molecular Genetics, 15(Spec No 2), R271–277. Kondo, T., Strayer, C. A., Kulkarni, R. D., Taylor, W., Ishiura, M., Golden, S. S., et al. (1993). Circadian rhythms in prokaryotes: Luciferase as a reporter of circadian gene expression in cyanobacteria. Proceedings of the National Academy of Sciences of the United States of America, 90, 5672–5676. Konopka, R. J., & Benzer, S. (1971). Clock mutants of Drosophila melanogaster. Proceedings of the National Academy of Sciences of the United States of America, 68, 2112–2116. Koskenvuo, M., Hublin, C., Partinen, M., Heikkila, K., & Kaprio, J. (2007). Heritability of diurnal type: A nationwide study of 8753 adult twin pairs. Journal of Sleep Research, 16, 156–162. Lee, C., Etchegaray, J. P., Cagampang, F. R., Loudon, A. S., & Reppert, S. M. (2001). Posttranslational mechanisms regulate the mammalian circadian clock. Cell, 107, 855–867. Lee, C., Weaver, D. R., & Reppert, S. M. (2004). Direct association between mouse PERIOD and CKIepsilon is critical for a functioning circadian clock. Molecular and Cellular Biology, 24, 584–594. Lewy, A. J., Bauer, V. K., Cutler, N. L., Sack, R. L., Ahmed, S., Thomas, K. H., et al. (1998). Morning vs evening light treatment of patients with winter depression. Archives of General Psychiatry, 55, 890–896. Lewy, A. J., Sack, R. L., Miller, L. S., & Hoban, T. M. (1987). Antidepressant and circadian phaseshifting effects of light. Science, 235, 352–354. Liu, A. C., Welsh, D. K., Ko, C. H., Tran, H. G., Zhang, E. E., Priest, A. A., et  al. (2007). Intercellular coupling confers robustness against mutations in the SCN circadian clock network. Cell, 129, 605–616. Liu, C., Weaver, D. R., Strogatz, S. H., & Reppert, S. M. (1997). Cellular construction of a circadian clock: Period determination in the suprachiasmatic nuclei. Cell, 91, 855–860. Maywood, E. S., Reddy, A. B., Wong, G. K., O’Neill, J. S., O’Brien, J. A., McMahon, D. G., et al. (2006). Synchronization and maintenance of timekeeping in suprachiasmatic circadian clock cells by neuropeptidergic signaling. Current Biology, 16, 599–605. Meijer, J. H., & Schwartz, W. J. (2003). In search of the pathways for light-induced pacemaker resetting in the suprachiasmatic nucleus. Journal of Biological Rhythms, 18, 235–249. Melyan, Z., Tarttelin, E. E., Bellingham, J., Lucas, R. J., & Hankins, M. W. (2005). Addition of human melanopsin renders mammalian cells photoresponsive. Nature, 433, 741–745. Millar, A. J., Carre, I. A., Strayer, C. A., Chua, N. H., & Kay, S. A. (1995). Circadian clock mutants in Arabidopsis identified by luciferase imaging. Science, 267, 1161–1163. Millar, A. J., Short, S. R., Chua, N. H., & Kay, S. A. (1992). A novel circadian phenotype based on firefly luciferase expression in transgenic plants. The Plant Cell, 4, 1075–1087. Nagoshi, E., Saini, C., Bauer, C., Laroche, T., Naef, F., & Schibler, U. (2004). Circadian gene expression in individual fibroblasts: Cell-autonomous and self-sustained oscillators pass time to daughter cells. Cell, 119, 693–705.

Studying Human Circadian Behaviour Using Peripheral Cells

301

Nakajima, M., Imai, K., Ito, H., Nishiwaki, T., Murayama, Y., Iwasaki, H., et  al. (2005). Reconstitution of circadian oscillation of cyanobacterial KaiC phosphorylation in  vitro. Science, 308, 414–415. Panda, S., Nayak, S. K., Campo, B., Walker, J. R., Hogenesch, J. B., & Jegla, T. (2005). Illumination of the melanopsin signaling pathway. Science, 307, 600–604. Pando, M. P., Morse, D., Cermakian, N., & Sassone-Corsi, P. (2002). Phenotypic rescue of a peripheral clock genetic defect via SCN hierarchical dominance. Cell, 110, 107–117. Peirson, S., & Foster, R. G. (2006). Melanopsin: Another way of signaling light. Neuron, 49, 331–339. Pereira, D. S., Tufik, S., Louzada, F. M., Benedito-Silva, A. A., Lopez, A. R., Lemos, N. A., et al. (2005). Association of the length polymorphism in the human Per3 gene with the delayed sleep-phase syndrome: Does latitude have an influence upon it? Sleep, 28, 29–32. Plautz, J. D., Kaneko, M., Hall, J. C., & Kay, S. A. (1997). Independent photoreceptive circadian clocks throughout Drosophila. Science, 278, 1632–1635. Preitner, N., Damiola, F., Lopez-Molina, L., Zakany, J., Duboule, D., Albrecht, U., et al. (2002). The orphan nuclear receptor REV-ERBalpha controls circadian transcription within the positive limb of the mammalian circadian oscillator. Cell, 110, 251–260. Price, J. L., Blau, J., Rothenfluh, A., Abodeely, M., Kloss, B., & Young, M. W. (1998). double-time is a novel Drosophila clock gene that regulates PERIOD protein accumulation. Cell, 94, 83–95. Pulivarthy, S. R., Tanaka, N., Welsh, D. K., De Haro, L., Verma, I. M., & Panda, S. (2007). Reciprocity between phase shifts and amplitude changes in the mammalian circadian clock. Proceedings of the National Academy of Sciences of the United States of America, 104, 20356–20361. Qiu, X., Kumbalasiri, T., Carlson, S. M., Wong, K. Y., Krishna, V., Provencio, I., et  al. (2005). Induction of photosensitivity by heterologous expression of melanopsin. Nature, 433, 745–749. Ralph, M. R., Foster, R. G., Davis, F. C., & Menaker, M. (1990). Transplanted suprachiasmatic nucleus determines circadian period. Science, 247, 975–978. Reppert, S. M., & Weaver, D. R. (2002). Coordination of circadian timing in mammals. Nature, 418, 935–941. Rosenthal, N. E., Sack, D. A., Gillin, J. C., Lewy, A. J., Goodwin, F. K., Davenport, Y., et  al. (1984). Seasonal affective disorder. A description of the syndrome and preliminary findings with light therapy. Archives of General Psychiatry, 41, 72–80. Salmon, P., & Trono, D. (2007). Production and titration of lentiviral vectors. Current Protocols in Human Genetics, 54, 12.10.1–12.10.24 Schwartz, P. J., Rosenthal, N. E., Turner, E. H., Drake, C. L., Liberty, V., & Wehr, T. A. (1997). Seasonal variation in core temperature regulation during sleep in patients with winter seasonal affective disorder. Biological Psychiatry, 42, 122–131. Silver, R., LeSauter, J., Tresco, P. A., & Lehman, M. N. (1996). A diffusible coupling signal from the transplanted suprachiasmatic nucleus controlling circadian locomotor rhythms. Nature, 382, 810–813. Souetre, E., Salvati, E., Belugou, J. L., Pringuey, D., Candito, M., Krebs, B., et  al. (1989). Circadian rhythms in depression and recovery: Evidence for blunted amplitude as the main chronobiological abnormality. Psychiatry Research, 28, 263–278. Souetre, E., Salvati, E., Wehr, T. A., Sack, D. A., Krebs, B., & Darcourt, G. (1988). Twenty-four-hour profiles of body temperature and plasma TSH in bipolar patients during depression and during remission and in normal control subjects. The American Journal of Psychiatry, 145, 1133–1137. Stehle, J. H., von Gall, C., & Korf, H. W. (2003). Melatonin: A clock-output, a clock-input. Journal of Neuroendocrinology, 15, 383–389. Stokkan, K. A., Yamazaki, S., Tei, H., Sakaki, Y., & Menaker, M. (2001). Entrainment of the circadian clock in the liver by feeding. Science, 291, 490–493. Sun, Z. S., Albrecht, U., Zhuchenko, O., Bailey, J., Eichele, G., & Lee, C. C. (1997). RIGUI, a putative mammalian ortholog of the Drosophila period gene. Cell, 90, 1003–1011. Szuba, M. P., Guze, B. H., & Baxter, L. R., Jr. (1997). Electroconvulsive therapy increases circadian amplitude and lowers core body temperature in depressed subjects. Biological Psychiatry, 42, 1130–1137.

302

L. Pagani et al.

Takahashi, J. S. (1995). Molecular neurobiology and genetics of circadian rhythms in mammals. Annual Review of Neuroscience, 18, 531–553. Thompson, J. F., Hayes, L. S., & Lloyd, D. B. (1991). Modulation of firefly luciferase stability and impact on studies of gene regulation. Gene, 103, 171–177. Toh, K. L., Jones, C. R., He, Y., Eide, E. J., Hinz, W. A., Virshup, D. M., et al. (2001). An hPer2 phosphorylation site mutation in familial advanced sleep phase syndrome. Science, 291, 1040–1043. Vanselow, K., Vanselow, J. T., Westermark, P. O., Reischl, S., Maier, B., Korte, T., et al. (2006). Differential effects of PER2 phosphorylation: Molecular basis for the human familial advanced sleep phase syndrome (FASPS). Genes & Development, 20, 2660–2672. Wagner, S., Castel, M., Gainer, H., & Yarom, Y. (1997). GABA in the mammalian suprachiasmatic nucleus and its role in diurnal rhythmicity. Nature, 387, 598–603. Welsh, D. K., Logothetis, D. E., Meister, M., & Reppert, S. M. (1995). Individual neurons dissociated from rat suprachiasmatic nucleus express independently phased circadian firing rhythms. Neuron, 14, 697–706. Welsh, D. K., Yoo, S. H., Liu, A. C., Takahashi, J. S., & Kay, S. A. (2004). Bioluminescence imaging of individual fibroblasts reveals persistent, independently phased circadian rhythms of clock gene expression. Current Biology, 14, 2289–2295. Whitmore, D., Foulkes, N. S., Strahle, U., & Sassone-Corsi, P. (1998). Zebrafish clock rhythmic expression reveals independent peripheral circadian oscillators. Nature Neuroscience, 1, 701–707. Xu, Y., Padiath, Q. S., Shapiro, R. E., Jones, C. R., Wu, S. C., Saigoh, N., et al. (2005). Functional consequences of a CKIdelta mutation causing familial advanced sleep phase syndrome. Nature, 434, 640–644. Yagita, K., Tamanini, F., van Der Horst, G. T., & Okamura, H. (2001). Molecular mechanisms of the biological clock in cultured fibroblasts. Science, 292, 278–281. Yamazaki, S., Numano, R., Abe, M., Hida, A., Takahashi, R., Ueda, M., et al. (2000). Resetting central and peripheral circadian oscillators in transgenic rats. Science, 288, 682–685. Yin, L., Wang, J., Klein, P. S., & Lazar, M. A. (2006). Nuclear receptor Rev-erbalpha is a critical lithium-sensitive component of the circadian clock. Science, 311, 1002–1005. Zylka, M. J., Shearman, L. P., Weaver, D. R., & Reppert, S. M. (1998). Three period homologs in mammals: Differential light responses in the suprachiasmatic circadian clock and oscillating transcripts outside of brain. Neuron, 20, 1103–1110.

Genome-Wide Expression Profiles of Amygdala and Hippocampus in Mice After Fear Conditioning Zheng Zhao and Yinghe Hu

Abstract  The amygdala and hippocampus are known to be involved in the formation of fear conditioning memories, and both contextual and cued fear memory requires activation of the NMDA receptors. However, the global molecular responses to fear conditioning in the amygdala and hippocampus are still poorly understood. We have systematically analyzed the gene expression profiles in these two brain regions of mice after fear conditioning treatment using high-density microarrays containing 11,000 genes and expressed sequence tags. A total of 222 genes in the amygdala and 145 genes in the hippocampus exhibit dynamic changes in their expression levels. The overall patterns of gene expression as well as the individual genes are drastically different in amygdala and hippocampus. However, a number of genes display similar regulatory responses in both brain regions. Based on the expression kinetics, the genes can be further grouped into eight and six unique clusters in amygdale and hippocampus, respectively. Our gene expression analysis demonstrates that different genomic responses are initiated in the amygdala and hippocampus, two brain regions that play distinct roles in associative memory formation. Keywords  High-density microarrays • Fear memory • Associative learning • Gene function • GluR1 Abbreviations AMPA Amino-3-hydroxy-5-methyl-4-isoxalone propionic acid APP Amyloid precursor protein

Y. Hu (*) Key Laboratory of Brain Functional Genomics, MOE & STCSM, Shanghai Institute of Brain Functional Genomics, East China Normal University, 3663 Zhongshan Road (N), Shanghai, 200062, China e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_12, © Springer Science+Business Media, LLC 2011

303

304

Z. Zhao and Y. Hu

ARF ADP-ribosylation factor CaMKII Calcium-calmodulin-kinase II CCT Chaperonin-containing TCP CS Conditioned stimulus Cx30 Connexin-30 ESTs Expressed sequence tags FXR1 Fragile-X-related gene 1 GABA Gamma-aminobutyric acid GABARAP GABA receptor-associated protein GAPDH Glyceraldehyde-3-phosphate dehydrogenase GluR1 Glutamate receptor 1 IAP Integrin-associated protein KLC1 Kinesin light chain 1 MAP Microtubule-associated protein MOBP Myelin-associated oligodendrocytic basic protein NF1 Neurofibromatosis type 1 NMDA N-methyl-d-aspartate NP25 Neuronal protein 25 NSG1 Neuron-specific gene family member 1 OSP Oligodendrocyte-specific protein PIP5K Phosphatidylinositol-4-phosphate-5-kinase PKC Protein kinase C PLD Phospholipase D PLP Proteolipid protein PP Protein phosphatases PP1 Protein phosphatases 1 PP2A Protein phosphatases 2A Stat3 Signal transducer and activator of transcription US Unconditioned stimulus VAMP Vesicle-associated membrane protein

1 Introduction The completion of the human genome sequence marks an exciting new era for studies of gene functions in biological and pathological processes. However, understanding how the genome functions as a whole in the complex behavioral situation presents a great challenge. It has been suggested that knowing when and where genes are expressed and regulated may provide crucial clues as to the molecular and cellular mechanisms of a given behavior. Furthermore, the patterns of gene expression in the brain could provide valuable information for understanding the molecular mechanisms of learning and memory. Recent progress in gene chip or DNA microarray technology makes it possible to monitor large-scale gene expression

Genome-Wide Expression Profiles of Amygdala and Hippocampus

305

regulated by a behavioral learning paradigm. This approach allows us to examine molecular and genetic mechanisms underlying many behavioral and cognitive changes during brain aging (Hu & Tsien, 2002). Emotional memories tend to be long lasting and play an important role in regulating the behavioral responses in both animals and humans. Most of the current knowledge of emotional memories comes from the study of classical fear conditioning. It is believed that contextual fear conditioning is dependent on the structural integrity of the hippocampus and amygdala, whereas cued fear conditioning is hippocampal-independent, but requires the amygdala (Falls, Miserendino, & Davis, 1992; Rodrigues, Schafe, & LeDoux, 2001; Rodrigues, Schafe, & LeDoux, 2004). Pharmacological and genetic experiments further suggest that the formation of many associative memories including contextual and cued fear memory requires the activation of NMDA receptors (Sara, 2000; Tang et  al., 1999; Tsien, 2000). After giving a brief description on the basic concepts of classical fear conditioning as well as DNA microarray technology, this chapter will focus mainly on the most recent advances in the application of gene chip technique for the analysis of the large-scale transcriptional responses that occur in the amygdala and hippocampus after fear conditioning.

2 Fear Conditioning Fear conditioning is one form of Pavlovian conditioning (Pavlov, 1927) and represents an essential component of many defensive behavior systems in ­mammals (Fanselow, 1994). It is nowadays still a powerful laboratorial approach to study the anatomical, cellular and molecular bases of the formation of ­emotional memories in the brain (Davis, 1997; Fendt & Fanselow, 1999; LeDoux, 2000; Rodrigues et  al., 2004). It states that an emotionally innocuous, conditioned stimulus (CS) (e.g., a tone or a light), when paired with a noxious, unconditioned stimulus (US) (e.g., a footshock) can elicit unconditioned ­emotional fear reactions. With the CS-US association formation, various innate physiological and behavioral fear responses, such as defensive behaviors, ­autonomic arousal, hypoalgesia, reflex potentiation and stress hormones, come under the control of the CS (LeDoux, 2000). In this regard, a human conditioned fear experiment, commonly referred to as “Little Albert and the White Rat,” by Watson and Rayner (1920) is perhaps the most frequently quoted example to demonstrate that fears can be learned (conditioned) through ordinary processes involving CS-US association (Watson et  al., 1920). Fear conditioning works throughout mammalian species and humans (LeDoux, 2000), it can occur very rapidly (one CS-US pairing is sufficient under some circumstances), and ­persists for a long period of time. Functioning as an interface of memory and emotion, fear conditioning has become a popular behavioral tool for ­understanding the neural substrates of associative learning and memory in mammals (Davis, 1997; LeDoux, 2000; Maren, 2001).

306

Z. Zhao and Y. Hu

2.1 Fear Conditioning and Amygdala A large body of evidence based on neurotoxic lesion, pharmacological, genetic and neurophysiological studies have accumulated indicating that there is an essential neuroanatomical circuit for fear learning and behavior that is centered on the amygdala. These studies show that amygdala plays essential roles in regulating the acquisition, storage, and expression of fear memory (Blair, Schafe, Bauer, Rodrigues, & LeDoux, 2001; Fanselow & LeDoux, 1999; Fendt & Fanselow, 1999; Maren, 2001; Pare, Quirk, & Ledoux, 2004; Walker & Davis, 2002). The amygdala consists of several anatomically and functionally distinct regions, each of which can be further classified to several subareas. Among them, the most relevance to the pathways of fear conditioning are the lateral (L), basal lateral (B), accessory basal (AB), and central (C) nuclei (Maren, 2001). The pathway via which the amygdala manipulates the formation of fear ­memory (CS-US association formation) involves acquisition of a CS, such as an auditory or sensory input, from various regions of the brain (e.g., thalamus, cortex, hippocampus), projection of the CS to the basolateral amygdaloid complex ­(containing L, B, and AB nuclei) and termination of the input in C nucleus of the amygdala (Aggleton, 2000; LeDoux, 2000; Pare et  al., 2004). It is generally accepted that plasticity in this basolateral amygdaloid complex mediates the ­formation and ­storage of a CS-US associative learning and projects this ­information to the C nucleus of the amygdala (Gale et  al., 2004). The C nucleus of the amygdala is the main region thought to output the projection to hypothalamic and brainstem regions that are believed to trigger the defensive and autonomic ­emotional responses to fear (Aggleton, 2000; Fendt & Fanselow, 1999; LeDoux, 2000; Maren, 2001). In contrast, studies in small animals such as rats show that fear responses can be triggered by the CS as well as by the contextual cues upon the CS-US association. For example, rats show themselves as being fear-conditioned when returned to a compartment where the pairing of the CS and the US takes place, or a compartment in which USs occur alone (Antoniadis & McDonald, 2001; LeDoux, 2000; Phillips & LeDoux, 1992). This is called contextual fear conditioning and involves the interactions between the hippocampus and B and AB nuclei of the amygdala. As mentioned above for CS conditioning, the C nucleus of the amygdala controls the expression of the fear responses (Fanselow & LeDoux, 1999; Frankland, Cestari, Filipkowski, McDonald, & Silva, 1998; Kim & Jung, 2006; LeDoux, 2000; Maren, Aharonov, & Fanselow, 1997).

2.2 Fear Conditioning and Hippocampus In addition to it being well recognized that the amygdala generally serves as a ­critical locus of plasticity for the formation and storage of associative learning, a number of findings suggest that the hippocampus is also normally involved in the

Genome-Wide Expression Profiles of Amygdala and Hippocampus

307

encoding and retrieval of specific types of information (e.g., spatial and episodic memories). It has been demonstrated that electrolytic lesions of the dorsal ­hippocampus (DH) in rats selectively interfere with the acquisition of context fear memory but not tone fear memory (Kim & Fanselow, 1992; Maren et  al., 1997; Phillips & LeDoux, 1992; Selden, Everitt, Jarrard, & Robbins, 1991). Similar observations have been made using a within-subjects design and neurotoxic hippocampal lesions, showing that the hippocampus plays a time-limited role in the consolidation of contextual fear conditioning (Anagnostaras, Maren, & Fanselow, 1999; Kim & Fanselow, 1992; Maren et al., 1997). Many analogous results from transgenic or knockout-based studies are also supportive of the role of the ­hippocampus in contextual fear conditioning (Abeliovich et al., 1993; Bourtchuladze et al., 1994; Huerta, Sun, Wilson, & Tonegawa, 2000; Sara, 2000; Tang et al., 1999; Tsien, 2000; Tsien, Huerta, & Tonegawa, 1996).

3 DNA Microarrays DNA microarrays represent a unique and high-throughput technology for largescale identification of alterations in gene expression in various cell and tissue types. The microarray-based technologies are dominated by two common platforms: cDNA arrays (also known as spotted arrays) (Cirelli & Tononi, 1999; Schena, Shalon, Davis, & Brown, 1995; Wang, Gan et al., 1999) and Affymetrix oligonucleotide arrays (also referred to as genechips) (Lockhart et al., 1996). In the Affymetrix approach, the hybridization reactions are performed on separate arrays using multiple probes for each gene. The multiple probes are designed to be paired by a perfect-match (PM) probe and a corresponding mismatch (MM) probe. The sequence of the MM oligonucleotide is the same as that of the corresponding PM, except that it contains a single mismatched base in a central position (Lipshutz, Fodor, Gingeras, & Lockhart, 1999). The use of multiple probe pairs in oligonucleotide arrays makes the technique greatly improved in the selectivity and specificity in identifying close members of a gene family (or different splice variants of a gene) from each other, and in accuracy in outputting quantitative expression ­information (Luo & Geschwind, 2001). We have applied high-density oligonucleotide microarrays with more than 11,000 mouse transcripts to analyze gene expression in the cortex of mice after exposure to enriched environments (Rampon et al., 2000). The study reveals that the expression of a large number of genes is altered in response to enrichment ­training, including genes associated with neuronal structure, synaptic plasticity, and signal transduction. Further studies should be able to reveal whether these genes play important roles in modulating learning and memory processes. In another study, we have used the same Affymetrix high-density gene chip approach to examine age-related changes in gene expression in the hypothalamus and cortex of young and aged mice (Jiang, Tsien, Schultz, & Hu, 2001). The results show that a number of key genes involved in neuronal structure and

308

Z. Zhao and Y. Hu

s­ ignaling, such as synaptotagmin I, cAMP-dependent protein kinase C (PKC) b, apolipoprotein E and protein phosphatase 2A, are differentially expressed in both the aged hypothalamus and cortex. The differential regulation of these proteins during the aging process may contribute to age-related memory deficits and ­neurodegenerative diseases. Furthermore, the expression of many proteases, known to be critical in regulating neuropeptide metabolism, amyloid precursor protein (APP) processing, and neuronal apoptosis, are also found to be ­up-­regulated in the aged brain. Interestingly, a subset of these genes whose expression is affected by aging are oppositely regulated by exposure of mice to an enriched ­environment (Rampon et al., 2000).

4 Gene Expression Profiling in Amygdala and Hippocampus After Fear Conditioning 4.1 Methodologies 4.1.1 Animals and Fear Conditioning Training C57BL6/CBAF1 adult mice (5-month-old) were housed in an environment of 23 ± 0.5°C with a relative humidity of 50 ± 10%. Every cage had a complete exchange of air 15–18 times per hour and a 12-h light–dark cycle with no twilight. Water and food were continuously available. A single CS-US pairing paradigm were used to create contextual and cued fear memories in mice (Rampon et  al., 2000; Tang et  al., 1999). Before training, the experimental mice were individually handled for 1 week followed by adaptation to the chamber for 1 day. The experiments were performed for 5 min per session and three sessions in total. For the fear conditioning training, an 85  dB sound at 2,800  Hz and a continuous scrambled foot shock at 0.75  mA were used as the CS-US pair, with a constant tape recording of radio noise (68  dB) as the background white noise. During a single training session for both contextual and cued conditioning, the mice were put individually into the chamber and allowed to explore the environment freely for 3 min. After that, the animals were exposed to the CS for 30 s. At the last 2 s of the CS, the US was delivered for 2 s. After the CS-US conditioning, the mice were allowed to stay in the chamber for another 30 s and then returned to their home cages immediately. Both short-term and long-term contextual fear conditioning retentions were first determined at 1 and 24  h after training by placing separate groups of animals (10 mice/time point) into the same shock chamber for 5 min. The cued fear memory retention was then tested for the same mice using a different chamber. The formation of fear memory (CS-US association formation) was measured as a percentage of freezing responses, and a statistical analysis of the differences in the responses between untrained and conditioned animals was performed to assess if the fear conditioning in the animals

Genome-Wide Expression Profiles of Amygdala and Hippocampus

309

succeeded (Fig.  1) (Mei et  al., 2005). All data from the subsequent experiments including gene expression, real-time RT-PCR and immunohistochemistry were manipulated by comparison and statistical analysis between the trained and untrained control mice. 4.1.2 Tissue Collection and RNA Extraction For gene expression profiling studies, the amygdalas and hippocampi were ­dissected and collected at the time points of 0.5, 6 and 24 h after the conditioning (10–14 trained mice per time point) followed by separating them into two ­independent tissue pools (5–7 mice per pool) for the extraction of poly(A) mRNA and probe labeling. The pooled tissues were immediately frozen in liquid ­nitrogen, and then stored at −80°C. Total RNA was isolated from tissue using an RNA Extraction Kit (Amersham Pharmacia). Briefly, 60–120  mg of tissue was manually homogenized in 2 ml of prewarmed extraction buffer. The homogenate was centrifuged for 5  min at room temperature to remove cellular debris. The supernatant was transferred to a fresh sterile tube and then sheared by passing through a 23-gauge needle and syringe several times. This homogenate was ­layered over cesium trifluoroacetates, and then centrifuged overnight at 125,000g at 15°C. Following centrifugation, the supernatant was aspirated, and the RNA pellet at the bottom of the tube was resuspended, followed by ethanol ­precipitation. RNA concentration was determined using spectrophotometer at 260 nm and the samples were stored at −80°C. The samples from the untrained control mice were handled in the same manner. 70 60

Contextual Memory

*

Freezing (%)

50 40

Cued Memory

1 h Retention 24 h Retention

*

** *

30 20 10 0

Immediate Contextual Recall

Pre-Tone

Cued Recall

Fig. 1  Formations of short-term (1 h) and long-term (1 day) contextual and cued fear memories in the trained mice compared to the untrained control mice. Memory was measured as percentage of freezing responses. In both contextual fear memory retention and cued fear memory retention, the mice exhibit robust freezing responses when tested at both 1 and 24 h after training

310

Z. Zhao and Y. Hu

4.1.3 Gene Expression Analysis The high-density oligonucleotide microarray (gene chip) analysis was conducted as previously described (Mody et  al., 2001; Rampon et  al., 2000). Double-stranded DNA was synthesized from 5  mg of total RNA using the SuperScript Choice System (Life Technologies, Rockville, MD). In vitro transcription was carried out with double-stranded cDNA as a template in the presence of biotinylated UTP and CTP using Enzo BioArray High Yield RNA Transcript Labeling Kit (Enzo Diagnostics). Biotin-labeled cRNA was purified, fragmented, and hybridized to the arrays in 1× Mes buffer (100  mM Mes, pH 7.4/1  M NaCl/20  mM EDTA/0.01% Tween 20). The arrays were washed and stained with streptavidin-phycoerythrin and then scanned with an Affymetrix GeneArray Scanner. Mu11KsubA and Mu11KsubB (GeneChip, Affymetrix, Santa Clara, CA) containing 13,069 probe sets corresponding to more than 11,000 genes and expressed sequence tags (ESTs) were used for the experiments. Data were analyzed with the Affymetrix GeneChip Expression Analysis Software (version 3.1) as described by Lipshutz et al. (1999). Genes with fold change ³2 and p £ 0.01 in ANOVA analysis between conditioned and control mice were extracted for advanced analysis. The extracted genes were further annotated and categorized by GENMAPP and MAPPFINDER software packages (www.genmapp.org). The data were loaded and clustered by GENECLUSTER 1.0 (MIT, Cambridge, MA), and then visualized by Treeview (Eisen, Spellman, Brown, & Botstein, 1998). The functional classification was based on the Gene Ontology (GO) EASE software (Hosack, Dennis, Sherman, Lane, & Lempicki, 2003). In order to ensure the reliability of the data, we performed the hybridization experiments in duplicates consisting of two independent mRNAs and two sets of duplicate microarrays (Mody et al., 2001; Rampon et al., 2000). The separate data from these two sets of microarrays were analyzed and only commonly regulated genes were chosen for further analysis. 4.1.4 Real-Time RT-PCR and Immunohistochemistry Although microarray technology has been extensively used for the studies on genome-wide expression analysis, considerable false positive and false negative results exist (Lee, Kuo, Whitmore, & Sklar, 2000). Duplicate hybridization of the same sample on separate arrays can significantly improve the reliability of ­expression data by ensuring the reproducibility of the methodology. It is widely accepted that the alterations in gene expression identified via microarray must be confirmed and validated by follow-up studies such as real-time quantitative RT-PCR, northern blot analysis, western blot analysis and immunohistochemistry (Hakak et al., 2001; Lee et al., 2000; Livesey, Furukawa, Steffen, Church, & Cepko, 2000; Nisenbaum, 2002; Pongrac, Middleton, Lewis, Levitt, & Mirnics, 2002; Sandberg et al., 2000; Thibault et al., 2000; Wada, Tifft, & Proia, 2000; Wang et al., 2006; Yoshikawa et al., 2000). Furthermore, confirmation of the DNA microarray transcriptome profiling results at the protein level could provide ­important information­ for the functional analysis in the system under investigation (Ho et al., 2001; Wada et al., 2000).

Genome-Wide Expression Profiles of Amygdala and Hippocampus

311

Validation of selected genes was performed using a TaqMan real-time RT-PCR approach. Total RNA was extracted and reverse-transcribed using the SuperScript First-Strand Synthesis System (Life Technologies, Invitrogen, Carlsbad, CA). A house-keeping gene (e.g., glyceraldehyde-3-phosphate dehydrogenase [GAPDH]) was used as the endogenous control. The sequences of primers and TaqMan probes of these selected genes and GAPDH are listed in Table 1. The PCR reaction was carried out using 50 ml of total reaction mixture volume and 2 ml of cDNA reaction products at 95°C for 10 min, with 40 cycles of 95°C for 30 s and 55°C for 1 min. All TaqMan probes (Perkin-Elmer) were labeled with the fluorescent dyes 6-­carboxyfluorescein (FAM) and 6-carboxy-tetramethyl rhodamine (TAMRA). The TaqMan reaction buffer consisted of 5.5 mM MgCl2; 200 nM each of dATP, dCTP, and dGTP; 400 nM dUTP, 0.5 U of uracyl DNA glycosylase, and 1.25 U of AmpliTaq. TaqMan probe concentrations were maintained at 100 nM, while PCR primer concentrations were systematically varied in all combinations. The fold change in expression levels of the targets genes was calculated using the 2−DDCt methods (Livak & Schmittgen, 2001). Mice were anesthetized with avertin (0.2 ml per 10 g body weight, i.p.) 6 and 24 h after fear conditioning as previously described (Rampon et al., 2000). Brains were removed and fixed in 4% paraformaldehyde in 0.1 M PBS for 3 h and then cryoprotected in 30% sucrose in 0.1 M PBS. Twenty-five mM free-floating slices through amygdala were obtained using a Leica cryostat (Germany). After blocking in 0.1 M PBS containing 3% normal goat serum, slices were incubated at 4°C overnight in 0.1 M PBS containing 3% goat serum, 0.25% Triton X-100, and 0.25 mg/ ml rabbit anti-glutamate receptor 1 (CHEMCOM, Temecula, CA). The samples were incubated in 0.1  M PBS containing 1:200 biotinylated goat anti-rabbit IgG (Vector Laboratories, Burlingame, CA) at room temperature for 1  h. Slices were then incubated with avidin/biotinylated enzyme complex for 5  min followed by incubation with nickel-3-3 diamino benzidine (nickel-DAB). Finally, the samples were mounted on gelatinized slides, stained with cresyl violet, and coverslipped. The staining intensities in each of the amygdala nuclei were measured using ImagePro (Media Cybernetics, Silver Spring, MD). The faint signal in corpus ­collosum was used as the internal background for subtraction as well as for ­standardization between cross-sections and cross-animal comparisons.

4.2 Distinct Gene Expression Patterns Induced by Fear Conditioning in Amygdala and Hippocampus The dynamical changes of gene expression in the amygdale and hippocampus between fear conditioned and no conditioning mice were compared at three time points (0.5, 6, and 24 h) (Mei et al., 2005). The results of the gene chip analysis revealed that the expressions of 222 genes in the amygdala and 145 genes in the hippocampus from the fear-conditioned mice were consistently and reproducibly changed by more than twofold as compared with the control mice. Some of the representative genes are summarized in Tables 2–4.

Enkephalin GABAR-AP Calmodulin synthesis

Clathrin

MAP4

NCSP F3 CCT NAP-22

Pantophysin Synaptotagmin MOBP

PKC-d U2AF65 PLC-a CaMKII 14-3-3

AATTTCCTGGCGTGCACACT GCGCCCTTCCTGCTTGT TGCCATGAGCAGGTTTTCTTT

CCAGAAGATGCAAGCTT TTGTTT TGACCATGCAAGACTGTCAATG

CAAAGGCCGCTTCGAACTC CATCACCCCAATGCAGTACAAG GGGTCATGATGGTGGCAAAG CTACGGTGGCATCCATGATG GGTGGCCTACAAAAAC GTGGTA TGCATTGCTGCCCTTCTG CAAAAGTCCACCGGAAAACC ACGAGCACTCAGGATAGG CTTT TTATGCTCGACGCCAATGG ACTTGCTTGTGGTGGGATAGCT GGGAGGGAGGCGTTTGA

CTTGCAGGTCTCCCAGATTTTG GTGGAGGCTCTGGCAGAGTAA GGTACTCGGACACTATTTTTT TGTACTG

GGAGAGCGCTGTGATCCAAT

CCGGCCTGCGTGTTTC CCCTGCATGTCCCAAACAG AATACTGCTTTCATAAAGATC TGATTGC CATCACGGTGCTCGTTAAAGG

TGGGCAGTTTGCGACTATCC CCGAGTATGGCACCTTGAAAG GTAGCCCATTGGTGAGACAAATT

TGCAGTCCGCAGATGATCTC GAAGGGCAGTGGCTGGAAT GTTTTACGGCTAGCTACAGCAAAGT CCGGGCGTTGAATTTGC CTTCTGCTCAATGCTCGAGATC

Table 1  Primers and probes used in the real-time RT-PCR experiments Gene name Forward primer Reverse primer GAPDH CTTCACCACCATGGAGAAGGC GGCATGGACTGTGGTCATGAG NF1 CCCTTGTTCTCAGTGGGATGA TGGTAGAGTAAATGCCGGG Vimentin CAGATCGATGTGGACGTTTCC ATACTGCTGGCGCGCACATCA APP CGTCCGTGCCGAACAGA GGTCCACCATGCGCACAT

CAGCCTCAGCAAGTGTTAGATAC TGACCAGG CGACAAAGACAACACTAATCA GCATATCCT AATGTGAAGGACAGCTGCCTTCTT TCTCTTGTAAATAACTGGCTGTTCTCA AGTGGCTAAACAAAGTTTAAAAAGC AAGTAACA

AACTGCTCATCCGAAATGCGCAGC TTCCTTTGATGACCTGAATCCTG TCTGCCACAGGACCAGTGCCA

TCTATGTTGGTTACACGAACCTCTACC TCAATCCAGTCTTCAATGAACAGTTT TGGATTGGATGAAGCTAGCCTGG

TaqMan probe CCTGGCCAAGGTCATCCATGACAACTTT TGGCTCGGGTTCTGGTCACTCTGTT AGCCTGACCTCACTGCTGCCC AGACAGACAGCACACCCCTAAAC CATTTTG ACCGGGCTACGTTTTATGCAGC CCATGCAAGCTGCGGGTCA AATTCCTTGATGCTGGACACAAACTC ATCGCCAAGAGACGGTAGAGTGCTT CCGCAGTCCGCCTGGAGGG

312 Z. Zhao and Y. Hu

−3.15   3.5

  2.25   4.05

  4.05

−2.2

Genes showing opposite regulation U60150 Vesicle-associated membrane protein VAMP-2 X61435 Kinesin heavy chain U07950 GDP-dissociation inhibitor X61434 cAMP-dependent protein kinase C b-subunit W57404 Carboxypeptidase H precursor L04961 Nuclear-localized inactive X-specific Transcript AA388848 Splicing factor, arginine/serine-rich 5 L13171 Myocyte-specific enhancer factor 2

−3.4

Genes showing similar regulation, but with different time onset AA059550 Ectonucleotide pyrophosphatase/phosphodiesterase 2 X79082 Kinase 1 X59728 Gas5 growth arrest specific protein AA408185 Splicing factor, arginine/serine-rich 7 AA655109 Ribosomal protein S3   3.1   4.35   3.55   3.15

 3   2.55 −3.6 −3.3 −3 −2.5

Fold change Amygdala (h) 0.5 6

Accession number Gene name Genes showing the same regulation X51438 Vimentin M13366 Glycerophosphate dehydrogenase M27844 Calmodulin M73329 Phospholipase C-alpha AA655109 Eukaryotic translation initiation factor 3 L04280 Ribosomal protein (Rpl12)

Table 2  Common set of genes regulated in both the amygdala and hippocampus

8.5 3.2 2.5 3.55 5.65 4.25

24

−3.3 −2.2 −2.1

 2  2  2

−2.0

 2

−2.7 −2.5 −2.9 −2.6 −2.4

−2

 3   2.55 −3.6 −3.3 −3 −2.5

Hippocampus (h) 0.5 6

−10.7

  −5

  −3.2

  2

24

Genome-Wide Expression Profiles of Amygdala and Hippocampus 313

314

Z. Zhao and Y. Hu

Table 3  Representative genes changed in the hippocampus after fear conditioning Fold change (h) Accession Cluster number Gene name 0.5 6 I L25274 Glycoprotein (DM-GRASP)   2 ET63017 Cadherin-8   2.8 U56649 Phosphodiesterase (PDE1A2)   2.6 II D30730 GTPase-activating protein (NF1)   5.4 X51438 Vimentin   1.9 X60304 Protein kinase C-d   6.2 X64587 Splicing factor U2AF65   4.8 III X56007 Adhesion molecule on glia (AMOG)   −2.7 −3.2 AA015415 Kinesin light chain 1 (KLC1)   −5.8 −9.7 AA066354 JAK1 protein tyrosine kinase   −6.8 −6.3 AA002629 Calcineurin B   −3 −4.4 AA108330 Astrocytic phosphoprotein PEA-15   −2.3 −3.2 AA000227 Diacylglycerol kinase −19.6 −7.6 IV M73329 Phospholipase C-a (PLC-a)   2   2.3 V AA107895 Cathepsin D   −2.8 AA008502 Neuron specific gene family member 1 −11.9 VI AA066335 Amyloid precursor protein (APP) −2.5 D21165 Visinin-like Ca2+-binding protein −2.2 M27844 Calmodulin synthesis (CaM) −2.2 W12204 Ca2+/calmodulin-dependent protein −3.2 kinase II isom g-b (CaMKII) X61434 Protein kinase C b subunit −2.6 AA218341 Protein phosphatase type 1-a (PP1) Z67745 Protein phosphatase 2A (PP2A) −2 W46019 Protein kinase regulator 14-3-3 −2.8 M63436 GABAA receptor alpha-1 subunit −2.2 D87898 ADP-ribosylation factor (ARF1) −2.1 AA067362 Vesicle protein pantophysin −3.8 D37792 SynaptotagminI/65 −3.2 U58886 Endophilin I −3.3 AA118297 Neuronal protein 25 (NP25) −2.6 X16314 Glutamine synthetase None U60150 Vesicle-associated membrane protein 2   −3.3 −2.7

24

  2.6 −3

The overall patterns of gene expression as well as the individual genes for the amygdala and hippocampus were drastically different in the conditioned mice ­compared to no conditioning (Fig.  2). Surprisingly, only six genes (Table  2) showed the same up- or down-regulation at the same time point(s) in both regions. Another five genes showed similar trends in the changes of expression levels (either increase or decrease), but differed in their time points. Among 145 genes from the hippocampus whose expressions were altered by fear conditioning ­paradigm, 33 genes (23%) were up-regulated while 112 genes (77%) were downregulated over the period of 0.5–24 h (Fig. 2). In the amygdala, a total of 222 genes

Genome-Wide Expression Profiles of Amygdala and Hippocampus Table 4  Representative genes changed in the amygdala after fear conditioning Fold change (h) Accession Cluster number Gene name 0.5 6 I AF026124 Schwannoma-associated protein 2.35 (SAM-9) AB006361 Prostaglandin D synthetase 3.85 I31397 Dynamin 3.1 D38613 921-L presynaptic protein 3.05 AF026489 b-III spectrin 4.1 II R75491 Myelin-associated oligodendrocytic 2.85 basic protein (MOBP) X94310 L1 2.4 X14943 Neuronal cell surface protein F3 3.25 Z31557 Chaperonin-containing TCP-1 (CCT) 2.6 X90875 FXR1 4.5 AA031158 NAP-22 2.3 M72414 Microtubule-associated protein 3.3 4 (MAP4) X70398 P311 2.25 III AA139495 Clathrin, heavy polypeptide −2.6 M73329 Phospholipase C-alpha −3.35 M13227 Enkephalin −5.15 −2.75 −2.9 IV AA239103 GABAA receptor-associated protein-like 2 AA409978 Calmodulin synthesis −2.9 −2.15 V W76777 a-Actinin 6.4 2.55 W89940 A-X actin 41.9 10.6 X57497 Glutamate receptor 1 (GluR1) 6.5 3.55 VI D86177 Phosphatidylinositol 4-phosphate 5-kinase U20365 g-Actin VII U19582 Oligodendrocyte-specific −2.95 protein (OSP) L02526 MEK1 protein kinase −2.1 U60001 Protein kinase C inhibitor −2.55 Z70023 Connexin-30 −2.15 X07215 Proteolipid protein (PLP) −2.05 VIII AA111149 a-Tubulin 3.95 AA059763 b-Tubulin 4.7 AA122619 Protein phosphatase 2A inhibitor 2.3 (PP2AI) X61434 Protein kinase C b-subunit 2.25 U27106 Clathrin-associated AP-2, AP50 2.45 subunit U60150 Vesicle-associated membrane 4.05 protein VAMP2 U06922 Signal transducer and activator of 3.05 3.15 None transcription (Stat3) Z25524 Integrin-associated protein (IAP) −2.25

315

24

6.3 25.15 4.8 3.05 9.9

5.35 5.15 2.05 3.55 3 8.5

−2.05

Z. Zhao and Y. Hu

Number of genes

316 110 100 90 80 70 60 50 40 30 20 10 0 -10 -20 -30 -40 -50 -60 -70 -80 -90 -100 -110

Amygdala (222 genes)

Hippocampus (145 genes)

24 h

24 h 0.5 h

6h

0.5 h

Increased Decreased 6h

Fig.  2  Fear conditioning-induced expressional changes at 0.5, 6, and 24  h time points in 222 amygdala genes and 145 hippocampus genes from the conditioned mice compared to no ­conditioning mice. Those changes were reflected in both directions, either up-regulated or down­regulated. A total of 11,000 genes and ESTs were screened in duplicate. The cut-off threshold is twofold changes in the expression level above the zero point

showed changes in their expression after fear conditioning. About 123 of these genes (55%) exhibited an up-regulated expression, while 99 genes (45%) showed decreased expression (Fig.  2). The cluster analysis of the expression alterations triggered by fear conditioning further revealed the intrinsic expression kinetics of these genes over the three time points. Most of the 145 genes from the hippocampus were allocated into six clusters (Fig.  3a, c, and some of them are shown in Table  3). The majority of these genes were assigned to the clusters showing a reduced expression level either at 6 h only (Cluster III: 78 genes) or at both 0.5 and 6 h (Cluster IV: 25 genes). In contrast, most of the 222 genes from the amygdala could be grouped into eight different clusters (Fig. 3b, d, and some important ones are listed in Table 4). Distinct genomic responses between the amygdala and the hippocampus further became apparent based on their cellular functions (Fig. 4). The hippocampus held a higher percentage of signaling genes (34%) with increased expression induced by the conditioning, compared to that of the amygdala (23%) (Fig.  4a, b). Of 222 genes with increased expression levels in the amygdala, about 22% encode ­structural/cytoskeleton protein genes, compared to only 6% of 145 genes in the hippocampus that encode structural proteins (Fig.  4a, b). In addition, larger ­numbers of genes (33%) up-regulated in the hippocampus were related to DNA/ RNA regulation, whereas in the amygdala, only 5% of the genes belonged to this

Genome-Wide Expression Profiles of Amygdala and Hippocampus

a

0 I

0.5

6

24

b

0

0.5

6

24

c

I

317

Cluster I (13 genes)

Cluster II (13 genes)

Cluster III (78 genes)

0 0.5 6 24

0 0.5 6 24

0 0.5 6 24

Cluster IV (25 genes)

Cluster V (6 genes)

Cluster VI (3 genes)

0 0.5 6 24

0 0.5 6 24

0 0.5 6 24

Cluster I (16 genes)

Cluster II (43 genes)

Cluster III (11 genes)

Cluster IV (26 genes)

0 0.5 6 24

0 0.5 6 24

0 0.5 6 24

0 0.5 6 24

Cluster V (6 genes)

Cluster VI (29 genes)

Cluster VII (57 genes)

Cluster VIII (27 genes)

0 0.5 6 24

0 0.5 6 24

0 0.5 6 24

0 0.5 6 24

II II III

III IV V

IV VI

VI

V 10 or more 8 6 VII 4 2 0 �2 �4 VIII �5 �8 �10 or more

d

Fig.  3  Drastically different gene expression profiles triggered by fear conditioning in the hippocampus (a) and amygdala (b) as compared with the control mice. Numbers in the X-axis represent time points in hours (0, 0.5, 6, and 24). The Y-axis represents the number of genes that show dynamic changes (coded in color). The Roman numbers (I–VI) along the Y-axis represent the gene clusters with similar expression kinetics. The color bar on the left corner represents the scale of change in expression. (c) Most of the 145 genes identified in the hippocampus can be grouped into six clusters (I–VI). (d) Most of the 222 genes with altered expression levels in the amygdala could be placed into eight clusters (I–VIII). For illustration, changes of gene expression in each cluster (in c, d) are simply represented in a binary mode in the Y-axis. The X-axis shows the four time points (0, 0.5, 6, and 24 h after training). Numbers of genes in each particular cluster is shown at the top center of each cluster

category. For those genes down-regulated in the hippocampus after fear ­conditioning, the largest group, about 29%, were composed of signaling molecules (Fig.  4c) while the second largest group of the genes consisted of transcription factors. Interestingly, in the amygdala, the largest proportion of down-regulated genes have roles in regulating transcription (Fig. 4d). These microarray-based results were further validated by means of real-time quantitative PCR of selected genes and immunohistochemical analysis of a selected protein, GluR1. The real-time PCR results of a total of 20 genes selected from the hippocampus and amygdala from the conditioned and control mice are listed in the Table 5, showing an expression pattern similar to that seen in the gene chip experiments. Quantitative immunohistological analysis of the GlutR1 protein, which is

318

Z. Zhao and Y. Hu

a

b

6% 18%

16%

22%

6% 9%

34% 14% 23% 6%

33%

5%

c

d

7% 23%

8% 12%

20% 14%

29%

4%

5% 9% 16% 15%

5%

7% 12%

2%

20%

Structural/cytoskeleton

Signaling

Transcription factors

DNA/RNA regulation

Protein regulation

Metabolic and Catabolic enzymes

Fig. 4  Distinct genomic responses in the hippocampus and the amygdala of the conditioned mice defined upon the distribution of genes according to their cellular functions. Illustrated in pie graphs are the percentage of genes in each functional class with increased expression levels in the hippocampus (a) and amygdala (b), as well as those genes with reduced expressions in the ­hippocampus (c) and amygdala (d)

one of the amino-3-hydroxy-5-methyl-4-isoxalone propionic acid (AMPA) receptor subunits, and which has increased mRNA expression in amygdala at 0.5, 6, and 24 h after fear conditioning, also demonstrated a consistent increased expression in the amygdala at both 6 and 24 h after training as compared with the control (Fig. 5). Therefore, this result extends the regulation of GluR1 mRNA to the protein level. More importantly, it also provides a potential mechanistic explanation for the reported findings that fear conditioning enhances AMPA-mediated transmission in amygdala neurons (McKernan & Shinnick-Gallagher, 1997; Rogan, Staubli, & LeDoux, 1997).

Amygdala

Tissues Hippocampus

X14943 Z31557 AA031158 M72414 AA139495 M73329 M13227 AA239103 AA409978

W46019 AA067362 D37792 R75491

Accession number D30730 X51438 AA066335 X60304 X64587 M73329 W12204

Gene name GTPase-activating protein (NF1) Vimentin Amyloid precursor protein (APP) Protein kinase C-d Splicing factor U2AF65 Phospholipase C-a (PLC-a) Ca2+/calmodulin-dependent protein kinase II isom g-b (CaMKII) Protein kinase regulator 14-3-3 Vesicle protein pantophysin SynaptotagminI/65 Myelin-associated oligodendrocytic basic protein (MOBP) Neuronal cell surface protein F3 Chaperonin-containing TCP-1 (CCT) NAP-22 Microtubule-associated protein 4 (MAP4) Clathrin, heavy polypeptide Phospholipase C-alpha Enkephalin GABAA receptor-associated protein-like 2 Calmodulin synthesis

Table 5  Validation of microarray-based expression profile by real-time RT-PCR

9.95 6.69 8.52 11.37 −9.15 −6.26 −9.3 −10.63 −8.53

−7 −7.97 −3.69 11.65

Fold change Real-time RT-PCR 9.95 6.47 −7.87 6.16 6.82 5.34 −5.47

3.25 2.6 2.3 3.3 −2.6 −3.35 −5.15 −2.9 −2.15

−2.8 −3.8 −3.2 2.85

Microarray 5.4 1.9 −2.5 6.2 4.8 2.3 −3.2

Genome-Wide Expression Profiles of Amygdala and Hippocampus 319

320

Z. Zhao and Y. Hu

a

b

c

d

e

Control 6 hour 24 hour

0.25

Optical Density

0.20 0.15 0.10 0.05 0.00 Lateral Nucleus

Basal Nucleus

Accessory Basal Nucleus

Central Nucleus

Fig. 5  Increase in GluR1 protein after fear conditioning. (a) A coronal section of mouse brain showing amygdala nuclei is enclosed in the dot-lined box. (b) The area containing amygdala nuclei from control mouse is shown. Amygdala nuclei are divided into the lateral nucleus (L), basal lateral nucleus (B), accessory basal nucleus (AB), and central nucleus (C). Increased GluR1 staining in the amygdala nuclei at 6 h (c) and 24 h (d) after fear conditioning. (e) Quantitative measurement of GluR1 staining in the subnucleus of the amygdala. Asterisks indicates the statistically significant increases in the staining intensities of GluR1 protein in the subregions of amygdala nucleus from the trained mice

Genome-Wide Expression Profiles of Amygdala and Hippocampus

321

4.2.1 Fear Conditioning-Triggered Gene Expression Profiles in the Amygdala Perhaps the most noticeable feature for the amygdala gene profiles was that genes encoding structural and cell adhesion proteins represented the major percentage (22%) of total genes regulated in response to fear conditioning. For example, a number of these genes have been clearly linked with synaptic, dendritic and axonal structures such as actin, brain spectrin, tubulin, and microtubule-associated proteins (MAPs) (van Rossum & Hanisch, 1999). b-III spectrin, a Golgi- and vesicle-­ associated protein that interacts with the NMDA receptor, showed an increase at 0.5 h. Also, a microtubule-associated protein 4 (MAP4), b-tubulin, a-tubulin and cytosolic chaperonin-containing TCP (CCT) were all up-regulated after training. CCT is a molecular chaperone required for the folding of a-actin and b-tubulin (Llorca et  al., 2001). The up-regulation of CCT after learning provides further ­evidence for the requirement of functional a-actin and b-tubulin to undergo ­learning-induced changes of cytoskeletal structure. Several genes regulated in the amygdala are involved in the dynamic turnover and physical regulation of ionotropic receptors. The GABAA receptor-associated protein (GABARAP), known to bind to GABAA receptors both in vitro and in vivo (Wang, Bedford, et al., 1999), was down-regulated at the 0.5 and 6 h time points. GABARAP can modulate channel kinetics by promoting the clustering of GABAA receptors through microtubules (Chen, Wang, Vicini, & Olsen, 2000). The downregulation of GABARAP after fear conditioning suggests that the inhibitory effect of the GABAA receptors is decreased to allow for an increase in synaptic excitability. It has been postulated that the activity-dependent trafficking of AMPA receptors at synapses is responsible for altering synaptic strength (Carroll, Beattie, von Zastrow, & Malenka, 2001; Malinow & Malenka, 2002; van Rossum & Hanisch, 1999). Indeed, the microarray study showed an increased expression of the AMPA receptor subunit, GluR1, in the amygdala of the conditioned mice by 6.5, 3.55, and 4.8-fold at the 0.5, 6, and 24  h time points, respectively, compared to that of no conditioning (Table  4). The gene expression profiling in amygdala also revealed that there were genes implicated in the insertion and removal of AMPA receptors. For example, AMPA receptors are specifically regulated by dynamin-dependent endocytosis (Carroll et al., 2001). The expression of dynamin, a GTPase known to be involved in the fission step of vesicle formation (Scales & Scheller, 1999), was increased at 0.5  h after the conditioning. The expression of clathrin-associated AP-2 protein was also triggered to be increased after 0.5 and 24 h of fear conditioning, further suggesting that clathrin-coated pits play an important role in regulating the ligand-induced endocytosis of AMPA receptors (Scales & Scheller, 1999). Furthermore, it has been shown that Calcium-calmodulin-kinase II (CaMKII)– actinin–actin complex provides an additional physical interacting site for anchoring AMPA receptors at synapses (Carroll et al., 2001). Indeed, the expressions of both a-actinin and actin in the fear-conditioned amygdala were drastically increased at all time points. Given the fact that the fear learning in rats is interfered with CNQX,

322

Z. Zhao and Y. Hu

an AMPA receptor antagonist (Li, Stutzmann, & LeDoux, 1996), these studies strongly suggest the involvement of AMPA regulation in the formation of ­associative learning and memory. It has been reported that protein phosphatases (PP) (including PP1 and PP2A) can differentially regulate the Ca2+-independent activity of CaMKII by dephosphorylation of the soluble and postsynaptic density-associated forms of CaMKII (Lisman & Zhabotinsky, 2001; Strack, Barban, Wadzinski, & Colbran, 1997). The increased expression of the PP2AI gene (PP2A inhibitor) in the amygdala at 0.5 and 24 h after fear conditioning suggested a possible alteration in the signal ­transduction. Consistent with the view of the involvement of neuronal signal ­molecules in the formation of associative memory, the gene expression profiling in the amygdala showed that the expressions of SAM-9 and phosphatidylinositol-4-phosphate-5-kinase (PIP5K) were up-regulated at 0.5 and 24 h, respectively. SAM-9 is a ­membrane-associated member of the phospholipase D (PLD) ­superfamily. PLD enzymes have been implicated in signal transduction pathways associated with cell growth and membrane ­trafficking in mammalian cells (Singer, Brown, & Sternweis, 1997), whereas PIP5K catalyzes the synthesis of ­phosphatidylinositol 4,5-­bisphosphate (PIP2) and has been implicated in cellular signaling processes (Oude Weernink, Schmidt, & Jakobs, 2004). It has been reported that the expression of PIP5K is reduced by 50% in the frontal cortex of Alzheimer’s patients (Jolles, Bothmer, Markerink, & Ravid, 1992). Fear conditioning triggered alterations in expression profiling in the amygdala also involved a number of genes known to be linked with human mental retardation. The fragile-X-related gene 1 (FXR1), which encodes a ribosome-associated, RNA-binding protein (Van Dam et  al., 2000), was up-regulated at 6  h after fear ­conditioning. It has been shown that FXR1 knockout mice displays significant ­deficit in fear conditioning (Paradee et  al., 1999). Because FXR1 is located in spines and dendrites of neurons and is thought to play a role in translational ­regulation of selective messenger RNA transcripts, the increase in FXR1 ­expression induced by fear conditioning suggested that FXR1 is an important molecule in memory formation. A study has recently demonstrated that glia can control synapse number in vitro and may play a role in the changes underlying synaptic plasticity (Ullian, Sapperstein, Christopherson, & Barres, 2001). In support of this finding, the expression profiling in the amygdala revealed several glia-enriched or ­glial-specific genes whose expressions were altered after fear conditioning. For example, ­expression of astrocyte-specific connexin-30 (Cx30) (Rash, Yasumura, Dudek, & Nagy, 2001) was down-regulated by fear conditioning at 0.5  h. Also, the ­expressions of several genes known to be important in myelin formation or stabilization were changed in the fear conditioned amygdala, including ­myelin-associated ­oligodendrocytic basic protein (MOBP) (McCallion, Stewart, Montague, Griffiths, & Davies, 1999), proteolipid protein (PLP) (Yool et  al., 2001), and oligodendrocyte-specific protein (OSP) (Bronstein, Popper, Micevych, & Farber, 1996).

Genome-Wide Expression Profiles of Amygdala and Hippocampus

323

4.2.2 Fear Conditioning-Triggered Gene Expression Profiles in the Hippocampus The Affymetrix microarray assay revealed that 27 out of the 38 signaling-­associated genes in the hippocampus were down-regulated after fear conditioning compared to no conditioning. Among them, repressed expressions of 18 genes was observed at 6 h, but not at 0.5 or 24 h. For instance, the protein kinase regulator 14-3-3, which has previously been identified as a learning-related gene in Drosophila (Skoulakis & Davis, 1996) and appears to interact with the C-terminus of GABAB receptors in cultured neurons (Couve et al., 2001), showed a reduced expression at 6 h during the conditioning. Similarly, the GABAA receptor alpha-1 subunit also displayed a decreased expression at that time point. This raised the interesting ­possibility that the formation of contextual fear memory involves modifications of gamma-aminobutyric acid (GABA) receptor-mediated inhibition of hippocampal circuit’s excitability. Pharmacological evidence also indicated that the GABAA-specific antagonist picrotoxin mediated currents that contributed to synaptic ­integration of US and CS in fear conditioning (Rosis, Johnson, & LeDoux, 2003). The gene expression profiling in the fear conditioned hippocampus also revealed several down-regulated genes, such as vesicle-associated membrane protein (VAMP) and synaptotagmin, which have been demonstrated to be involved in ­synaptic vesicle trafficking and neurotransmitter release. The expression of ­pantophysin, a vesicle protein related to synaptophysin and usually co-distributed with VAMP (Windoffer et al., 1999), was reduced at the 6 h time point during fear conditioning. In addition, other proteins involved in vesicle formation and ­assembly, for example, endophilin I (Schmidt et  al., 1999) and ADP-ribosylation factor (ARF1) (Faundez, Horng, & Kelly, 1997) also exhibited changes in a similar manner. Modifications of synaptic plasticity, such as depotentiation and long-term depression, are known to involve serine/threonine PP1 and PP2A (Mulkey, Endo, Shenolikar, & Malenka, 1994). These phosphatases interact closely with a variety of kinases including PKC and CaMKII. In support of these observations, identified by various experimental approaches, our gene chip assay also revealed that the expression levels of PP1, PP2A, PKC, and CaMKII were all similarly down-­ regulated in the hippocampus at 6 h after fear conditioning. Several genes in the hippocampus showing learning-related changes have ­previously been implicated in learning and memory disorders in humans. For example, APP, an Alzheimer’s disease-related gene, was decreased in its ­expression in the hippocampus after fear conditioning. Moreover, two additional proteins, kinesin light chain 1 (KLC1) and the splicing factor, U2AF65, known to interact with APP, also showed changes in their expression levels. Another gene involved in learning disorders was the neurofibromatosis type 1 gene (NF1). Loss of NF1 function can lead to memory deficits in both mutant mice (Silva et al., 1997) and Drosophila (Guo, Tong, Hannan, Luo, & Zhong, 2000). The fear conditioningtriggered overexpression of NF1 at the time point of 6  h in the hippocampus ­suggested that NF1 is tightly regulated during memory processes.

324

Z. Zhao and Y. Hu

In this study, genes previously known to be important in direct regulation of neuronal functions were found to be regulated by fear conditioning in the ­hippocampus. In addition, a number of genes, frequently expressed in non-neuronal cells, also exhibited fear conditioning-induced changes in their expression levels in the hippocampus. For example, the glial enzyme glutamine synthetase showed an up-regulated level of expression at 24 h during the conditioning. Inconsistent with this observation, it was reported that in patients with Alzheimer’s disease the expression of the gene was reduced in astrocytes (Robinson, 2000). Vimentin, a gene that encodes an intermediate filament protein, also showed up-regulation after the fear learning. Since the intermediate filaments are major components of the cytoskeleton in astrocytes (Menet et al., 2001), the observed increase in its ­expression also indicated a role of fear conditioning in evoking structural changes in astrocytes.

5 Summary In an effort towards understanding the molecular genetic responses produced by fear conditioning training, we applied gene chip technology to simultaneously examine large-scale gene expression changes in the amygdala and hippocampus after paired fear conditioning. Drastically different profiles of the gene expression in the amygdala and the hippocampus induced by fear conditioning were identified. The expression kinetics over three time points (0.5, 6 and 24  h) (Fig. 3) and the biological functions (Fig.  4) of these genes in these two brain regions suggested their distinct roles in acquisition, storage, and expression of associative memory. It is worth noting that the expressions of a number of genes were observed to be decreased in response to the fear conditioning. In particular, there were 78 genes in the hippocampus exhibiting the conditioning-triggered reduction in expression at the time point of 6  h, indicating that a coordinated down-regulation in gene ­expression is not only an important part of the overall genomic program but also an active and integral part of transcriptional processes in response to external stimuli. It is also worth mentioning that some genes in the amygdala and the hippocampus with altered expression induced by fear memory have also been identified ­previously using the same gene chip approach (Jiang et al., 2001; Rampon et al., 2000). These previous studies examined the gene expression profiles of other ­cognitive processes that affect memory in animals subjected to an enriched ­environment (which improves learning and memory) (Rampon et  al., 2000) or aging brains (which are typically associated with decline in learning and memory) (Jiang et  al., 2001). For example, genes similarly regulated by environmental enrichment and fear memory include clathrin, synaptogamin I/65, VAMP-2, GluR1, and ryanodine receptor type 2 (Rampon et al., 2000), whereas genes regulated in both aging and fear memory include dyanmin, clathrin-associated AP-2 protein, prostaglandin D synthetase, and PP2A (Jiang et al., 2001). The transcriptional convergence of these genes in several memory-related behavioral paradigms points to their potential role in the critical regulation of learning behaviors.

Genome-Wide Expression Profiles of Amygdala and Hippocampus

325

Acknowledgements  This work was supported in part by grants from the Shanghai Municipal Education Commission and the Ministry of Science and Technology of China (2003AA221061), and grants from the Science and Technology Commission of Shanghai Municipality (05PJ14044 and 06DZ19002).

References Abeliovich, A., Paylor, R., Chen, C., Kim, J. J., Wehner, J. M., & Tonegawa, S. (1993). PKCg mutant mice exhibit mild deficits in spatial and contextual learning. Cell, 75, 1263–1271. Aggleton, J. P. (2000). The amygdala: A functional analysis. Oxford: Oxford University Press. Anagnostaras, S. G., Maren, S., & Fanselow, M. S. (1999). Temporally graded retrograde amnesia of contextual fear after hippocampal damage in rats: Within-subjects examination. The Journal of Neuroscience, 19, 1106–1114. Antoniadis, E., & McDonald, R. (2001). Amygdala, hippocampus, and unconditioned fear. Experimental Brain Research, 138, 200–209. Blair, H. T., Schafe, G. E., Bauer, E. P., Rodrigues, S. M., & LeDoux, J. E. (2001). Synaptic plasticity in the lateral amygdala: A cellular hypothesis of fear conditioning. Learning & Memory, 8, 229–242. Bourtchuladze, R., Frenguelli, B., Blendy, J., Cioffi, D., Schutz, G., & Silva, A. J. (1994). Deficient long-term memory in mice with a targeted mutation of the camp-responsive elementbinding protein. Cell, 79, 59–68. Bronstein, J. M., Popper, P., Micevych, P. E., & Farber, D. B. (1996). Isolation and characterization of a novel oligodendrocyte-specific protein. Neurology, 47, 772–778. Carroll, R. C., Beattie, E. C., von Zastrow, M., & Malenka, R. C. (2001). Role of AMPA receptor endocytosis in synaptic plasticity. Nature Reviews. Neuroscience, 2, 315–324. Chen, L., Wang, H., Vicini, S., & Olsen, R. W. (2000). The g-aminobutyric acid type A (GABAA) receptor-associated protein (GABARAP) promotes GABAA receptor clustering and modulates the channel kinetics. Proceedings of the National Academy of Sciences of the United States of America, 97, 11557–11562. Cirelli, C., & Tononi, G. (1999). Differences in gene expression during sleep and wakefulness. Annals of Medicine, 31, 117–124. Couve, A., Kittler, J. T., Uren, J. M., Calver, A. R., Pangalos, M. N., Walsh, F. S., et al. (2001). Association of GABAB receptors and members of the 14-3-3 family of signaling proteins. Molecular and Cellular Neurosciences, 17, 317–328. Davis, M. (1997). Neurobiology of fear responses: The role of the amygdala. The Journal of Neuropsychiatry and Clinical Neurosciences, 9, 382–402. Eisen, M. B., Spellman, P. T., Brown, P. O., & Botstein, D. (1998). Cluster analysis and display of genome-wide expression patterns. Proceedings of the National Academy of Sciences of the United States of America, 95, 14863–14868. Falls, W. A., Miserendino, M. J., & Davis, M. (1992). Extinction of fear-potentiated startle: Blockade by infusion of an NMDA antagonist into the amygdala. The Journal of Neuroscience, 12, 854–863. Fanselow, M. S. (1994). Neural organization of the defensive behavior system responsible for fear. Psychonomic Bulletin & Review, 1, 429–438. Fanselow, M. S., & LeDoux, J. E. (1999). Why we think plasticity underlying Pavlovian fear conditioning occurs in the basolateral amygdala. Neuron, 23, 229–232. Faundez, V., Horng, J.-T., & Kelly, R. B. (1997). ADP ribosylation factor 1 is required for ­synaptic vesicle budding in PC12 cells. The Journal of Cell Biology, 138, 505–515. Fendt, M., & Fanselow, M. S. (1999). The neuroanatomical and neurochemical basis of ­conditioned fear. Neuroscience and Biobehavioral Reviews, 23, 743–760.

326

Z. Zhao and Y. Hu

Frankland, P. W., Cestari, V., Filipkowski, R. K., McDonald, R. J., & Silva, A. (1998). The dorsal hippocampus is essential for context discrimination, but not for contextual conditioning. Behavioral Neuroscience, 112, 863–874. Gale, G. D., Anagnostaras, S. G., Godsil, B. P., Mitchell, S., Nozawa, T., Sage, J. R., et al. (2004). Role of the basolateral amygdala in the storage of fear memories across the adult lifetime of rats. The Journal of Neuroscience, 24, 3810–3815. Guo, H.-F., Tong, J., Hannan, F., Luo, L., & Zhong, Y. (2000). A neurofibromatosis-1-regulated pathway is required for learning in drosophila. Nature, 403, 895–898. Hakak, Y., Walker, J. R., Li, C., Wong, W. H., Davis, K. L., Buxbaum, J. D., et  al. (2001). Genome-wide expression analysis reveals dysregulation of myelination-related genes in chronic schizophrenia. Proceedings of the National Academy of Sciences of the United States of America, 98, 4746–4751. Ho, L., Guo, Y., Spielman, L., Petrescu, O., Haroutunian, V., Purohit, D., et al. (2001). Altered expression of a-type but not b-type synapsin isoform in the brain of patients at high risk for Alzheimer’s disease assessed by DNA microarray technique. Neuroscience Letters, 298, 191–194. Hosack, D., Dennis, G., Sherman, B., Lane, H., & Lempicki, R. (2003). Identifying biological themes within lists of genes with EASE. Genome Biology, 4, R70. Hu, Y., & Tsien, J. Z. (2002). Functional genomics: profiling the brain and cognition with DNA microarrays. Applied Genomics and Proteomics, 1(4), 219–226. Huerta, P. T., Sun, L. D., Wilson, M. A., & Tonegawa, S. (2000). Formation of temporal memory requires NMDA receptors within CA1 pyramidal neurons. Neuron, 25, 473–480. Jiang, C. H., Tsien, J. Z., Schultz, P. G., & Hu, Y. (2001). The effects of aging on gene expression in the hypothalamus and cortex of mice. Proceedings of the National Academy of Sciences of the United States of America, 98, 1930–1934. Jolles, J., Bothmer, J., Markerink, M., & Ravid, R. (1992). Phosphatidylinositol kinase is reduced in Alzheimer’s disease. Journal of Neurochemistry, 58, 2326–2329. Kim, J. J., & Fanselow, M. S. (1992). Modality-specific retrograde amnesia of fear. Science, 256, 675–677. Kim, J. J., & Jung, M. W. (2006). Neural circuits and mechanisms involved in Pavlovian fear conditioning: A critical review. Neuroscience and Biobehavioral Reviews, 30, 188–202. LeDoux, J. E. (2000). Emotion circuits in the brain. Annual Review of Neuroscience, 23, 155–184. Lee, M.-L. T., Kuo, F. C., Whitmore, G. A., & Sklar, J. (2000). Importance of replication in microarray gene expression studies: Statistical methods and evidence from repetitive cDNA hybridizations. Proceedings of the National Academy of Sciences of the United States of America, 97, 9834–9839. Li, X. F., Stutzmann, G. E., & LeDoux, J. E. (1996). Convergent but temporally separated inputs to lateral amygdala neurons from the auditory thalamus and auditory cortex use different ­postsynaptic receptors: In vivo intracellular and extracellular recordings in fear conditioning pathways. Learning & Memory, 3, 229–242. Lipshutz, R. J., Fodor, S. P. A., Gingeras, T. R., & Lockhart, D. J. (1999). High density synthetic oligonucleotide arrays. Nature Genetics, 21, 20–24. Lisman, J. E., & Zhabotinsky, A. M. (2001). A model of synaptic memory: A CaMKII/PP1 switch that potentiates transmission by organizing an AMPA receptor anchoring assembly. Neuron, 31, 191–201. Livak, K. J., & Schmittgen, T. D. (2001). Analysis of relative gene expression data using real-time quantitative PCR and the 2−DDCt method. Methods, 25, 402–408. Livesey, F. J., Furukawa, T., Steffen, M. A., Church, G. M., & Cepko, C. L. (2000). Microarray analysis of the transcriptional network controlled by the photoreceptor homeobox gene Crx. Current Biology, 10, 301–310. Llorca, O., Martin-Benito, J., Grantham, J., Ritco-Vonsovici, M., Willison, K. R., Carrascosa, J. L., et al. (2001). The “sequential allosteric ring” mechanism in the eukaryotic chaperonin-assisted folding of actin and tubulin. The EMBO Journal, 20, 4065–4075. Lockhart, D. J., Dong, H., Byrne, M. C., Follettie, M. T., Gallo, M. V., Chee, M. S., et al. (1996). Expression monitoring by hybridization to high-density oligonucleotide arrays. Nature Biotechnology, 14, 1675–1680.

Genome-Wide Expression Profiles of Amygdala and Hippocampus

327

Luo, Z., & Geschwind, D. H. (2001). Microarray applications in neuroscience. Neurobiology of Disease, 8, 183–193. Malinow, R., & Malenka, R. C. (2002). AMPA receptor trafficking and synaptic plasticity. Annual Review of Neuroscience, 25, 103–126. Maren, S. (2001). Neurobiology of pavlovian fear conditioning. Annual Review of Neuroscience, 24, 897–931. Maren, S., Aharonov, G., & Fanselow, M. S. (1997). Neurotoxic lesions of the dorsal hippocampus and pavlovian fear conditioning in rats. Behavioural Brain Research, 88, 261–274. McCallion, A. S., Stewart, G. J., Montague, P., Griffiths, I. R., & Davies, R. W. (1999). Splicing pattern, transcript start distribution, and DNA sequence of the mouse gene (mobp) encoding myelin-associated oligodendrocytic basic protein. Molecular and Cellular Neurosciences, 13, 229–236. McKernan, M. G., & Shinnick-Gallagher, P. (1997). Fear conditioning induces a lasting potentiation of synaptic currents in vitro. Nature, 390, 607–611. Mei, B., Li, C., Dong, S., Jiang, C. H., Wang, H., & Hu, Y. (2005). Distinct gene expression profiles in hippocampus and amygdala after fear conditioning. Brain Research Bulletin, 67, 1–12. Menet, V., Gimenez, Y., Ribotta, M., Chauvet, N., Drian, M. J., Lannoy, J., et  al. (2001). Inactivation of the glial fibrillary acidic protein gene, but not that of vimentin, improves ­neuronal survival and neurite growth by modifying adhesion molecule expression. The Journal of Neuroscience, 21, 6147–6158. Mody, M., Cao, Y., Cui, Z., Tay, K.-Y., Shyong, A., Shimizu, E., et al. (2001). Genome-wide gene expression profiles of the developing mouse hippocampus. Proceedings of the National Academy of Sciences of the United States of America, 98, 8862–8867. Mulkey, R. M., Endo, S., Shenolikar, S., & Malenka, R. C. (1994). Involvement of a calcineurin/ inhibitor-1 phosphatase cascade in hippocampal long-term depression. Nature, 369, 486–488. Nisenbaum, L. K. (2002). The ultimate chip shot: Can microarray technology deliver for neuroscience? Genes, Brain, and Behavior, 1, 27–34. Oude Weernink, P. A., Schmidt, M., & Jakobs, K. H. (2004). Regulation and cellular roles of phosphoinositide 5-kinases. European Journal of Pharmacology, 500, 87–99. Paradee, W., Melikian, H. E., Rasmussen, D. L., Kenneson, A., Conn, P. J., & Warren, S. T. (1999). Fragile X mouse: Strain effects of knockout phenotype and evidence suggesting ­deficient amygdala function. Neuroscience, 94, 185–192. Pare, D., Quirk, G. J., & Ledoux, J. E. (2004). New vistas on amygdala networks in conditioned fear. Journal of Neurophysiology, 92, 1–9. Pavlov, I. P. (1927). Conditioned reflexes: An investigation of the physiological activity of the cerebral cortex. London: Oxford University Press. Phillips, R. G., & LeDoux, J. E. (1992). Differential contribution of amygdala and hippocampus to cued and contextual fear conditioning. Behavioral Neuroscience, 106, 274–285. Pongrac, J., Middleton, F. A., Lewis, D. A., Levitt, P., & Mirnics, K. (2002). Gene expression profiling with DNA microarrays: Advancing our understanding of psychiatric disorders. Neurochemical Research, 27, 1049–1063. Rampon, C., Jiang, C. H., Dong, H., Tang, Y.-P., Lockhart, D. J., Schultz, P. G., et  al. (2000). Effects of environmental enrichment on gene expression in the brain. Proceedings of the National Academy of Sciences of the United States of America, 97, 12880–12884. Rash, J. E., Yasumura, T., Dudek, F. E., & Nagy, J. I. (2001). Cell-specific expression of ­connexins and evidence of restricted gap junctional coupling between glial cells and between neurons. The Journal of Neuroscience, 21, 1983–2000. Robinson, S. R. (2000). Neuronal expression of glutamine synthetase in Alzheimer’s disease indicates a profound impairment of metabolic interactions with astrocytes. Neurochemistry International, 36, 471–482. Rodrigues, S. M., Schafe, G. E., & LeDoux, J. E. (2001). Intra-amygdala blockade of the NR2B subunit of the NMDA receptor disrupts the acquisition but not the expression of fear ­conditioning. The Journal of Neuroscience, 21, 6889–6896.

328

Z. Zhao and Y. Hu

Rodrigues, S. M., Schafe, G. E., & LeDoux, J. E. (2004). Molecular mechanisms underlying emotional learning and memory in the lateral amygdala. Neuron, 44, 75–91. Rogan, M. T., Staubli, U. V., & LeDoux, J. E. (1997). Fear conditioning induces associative longterm potentiation in the amygdala. Nature, 390, 604–607. Rosis, S., Johnson, L. R., & LeDoux, J. E. (2003). gabaa receptors contribute to synaptic integration in lateral amygdala neurons. Program No. 623.2. 2003 Abstract Viewer/Itinerary Planner. Washington, DC: Society for Neuroscience. (http://www.sfn.org/index.cfm? pagename=abstracts_archive&task=view&controlID=10796&year=2003&CFID= 5270238&CFTOKEN=82264079&jsessionid=b430adb20d7a27561761) Sandberg, R., Yasuda, R., Pankratz, D. G., Carter, T. A., Del Rio, J. A., Wodicka, L., et al. (2000). Regional and strain-specific gene expression mapping in the adult mouse brain. Proceedings of the National Academy of Sciences of the United States of America, 97, 11038–11043. Sara, S. J. (2000). Retrieval and reconsolidation: Toward a neurobiology of remembering. Learning & Memory, 7, 73–84. Scales, S. J., & Scheller, R. H. (1999). Cell biology: Lipid membranes shape up. Nature, 401, 123–124. Schena, M., Shalon, D., Davis, R. W., & Brown, P. O. (1995). Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science, 270, 467–470. Schmidt, A., Wolde, M., Thiele, C., Fest, W., Kratzin, H., Podtelejnikov, A. V., et  al. (1999). Endophilin I mediates synaptic vesicle formation by transfer of arachidonate to lysophosphatidic acid. Nature, 401, 133–141. Selden, N. R. W., Everitt, B. J., Jarrard, L. E., & Robbins, T. W. (1991). Complementary roles for the amygdala and hippocampus in aversive conditioning to explicit and contextual cues. Neuroscience, 42, 335–350. Silva, A. J., Frankland, P. W., Marowitz, Z., Friedman, E., Lazlo, G., Cioffi, D., et  al. (1997). A mouse model for the learning and memory deficits associated with neurofibromatosis type I. Nature Genetics, 15, 281–284. Singer, W. D., Brown, A. H. A., & Sternweis, P. C. (1997). Regulation of eukaryotic phosphatidylinositol-specific phospholipase C and phospholipase D. Annual Review of Biochemistry, 66, 475–509. Skoulakis, E. M. C., & Davis, R. L. (1996). Olfactory learning deficits in mutants for leonardo, a Drosophila gene encoding a 14-3-3 protein. Neuron, 17, 931–944. Strack, S., Barban, M. A., Wadzinski, B. E., & Colbran, R. J. (1997). Differential inactivation of postsynaptic density-associated and soluble Ca2+/calmodulin-dependent protein kinase II by protein phosphatases 1 and 2a. Journal of Neurochemistry, 68, 2119–2128. Tang, Y.-P., Shimizu, E., Dube, G. R., Rampon, C., Kerchner, G. A., Zhuo, M., et  al. (1999). Genetic enhancement of learning and memory in mice. Nature, 401, 63–69. Thibault, C., Lai, C., Wilke, N., Duong, B., Olive, M. F., Rahman, S., et al. (2000). Expression profiling of neural cells reveals specific patterns of ethanol-responsive gene expression. Molecular Pharmacology, 58, 1593–1600. Tsien, J. Z. (2000). Linking hebb’s coincidence-detection to memory formation. Current Opinion in Neurobiology, 10, 266–273. Tsien, J. Z., Huerta, P. T., & Tonegawa, S. (1996). The essential role of hippocampal CA1 NMDA receptor-dependent synaptic plasticity in spatial memory. Cell, 87, 1327–1338. Ullian, E. M., Sapperstein, S. K., Christopherson, K. S., & Barres, B. A. (2001). Control of ­synapse number by glia. Science, 291, 657–661. Van Dam, D., D’Hooge, R., Hauben, E., Reyniers, E., Gantois, I., Bakker, C. E., et  al. (2000). Spatial learning, contextual fear conditioning and conditioned emotional response in FMR1 knockout mice. Behavioural Brain Research, 117, 127–136. van Rossum, D., & Hanisch, U.-K. (1999). Cytoskeletal dynamics in dendritic spines: Direct modulation by glutamate receptors? Trends in Neurosciences, 22, 290–295. Wada, R., Tifft, C. J., & Proia, R. L. (2000). Microglial activation precedes acute neurodegeneration in sandhoff disease and is suppressed by bone marrow transplantation. Proceedings of the National Academy of Sciences of the United States of America, 97, 10954–10959.

Genome-Wide Expression Profiles of Amygdala and Hippocampus

329

Walker, D. L., & Davis, M. (2002). The role of amygdala glutamate receptors in fear learning, fearpotentiated startle, and extinction. Pharmacology, Biochemistry and Behavior, 71, 379–392. Wang, H., Bedford, F. K., Brandon, N. J., Moss, S. J., & Olsen, R. W. (1999). GABAA-receptor-associated protein links GABAA receptors and the cytoskeleton. Nature, 397, 69–72. Wang, K., Gan, L., Jeffery, E., Gayle, M., Gown, A. M., Skelly, M., et al. (1999). Monitoring gene expression profile changes in ovarian carcinomas using cDNA microarray. Gene, 229, 101–108. Wang, Y., Barbacioru, C., Hyland, F., Xiao, W., Hunkapiller, K., Blake, J., et  al. (2006). Large scale real-time PCR validation on gene expression measurements from two commercial longoligonucleotide microarrays. BMC Genomics, 7, 59. Watson, J. B., & Rayner, R. (1920). Conditioned emotional reactions. Journal of Experimental Psychology, 3, 1–14. (http://en.wikipedia.org/wiki/Little_Albert_experiment). Windoffer, R., Borchert-Stuhlträger, M., Haass, N. K., Thomas, S., Hergt, M., Bulitta, C. J., et al. (1999). Tissue expression of the vesicle protein pantophysin. Cell and Tissue Research, 296, 499–510. Yool, D. A., Klugmann, M., McLaughlin, M., Vouyiouklis, D. A., Dimou, L., Barrie, J. A., et al. (2001). Myelin proteolipid proteins promote the interaction of oligodendrocytes and axons. Journal of Neuroscience Research, 63, 151–164. Yoshikawa, T., Nagasugi, Y., Azuma, T., Kato, M., Sugano, S., Hashimoto, K., et  al. (2000). Isolation of novel mouse genes differentially expressed in brain using cDNA microarray. Biochemical and Biophysical Research Communications, 275, 532–537.

Part IV

Psychiatric Disorders

Genetic Studies of Schizophrenia Brien Riley

Abstract  Both genetic and non-genetic risk factors are thought to contribute to liability for, and the development of, schizophrenia. Genetic epidemiology consistently supports the involvement of genes in liability. Molecular genetic studies have made slow progress in identifying specific liability genes, but recent progress suggests that a number of specific genes contributing to risk have been identified. These collective results are complex and inconsistent with a single common DNA variant in any gene influencing risk across human populations. No specific genetic variant influencing risk has yet been unambiguously identified. Contemporary approaches hold great promise to further elucidate liability genes and their potential inter-relationship. In order to understand why researchers have come to these conclusions, we will review what is known about the genetic epidemiology and molecular genetics of schizophrenia in some detail. We will also consider how this field of study informs our understanding of the potential structure of non-genetic risk factors. Keywords  Schizophrenia • Genetic epidemiology • Family • Twin • Adoption • Linkage • Association • Molecular genetics

1 Introduction Genetic study of schizophrenia is based on three key areas of research (specialist genetic terminology is shown in bold and defined throughout). First, genetic epidemiology asks whether there is risk in excess of the population baseline in the relatives of cases, and, if so, whether the excess risk is attributable to the genetic B. Riley (*) Departments of Psychiatry and Human & Molecular Genetics, and Virginia Institute for Psychiatric and Behavioral Genetics, Virginia Commonwealth University, Richmond, VA, USA e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_13, © Springer Science+Business Media, LLC 2011

333

334

B. Riley

material or the environments they share. The answers to these questions explain why we are looking for schizophrenia liability genes and what we think we might find, both critical for appropriate study design. Study of the genetic epidemiology of schizophrenia has consistently demonstrated that (1) in aggregate, genetic effects are important risk factors for this disorder, (2) schizophrenia is not completely attributable to genetics: it is a multifactorial trait, with non-genetic (or environmental) factors also contributing to risk, and (3) patterns of transmission and linkage data in families are inconsistent with the action of individual genetic risk factors of large effect size (like Mendelian mutations) in one or a few single genes. Schizophrenia is a complex trait, one influenced by a large number of risk factors, most of which seem more likely to be within the range of normal human variation and to produce much more modest individual increases in risk. There is substantial difference in DNA sequence between individuals at variable sites known as genetic polymorphisms or markers, which have multiple detectable specific forms, or alleles (the specific DNA sequence at such a variable position). The second area, molecular genetics, asks whether certain alleles are more common in affected than in unaffected individuals. The most common approaches in human molecular genetics are studies of linkage (which ask whether a trait and specific large chromosome segments cosegregate − or are inherited together − in families) and association (which ask whether DNA variation in specific smaller chromosome segments, often individual genes, is more common in affected than unaffected individuals in populations). We will discuss the underlying causes of these two genetic phenomena, the methods for detecting them and the limitations of each. These are essential for a critical assessment of the large body of evidence, as they describe how we attempt to identify liability genes, and how well we succeed. Molecular genetic studies have, until very recently, been less successful in identifying the relationship between the aggregate genetic risks and specific DNA variants, protein molecules or biological processes for complex traits generally. A number of features of complex traits like schizophrenia contribute to an overall reduction in power, the probability that a true effect will be detected, particularly for linkage studies. First, schizophrenia is thought to be influenced by multiple common alleles of small effect. Both linkage and association study designs rely heavily on analytic approaches that assess single genes, which are less powerful for detecting multiple risk factors of small effect. Second, environmental factors and interactions appear necessary to account for patterns of risk. These remain unknown and generally untested. Third, the disorder is common, and genetic liability variants seem likely to be common, although increased rates of rare deletions and duplications (structural or copy number variants) in cases have been observed multiple times and suggest that rare variation may also contribute to disease in a proportion of cases. The common risk variants are expected to occur with relatively high frequency in the general population, reducing contrast between affected and unaffected individuals and reducing power. The impact of individual rare structural variants in the subset of cases where they are observed is harder to assess currently, but the observation of an aggregate increase appears robust, further increasing the

Genetic Studies of Schizophrenia

335

apparent etiological complexity. Fourth, the expected frequency of risk alleles and the clinical variability in presentation, course and outcome suggest that the etiology of individual cases may be heterogeneous, derived from different specific genes or alleles between individuals. Fifth, diagnostic boundaries are difficult to draw and there is some evidence for genes influencing risk of both schizophrenia and bipolar disorder, disease entities considered separate. Incorrect specification of affected status can substantially reduce power. Linkage methods, which deliver relatively complete coverage of the genome, have great power to identify single genes causing Mendelian disorders but are poorly suited to the genetic architecture of complex traits. Association methods are undeniably more powerful in such situations, but our limited understanding of the underlying disease neurobiology in the brain means that candidate gene selection has been weak, and such studies of schizophrenia have not generally produced robust and replicable results. Affordable technologies to deliver higher density data required for unbiased genomewide association testing have only recently become available and been applied to multiple large case/control samples. In spite of these limitations, numerous regions of the human genome give consistent, though by no means unanimous, support for linkage. The precise nature of the linkage signals is not yet understood, and power to position the effects is poor, but meta-analyses show the co-occurrence is unlikely to be due to chance. Combined approaches utilizing linkage for genomewide coverage and association for fine-scale follow-up have identified several promising positional candidate genes. A number of these reported associations have been widely, though again not unanimously, replicated. A clear definition of replication in a complex trait remains difficult to achieve, but the association data do not satisfy the most rigorous definition of replication, because the same alleles are not observed to have the same effect on risk across samples. The field is set to be advanced further by the current round of genomewide association studies (GWAS). The third key area, molecular biology, undertakes functional studies of the effects of genes and gene variants to elucidate how a gene functions and is normally regulated, how specific DNA variants alter normal processes and how these alterations contribute to specific disease states. This chapter provides a broad overview of the first two areas, an integration of the collected results of many studies and a discussion of the most promising current areas.

2 Genetic Epidemiology: Why Are We Looking for Genes Contributing to Schizophrenia? A large body of data collected from families, twins and adoptees over many years has consistently supported the involvement of a major, complex genetic component in liability to schizophrenia and schizophrenia spectrum disorders. These results indicate only that genes contribute to risk in aggregate, and provide no information about the specific genes involved or their number.

336

B. Riley

2.1 Family Studies: Does Risk Aggregate in Relatives? The first systematic family study found that dementia praecox was more common among the siblings of probands than in the general population (Rudin, 1916). A large study of over 1,000 schizophrenic probands showed that both siblings and offspring had increased rates of the disorder (Kallmann, 1938). There was early recognition of the need for systematic ascertainment (unselected sampling of cases, e.g., from consecutive admissions referred to a clinic) of probands (the index cases) for family studies so that the affected individuals sampled were maximally representative of schizophrenia as a whole, but some early studies lack this critical feature. In order to make accurate estimates of the lifetime morbid risk to various classes of relatives, family studies also need to correct for the age of the subjects. Only those who are beyond the age of risk can be unequivocally classified as unaffected, and some relatives unaffected at the time of study may develop the disorder in the future. Lifetime morbid risks are therefore calculated by dividing the number of affected by an age-corrected total of lifetimes at risk. The combined results of many European studies published between 1921 and 1987 (Gottesman, 1991) are shown in Fig. 1. The lifetime morbid risk (MR) in the general population is about 1%, but approximately 10 times that in the siblings or offspring of patients with schizophrenia. Smaller but consistent increases in risk are seen in second and third degree relatives. Two anomalies in the family data merit some discussion. Lower risk in parents (6%) compared to siblings (9%), both of whom share 50% of their genetic material with probands, may be due to the substantial Lifetime Morbid Risk of Schizophrenia in Various Classes of Relatives

60% 50% 48% 46%

MR

40% 30% 20%

17% 17%

13%

4%

0% Half-siblings

Parents

Siblings

Children

Offspring with 1 Affected Parent

DZ Twins

Offspring of Dual Matings

MZ Twins

Relative Type Fig. 1  Risks for schizophrenia in various classes of relatives

2%

2%

1% Baseline

5%

First Cousins

6%

Uncles/Aunts

6%

Nephews/Nieces

9%

Grand children

10%

Genetic Studies of Schizophrenia

337

reduction in fertility observed in schizophrenia: if affected individuals are less likely to be parents, then they are less likely to be the parents of probands, thus reducing the estimated parental MR. Lower risk in siblings (9%) compared to fraternal or dizygotic (DZ) twins (17%), again both of whom share 50% of their genetic material with probands, has not been explained, but suggests that some environmental factors more highly shared between twins than between singleton siblings, such as the intrauterine environment, may be involved. Neurodevelopment, a potential target of prenatal effects and a time of critical regulation of gene expression, has long been thought to be involved in the pathogenesis of schizophrenia. Criticisms of the methodology of early family studies included the lack of systematic ascertainment discussed above, the lack of proper controls (since rates of schizophrenia were not assessed in the families of unaffected individuals), lack of standardized diagnostic criteria and failure to diagnose family members blind to the status of the proband (which can produce bias in the diagnoses). A combined analysis of data from seven studies designed to avoid these weaknesses yielded totals of 15 cases of schizophrenia in 3,035 lifetimes at risk in the control families, and 116 cases of schizophrenia in 2,418 lifetimes at risk in the patient families. These translate into average morbid risk for narrowly defined schizophrenia of 0.5% for relatives of controls and 4.8% for relatives of a proband with schizophrenia (Kendler & Diehl, 1993). The more recent, methodologically stronger studies replicate the early findings that a close relative of a schizophrenic patient has an average ten times the baseline population risk of the disorder.

2.2 Twin Studies: How Large Is the Genetic Component of Risk? Studies of twins provide a way to estimate the relative importance of shared genetic material and shared environment in the development of a trait. The main strategy in twin studies is to compare the concordance for the disease between members of monozygotic (MZ) twin pairs and members of dizygotic (DZ) twin pairs. Since MZ twins are genetically identical whereas DZ twins share on average 50% of their genes, greater MZ than DZ concordance will reflect genetic influence, providing both MZ and DZ twins share their environment to approximately the same extent. Systematic ascertainment is also essential in twin studies, as bias here can result from the preferential selection of the most prominent twin pairs, which are likely to be MZ and concordant for the disorder. Since individual twins in a pair may be ascertained independently as two separate probands, the proband-wise concordance rate (where concordant pairs are counted twice if both co-twins are independently ascertained) is statistically preferable to the pair-wise concordance rate (where the pair is only counted once regardless of dual ascertainment), and provides an unbiased assessment of morbid risk for each co-twin that can be compared with the morbid risk to other relatives and the general population.

338

B. Riley

Twin studies of schizophrenia show consistent evidence of a genetic effect with higher concordance in MZ (35–58%) than DZ (9–27%) twins. Individual twin studies (Cardno et al., 1999; Farmer, McGuffin, & Gottesman, 1987) and meta-analyses of twin studies (Sullivan, Kendler, & Neale, 2003) estimate the heritability of liability to schizophrenia to be approximately 80%. Twin studies of schizophrenia have been reviewed in detail by Cardno & Gottesman (2000), whose meta-analysis of rigorously conducted twin study data suggested somewhat lower overall concordance rates, but similarly high heritability. Twin studies have been criticized for making the assumption that the environments are equal between members of MZ and DZ twin pairs because MZ twins are more likely to act and be treated more similarly, and so may have a greater degree of shared environment than DZ twins. However, a search for pairs of MZ twins reared apart in a review of all systematic twin studies of schizophrenia found that 7/12 (58%) such pairs of MZ twins (who do not share their environment) were concordant for schizophrenia, a rate similar to MZ twins reared together (Gottesman & Shields, 1982). This argues against MZ-specific increases in shared environment as the source of the increased concordance compared to DZ twins. The results from studies of twins argue that liability to schizophrenia is largely genetically mediated, because of the substantially higher risk in MZ compared to DZ co-twins, but not genetically determined, because the risk to an MZ co-twin of a proband is less than 100%.

2.3 Adoption Studies: Is Familial Aggregation Due to Shared Environment? Family members share environments, in addition to sharing genetic material, so the effects of shared genes and shared environments are not truly separable. Adoption studies ask whether the increased risk observed in family members is still present when the relatives do not share their environments. Across all adoption studies performed, the increased risk of schizophrenia was present in the biological relatives of patients with schizophrenia (Prescott & Gottesman, 1993). The first adoption study found that 5/47 (16.6%) adopted-away offspring of schizophrenic mothers had schizophrenia compared to 0 of 50 adopted-away offspring of control mothers (Heston, 1966). In Finland, a much larger study of adopted away offspring of schizophrenic and control mothers found that 13/144 (9.1%) children of schizophrenic mothers had a schizophrenia spectrum disorder and 7/144 (4.9%) had schizophrenia, while 2/178 control offspring (1.1%) had schizophrenia (Tienari, 1991). In studies of adoptees in Denmark, schizophrenia was found to be significantly more common in the biological relatives of schizophrenic adoptees than in the biological relatives of control adoptees in both urban and non-urban samples (Kety, Rosenthal, Wender, Schulsinger, & Jacobsen, 1968; Kety et al., 1994). The rates of schizophrenia were low and not different in the adoptive families of both affected and control groups.

Genetic Studies of Schizophrenia

339

2.4 Transmission Models: How Is the Genetic Risk Transmitted? Transmission models are mathematical expressions of a hypothetical genetic architecture for a trait and make specific predictions about risk in various classes of relatives. Such models are tested by assessing how well their predictions match observed patterns of risk. The risk patterns for single gene Mendelian disorders are expressed using the generalized single major locus (SML) model, which assumes that a single gene is responsible for all liability. These two terms, gene and locus (pl. loci) are often used interchangeably, though they have specific meanings. A locus is simply a position on a chromosome (e.g., a polymorphic marker) which may or may not be within a gene, itself defined as the specific DNA template for a polypeptide or functional RNA molecule. An analysis of the data from a carefully selected number of European studies demonstrated that the SML model is inadequate to explain observed patterns of risk for SZ (McGue, Gottesman, & Rao, 1985). Both segregation analysis and the pattern of risk in families (~10% in firstdegree relatives and ~50% in MZ co-twins of probands) are inconsistent with the action of highly penetrant mutations in single genes. The offspring of unaffected MZ co-twins have an elevated risk of illness, and there are no pedigrees in which schizophrenia is transmitted as a classical Mendelian disorder, unlike Alzheimer’s disease or breast cancer, where a well-recognized series of pedigrees exist in which the disorder segregates in a Mendelian manner. In common with other complex traits, the conclusion has generally been that multiple liability, susceptibility or predisposing alleles seem most likely to account for the aggregate genetic effects observed in genetic epidemiology studies. One consequence of such a model is that, if the number of genes contributing to a trait rises, the frequencies of the risk alleles must also rise in the population in order to maintain the observed prevalence. Risk alleles for most complex traits are expected to be common, and to be found at appreciable frequencies in controls or in the population (and elevated frequencies in cases). Data from studies of numerous complex traits are consistent in showing little evidence for risk alleles of major effect size. Effect sizes of single gene mutations are described by penetrance (the proportion of individuals with a mutation who develop a disease, usually 1 for Mendelian mutations, where the presence of the mutation determines disease state in a 1:1 correspondence) while the odds ratio (OR), which compares the risk in those carrying a particular variant to those not carrying the variant, is currently more widely used. The OR can take values from less than 1 (indicating reduced risk in carriers and protective effects) to 1 (indicating no difference between carriers and non-carriers) to greater than 1 (indicating increased risk in carriers and liability effects). The alleles predisposing individuals to complex traits appear unlikely to have odds ratios greater than 1.1–1.5 (a 10–50% increase in risk relative to baseline). This has been well validated by the initial round of GWAS of complex traits (Altshuler & Daly, 2007; Petretto, Liu, & Aitman, 2007). Thus, we expect the genetic risk for schizophrenia to be mediated by common alleles of small effect in multiple genes.

340

B. Riley

The mutations that cause most Mendelian disorders are sufficiently rare that, for all practical purposes, the same disease gene will be responsible for all cases of illness in a family. Between families, the pattern with Mendelian disorders is variable. In most disorders, mutations at a single gene are responsible for all known cases of illness (e.g., Huntington’s disease, cystic fibrosis). However, for some Mendelian syndromes (e.g., limb-girdle muscular dystrophy and retinitis pigmentosa) a number of distinct genes, usually on different chromosomes, have been found in different subsets of families. Critically, within an individual family, the same gene and mutation are responsible for all cases of disease. Epidemiological data from schizophrenia suggest that the population frequencies of liability variants are likely to be orders of magnitude greater than the frequency of even the most common single-gene mutations (like cystic fibrosis which has a high population allele frequency of 2.5%). Schizophrenia is much more common and clinically highly variable. It is widely believed that variants in many genes can increase liability to schizophrenia (a polygenic model of total population risk). These are not thought to all be necessary for disease but rather that a few of these genes predispose an individual to schizophrenia (an oligogenic model of individual risk). A consequence of this model is that the specific genetic risk factors are likely to differ between individuals. The implication is that between cases, even those within families, variants in both shared and distinct genes may predispose to illness. The expected frequency of risk alleles and the clinical variability in presentation, course and outcome suggest that heterogeneity in etiology is likely at the level of the gene (different risk genes in different individuals) and possible at the level of the allele (different risk alleles in the same gene in different individuals). The multifactorial threshold (MT) model is less genetically deterministic, and assumes a continuum of liability due to multiple genetic and environmental factors, based on the collected data above. Individuals with liability in the upper reaches of this continuum have substantial risk of developing schizophrenia. The MT model, which has great intuitive appeal but poor predictive power, assumes that the effects of multiple genes on risk combine additively (the total liability from n genes is equal to the sum of the n individual liabilities) and that there is no interaction between them. The segregation analysis above also demonstrated that the MT model is inadequate to explain observed patterns of risk (McGue et al., 1985). A key predictive failure of the MT model is that the observed concordance rate in monozygotic twins (~50%) is too high relative to the risk in siblings and dizygotic twins (9–27%). Such a pattern is more consistent with some degree of gene–gene (G×G) or epistatic interaction between genes, where the total liability from n genes is greater than the sum of the n individual liabilities. The fall-off in concordance rates for schizophrenia in first, second and third degree relatives in the families of patients with schizophrenia is also most consistent with multiple epistatically interacting genes (Risch, 1990) although this study did not model the impact of reduced fertility, which is large in individuals with schizophrenia and could substantially bias the results. Conversely, if the inheritance model was fully epistatic, then the tetrachoric correlation (the correlation in the normally distributed liability to illness) in MZ twins should be substantially more than twice that seen in DZ twins, which is

Genetic Studies of Schizophrenia

341

not observed. Additionally, recent studies have suggested that gene–environment interactions (G×E) are important components of the overall risk for conduct disorder (Caspi et  al., 2002; Foley et  al., 2004) and depression (Caspi et  al., 2003; Kendler, Kuhn, Prescott, Vittum, & Riley, 2005) and possibly also for schizophrenia (Tienari et al., 2004). Although segregation analyses have not provided a definitive inheritance model for the genetic risk for schizophrenia, they suggest collectively that numerous kinds of influences (including additive and epistatic genetic risks, environmental risks and G×E interactions) are involved, and substantially reduce the power of genetic study designs. Common risk alleles with small effect sizes reduce the contrast between affected and unaffected individuals, whether they are siblings in a linkage study or cases and controls in an association study. Large samples are needed to detect such effects, and many past studies have clearly been underpowered. Environmental risk factors also appear critical; several, including obstetric complications and intrauterine influenza infections, have been suggested to increase risk of illness, and some convincing epidemiological support for such risk factors is emerging. The evidence for one such factor is particularly compelling: subjects who were in utero during the very severe famine in Holland in the winter of 1944−1945 had a two-fold increase in their risk for schizophrenia (Susser & Lin, 1992). While intriguing, it was hard to see how such results could be easily verified. However, in 2005, using data from the severe famine in regions of China in 1959−1961, the results of the earlier Dutch study were robustly replicated (St Clair et al., 2005). We would expect the products of interacting genes (or their functions) to be biologically related, functionally (e.g., the subunits of heteromultimeric neurotransmitter receptors), temporally (e.g., in neurodevelopmental cascades) or spatially (e.g., co-localized cell surface receptors). Epistatic interaction is only possible if the risk allele or genotype (the pair of alleles at a variable site present on an individual’s pair of chromosomes) at one gene moderates the effect of the genotype at another on the phenotype (the observable effect of the genotypes, here risk for schizophrenia). Where unbiased, genomewide data are available for analysis, pathway and systems analysis approaches may help to isolate potentially interacting subsets of genes. It is less clear how environmental risk factors may interact with genetic vulnerabilities, and what kind of relationship we would expect between them.

2.5 Spectrum Disorders: How Broad Is the Range of Psychiatric Illness Transmitted and Who Do We Consider Affected? Kendler & Diehl (1993) also reviewed results from studies of other illnesses in relatives of patients with schizophrenia. Again, only studies that used rigorous methodology, including personal interviews, structured diagnostic criteria and blind diagnoses, were considered. The results are extremely variable across studies and

342

B. Riley

different conditions examined. In 5 of 7 studies, the risk of schizotypal or paranoid personality disorders in relatives of patients with schizophrenia are consistent at 4−4.5 times that in the control families. In studies (overlapping those above) which meet the same criteria and examine the risk of schizoaffective, schizophreniform and delusional disorders, and atypical psychosis in the relatives of patients, 5 of 7 also showed significantly higher risks of these conditions. The risk of psychotic spectrum disorders is clearly elevated in the relatives of schizophrenia probands. Two other studies, examining the converse of the question, found that the risk of schizophrenia is significantly higher in the relatives of individuals with schizoaffective and schizophreniform disorders than in controls (Kendler, Gruenberg, & Tsuang, 1986; Kendler et al., 1993). Studies of wider spectra of psychopathology, including unipolar and bipolar affective illness, anxiety disorder and alcoholism show more ambiguous results. In studies examining the risk of affective disorders, 6 of 9 find no significant difference between relatives of patients with schizophrenia or schizoaffective disorders and relatives of controls, consistent with the generally accepted dichotomy between psychotic and affective illness, but it is important to note that a third of studies do detect excess risk of affective disorders. There is evidence emerging from the molecular studies described below that individual genes associated with schizophrenia may also contribute to risk for other mental illness, particularly bipolar disorder. In studies of other conditions, 5 of 6 studies of anxiety disorder and 4 of 5 studies of alcoholism found no significant increase in risk in the relatives of schizophrenic probands. A twin study which explored the DSM III definition of schizophrenia found evidence of an increased degree of genetic determination when concordance was broadened to include categories such as schizophreniform and atypical psychosis. However, when co-twins with a broader range of conditions including major depression were included as “affected” the evidence of a genetic effect, as reflected in the monozygotic to dizygotic concordance ratio, fell markedly (Farmer et  al., 1987). Answers to this question are particularly important for molecular genetic investigation of schizophrenia, since they specify who is considered affected. Misclassification of affected individuals causes great loss in power to detect genetic effects. As is clear from the summary above, the boundaries of psychiatric illnesses are unclear, and consequently, we do not know exactly where to set definitions of illness for classifying individuals. There is obviously a great difference between calling a pair of siblings affected if both are diagnosed with schizophrenia and calling the same pair affected if one is diagnosed with schizophrenia and the other with a personality disorder, or one with schizophrenia and one with alcohol dependence. Though the field as a whole concurs that we do not know where to set the boundary, the approximation we choose in the meantime is the source of considerable controversy. It has become reasonably common to perform several analyses of data using a number of different definitions of illness. This has the benefits of allowing different genes to influence different ranges of pathology, and making no assumptions about the range of pathology to which any gene predisposes a person.

Genetic Studies of Schizophrenia

343

3 Methods: Where and How Will Such Genes Be Found? 3.1 Approaches: What Are Linkage and Association? Linkage and association are two fundamentally different genetic effects with quite different strategies for their detection. Both are widely used in the search for schizophrenia susceptibility genes, but because of the differences in the effects and their detection, they were applied to very different research questions, and so have generally been considered separately. Association studies tended in the past to be focused on candidate genes, and the grounds on which candidates were selected were often relatively weak. Linkage studies have always collected data genomewide and implicitly make no assumption about specific genes involved in etiology, but linkage is relatively weak when applied to complex traits. In the past 10 years, a first important shift in the field was that association studies have tended more and more to focus on chromosomal regions supported by multiple linkage studies. The sequential application of these two approaches has produced the most exciting current results, including a small but growing number of specific genes, which we will focus on, for which multiple groups have found support (note that additional genes may have been reported to be associated in the genomic regions we discuss below, but only those with substantial positive replication evidence are included here). In a few cases, genes with known molecular interactions with the candidates have also generated replicated association evidence. A number of these best-supported candidate genes also provide some degree of evidence for association with other psychiatric phenotypes, particularly bipolar disorder, suggesting that there is some etiological overlap between conditions. It is only in the last 2 years that association studies of schizophrenia have begun to assess genomewide data completely independent of linkage studies.

3.2 Linkage Classical single-gene, or Mendelian, genetic illnesses are caused by mutations in a single gene, located at a single place on a chromosome. Because these illnesses are rare, the rare risk allele must segregate from parents with a family history into affected offspring, or arise as an even rarer de  novo mutation. By following the segregation of marker alleles from the affected lineage into offspring, chromosome regions in which affected offspring inherit one marker allele and unaffected offspring the other can be identified. This phenomenon of chromosomal regions segregating with a trait in multiple families is called linkage. It is detectable because there is crossover, a physical exchange of material, between the chromosome pairs during meiosis, the cell division resulting in the production of eggs or sperm. Recombination, the occurrence of chromosomes with new combinations of alleles, is observed genetically and is the result of this crossover (Fig. 2). It also provides a

Loci far apart

a

Loci close together

A

A

a

a

B

B

b

b

A B

a b

A B

a b

b

A

A

a

a

B

B

b

b

A

A

a

a

B

b

B

b

A B

A B

a b

a b

A B

A B

a b

a b

c

Meiosis I

d

A

A

a

a

B

b

B

b

A B

A B

a b

a b

Recombinants

Fig. 2  Recombination between homologous chromosomes in meiosis. The left diagram illustrates two loci which are far apart on the same chromosome. These loci have an even chance that they will be separated by crossovers at meiosis. On the right, the two loci are close together so they are less likely to recombine. (a) Homologous chromosomes condensed prior to meiosis. (b) Pairing of homologous chromosomes. (c) Crossover and physical exchange of material. (d) Recombinant chromosomes resulting from crossover (left) and non-recombinant chromosomes (right)

Genetic Studies of Schizophrenia

345

common measure of genetic distance, the centimorgan (cM), which is equivalent to 1% recombination between two loci. If two genetic loci are on different chromosomes, the probability that they are inherited together will be 0.5. This means that the theoretical maximum genetic distance is 50 cM. This phenomenon of independent assortment, as Mendel described it, is also true for two loci far apart on the same chromosome when there is a 50% chance that they will be separated by crossovers at meiosis. On the other hand, linkage is observed between two loci when they are in such close proximity on the same chromosome that their alleles are separated by crossover less than half the time. This departure from independent assortment can be statistically tested for.

3.3 Linkage Analysis Methods Two different forms of linkage analyses are in common use. Parametric analyses require the specification of a genetic model. Where these models can be specified accurately, parametric linkage is a powerful approach. However, the difficulties discussed above in defining a transmission model and specifying who should be considered affected suggest that it is much less powerful for analysis of schizophrenia. The parametric linkage statistic is the LOD score – the Logarithm of the Odds. A LOD score is the logarithm, to the base 10, of the ratio of the likelihood of the observed data given linkage divided by the likelihood of that same data given no linkage. A LOD score of +3 (which corresponds to a genome-wide significance level of p = 0.05; Lander & Kruglyak, 1995; Morton, 1955) thus means that the likelihood that the observed families are linked is 1,000 times the likelihood that they are unlinked (and the likelihood ratio is 103 or 1,000:1); a LOD score of −2 (which is generally accepted as significant evidence against linkage) means that the likelihood that the observed families are unlinked is 100 times the likelihood that they are linked (and the likelihood ratio is thus 10−2 or 0.01:1). Because linkage analysis is likelihood-based and likelihood-based tests can be maximized over one or more parameters, a common technique is to vary parameters and use likelihood maximization to choose the best parameter value (e.g., the recombination fraction, Q, which allows the assessment of the likelihood of linkage at several different genetic distances from a marker). Thus, the focus of linkage analysis is often comparison of the relative likelihood of the data under one parameter value compared to another. Variations of the LOD score which allow for heterogeneity, or which assess only the affected individuals in a sample, have been applied to address some of the problems outlined above, but we will not distinguish between these in the discussion that follows. An alternative approach, widely used in complex traits, is nonparametric linkage analysis (Kruglyak, Daly, Reeve-Daly, & Lander, 1996). All classes of relatives have predefined probabilities of sharing zero, one or two alleles at a random marker locus. These nonparametric statistics are based on testing for deviations from expected allele sharing distributions, and avoid the problem of specifying a transmission model, which, as we saw above, is very difficult for schizophrenia.

346

B. Riley

3.4 Association Association studies examine whether individuals affected by a disease more frequently have particular genetic variation in a gene than individuals not affected by the disease. This association can occur for two reasons. Either the allele being studied directly influences risk for the disorder or, more commonly, the allele is in linkage disequilibrium (LD) with the disease-predisposing mutation. Linkage disequilibrium means that specific alleles at two nearby loci tend to occur together in an entire population. Linkage (the co-segregation of a chromosome region and a disease observed in families) occurs at scales of tens of millions of base pairs, while association (and LD) are seen at scales of tens of thousands of base pairs. LD is a reflection of an evolutionary history in which, because of these very small distances, recombination occurs very rarely between the two loci. LD occurs because a new variant (i.e., polymorphic marker or mutation) always arises on a specific background chromosome, and will, until separated by recombination, only exist in conjunction with the other alleles present on that background. Over time, the original LD (and thus the genetic association) between more distant loci decay as a result of recombination events, while the rarity of recombination between nearby loci preserves some or most of the original LD and association.

3.5 Association Analysis Methods The simplest genetic association tests ask if specific alleles, genotypes or haplotypes, the combination of alleles on one of the pair of chromosomes, are more common in cases than in controls. Association approaches have two important advantages when compared to linkage. First, individual patients can be studied rather than families. Second, under many circumstances, a sample of equal size has considerably more power to detect association than to detect linkage for a gene of modest effect (Risch & Merikangas, 1996). However, they have two potential disadvantages. First, association studies have in the past only been able to examine much smaller regions of the genome than linkage studies. Because of the very different scales on which the two effects occur, association studies require more than 1,000 times the density of data in linkage studies. Practically, this means that association studies were used for the assessment of candidate genes or regions only, although this has changed dramatically with the arrival of affordable methods for the genomewide association study (GWAS), which provides unbiased assessment of association across the entire genome. Second, association can occur for spurious reasons unrelated to disease etiology, such as population stratification, where the cases and controls come from different population groups or sub-groups, and observed genotypic differences are due to this population difference, rather than to true association between marker and phenotype. This particular issue, although real, appears less significant than once thought provided controls are sampled from the same population as cases.

Genetic Studies of Schizophrenia

347

A number of analytical developments have improved association studies. Transmission disequilibrium testing (TDT) (Spielman & Ewens, 1996) assesses whether certain alleles or haplotypes, are transmitted to affected individuals more often than expected by chance. Pedigree disequilibrium testing (PDT) (Martin, Monks, Warren, & Kaplan, 2000) and family-based association testing (FBAT) (Laird, Horvath, & Xu, 2000) utilize both the “vertical” transmission information and the “horizontal” information in concordant and discordant siblings. Both of these have furthered the pursuit of association studies in family samples originally collected for linkage studies. The problem of stochastic sampling variation, the random variation of the predisposing genes represented in different family or casecontrol samples, and heterogeneity (among others) all suggest that the best place to follow up a linkage finding is in the family sample which produced it. Association studies in schizophrenia (and other psychiatric disorders) have tended in the past to focus on a limited set of “usual suspect” genes, generally those coding for receptors, transporters and synthetic or degradatory enzymes in neurotransmitter pathways. Results from these studies have generally not been particularly robust and few replicated findings have emerged to date. While association studies remain a major interest in attempting to clarify the nature of the genetic liability to schizophrenia, a powerful, widely replicated finding has not emerged from the older applications of this technique alone.

3.6 Limitations of Linkage and Association Linkage is a powerful method for Mendelian disorders where a small number of families can usually unambiguously produce strong evidence for linkage to a small chromosomal region. As we saw above, schizophrenia (in common with many complex traits) differs from Mendelian disorders in many critical ways, all of which reduce the power of linkage approaches: (1) risk cannot generally be explained by the action of rare highly penetrant alleles in single genes; (2) environmental factors are critical to account for observed patterns of risk, and G×G and/or G×E interactions seem likely; (3) risk variants generally seem likely to be common; (4) both locus and allelic heterogeneity seem likely; and (5) diagnostic boundaries are unclear and risk from a predisposing allele may not be limited to schizophrenia or schizophrenia spectrum disorders. Association is more powerful generally for detecting genes of small effect (Risch & Merikangas, 1996), but the specific features listed above all reduce the power of association studies. Current approaches analyze genes individually, but the risk associated with an individual gene is likely to be small and may depend on interactions with other genes. Replication studies are hampered by two additional issues. Inadequate sample sizes have been used in many studies: in replication, the follow-up sample must be larger than the original discovery sample to maintain power. Under a polygenic model of risk in the population, substantial stochastic sampling variation is expected: each sample tested will vary in the extent to which any specific risk factor is present (and association

348

B. Riley

detectable). All of these issues should be borne in mind when assessing the evidence for the specific chromosomal locations that are discussed below.

3.7 DNA Polymorphisms Review of genetic marker types and their designations facilitates understanding the material that follows. Most linkage studies use a particular kind of DNA polymorphism, the tandem repeat. Tandem repeats are variable numbers of a repeating unit of nucleotides, from mononucleotides (rarely used due to the difficulty of accurate genotyping) through dinucleotides (the most common and most commonly used), tri-, tetra- and pentanucleotides. The frequency of polymorphisms of various unit sizes in the genome is generally inversely correlated with the length of the tandem repeat unit. They are referred to variously as microsatellites, short tandem repeats (STRs), simple sequence repeats (SSRs) or AC repeats (after the most common dinucleotide repeat sequence). The alleles of these markers are different segment lengths due to different numbers of the tandemly repeated unit. This marker type is very common and tends to be extremely polymorphic (i.e., to have many alleles) and therefore to have high heterozygosity (the proportion of individuals who have two different alleles at the marker locus). As a result of the high heterozygosity, they also tend to be extremely informative, where information content is defined as the probability that the allele transmitted to a given offspring from a given parent can be unambiguously determined. Nomenclature for markers of this type are “D numbers” which identify the chromosome on which the marker locus maps and the historical order in which the marker was identified. Thus, the 278th microsatellite identified on chromosome 22 (to be discussed shortly) has the identifier D22S278. In contrast, single or simple nucleotide polymorphisms (SNPs) are genetic variations involving common transition (purine to purine or pyrimidine to pyrimidine) or rarer transversion (purine to pyrimidine or pyrimidine to purine) changes at a single base or insertion/deletion variation up to a few nucleotides in size. SNPs generally have only two alleles, have lower heterozygosity and lower information content. Large numbers of alleles (and therefore high information content) are useful for linkage as they maximize the number of informative meioses. Association and LD studies tend, in general, to use SNPs as the marker of choice, because alleles of these markers evolve more slowly than microsatellites and preserve more of the evolutionary relationships on which LD and association are based. Standard nomenclature for these markers is the “rs number” catalog entry in dbSNP. The “common allele-position-rare allele” convention is common in older literature (e.g., the T102C polymorphism in the serotonin 2a-receptor (5HT2A) gene). Where SNPs occur in coding sequence and alter amino acid sequence, the amino acid change and position (now in the polypeptide) are often used (e.g., Ser311Cys polymorphism in the dopamine receptor 2 (DRD2) gene). We follow standard nomenclature for distinguishing between human genes and their protein products. Proteins are given as either the full name (e.g., catechol-O-methyltransferase) or abbreviation (COMT) in upper case Roman font. Where a gene or protein has more

Genetic Studies of Schizophrenia

349

than one name in current use in the literature (e.g., the approved nomenclature d-amino acid oxidase activator (DAOA) and the widely-used G72), both will be given. Gene symbols, usually the same as the protein abbreviations, are given in upper case italics (COMT).

4 Results: Where Is the Evidence Strongest for Schizophrenia Susceptibility Genes? Through 2004, 25 complete or nearly complete genome scans for schizophrenia (in which about 400 individual genetic markers are genotyped at regular intervals over the entire human genome) were published, and none revealed evidence for a gene of major effect for schizophrenia, consistent with the evidence reviewed above. A smaller number of genome scans examining specific clinical features of the illness (such as neuropsychological deficits, age at onset and positive, negative and disorganized symptoms) have been published. Finally, two meta-analyses of genome scan data using different statistical approaches and six meta-analyses of specific chromosomal regions have been published (reviewed in Riley & McGuffin, 2000; Sullivan, 2005). Some tentative evidence for replicated linkage for schizophrenia susceptibility genes has emerged from these studies. A number of promising genes have emerged from sequential linkage and association studies and multiple replication reports: in historical order, these are 22q12-q13, 8p22-p21, 6p24-p22, and 13q14-q32 and 1q32-42. Two additional regions with little support in the primary literature, 2p11.1-q21.1 and 3p25.3-p22.1, were among the most significant in a meta-analysis of schizophrenia genome scans. A number of other regions (including 5q22-q31, 10p15-p11, 6q21-q22 and 15q13-q14) have less strong summary evidence but have been reported multiple times. We must note that the interpretation of these results is controversial, particularly as the definition of replication for linkage to a complex trait remains uncertain (Baron, 1996; Kendler, Straub, MacLean, & Walsh, 1996). In the interests of brevity, studies that do not find evidence for linkage in these selected regions have been omitted, but it is important to bear this selective bias in mind when considering the data that follow. More detailed information about putative linkage regions can be found elsewhere (Riley & McGuffin, 2000; Sullivan, 2005). The discussion of associated genes in these regions includes only those with substantial support from replication.

4.1 Chromosome 22q Linkage Studies Initial evidence for linkage to chromosome 22q came from three markers spanning ~23 cM in the 22q13.1 region in the Maryland family sample (Pulver, Karayiorgou, Wolyniec, et al., 1994). A collaborative replication study in a total of 217 multiplex pedigrees did not confirm the linkage in the new samples (Pulver, Karayiorgou,

350

B. Riley

Lasseter, et al., 1994) but two other replication samples were positive (Coon et al., 1994; Polymeropoulos et al., 1994). Eleven groups contributed data from the most significant marker in the original sample and one replication, the dinucleotide repeat polymorphism, D22S278, to a large collaborative schizophrenia linkage study. There was excess sharing of alleles in these 620 affected sibling pairs (p = 0.006), particularly in 296 pairs with data available from both parents (p = 0.001). It is important to note, however, that the authors calculated that this locus is likely to account for no more than 2% of total variance in liability (Gill et al., 1996). The role of this region in liability to schizophrenia remains unclear, although the multiple positive findings seem unlikely to have occurred by chance. Additional interest in this region of 22q came from a known chromosomal rearrangement. The cosegregation of chromosomal anomalies or rearrangements with phenotypes resembling a particular disease has provided useful clues to the locations of the gene(s) involved, most notably in the positional cloning of the dystrophin (or DMD) gene. Velo-cardio-facial syndrome (VCFS) is caused by deletions at 22q11, near these linkage results for schizophrenia. Historically, about 10% of VCFS patients were thought to present with a psychotic phenotype, but more recent studies suggest much higher rates of 25–29% (Murphy, Jones, & Owen, 1999; Pulver, Nestadt, Goldberg, et  al., 1994). Conversely, preliminary results suggest that about 2% of adult onset and 6% of childhood onset schizophrenic patients have microdeletions in this region, in excess of the estimated general population frequency of such deletions of 0.025% (Karayiorgou et al., 1995). Although statistically significant, this excess of deletions is probably not enough to explain significant population attributable risk, but variation in genes in this region in individuals without a deletion may contribute to liability (see also Structural Variation below).

4.2 Chromosome 22q Candidate Genes 4.2.1 COMT The VCFS critical region contains the gene for COMT located at 22q11, involved in the degradation of catecholamines, and is genetically and functionally polymorphic with a variable amino acid, Val158Met. Val and Met alleles are of almost identical frequency. Studies of the COMT gene show mixed results, recently reviewed in Williams, Owen, & O’Donovan (2007). One study suggests that the high activity (Val) allele, through increased catabolism of dopamine in the pre-frontal cortex, may slightly increase the risk of schizophrenia and may explain some of the observed differences in cognitive performance and pre-frontal cortical functioning between cases and controls (Egan et al., 2001). Another study of COMT in a homogeneous population of Ashkenazi Jews used the largest case-control sample for schizophrenia then reported (Shifman et  al., 2002). Three SNPs in the COMT gene showed significant association with schizophrenia in ~720 cases and 2,000–4,000 controls.

Genetic Studies of Schizophrenia

351

In agreement with the study above, an association was found with the homozygous high activity genotype (Val/Val) and with the two other SNPs tested.

4.3 Chromosome 8p22-p21 Linkage Studies The Maryland family sample also gave the first evidence of linkage to chromosome 8p22-p21 (Pulver et al., 1995). A multicenter collaborative linkage study supported this putative locus with excess allele sharing at D8S261 (Levinson et  al., 1996). Data from pedigrees from numerous different ethnic backgrounds all support a locus on 8p as did a statistically robust meta-analysis (Lewis et al., 2003). These replication results are spread across about 15 Mb of sequence. One of the key points to note is that although numerous samples support a locus on this chromosome, comparison between individual studies is consistent with the possibility of multiple genes in the region, a feature of a number of linkage regions.

4.4 Chromosome 8p22-p21 Candidate Genes 4.4.1 NRG1 Following linkage evidence to chromosome 8p in Icelandic families, fine mapping with 50 markers across a 30-cM interval identified two risk haplotypes spanning a region of ~1 Mb within the gene for neuregulin 1 (NRG1) (Stefansson et al., 2002). Case-control samples from Scotland (Stefansson et al., 2003) and Ireland (Corvin et  al., 2004) have provided additional support for this locus and for haplotypes identical or closely related to those identified in the Icelandic cases. In replication attempts, 13 independent studies in multiple populations have provided support for association (Bakker et al., 2004; Fukui, Muratake, Kaneko, Amagane, & Someya, 2006; Georgieva et al., 2008; Hall et al., 2006; Li et al., 2004; Norton et al., 2006; Petryshen et al., 2005; Tang et al., 2004; Thomson et al., 2007; Turunen et al., 2007; Williams et al., 2003; Yang et al., 2003; Zhao et al., 2004), though not always with the specific haplotypes originally reported, while nine studies have not supported involvement of NRG1 in schizophrenia (Duan, Martinezk, et al., 2005; Hall, Gogos, & Karayiorgou, 2004; Ikeda et al., 2008; Ingason et al., 2006; Iwata et al., 2004; Rosa et  al., 2007; Thiselton et  al., 2004; Vilella et  al., 2007; Walss-Bass et  al., 2006). Studies of NRG1 have been recently reviewed (Tosato, Dazzan, & Collier, 2005). A meta-analysis of studies of NRG1 supported involvement of the gene in schizophrenia liability, but did not provide evidence supporting association of the most prominent marker in the original studies (Munafo, Thiselton, Clark, & Flint, 2006). In a pattern observed for a number of the best-supported schizophrenia genes, several studies have also shown association between NRG1 and bipolar disorder (Georgieva et al., 2008; Green et al., 2005; Thomson et al., 2005).

352

B. Riley

4.4.2 NRG1 and ERBB4 ErbB4, encoded by the ERBB4 gene, is a receptor for NRG1 and has important roles in neurodevelopment and the modulation of NMDA receptor functioning. Both activation of ErbB4 and suppression of NMDA receptor activation by NRG1 are increased in the prefrontal cortex in individuals with schizophrenia compared to controls (Hahn et al., 2006). This functional relationship prompted assessment of ERBB4 for association with schizophrenia. Both primary association in ERBB4 and evidence of interaction with NRG1 have been reported (Benzel et  al., 2007; Nicodemus et al., 2006; Norton et al., 2006; Silberberg, Darvasi, Pinkas-Kramarski, & Navon, 2006). Associated alleles in ERBB4 alter splice-variant expression (Law, Kleinman, Weinberger, & Weickert, 2007), and both NRG1 and ErbB4 protein are increased in the brain in schizophrenia. These results confirming potential interaction between genes in schizophrenia may be of particular importance.

4.5 Chromosome 6p24-p22 Linkage Studies The first evidence for linkage of schizophrenia to the 6p region came from studies of Irish families with a high density of disease (Straub et al., 1995). In data from 16 markers, evidence for linkage was modest under a narrow diagnostic model, but increased substantially as the diagnostic definition broadened to include spectrum disorders. Evidence for linkage fell when the definition was broadened further to include non-spectrum disorders, in keeping with the risk in relatives for these traits discussed above. Multiple independent reports of analyses of this region of 6p have been published. Studies of German and mixed German and Israeli pedigrees supported linkage to 6p24-p22 (Moises et  al., 1995; Schwab et  al., 1995). A family sample from Quebec found supportive evidence for a schizophrenia susceptibility locus in some but not all families (Maziade et al., 1997). A large, multigeneration family from Sweden supported this linkage in a single branch of the family; a haplotype of markers within the putative linked segment was found to segregate with schizophrenia (Lindholm et al., 1999). A large, multicenter collaboration detected significant excess allele sharing in this region (Levinson et al., 1996) as did the best meta-analysis (Lewis et al., 2003).

4.6 Chromosome 6p24-p22 Candidate Genes 4.6.1 DTNBP1 It is useful to understand the extent of diversity in reported association findings in studies of complex traits. The association of the dystrobrevin binding protein 1 or dysbindin (DTNBP1) gene was first reported in the Irish high-density family set

Genetic Studies of Schizophrenia

353

5’---------------3’ r r r r r r r r s s s s s s s s 1 1 2 7 2 2 1 3 4 0 6 6 0 6 0 2 7 1 1 0 0 1 1 1 4 8 9 7 5 9 1 3 6 3 5 6 9 5 3 2 0 8 2 1 7 2 1 0 5 1 2 6 8 3 7 A C T C G G G A / 1 1 1 1 1 1 1 1 | HT 1, 0.733 | /----------| | HT | /--------------------------------------8 | | | 11 /------------9 \----------| | | ATGCAAGA HT +--------10 \-------------------------- 12212211 | | ACTTGGGA HT 6, 0.015 | \ 11121111 | HT 4, 0.010 | | ACTCGGAA \---------- 11111121 HT 3, 0.071

GCGTAAGG 21222212 2, 0.058 GTGTAAGA 22222211 5, 0.060

Fig. 3  Original evolutionary tree produced from DTNBP1 SNP data in the Irish Study of High Density Schizophrenia Families (ISHDSF) sample (van den Oord et  al., 2003). Haplotype ID number (HT 1–6) and population frequency in the ISHDSF sample are shown for each individual haplotype. The SNPs are arranged in genome sequence order and 3¢–5¢ in DTNBP1. The underlined haplotypes (1, 2, 5 and 6) have all been associated with schizophrenia in multiple independent samples. 1 under a nucleotide indicates it is the common or major allele at that position, while 2 indicates the rare or minor allele

(Straub et al., 2002). Subsequent reanalysis of the genotypic data showed that eight markers (rs1474605/p1792–rs3213207/p1635) covering 30.1 Kb from intron 1 to intron 4 yield a stable haplotype structure (Fig. 3) of six common (frequency >1%) haplotypes accounting for 94.7% of all haplotypes observed. A single haplotype (haplotype 2, Fig.  3) was associated with schizophrenia in the Irish high-density family sample (van den Oord, Sullivan, Chen, Kendler, & Riley, 2003). Given the multiple ethnicities represented in published studies, it is not surprising that there is some variation in the details of the DTNBP1 LD structures reported, particularly in the haplotype complement observed due to fluctuating frequencies of rarer haplotypes around the cutoff value of 1%. Overall, however, the region of LD, the pattern of specific haplotypes and their relative frequencies are generally maintained across other reported samples where sufficient SNPs are genotyped to allow assessment. DTNBP1 has been widely associated with schizophrenia in samples from diverse ethnic backgrounds: 13 published studies of 15 independent samples reported significant positive associations with SNPs and/or haplotypes in the gene (Funke et al., 2004; Kirov et  al., 2004; Li et  al., 2005; Numakawa et  al., 2004; Schwab et  al., 2003; Straub et al., 2002; Tang et al., 2003; Tochigi et al., 2006; Tosato et al., 2007; Van Den Bogaert et  al., 2003; van den Oord et  al., 2003; Vilella et  al., 2007; Williams, Preece, Morrisk, et al., 2004), while 14 published studies of 18 independent samples showed no evidence for association of this gene with schizophrenia

354

B. Riley

140 kb 30 kb

1765

T G

C T

G A

G A

Irish

G

C

German Swedish German Polish Chinese UK UK Irish Bulgarian US (C) US (H)

C T

T T C C T

A A

A A G G

A

G A A G

C G

C G T A

C T

G G A A A

C

A

G G G G

A

C

T

Adapted from Funke et al, 2004 Straub et al, 2002 Van den Oord et al, 2003 Schwab et al, 2002

A A A

G T T T

A

G G G G

T

G

A A A A A

T T T

C C

Tang et al, 2003 Williams et al, 2004 Kirov et al, 2004 Funke et al, 2004

G G

US (C) Afrikaans Chinese

G G

Numakawa et al, 2004 Hall et al, 2004

G

G G

A Li et al, 2005

Scottish C

A

Canadian Australian Indian Japanese

De Luca et al, 2005 Holliday et al, 2006 A G

Korean Finnish Italian

10

Van den Bogaert et al, 2003

G Japanese

G G T T

9 1328

1757

C T

8 1287 1333

1320

A G

7

1655

1763

G A

6

1635

1578

Major A Minor T

5

1325

1792

4

1583

2 3

SNP A

1

C C

G G

G G

A A

G C

C T

Tochigi et al, 2006 Joo et al, 2006 Turunen et al, 2007

Tosato et al, 2007 C A Spanish A Vilella et al, 2007 Taiwanese Liu et al, 2007 US (C) US (AA) Pedrosa et al, 2007 US Wood et al, 2007 Dutch Bakker et al, 2007 UK Datta et al, 2007 US Duan et al, 2007 Australian Peters et al, 2008 European Sanders et al, 2008 SNP A: rs2619538; p1583: rs909706; p1792: rs1474605; p1578: rs1018381; p1763: rs2619522; p1320: rs760761; p1757: rs2005976; p1765: rs2619528; p1325: rs1011313; p1635: rs3213207; p1655: rs2619539; p1287: rs760666; p1333: rs742105; p1328: rs742106;

Fig. 4  Summary of published association studies of DTNBP1 and schizophrenia. Alleles associated in single marker analyses are shown unshaded, and shaded lines are haplotypes associated, if any were reported. Dashes indicate markers were tested but not associated in a study, blank cells indicate markers were not tested in a study. Dotted vertical lines indicate the 8 markers (1792– 1635) forming the LD block and resulting haplotypes in Fig.  3. C Caucasian; H Hispanic; AA African American in the US studies

(Bakker et al., 2007; Datta et al., 2007; DeLuca, Voineskos, Shinkai, Wong, & Kennedy, 2005; Hall et  al., 2004; Holliday et  al., 2006; Joo et  al., 2006; Liu et  al., 2007; Morris et al., 2003; Pedrosa et al., 2007; Peters et al., 2008; Sanders et al., 2008; Turunen et al., 2007; Van Den Bogaert et al., 2003; Wood, Pickering, & Dechairo, 2007). One sample which showed no evidence in a first report (Morris et al., 2003) was positive when additional SNPs were typed (Williams, Preece, Morris, et  al., 2004). The specific details of reported associations have varied considerably and are summarized in Fig. 4.

Genetic Studies of Schizophrenia

355

One sample of 670/588 Japanese cases/controls gave results consistent with the original studies and identified association with the minor allele of p1635 (rs3213207) (which tags the Irish family risk haplotype) as well as other alleles and a 2-marker haplotype also consistent with the Irish family results(Numakawa et al., 2004). Three of six samples in two further replication studies unambiguously identified association with haplotype 5 in Fig. 3: 142/272 Swedish cases/controls (Van Den Bogaert et al., 2003), and 258/467 US Caucasian and 51/32 Hispanic cases/ controls (Funke et al., 2004). This is the closest neighbor haplotype to that associated in the Irish multiplex families and the Japanese case/control sample, and shares substantial variation with it. However, when considering all evidence currently available (including the reported frequencies of associated haplotypes and the specific patterns of alleles outside the LD block), the most consistent results are those from the seven independent samples showing association with common alleles and high-frequency, common-allele haplotype 1 in Fig.  3: 78 sibling pair families and 125 triads from Germany, Israel and Hungary (Schwab et al., 2003), 233 Han Chinese triads (Tang et al., 2003), samples of 708/711 UK cases/controls and 219/231 Irish case/controls (Williams, Preece, Morris, et al., 2004), 488 Bulgarian trios (Kirov et al., 2004), 638 Han Chinese families (Li et  al., 2005) and 314/314 Japanese cases/controls (Tochigi et al., 2006). Two further samples, 580/620 Scottish case/controls (Li et al., 2005) and 80/108 Italian case/controls (Tosato et al., 2007) show evidence for association consistent with haplotype 6 (Fig. 3), although the significance of association with a haplotype of only 1.5% frequency is unclear. No individual markers were associated in the Scottish sample and only haplotypes of two markers were analyzed. No evidence of association with any SNP or haplotype was seen in samples of 418/285 German or 294/113 Polish cases/controls (Van Den Bogaert et al., 2003), 219 US triads or 230 Afrikaans families (Hall et al., 2004), 117 Canadian families (DeLuca et al., 2005), 41 Australian families, 194/180 Australian cases/controls or 197 Indian families (Holliday et al., 2006), 441 Finnish families (Turunen et al., 2007), 693 Taiwanese families (Liu et  al., 2007), 194/351 Korean cases/controls (Joo et  al., 2006), 172/165 Caucasian cases/controls and 152/129 African-American cases/ controls (Pedrosa et  al., 2007), 451/291 US cases/controls (cases in this sample included 140 diagnosed with schizoaffective disorder) (Wood et al., 2007), 273/580 Dutch cases/controls (Bakker et al., 2007), 450/450 UK cases/controls (Datta et al., 2007), 336/172 Australian cases/controls (Peters et  al., 2008) and 1870/2002 European-ancestry cases/controls (Sanders et  al., 2008). Evidence of association was observed only in the 3¢ region of the gene in 135 European-descended Caucasian families (Duan et al., 2007). A further four studies have also provided positive evidence for association of DTNBP1 with bipolar disorder (Breen et al., 2006; Fallin et al., 2005; Joo et al., 2007; Pae et al., 2007). The function of DTNBP1 protein in brain is unknown. It was first identified as a binding partner of both a- and b-dystrobrevins (Benson, Newey, Martin-Rendon, Hawkes, & Blake, 2001), which are binding partners of dystrophin, a large, membraneassociated protein expressed at highest levels in muscle and brain and mutated in

356

B. Riley

Duchenne and Becker muscular dystrophy (Ray et al., 1985). In muscle, dystrophin is associated with the dystrophin-associated protein complex (DPC), which spans the membrane and links the cytoskeleton and the extra-cellular matrix. However, several lines of evidence suggest that the mechanisms of action for these molecules are different in neuronal and muscle tissue, and may be different in different regions of the brain. Some evidence suggests that the expression of DTNBP1 is reduced in certain brain regions of patients with schizophrenia at both the RNA (Weickert et al., 2004) and protein (Talbot et al., 2004) levels. 4.6.2 Chromosome 13q14-q32 Linkage Studies Data from a mixed sample of UK and Japanese families initially suggested linkage to chromosome 13q14.1-q32 (Lin et al., 1995), of interest as the region contains the serotonin receptor 2A gene, HTR2A. Preliminary data from the Maryland and UK/ Icelandic samples gave some initial support (Antonarakis et al., 1996; Kalsi et al., 1996). An attempt by the original group to replicate in an independent sample of Taiwanese and UK families supported the finding only in the European families (Lin et al., 1997). Further analyses of the European sample using slightly different methods yielded positive data at two markers located at 13q32, but they were separated by a region where the values of the statistics dropped almost to zero. Genome scan data from a mixed UK/US sample gave positive evidence, but extremely distant from other findings in the region (Shaw et  al., 1998). The Maryland family sample gave modest evidence for linkage under a recessive model; nonparametric analysis of the same data was highly significant. Marker data in narrowly defined Canadian pedigrees gave fairly strong evidence for linkage (Brzustowicz et al., 1999). The results from chromosome 13 are particularly difficult to interpret because of the very large distances between positive markers. Unlike chromosome 6, where two distinct regions have been detected in different samples, there has been little agreement about the site of greatest evidence on 13q. Overall, the combined linkage reports are spread over a region of ~60 Mb, containing ~120 known or putative genes. On the other hand, although locations are much less certain on chromosome 13q than in other linkage regions, this chromosome has produced some of the most significant linkage evidence seen in the studies of schizophrenia.

4.7 Chromosome 13q14-q32 Candidate Genes 4.7.1 G72/DAOA A study of ~200 SNPs tested across the distal 5  Mb of this broad linkage region identified two regions of association (Chumakov et  al., 2002). In one of these regions, two genes (initially called G72, now DAOA and G30) were investigated.

Genetic Studies of Schizophrenia

357

Of note, the exons of these genes could not be predicted by any computational method tested, suggesting that they are highly novel in their sequence and organization. Both genes show alternative transcripts in brain and other tissues. In 12 published association studies, SNPs within G72/DAOA have also provided substantial positive replication evidence (Addington et al., 2004; Corvin et al., 2007; Hong, Hou, Yen, Liou, & Tsai, 2006; Korostishevsky et al., 2004; Korostishevsky et al., 2006; Ma et al., 2006; Schumacher et al., 2004; Shin et al., 2007; Shinkai et al., 2007; Wang et  al., 2004; Yue et  al., 2006; Zou et  al., 2005). A further seven studies have not provided supportive evidence of association (Goldberg et al., 2006; Hall et al., 2004; Liu, Fann, Liu, Chang, et al., 2006a; Mulle, Chowdari, Nimgaonkar, & Chakravarti, 2005; Sanders et  al., 2008; Vilella et  al., 2007; Wood et  al., 2007). Three metaanalyses of the collected data showed weak positive (Li & He, 2007) and strong positive (Detera-Wadleigh & McMahon, 2006; Shi, Badner, Gershon, & Liu, 2008) evidence of association between G72/DAOA and schizophrenia. G72/DAOA has also been associated with bipolar disorder in four studies (Chen, Akula et  al., 2004; Goldberg et al., 2006; Hattori et al., 2003; Prata et al., 2008) and in one (DeteraWadleigh & McMahon, 2006) but not a second (Shi et al., 2008) meta-analysis. 4.7.2 G72/DAOA and DAO d-amino acid oxidase (DAO) was identified as a binding partner of, and is activated by, the protein product of G72/DAOA. The DAO gene, on chromosome 12q24, was screened for association evidence in the original study (Chumakov et al., 2002), and four SNPs tested were significantly associated. Results of this kind (showing association in two interacting genes in the same sample) are rare, so this study had a unique opportunity to test for an epistatic genetic interaction. Evidence for epistasis was observed for one pair of DAO and G72/DAOA genotypes, supporting a potential interaction between them in risk for schizophrenia. Fewer replication studies have assessed the role of DAO, and these are less clear in their support for the reported association, with four studies supporting (Corvin et al., 2007; Liu et al., 2004; Schumacher et al., 2004; Wood et al., 2007) and four showing no support for association with schizophrenia (Goldberg et al., 2006; Liu, Fann, Liu, Chang, et al., 2006a; Shinkai et al., 2007; Vilella et al., 2007). One study of DAO suggested association with bipolar disorder (Fallin et al., 2005).

4.8 Chromosome 1q32-q42 Linkage Studies 4.8.1 1q41-q42 and DISC1 Some of the strongest findings suggesting the involvement of genes on chromosome 1 in schizophrenia began with reports of a balanced 1:11 translocation segregating with serious mental illness in a large pedigree from Scotland (St Clair et al., 1990).

358

B. Riley

The chromosome 1 breakpoint lies at 1q42.1, and two groups reported suggestive linkage findings in this region, in national and population isolate samples from Finland (Ekelund et al., 2000; Hovatta et al., 1999) and in the Maryland sample. Ongoing work in the Scottish pedigree has now shown that the breakpoint directly disrupts a novel gene, Disrupted in Schizophrenia 1 (DISC1) (Millar et al., 2000). There are now nine positive reports of association of DISC1 with schizophrenia (Callicott et al., 2005; Cannon et al., 2005; Hennah et al., 2003; Hodgkinson et al., 2004; Liu, Fann, Liu, Chen, et  al., 2006b; Palo et  al., 2007; Qu et  al., 2007; Thomson et  al., 2005; Zhang et  al., 2006) and two of association with positive symptoms (DeRosse et al., 2007; Szeszko et al., 2008) suggesting that this gene is relevant to schizophrenia in the general population. Other rare variants in this gene besides the breakpoint have also been reported to be associated with schizophrenia (Sachs et al., 2005; Song et al., 2008) and association has been reported for additional psychiatric diagnoses, reviewed in (Hennah et al., 2009) and for bipolar disorder (Perlis et  al., 2008). A smaller number of negative reports have also been published (Chen, Chen et  al., 2007; Kim et  al., 2008; Kockelkorn et  al., 2004; Sanders et al., 2008; Zhang, Tochigi, et al., 2005).

4.9 1q23-q32 linkage and RGS4 Genome scan data in families from Canada (Brzustowicz, Hodgkinson, Chow, Honer, & Bassett, 2000) and follow-up fine-mapping data in the population isolate samples from Finland (Ekelund et al., 2001) have also provided evidence for linkage in a more centromeric position on chromosome 1, although the chromosomal position varies between these studies making interpretation difficult. The latter provided very strong LOD scores but was not replicated by a large collaborative study (Levinson et al., 2002). Microarray studies of post-mortem schizophrenic brain suggested that RGS4, the regulator of G-protein signaling 4 gene, showed altered expression in schizophrenia (Mirnics, Middleton, Lewis, & Levitt, 2001). RGS4 maps to the chromosome 1 linkage region, and in a subsequent study in mixed US pedigrees and samples from India, the same markers in the same 10 kb region were associated in both samples, although different specific marker haplotypes gave this evidence in the US compared to the Indian families (Chowdari et al., 2002). In replication studies, six have provided supportive evidence for association of the RGS4 locus with schizophrenia liability (Bakker et al., 2007; Chen, Dunham et al., 2004; Fallin et al., 2005; Morris et al., 2004; So et al., 2008; Williams, Preece, Spurlock, et al., 2004) while nine have not (Cordeiro et al., 2005; Guo et al., 2006; Ishiguro et al., 2007; Liu, Shen-Jang, et al., 2006d; Rizig et al., 2006; Sanders et al., 2008; Sobell, Richard, Wirshing, & Heston, 2005; Vilella et al., 2007; Wood et al., 2007). One study observed association in a Scottish sample but not in a Chinese sample (Zhang, St Clair, et  al., 2005). Results of meta-analyses have both supported (Talkowski et al., 2006) and failed to support (Li & He, 2006) the involvement of RGS4 in schizophrenia.

Genetic Studies of Schizophrenia

359

4.10 Other Chromosomal Regions and Genes A number of additional chromosome regions have provided multiple signals, although the evidence for linkage to schizophrenia is less well replicated and less certain. These include: 5q22-q31, where association was recently reported in the interleukin-3 (IL3) gene (Chen, Wang, et  al., 2007), which awaits replication; 10p15-q21, where evidence of association in the phophatidyl inositol-phosphate 5 kinase 2a (PIP5K2A) (Schwab et  al., 2006) followed by mixed positive (Bakker et al., 2007; He et al., 2007; Saggers-Gray et al., 2008) and negative (Jamra et al., 2006) replication results; 6q21-q22, where association with the trace amine associated receptor 6 (TAAR6, previously known as TRAR4) gene located at 6q23.2 (Duan et al., 2004) has mixed positive (Pae et al., 2008; Vladimirov et al., 2007) and negative (Duan, Du et al., 2005; Ikeda et al., 2005; Sanders et al., 2008) replications; and 15q13-q14, where evidence for linkage to an evoked potential abnormality common in patients and relatively rare in controls (Freedman et al., 1997) was supported by five additional studies reporting positive linkage evidence for schizophrenia in the same narrow region (Gejman, Sanders, Badner, Cao, & Zhang, 2001; Liu et al., 2001; Riley et al., 2000; Tsuang et al., 2001; Xu et al., 2001). A number of other high profile candidate genes such as PRODH2 (Liu et  al., 2002) and PPP3CC (Gerber et al., 2003), identified through other means, have not replicated well. One exception to this pattern is the evidence for involvement of the AKT1 gene in schizophrenia (Emamian et al., 2004), which has similar numbers of positive (Bajestan et al., 2006; Ikeda et al., 2004; Schwab et al., 2005; Thiselton et  al., 2008; Xu et  al., 2007) and negative (Ide et  al., 2006; Liu, Fann, Liu, Wu, et al., 2006c; Norton et al., 2007; Ohtsuki, Inada, & Arinami, 2004; Sanders et al., 2008; Turunen et al., 2007) replications.

4.11 Meta-Analyses of Linkage and Association Data Meta-analysis of whole genome screen data offers a different kind of insight into the mechanisms of complex trait genetics: because many samples, and therefore more data, are included, they represent a first approximation of a very large, multisample genome screen. Two different statistical approaches for such meta-analyses have been published. One combines the significance levels reported in the original genome screens after correcting each value for the size of the suggested region. Results from the first approach were significant for chromosomes 8p, 13q and 22q (Badner & Gershon, 2002). The major limitations of this analysis are that it relies on published results, which prevents critical standardization across studies, and, unlike the second approach below, no new information about potential regions of interest can be extracted with it. The second method ranks 30-cM bins of the genome from most positive to least positive for each study, and then sums the ranks for each bin. Significance levels are

360

B. Riley

calculated by simulation. Since this method uses not significance levels but the actual marker LOD scores, it is possible to identify regions of the genome which are of potential interest on the basis of modest positive results occurring in the same region across many studies but which may have been overlooked due to stronger signals in the individual sample analyses. Results of the second approach, which is methodologically the stronger of the two, supported linkage to chromosomes 6p, 8p and 10p of the previously identified regions discussed above (Lewis et al., 2003). However, the strongest evidence for a potential locus was on chromosome 2p11.1-q21.1, a region suggested by only a few studies and not widely followed up, and on 3p, the site of an early linkage finding in the Maryland sample which could never be replicated by subsequent studies. Finally, significant evidence of linkage was also detected for two regions never previously implicated by an individual study, on chromosomes 11q and 14p. Meta-analyses of association data have tended to assess one gene at a time. A recent effort has been made to systematize the collection and archiving of association data from studies of schizophrenia, and to provide a framework for continuous updating of both the data and the meta-analytic results (Allen et  al., 2008). The resulting SzGene database (http://www.szgene.org/) is regularly updated and publicly available. Meta-analyses of the data contained in this resource provided support of varying degrees for 24 SNPs in 16 previously reported genes, including older candidate genes (e.g., DRD2), those resulting from association-based follow-up of linkage data (e.g., DTNBP1) and those suggested by more recent genomewide studies (e.g., PLXNA2).

4.12 Summary of Current Gene Findings Currently, all of the regions and candidate genes discussed in the sections above remain promising, but further assessment of each is still needed to identify additional associated genes, clarify patterns of association and elucidate their contribution to the neurobiology of schizophrenia. The reported associations for several of the current candidates, DTNBP1, NRG1, DAOA/G72, DAO, DISC1 and RGS4, have been replicated in multiple samples, and positive replications outnumber negative ones for most. Other candidates, such as IL3, await the collection of sufficient data to interpret the validity of the original findings. Still others, like TAAR6, already have substantial data collected, but without strong or widespread replication. One particularly exciting shared feature of many of the candidates discussed above is that they can be related to potential pathophysiology through dysfunction in glutamatergic neurotransmission, which may be an important systemic element in the etiology of schizophrenia. Although a detailed discussion of this theory is outside the scope of this chapter, recent reviews of the genetic (Harrison & Owen, 2003) and neuroscience (Moghaddam, 2003) data and evidence from other studies highlight the positions of the gene products of NRG1, COMT, DAO, DAOA/G72, RGS4 and possibly DTNBP1 among others, in the biochemical and functional pathways influencing the glutamatergic system.

Genetic Studies of Schizophrenia

361

4.13 Genomewide Association Studies The field has been advanced further by the current round of GWAS, which have a number of distinct advantages over previous methods and have already revolutionized the understanding of other complex traits such as type 2 diabetes (Frayling et al., 2007; Saxena et al., 2007; Scott et al., 2007; Sladek et al., 2007; Steinthorsdottir et al., 2007; Zeggini et al., 2007). Assaying 500,000–1,000,000 DNA variants in a single experiment provides unbiased genomewide coverage and so avoids the weakness of selecting candidate genes when little is known about underlying disease processes. They use an association framework for analysis and so avoid the weaknesses of linkage in complex traits. They impose stringent criteria due to the number of tests performed (p < 10−7 for genomewide significance, although in practice p < 10−5 is often considered strong enough to replicate or report). Finally, they hold enormous potential to move beyond the identification of single genes (which may show small effects and be difficult to detect individually) toward the simultaneous identification of multiple genes and their functional involvement in pathways, systems or processes. Seven GWAS of schizophrenia have been published to date, with an eighth in analysis at the time of writing. The first four of these have had relatively small sample sizes and consequently modest power. The first (of 320 cases and 325 controls) was small, of limited density as it genotyped only 25,000 SNPs in 14,000 known genes, and did not detect any association that reached genomewide significance (Mah et  al., 2006); nominal association was reported in the plexin A2 (PLXNA2) gene. Replication studies are currently limited but only one of four samples tested in three independent studies has confirmed the association (Budel et  al., 2008; Fujii et  al., 2007; Takeshita et  al., 2008). The second (which was extremely underpowered with 178 cases and 144 controls) identified one genomewide significant association in the X/Y pseudoautosomal region (a homologous region of the sex chromosomes where recombination can occur), near the colony stimulating factor receptor 2 alpha (CSF2RA) and interleukin 3 receptor (IL3R) genes (Lencz et al., 2007). Cytokines have previously been suggested as possible candidates and one replication attempt supported association in IL3R (Sun et al., 2008). The third, using the CATIE (Lieberman et al., 2005) sample (738 cases and 733 controls), did not detect any genomewide significant results in its stage 1 analysis (Sullivan et al., 2008). The fourth, using a multi-stage design of discovery (in 479 cases and 2,937 controls) and targeted replication (in 6,666 cases and 9,897 controls) samples, identified one genomewide significant SNP in the zinc-finger protein (and likely transcription factor) ZNF804A gene (O’Donovan et al., 2008). It is critical to note that this SNP did not yield genomewide significant results in the replication samples independently, but only in the meta-analysis of data from the entire study. The strongest results from the type 2 diabetes GWAS come from combined analyses of multiple independent complete GWAS datasets, rather than from a single GWAS and targeted replication, and several genes were not supported at the threshold significance levels in every individual study (Saxena et  al., 2007;

362

B. Riley

Scott et al., 2007; Sladek et al., 2007; Zeggini et al., 2007). Although multi-stage designs are cost-effective, the issue of stochastic sampling variation may be critical in determining appropriate replication design. Three more recent studies have been published with substantially larger sample sizes and consequently increased power to detect association, particularly in metaanalysis. The first included 2,663 cases and 13,498 controls from eight sites in Europe (Stefansson et al., 2009), the second 3,322 cases and 3,587 controls from the International Schizophrenia Consortium (ISC) (Purcell et al., 2009), and the last included 2,681 cases and 2,653 controls of European descent, and 1,286 cases and 973 controls of African descent (Shi et al., 2009). None of the three studies individually detected any genomewide significant signal (p < 5 × 10−8), but meta-analysis of all three samples did detect genomewide significant evidence of aassociation in the major histocompatibility locus (MHC) region on chromosome 6p22.1. This region was studied extensively in the early era of schizophrenia genetics, but without yielding a robust signal. In addition to genes coding for proteins involved directly in immune function, the region also contains a cluster of histone genes, suggesting the possible involvement of chromatin modification and transcriptional regulation. Support for some prior loci was also observed, e.g., for ErbB4 in the African-American sample (Shi et al., 2009), and for ZNF804A in the ISC sample (Purcell et al., 2009). Although additional study of these data are certainly needed and meta-analyses of ~15,000 cases and 15,000 controls are planned for late 2009, the results of these three studies suggest that there are few (if any) risk loci for schizophrenia with ORs ~1.3 or greater, and overall do not provide strong support for the common disease/common variant hypothesis of schizophrenia.

4.14 Rare Structural Variation in Schizophrenia This chapter is written from the perspective of investigators who favor the common disease/common variant hypothesis of the genetic risks for complex traits, based on the epidemiological and genetic data outlined throughout. The current results of GWAS in other complex traits provided a major validation of this model. The alternative common disease/rare variant hypothesis of genetic risks for complex traits has been proposed in autism (Zhao et al., 2007) and schizophrenia recently (McClellan, Susser, & King, 2007), largely based on the reduction in fertility observed in cases. A key focus of research in this area has been the deletions, duplications and inversions of a few thousand (Kb) to a few million (Mb) base pairs collectively known as structural variants, an area of intense research interest generally since 2004 (Iafrate et al., 2004; Sebat et al., 2004; Tuzun et al., 2005). The most common mechanism by which these variants arise is recombination occurring between two distinct but highly similar repeat sequences physically close to each other on a chromosome, called non-allelic homologous recombination (Fig. 5).

Genetic Studies of Schizophrenia

363

Non-allelic homologous recombination a

Gene A

Duplication Deletion

b

Translocation

c

Inversion

Fig. 5  Repeat-mediated mechanisms of non-allelic homologous recombination (NAHR) and their results. (a) NAHR between adjacent repeat sequences produces duplications and deletions, depending on the specific pairing involved in the NAHR. (b) NAHR between repeats on nonhomologous chromosomes can produce translocations. In both cases, the copies created are in the same orientation as the original repeats. (c) By contrast, inversions can occur as a consequence of recombination between inverted intrachromosomal repeat sequences (Bailey & Eichler, 2006)

As a class, this type of variation is common, but the individual deletion, duplication and inversion events giving rise to the structural variation are rare. Genomic survey studies estimate that as much as 360 Mb or 12% of the genome is included in structural variation; 50% of variants were observed in more than one individual, consistent with stable inheritance in the population (Redon et al., 2006). No study has yet been published assessing these common structural variants in disease. A few such variants occur at high frequency due to apparent selection in certain populations or contexts, eg increases in copy number of the CCL3L1 gene (protective against HIV infection) in African primates including humans (Gonzalez et al., 2005) and the amylase (AMY1) gene in cereal farming groups, but not their pastoralist neighbors (Perry et al., 2007). However, studies of large samples are consistent in showing that the majority of structural variants are rare, often occurring in only one individual. This is broadly consistent with the widespread occurrence of repeat sequences in the human genome, which provide a basis for large numbers of randomly distributed, individually rare events. The aggregate rate of such rare structural variants is significantly increased in individuals with schizophrenia in all four studies reported (International Schizophrenia Consortium, 2008; Stefansson et al., 2008; Walsh et al., 2008; Xu et al., 2008). Critically, there is substantial overlap between studies in the regions where structural variation is observed in excess in cases, most notably on chromosomes 22q11, 15q13.3 and 1q21.1, with some evidence that regions containing genes expressed in neurodevelopment are overrepresented, as in (Rujescu et  al., 2009). The most extreme test of this hypothesis based on fertility argues that de novo events in affected individuals are the most likely to represent highly penetrant genetic lesions because they are not present in the parents of affected individuals, who have successfully reproduced. De novo events are also observed at  significantly higher frequency in cases of both autism (Sebat et  al., 2007)

364

B. Riley

and schizophrenia (Stefansson et al., 2008; Xu et al., 2008). A few relatively rare structural variants have been observed to be recurrent: one example is the VCFS deletion on chromosome 22q discussed above, which is present at elevated frequency in individuals with schizophrenia. However, even considered in aggregate, rare structural variants are observed in only 15%, and de novo structural variants in only 10%, of schizophrenia cases at the resolution and genomic coverage of experiments in 2008. The variants detected so far thus cannot account for a substantial fraction of the total population risk, and de novo events cannot account for any of the observed familial risk. A more difficult problem is that, because most are rare, the true impact of individual structural variants on schizophrenia is difficult to ­validate and interpret. The true importance of these findings in schizophrenia in the population is unclear at this time, although the replication of excess structural variation in cases on chromosomes 22q11, 15q13.3 and 1q21.1 is extremely encouraging.

5 Discussion As is clear from the final two sections above, the field of genetic studies of schizophrenia continues to change rapidly and may have altered considerably by the time this chapter is read. Certainly the most important development in the last several years has been the emergence of a number of replicated positional candidate genes in target regions. Given the vast number of statistical tests that are now performed in most studies (multiple markers, diagnostic or genetic models and analytic approaches), the true type I (or false positive) error rate emerging from any individual study is nearly impossible to quantify. It remains a major concern that highly statistically significant results could occur by chance alone because so many tests are performed. Therefore, replication is critical. However, there are many reasons why a “true” finding might not be replicated including variation between populations or samples and differences in statistical power, diagnostic methods and statistical approaches. Given the evidence of replication for linkage to a number of regions and association with genes in those regions, it seems increasingly unlikely that all of these results represent false positives. It is difficult to conceive of an inherent bias that would produce spuriously positive results across multiple groups (especially given the wide differences between studies described above) in the same gene or chromosomal region. In results from whole genome linkage scans for other complex disorders, including type 1 (Hashimoto et al., 1994) and type 2 (Busfield et al., 2002; Vionnet et al., 2000; Wiltshire et  al., 2001) diabetes mellitus, multiple sclerosis (Coraddu et  al., 2001; Sawcer et al., 1996), inflammatory bowel disease (Hugot et al., 2001; Ogura et al., 2001) and asthma (Laitinen et al., 2001), non-replication across groups is as frequent as replication. These results suggest that the difficulties in detecting replicable linkages for schizophrenia may not be unique to the psychiatric disorders, but rather may reflect a general pattern of problems associated with linkage studies in complex traits.

Genetic Studies of Schizophrenia

365

However, there appears to be cause for some optimism, as the candidates currently under most intense scrutiny have provided positive evidence across multiple samples. The results from GWAS of other complex traits have produced much better replicated initial results, but replication design may be critical to this success.

6 Conclusions The evidence is strong that schizophrenia is a familial disorder and that the familial aggregation of schizophrenia is due largely, although probably not entirely, to genetic factors. Whatever the familial predisposition that operates for schizophrenia, it not only influences the classic, deteriorating psychotic disorder but also increases liability to schizophrenia spectrum personality disorders and probably some other non-schizophrenic non-affective psychoses. Two decades of research using statistical methods have failed to clearly delineate the mode of transmission of schizophrenia, a result which is understandable given its likely complexity. In generalizing to other psychiatric and complex phenotypes, the broad conclusions from the study of schizophrenia seem likely to hold: (1) such phenotypes are genetically influenced but not genetically determined, (2) a number of genes (which may even vary between individual family members) are likely to be involved, (3) the liability variants in these genes are generally expected to be common, within the range of normal human variation and to have low risk associated with them individually, (4) some of the variants may interact with others or with environmental risk factors, and (5) some of the variants may predispose individuals to wider spectra of psychopathology. Advances in molecular and statistical genetics have opened up realistic opportunities to localize on the human genome the specific genes that influence the liability to schizophrenia. Association studies have yet to provide convincing evidence for the role of a range of candidate genes in the etiology of schizophrenia. Genome scan strategies have produced several regions, where multiple groups have found evidence for linkage, and more importantly, association in specific genes. While false positive findings cannot be ruled out, it seems likely than one or more of these are true susceptibility genes for schizophrenia. Further, given the power of genomewide association designs, it seems increasingly likely that, within several years, the field may have widely replicated susceptibility genes for schizophrenia. This would represent a true watershed event in the history of schizophrenia research. While unambiguous gene identification will itself represent a major advance, it is the identification of specific variants in these genes and their impact on disease which will initiate the most critical phase, that of translational research including (1) rational drug design based on knowledge of basic pathophysiology and personalized medicine, (2) characterization of genotype-phenotype relationships based on knowledge of specific pathogenic mutations, (3) identification of environmental risk factors that interact with specific genes, and (4) realistic prevention research given our ability to identify high-risk individuals.

366

B. Riley

References Addington, A. M., Gornick, M., Sporn, A. L., Gogtay, N., Greenstein, D., Lenane, M., et  al. (2004). Polymorphisms in the 13q33.2 gene G72/G30 are associated with childhood-onset schizophrenia and psychosis not otherwise specified. Biological Psychiatry, 55, 976–980. Allen, N. C., Bagade, S., McQueen, M. B., Ioannidis, J. P., Kavvoura, F. K., Khoury, M. J., et al. (2008). Systematic meta-analyses and field synopsis of genetic association studies in schizophrenia: The SzGene database. Nature Genetics, 40, 827–834. Altshuler, D., & Daly, M. (2007). Guilt beyond a reasonable doubt. Nature Genetics, 39, 813–815. Antonarakis, S. E., Blouin, J. L., Curran, M., Luebbert, H., Kazazian, H. H., Dombroski, B. A., et al. (1996). Linkage and sib-pair analysis reveal a potential schizophrenia susceptibility gene on chromosome 13q32. American Journal of Human Genetics, 59, A210. Badner, J. A., & Gershon, E. S. (2002). Meta-analysis of whole-genome linkage scans of bipolar disorder and schizophrenia. Molecular Psychiatry, 7, 405–411. Bailey, J. A., & Eichler, E. E. (2006). Primate segmental duplications: Crucibles of evolution, diversity and disease. Nature Reviews. Genetics, 7, 552–564. Bajestan, S. N., Sabouri, A. H., Nakamura, M., Takashima, H., Keikhaee, M. R., Behdani, F., et al. (2006). Association of AKT1 haplotype with the risk of schizophrenia in Iranian population. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 141, 383–386. Bakker, S. C., Hoogendoorn, M. L., Hendriks, J., Verzijlbergen, K., Caron, S., Verduijn, W., et al. (2007). The PIP5K2A and RGS4 genes are differentially associated with deficit and non-deficit schizophrenia. Genes, Brain, and Behavior, 6, 113–119. Bakker, S. C., Hoogendoorn, M. L., Selten, J. P., Verduijn, W., Pearson, P. L., Sinke, R. J., et al. (2004). Neuregulin 1: Genetic support for schizophrenia subtypes. Molecular Psychiatry, 9, 1061–1063. Baron, M. (1996). Linkage results in schizophrenia. American Journal of Medical Genetics, 67, 121–123. Benson, M. A., Newey, S. E., Martin-Rendon, E., Hawkes, R., & Blake, D. J. (2001). Dysbindin, a novel coiled-coil-containing protein that interacts with the dystrobrevins in muscle and brain. The Journal of Biological Chemistry, 276, 24232–24241. Benzel, I., Bansal, A., Browning, B. L., Galwey, N. W., Maycox, P. R., McGinnis, R., et al. (2007). Interactions among genes in the ErbB-Neuregulin signalling network are associated with increased susceptibility to schizophrenia. Behavioral and Brain Functions, 3, 31. Breen, G., Prata, D., Osborne, S., Munro, J., Sinclair, M., Li, T., et al. (2006). Association of the dysbindin gene with bipolar affective disorder. The American Journal of Psychiatry, 163, 1636–1638. Brzustowicz, L. M., Hodgkinson, K. A., Chow, E. W. C., Honer, W. G., & Bassett, A. S. (2000). Location of a major susceptibility locus for familial schizophrenia on chromosome 1q21-q22. Science, 288, 678–682. Brzustowicz, L. M., Honer, W. G., Chow, E. W. C., Little, D., Hodgkinson, K., & Bassett, A. (1999). Linkage of familial schizophrenia to chromosome 13q32. American Journal of Human Genetics, 65, 1096–1103. Budel, S., Shim, S. O., Feng, Z., Zhao, H., Hisama, F., & Strittmatter, S. M. (2008). No association between schizophrenia and polymorphisms of the PlexinA2 gene in Chinese Han Trios. Schizophrenia Research, 99, 365–366. Busfield, F., Duffy, D. L., Kesting, J. B., Walker, S. M., Lovelock, P., Good, D., et  al. (2002). A genomewide search for type 2 diabetes-susceptibility genes in indigenous Australians. American Journal of Human Genetics, 70, 349–357. Callicott, J. H., Straub, R. E., Pezawas, L., Egan, M. F., Mattay, V. S., Hariri, A. R., et al. (2005). Variation in DISC1 affects hippocampal structure and function and increases risk for schizophrenia. Proceedings of the National Academy of Sciences of the United States of America, 102, 8627–8632.

Genetic Studies of Schizophrenia

367

Cannon, T. D., Hennah, W., van Erp, T. G., Thompson, P. M., Lonnqvist, J., Huttunen, M., et al. (2005). Association of DISC1/TRAX haplotypes with schizophrenia, reduced prefrontal gray matter, and impaired short- and long-term memory. Archives of General Psychiatry, 62, 1205–1213. Cardno, A. G., & Gottesman, I. I. (2000). Twin studies of schizophrenia: From bow-and-arrow concordances to Star Wars Mx and functional genomics. American Journal of Medical Genetics, 97, 12–17. Cardno, A. G., Marshall, E. J., Coid, B., Macdonald, A. M., Ribchester, T. R., Davies, N. J., et al. (1999). Heritability estimates for psychotic disorders: The Maudsley twin psychosis series. Archives of General Psychiatry, 56, 162–168. Caspi, A., McClay, J., Moffitt, T. E., Mill, J., Martin, J., Craig, I. W., et al. (2002). Role of genotype in the cycle of violence in maltreated children. Science, 297, 851–854. Caspi, A., Sugden, K., Moffitt, T. E., Taylor, A., Craig, I. W., Harrington, H., et  al. (2003). Influence of life stress on depression: Moderation by a polymorphism in the 5-HTT gene. Science, 301, 386–389. Chen, Y. S., Akula, N., Tera-Wadleigh, S. D., Schulze, T. G., Thomas, J., Potash, J. B., et  al. (2004). Findings in an independent sample support an association between bipolar affective disorder and the G72/G30 locus on chromosome 13q33. Molecular Psychiatry, 9, 87–92. Chen, Q. Y., Chen, Q., Feng, G. Y., Lindpaintner, K., Wang, L. J., Chen, Z. X., et al. (2007). Casecontrol association study of disrupted-in-schizophrenia-1 (DISC1) gene and schizophrenia in the Chinese population. Journal of Psychiatric Research, 41, 428–434. Chen, X., Dunham, C., Kendler, S., Wang, X., O’Neill, F. A., Walsh, D., et al. (2004). Regulator of G-protein signaling 4 (RGS4) gene is associated with schizophrenia in Irish high density families. American Journal of Medical Genetics, 129B, 23–26. Chen, X., Wang, X., Hossain, S., O’Neill, F. A., Walsh, D., van den Oord, E., et  al. (2007). Interleukin 3 and schizophrenia: The impact of sex and family history. Molecular Psychiatry, 12, 273–282. Chowdari, K. V., Mirnics, K., Semwal, P., Wood, J., Lawrence, E., Bhatia, T., et  al. (2002). Association and linkage analyses of RGS4 polymorphisms in schizophrenia. Human Molecular Genetics, 11, 1373–1380. Chumakov, I., Blumenfeld, M., Guerassimenko, O., Cavarec, L., Palicio, M., Abderrahim, H., et  al. (2002). Genetic and physiological data implicating the new human gene G72 and the gene for D-amino acid oxidase in schizophrenia. Proceedings of the National Academy of Sciences of the United States of America, 99, 13675–13680. Coon, H., Holik, J., Hoff, M., Reimherr, F., Wender, P., Myles-Worsley, M., et al. (1994). Analysis of chromosome 22 markers in nine schizophrenia pedigrees. American Journal of Medical Genetics, 54, 72–79. Coraddu, F., Sawcer, S., D’Alfonso, S., Lai, M., Hensiek, A., Solla, E., et al. (2001). A genome screen for multiple sclerosis in Sardinian multiplex families. European Journal of Human Genetics, 9, 621–626. Cordeiro, Q., Talkowski, M. E., Chowdari, K. V., Wood, J., Nimgaonkar, V., & Vallada, H. (2005). Association and linkage analysis of RGS4 polymorphisms with schizophrenia and bipolar disorder in Brazil. Genes, Brain, and Behavior, 4, 45–50. Corvin, A., McGhee, K. A., Murphy, K., Donohoe, G., Nangle, J. M., Schwaiger, S., et al. (2007). Evidence for association and epistasis at the DAOA/G30 and D-amino acid oxidase loci in an Irish schizophrenia sample. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 144B, 949–953. Corvin, A. P., Morris, D. W., McGhee, K., Schwaiger, S., Scully, P., Quinn, J., et  al. (2004). Confirmation and refinement of an ‘at-risk’ haplotype for schizophrenia suggests the EST cluster, Hs. 97362, as a potential susceptibility gene at the Neuregulin-1 locus. Molecular Psychiatry, 9, 208–213. Datta, S. R., McQuillin, A., Puri, V., Choudhury, K., Thirumalai, S., Lawrence, J., et al. (2007). Failure to confirm allelic and haplotypic association between markers at the chromosome 6p22.3 dystrobrevin-binding protein 1 (DTNBP1) locus and schizophrenia. Behavioural and Brain Functions, 3, 50.

368

B. Riley

DeLuca, V., Voineskos, D., Shinkai, T., Wong, G., & Kennedy, J. L. (2005). Untranslated region haplotype in dysbindin gene: Analysis in schizophrenia. Journal of Neural Transmission, 112, 1263–1267. DeRosse, P., Hodgkinson, C. A., Lencz, T., Burdick, K. E., Kane, J. M., Goldman, D., et  al. (2007). Disrupted in schizophrenia 1 genotype and positive symptoms in schizophrenia. Biological Psychiatry, 61, 1208–1210. Detera-Wadleigh, S. D., & McMahon, F. J. (2006). G72/G30 in schizophrenia and bipolar disorder: Review and meta-analysis. Biological Psychiatry, 60, 106–114. Duan, S., Du, J., Xu, Y., Xing, Q., Wang, H., Wu, S., et  al. (2005). Failure to find association between TRAR4 and schizophrenia in the Chinese Han population. Journal of Neural Transmission, 113, 5. Duan, J., Martinez, M., Sanders, A. R., Hou, C., Burrell, G. J., Krasner, A. J., et  al. (2007). DTNBP1 (dystrobrevin binding protein 1) and schizophrenia: Association evidence in the 3¢ end of the gene. Human Heredity, 64, 97–106. Duan, J., Martinez, M., Sanders, A. R., Hou, C., Krasner, A. J., Schwartz, D. B., et al. (2005). Neuregulin 1 (NRG1) and schizophrenia: Analysis of a US family sample and the evidence in the balance. Psychological Medicine, 35, 1599–1610. Duan, J., Martinez, M., Sanders, A. R., Hou, C., Saitou, N., Kitano, T., et al. (2004). Polymorphisms in the trace amine receptor 4 (TRAR4) gene on chromosome 6q23.2 are associated with susceptibility to schizophrenia. American Journal of Human Genetics, 75, 624–638. Egan, M. F., Goldberg, T. E., Kolachana, B. S., Callicott, J. H., Mazzanti, C. M., Straub, R. E., et al. (2001). Effect of COMT Val108/158 Met genotype on frontal lobe function and risk for schizophrenia. Proceedings of the National Academy of Sciences of the United States of America, 98, 6917–6922. Ekelund, J., Hovatta, I., Parker, A., Paunio, T., Varilo, T., Martin, R., et al. (2001). Chromosome 1 loci in Finnish schizophrenia families. Human Molecular Genetics, 10, 1611–1617. Ekelund, J., Lichtermann, D., Hovatta, I., Ellonen, P., Suvisaari, J., Terwilliger, J. D., et al. (2000). Genome-wide scan for schizophrenia in the Finnish population: Evidence for a locus on chromosome 7q22. Human Molecular Genetics, 9, 1049–1057. Emamian, E. S., Hall, D., Birnbaum, M. J., Karayiorgou, M., & Gogos, J. A. (2004). Convergent evidence for impaired AKT1-GSK3beta signaling in schizophrenia. Nature Genetics, 36, 131–137. Fallin, M. D., Lasseter, V. K., Avramopoulos, D., Nicodemus, K. K., Wolyniec, P. S., McGrath, J. A., et  al. (2005). Bipolar I disorder and schizophrenia: A 440-single-nucleotide polymorphism screen of 64 candidate genes among Ashkenazi Jewish case-parent trios. American Journal of Human Genetics, 77, 918–936. Farmer, A. E., McGuffin, P., & Gottesman, I. I. (1987). Twin concordance for DSM-III schizophrenia. Scrutinizing the validity of the definition. Archives of General Psychiatry, 44, 634–640. Foley, D. L., Eaves, L. J., Wormley, B., Silberg, J. L., Maes, H. H., Kuhn, J., et  al. (2004). Childhood adversity, monoamine oxidase a genotype, and risk for conduct disorder. Archives of General Psychiatry, 61, 738–744. Frayling, T. M., Timpson, N. J., Weedon, M. N., Zeggini, E., Freathy, R. M., Lindgren, C. M., et  al. (2007). A common variant in the FTO gene is associated with body mass index and predisposes to childhood and adult obesity. Science, 316, 889–894. Freedman, R., Coon, H., Myles-Worsley, M., Orr-Urtreger, A., Olincy, A., Davis, A., et al. (1997). Linkage of a neurophysiological deficit in schizophrenia to a chromosome 15 locus. Proceedings of the National Academy of Sciences of the United States of America, 94, 587–592. Fujii, T., Iijima, Y., Kondo, H., Shizuno, T., Hori, H., Nakabayashi, T., et al. (2007). Failure to confirm an association between the PLXNA2 gene and schizophrenia in a Japanese population. Progress in Neuropsychopharmacology & Biological Psychiatry, 31, 873–877. Fukui, N., Muratake, T., Kaneko, N., Amagane, H., & Someya, T. (2006). Supportive evidence for neuregulin 1 as a susceptibility gene for schizophrenia in a Japanese population. Neuroscience Letters, 396, 117–120.

Genetic Studies of Schizophrenia

369

Funke, B., Finn, C. T., Plocik, A. M., Lake, S., DeRosse, P., Kane, J. M., et al. (2004). Association of the DTNBP1 locus with schizophrenia in a U.S. population. American Journal of Human Genetics, 75, 891–898. Gejman, P. V., Sanders, A. R., Badner, J. A., Cao, Q., & Zhang, J. (2001). Linkage analysis of schizophrenia to chromosome 15. American Journal of Medical Genetics, 105, 789–793. Georgieva, L., Dimitrova, A., Ivanov, D., Nikolov, I., Williams, N. M., Grozeva, D., et al. (2008). Support for neuregulin 1 as a susceptibility gene for bipolar disorder and schizophrenia. Biological Psychiatry, 64, 419–427. Gerber, D. J., Hall, D., Miyakawa, T., Demars, S., Gogos, J. A., Karayiorgou, M., et al. (2003). Evidence for association of schizophrenia with genetic variation in the 8p21.3 gene, PPP3CC, encoding the calcineurin gamma subunit. Proceedings of the National Academy of Sciences of the United States of America, 100, 8993–8998. Gill, M., Vallada, H., Collier, D., Sham, P., Holmans, P., Murray, R., et al. (1996). A combined analysis of D22S278 marker alleles in affected sib-pairs: Support for a susceptibility locus for schizophrenia at chromosome 22q12. American Journal of Medical Genetics, 67, 40–45. Goldberg, T. E., Straub, R. E., Callicott, J. H., Hariri, A., Mattay, V. S., Bigelow, L., et al. (2006). The G72/G30 gene complex and cognitive abnormalities in schizophrenia. Neuropsychopharmacology, 31, 2022–2032. Gonzalez, E., Kulkarni, H., Bolivar, H., Mangano, A., Sanchez, R., Catano, G., et al. (2005). The influence of CCL3L1 gene-containing segmental duplications on HIV-1/AIDS susceptibility. Science, 307, 1434–1440. Gottesman, I. I. (1991). Schizophrenia genesis. New York: W H Freeman. Gottesman, I. I., & Shields, J. (1982). Schizophrenia: The epigenetic puzzle. Cambridge: Cambridge University Press. Green, E. K., Raybould, R., Macgregor, S., Gordon-Smith, K., Heron, J., Hyde, S., et al. (2005). Operation of the schizophrenia susceptibility gene, neuregulin 1, across traditional diagnostic boundaries to increase risk for bipolar disorder. Archives of General Psychiatry, 62, 642–648. Guo, S., Tang, W., Shi, Y., Huang, K., Xi, Z., Xu, Y., et al. (2006). RGS4 polymorphisms and risk of schizophrenia: An association study in Han Chinese plus meta-analysis. Neuroscience Letters, 406, 122–127. Hahn, C. G., Wang, H. Y., Cho, D. S., Talbot, K., Gur, R. E., Berrettini, W. H., et  al. (2006). Altered neuregulin 1-erbB4 signaling contributes to NMDA receptor hypofunction in schizophrenia. Natural Medicines, 12, 824–828. Hall, D., Gogos, J. A., & Karayiorgou, M. (2004). The contribution of three strong candidate schizophrenia susceptibility genes in demographically distinct populations. Genes, Brain, and Behavior, 3, 240–248. Hall, J., Whalley, H. C., Job, D. E., Baig, B. J., McIntosh, A. M., Evans, K. L., et al. (2006). A neuregulin 1 variant associated with abnormal cortical function and psychotic symptoms. Nature Neuroscience, 9, 1477–1478. Harrison, P. J., & Owen, M. J. (2003). Genes for schizophrenia? Recent findings and their pathophysiological implications. Lancet, 361, 417–419. Hashimoto, L., Habita, C., Beressi, J. P., Delepine, M., Besse, C., Cambon-Thomsen, A., et al. (1994). Genetic mapping of a susceptibility locus for insulin-dependent diabetes mellitus on chromosome 11q. Nature, 371, 161–164. Hattori, E., Liu, C., Badner, J. A., Bonner, T. I., Christian, S. L., Maheshwari, M., et al. (2003). Polymorphisms at the G72/G30 gene locus, on 13q33, are associated with bipolar disorder in two independent pedigree series. American Journal of Human Genetics, 72, 1131–1140. He, Z., Li, Z., Shi, Y., Tang, W., Huang, K., Ma, G., et al. (2007). The PIP5K2A gene and schizophrenia in the Chinese population – A case-control study. Schizophrenia Research, 94, 359–365. Hennah, W., Thomson, P., McQuillin, A., Bass, N., Loukola, A., Anjorin, A., et al. (2009). DISC1 association, heterogeneity and interplay in schizophrenia and bipolar disorder. Molecular Psychiatry, 14, 865–873.

370

B. Riley

Hennah, W., Varilo, T., Kestila, M., Paunio, T., Arajarvi, R., Haukka, J., et al. (2003). Haplotype transmission analysis provides evidence of association for DISC1 to schizophrenia and suggests sex-dependent effects. Human Molecular Genetics, 12, 3151–3159. Heston, L. L. (1966). Psychiatric disorders in foster home reared children of schizophrenic mothers. The British Journal of Psychiatry, 112, 819–825. Hodgkinson, C. A., Goldman, D., Jaeger, J., Persaud, S., Kane, J. M., Lipsky, R. H., et al. (2004). Disrupted in schizophrenia 1 (DISC1): Association with schizophrenia, schizoaffective disorder, and bipolar disorder. American Journal of Human Genetics, 75, 862–872. Holliday, E. G., Handoko, H. Y., James, M. R., McGrath, J. J., Nertney, D. A., Tirupati, S., et al. (2006). Association study of the dystrobrevin-binding gene with schizophrenia in Australian and Indian samples. Twin Research and Human Genetics, 9, 531–539. Hong, C. J., Hou, S. J., Yen, F. C., Liou, Y. J., & Tsai, S. J. (2006). Family-based association study between G72/G30 genetic polymorphism and schizophrenia. NeuroReport, 17, 1067–1069. Hovatta, I., Varilo, T., Suvisaari, J., Terwilliger, J. D., Olikainen, V., Arajärvi, R., et al. (1999). A genome-wide screen for schizophrenia genes in an isolated Finnish subpopulation suggesting multiple susceptibility loci. American Journal of Human Genetics, 65, 1114–1124. Hugot, J. P., Chamaillard, M., Zouali, H., Lesage, S., Cezard, J. P., Belaiche, J., et  al. (2001). Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn’s disease. Nature, 411, 599–603. Iafrate, A. J., Feuk, L., Rivera, M. N., Listewnik, M. L., Donahoe, P. K., Qi, Y., et  al. (2004). Detection of large-scale variation in the human genome. Nature Genetics, 36, 949–951. Ide, M., Ohnishi, T., Murayama, M., Matsumoto, I., Yamada, K., Iwayama, Y., et  al. (2006). Failure to support a genetic contribution of AKT1 polymorphisms and altered AKT signaling in schizophrenia. Journal of Neurochemistry, 99, 277–287. Ikeda, M., Iwata, N., Suzuki, T., Kitajima, T., Yamanouchi, Y., Kinoshita, Y., et  al. (2004). Association of AKT1 with schizophrenia confirmed in a Japanese population. Biological Psychiatry, 56, 698–700. Ikeda, M., Iwata, N., Suzuki, T., Kitajima, T., Yamanouchi, Y., Kinoshita, Y., et al. (2005). No association of haplotype-tagging SNPs in TRAR4 with schizophrenia in Japanese patients. Schizophrenia Research, 78, 127–130. Ikeda, M., Takahashi, N., Saito, S., Aleksic, B., Watanabe, Y., Nunokawa, A., et al. (2008). Failure to replicate the association between NRG1 and schizophrenia using Japanese large sample. Schizophrenia Research, 101, 1–8. Ingason, A., Soeby, K., Timm, S., Wang, A. G., Jakobsen, K. D., Fink-Jensen, A., et al. (2006). No significant association of the 5¢ end of neuregulin 1 and schizophrenia in a large Danish sample. Schizophrenia Research, 83, 1–5. International Schizophrenia Consortium. (2008). Rare chromosomal deletions and duplications increase risk of schizophrenia. Nature, 455, 237–241. Ishiguro, H., Horiuchi, Y., Koga, M., Inada, T., Iwata, N., Ozaki, N., et al. (2007). RGS4 is not a susceptibility gene for schizophrenia in Japanese: Association study in a large case-control population. Schizophrenia Research, 89, 161–164. Iwata, N., Suzuki, T., Ikeda, M., Kitajima, T., Yamanouchi, Y., Inada, T., et al. (2004). No association with the neuregulin 1 haplotype to Japanese schizophrenia. Molecular Psychiatry, 9, 126–127. Jamra, R. A., Klein, K., Villela, A. W., Becker, T., Schulze, T. G., Schmael, C., et  al. (2006). Association study between genetic variants at the PIP5K2A gene locus and schizophrenia and bipolar affective disorder. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 141B, 663–665. Joo, E. J., Lee, K. Y., Jeong, S. H., Ahn, Y. M., Koo, Y. J., & Kim, Y. S. (2006). The dysbindin gene (DTNBP1) and schizophrenia: No support for an association in the Korean population. Neuroscience Letters, 407, 101–106. Joo, E. J., Lee, K. Y., Jeong, S. H., Chang, J. S., Ahn, Y. M., Koo, Y. J., et al. (2007). Dysbindin gene variants are associated with bipolar I disorder in a Korean population. Neuroscience Letters, 418, 272–275.

Genetic Studies of Schizophrenia

371

Kallmann, F. J. (1938). The genetics of schizophrenia. New York: Augustin. Kalsi, G., Chen, C. H., Smyth, C., Brynjolfsson, J., Sigmundson, T., Curtis, D., et  al. (1996). Genetic linkage analysis in an Icelandic/British sample fails to exclude the putative chromosome 13q14.1-q32 schizophrenia susceptibility locus. American Journal of Human Genetics, 59, A388. Karayiorgou, M., Morris, M. A., Morrow, B., Shprintzen, R. J., Goldberg, R., Borrow, J., et al. (1995). Schizophrenia susceptibility associated with interstitial deletions of chromosome 22q11. Proceedings of the National Academy of Sciences of the United States of America, 92, 7612–7616. Kendler, K. S., & Diehl, S. R. (1993). The genetics of schizophrenia: A current, genetic-epidemiologic perspective. Schizophrenia Bulletin, 19, 261–285. Kendler, K. S., Gruenberg, A. M., & Tsuang, M. T. (1986). A DSM-III family study of the nonschizophrenic psychotic disorders. The American Journal of Psychiatry, 143, 1098–1105. Kendler, K. S., Kuhn, J. W., Prescott, C. A., Vittum, J., & Riley, B. (2005). The interaction of stressful life events and a serotonin transporter polymorphism in the prediction of episodes of major depression: A replication. Archives of General Psychiatry, 62, 529–535. Kendler, K. S., McGuire, M., Gruenberg, A. M., Spellman, M., O’Hare, A., & Walsh, D. (1993). The Roscommon family study: II. The risk of nonschizophrenic nonaffective psychoses in relatives. Archives of General Psychiatry, 50, 645–652. Kendler, K. S., Straub, R. E., MacLean, C. J., & Walsh, D. (1996). Reflections on the evidence for a vulnerability locus for schizophrenia on chromosome 6p24-22. American Journal of Medical Genetics, 67, 124–126. Kety, S. S., Rosenthal, D., Wender, P. H., Schulsinger, F., & Jacobsen, B. (1968). The types and prevalence of mental illness in the biological and adoptive families of adopted schizophrenics. Journal of Psychiatric Research, 6, 345–362. Kety, S. S., Wender, P. H., Jacobsen, B., Ingraham, L. J., Jansson, L., Faber, B., et  al. (1994). Mental illness in the biological and adoptive relatives of schizophrenic adoptees: Replication of the Copenhagen study in the rest of Denmark. Archives of General Psychiatry, 51, 442–455. Kim, H. J., Park, H. J., Jung, K. H., Ban, J. Y., Ra, J., Kim, J. W., et al. (2008). Association study of polymorphisms between DISC1 and schizophrenia in a Korean population. Neuroscience Letters, 430, 60–63. Kirov, G., Ivanov, D., Williams, N. M., Preece, A., Nikolov, I., Milev, R., et al. (2004). Strong evidence for association between the dystrobrevin binding protein 1 gene (DTNBP1) and schizophrenia in 488 parent-offspring trios from Bulgaria. Biological Psychiatry, 55, 971–975. Kockelkorn, T. T., Arai, M., Matsumoto, H., Fukuda, N., Yamada, K., Minabe, Y., et al. (2004). Association study of polymorphisms in the 5¢ upstream region of human DISC1 gene with schizophrenia. Neuroscience Letters, 368, 41–45. Korostishevsky, M., Kaganovich, M., Cholostoy, A., Ashkenazi, M., Ratner, Y., Dahary, D., et al. (2004). Is the G72/G30 locus associated with schizophrenia? Single nucleotide polymorphisms, haplotypes, and gene expression analysis. Biological Psychiatry, 56, 169–176. Korostishevsky, M., Kremer, I., Kaganovich, M., Cholostoy, A., Murad, I., Muhaheed, M., et al. (2006). Transmission disequilibrium and haplotype analyses of the G72/G30 locus: Suggestive linkage to schizophrenia in Palestinian Arabs living in the North of Israel. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 141B, 91–95. Kruglyak, L., Daly, M. J., Reeve-Daly, M. P., & Lander, E. S. (1996). Parametric and nonparametric linkage analysis: A unified multipoint approach. American Journal of Human Genetics, 58, 1347–1363. Laird, N. M., Horvath, S., & Xu, X. (2000). Implementing a unified approach to family-based tests of association. Genetic Epidemiology, 19(suppl 1), S36–S42. Laitinen, T., Daly, M. J., Rioux, J. D., Kauppi, P., Laprise, C., Petays, T., et al. (2001). A susceptibility locus for asthma-related traits on chromosome 7 revealed by genome-wide scan in a founder population. Nature Genetics, 28, 87–91.

372

B. Riley

Lander, E., & Kruglyak, L. (1995). Genetic dissection of complex traits: Guidelines for interpreting and reporting linkage results. Nature Genetics, 11, 241–247. Law, A. J., Kleinman, J. E., Weinberger, D. R., & Weickert, C. S. (2007). Disease-associated intronic variants in the ErbB4 gene are related to altered ErbB4 splice-variant expression in the brain in schizophrenia. Human Molecular Genetics, 16, 129–141. Lencz, T., Morgan, T. V., Athanasiou, M., Dain, B., Reed, C. R., Kane, J. M., et  al. (2007). Converging evidence for a pseudoautosomal cytokine receptor gene locus in schizophrenia. Molecular Psychiatry, 12, 572–580. Levinson, D. F., Holmans, P. A., Laurent, C., Riley, B., Pulver, A. E., Gejman, P. V., et al. (2002). No major schizophrenia locus detected on chromosome 1q in a large multicenter sample. Science, 296, 739–741. Levinson, D. F., Wildenauer, D. B., Schwab, S. G., Albus, M., Hallmayer, J., Lerer, B., et  al. (1996). Additional support for schizophrenia linkage on chromosomes 6 and 8: A multicenter study. American Journal of Medical Genetics, 67, 580–594. Lewis, C. M., Levinson, D. F., Wise, L. H., DeLisi, L. E., Straub, R. E., Hovatta, I., et al. (2003). Genome scan meta-analysis of schizophrenia and bipolar disorder, part II: Schizophrenia. American Journal of Human Genetics, 73, 34–48. Li, D., & He, L. (2006). Association study of the G-protein signaling 4 (RGS4) and proline dehydrogenase (PRODH) genes with schizophrenia: A meta-analysis. European Journal of Human Genetics, 14, 1130–1135. Li, D., & He, L. (2007). G72/G30 genes and schizophrenia: A systematic meta-analysis of association studies. Genetics, 175, 917–922. Li, T., Stefansson, H., Gudfinnsson, E., Cai, G., Liu, X., Murray, R. M., et al. (2004). Identification of a novel neuregulin 1 at-risk haplotype in Han schizophrenia Chinese patients, but no association with the Icelandic/Scottish risk haplotype. Molecular Psychiatry, 9, 698–704. Li, T., Zhang, F., Liu, X., Sun, X., Sham, P. C., Crombie, C., et al. (2005). Identifying potential risk haplotypes for schizophrenia at the DTNBP1 locus in Han Chinese and Scottish populations. Molecular Psychiatry, 10, 1037–1044. Lieberman, J. A., Stroup, T. S., McEvoy, J. P., Swartz, M. S., Rosenheck, R. A., Perkins, D. O., et al. (2005). Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. The New England Journal of Medicine, 353, 1209–1223. Lin, M. W., Curtis, D., Williams, N., Arranz, M., Nanko, S., Collier, D., et al. (1995). Suggestive evidence for linkage of schizophrenia to markers on chromosome 13q14.1-q32. Psychiatric Genetics, 5, 117–126. Lin, M. W., Sham, P., Hwu, H. G., Collier, D., Murray, R., & Powell, J. F. (1997). Suggestive evidence for linkage of schizophrenia to markers on chromosome 13 in Caucasian but not Oriental populations. Human Genetics, 99, 417–420. Lindholm, E., Ekholm, B., Balciuniene, J., Johansson, G., Castensson, A., Koisti, M., et al. (1999). Linkage analysis of a large Swedish kindred provides further support for a susceptibility locus for schizophrenia on chromosome 6p23. American Journal of Medical Genetics, 88, 369–377. Liu, Y. L., Fann, C. S., Liu, C. M., Chang, C. C., Wu, J. Y., Hung, S. I., et al. (2006a). No association of G72 and D-amino acid oxidase genes with schizophrenia. Schizophrenia Research, 87, 15–20. Liu, Y. L., Fann, C. S., Liu, C. M., Chen, W. J., Wu, J. Y., Hung, S. I., et al. (2006b). A single nucleotide polymorphism fine mapping study of chromosome 1q42.1 reveals the vulnerability genes for schizophrenia, GNPAT and DISC1: Association with impairment of sustained attention. Biological Psychiatry, 60, 554–562. Liu, Y. L., Fann, C. S., Liu, C. M., Wu, J. Y., Hung, S. I., Chan, H. Y., et al. (2006c). Absence of significant associations between four AKT1 SNP markers and schizophrenia in the Taiwanese population. Psychiatric Genetics, 16, 39–41. Liu, X., He, G., Wang, X., Chen, Q., Qian, X., Lin, W., et al. (2004). Association of DAAO with schizophrenia in the Chinese population. Neuroscience Letters, 369, 228–233.

Genetic Studies of Schizophrenia

373

Liu, H., Heath, S. C., Sobin, C., Roos, J. L., Galke, B. L., Blundell, M. L., et al. (2002). Genetic variation at the 22q11 PRODH2/DGCR6 locus presents an unusual pattern and increases susceptibility to schizophrenia. Proceedings of the National Academy of Sciences of the United States of America, 99, 3717–3722. Liu, C. M., Hwu, H. G., Lin, M. W., Ou-Yang, W. C., Lee, S. F., Fann, C. S., et  al. (2001). Suggestive evidence for linkage of schizophrenia to markers at chromosome 15q13-14 in Taiwanese families. American Journal of Medical Genetics, 105, 658–661. Liu, C. M., Liu, Y. L., Fann, C. S., Yang, W. C., Wu, J. Y., Hung, S. I., et al. (2007). No association evidence between schizophrenia and dystrobrevin-binding protein 1 (DTNBP1) in Taiwanese families. Schizophrenia Research. doi:10.1016/j.schres.2007.02.003. Liu, Y. L., Shen-Jang, F. C., Liu, C. M., Wu, J. Y., Hung, S. I., Chan, H. Y., et  al. (2006d). Evaluation of RGS4 as a candidate gene for schizophrenia. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 141, 418–420. Ma, J., Qin, W., Wang, X. Y., Guo, T. W., Bian, L., Duan, S. W., et al. (2006). Further evidence for the association between G72/G30 genes and schizophrenia in two ethnically distinct populations. Molecular Psychiatry, 11, 479–487. Mah, S., Nelson, M. R., DeLisi, L. E., Reneland, R. H., Markward, N., James, M. R., et al. (2006). Identification of the semaphorin receptor PLXNA2 as a candidate for susceptibility to schizophrenia. Molecular Psychiatry, 11, 471–478. Martin, E. R., Monks, S. A., Warren, L. L., & Kaplan, N. L. (2000). A test for linkage and association in general pedigrees: The pedigree disequilibrium test. American Journal of Human Genetics, 67, 146–154. Maziade, M., Bissonnette, L., Rouillard, E., Martinez, M., Turgeon, M., Charron, L., et al. (1997). 6p24-22 region and major psychoses in the Eastern Quebec population. Le Groupe IREP. American Journal of Medical Genetics, 74, 311–318. McClellan, J. M., Susser, E., & King, M. C. (2007). Schizophrenia: A common disease caused by multiple rare alleles. The British Journal of Psychiatry, 190, 194–199. McGue, M., Gottesman, I., & Rao, D. C. (1985). Resolving genetic models for the transmission of schizophrenia. Genetic Epidemiology, 2, 99–110. Millar, J. K., Wilson-Annan, J. C., Anderson, S., Christie, S., Taylor, M. S., Semple, C. A. M., et al. (2000). Disruption of two novel genes by a translocation co-segregating with schizophrenia. Human Molecular Genetics, 9, 1415–1423. Mirnics, K., Middleton, F. A., Lewis, D. A., & Levitt, P. (2001). Analysis of complex brain disorders with gene expression microarrays: Schizophrenia as a disease of the synapse. Trends in Neurosciences, 24, 479–486. Moghaddam, B. (2003). Bringing order to the glutamate chaos in schizophrenia. Neuron, 40, 881–884. Moises, H. W., Yang, L., Kristbjarnarson, H., Wiese, C., Byerley, W., Macciardi, F., et al. (1995). An international two-stage genome-wide search for schizophrenia susceptibility genes. Nature Genetics, 11, 321–324. Morris, D. W., McGhee, K. A., Schwaiger, S., Scully, P., Quinn, J., Meagher, D., et al. (2003). No evidence for association of the dysbindin gene [DTNBP1] with schizophrenia in an Irish population-based study. Schizophrenia Research, 60, 167–172. Morris, D. W., Rodgers, A., McGhee, K. A., Schwaiger, S., Scully, P., Quinn, J., et  al. (2004). Confirming RGS4 as a susceptibility gene for schizophrenia. American Journal of Medical Genetics, 125B, 50–53. Morton, N. E. (1955). Sequential tests for the detection of linkage. American Journal of Human Genetics, 7, 277–318. Mulle, J. G., Chowdari, K. V., Nimgaonkar, V., & Chakravarti, A. (2005). No evidence for association to the G72/G30 locus in an independent sample of schizophrenia families. Molecular Psychiatry, 10, 431–433. Munafo, M. R., Thiselton, D. L., Clark, T. G., & Flint, J. (2006). Association of the NRG1 gene and schizophrenia: A meta-analysis. Molecular Psychiatry, 11, 539–546.

374

B. Riley

Murphy, K. C., Jones, L. A., & Owen, M. J. (1999). High rates of schizophrenia in adults with velo-cardio-facial syndrome. Archives of General Psychiatry, 56, 940–945. Nicodemus, K. K., Luna, A., Vakkalanka, R., Goldberg, T., Egan, M., Straub, R. E., et al. (2006). Further evidence for association between ErbB4 and schizophrenia and influence on cognitive intermediate phenotypes in healthy controls. Molecular Psychiatry, 11, 1062–1065. Norton, N., Moskvina, V., Morris, D. W., Bray, N. J., Zammit, S., Williams, N. M., et al. (2006). Evidence that interaction between neuregulin 1 and its receptor erbB4 increases susceptibility to schizophrenia. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 141, 96–101. Norton, N., Williams, H. J., Dwyer, S., Carroll, L., Peirce, T., Moskvina, V., et  al. (2007). Association analysis of AKT1 and schizophrenia in a UK case control sample. Schizophrenia Research, 91, 58–65. Numakawa, T., Yagasaki, Y., Ishimoto, T., Okada, T., Suzuki, T., Iwata, N., et al. (2004). Evidence of novel neuronal functions of dysbindin, a susceptibility gene for schizophrenia. Human Molecular Genetics, 13, 2699–2708. O’Donovan, M. C., Craddock, N., Norton, N., Williams, H., Peirce, T., Moskvina, V., et al. (2008). Identification of loci associated with schizophrenia by genome-wide association and followup. Nature Genetics, 40, 1053–1055. Ogura, Y., Bonen, D. K., Inohara, N., Nicolae, D. L., Chen, F. F., Ramos, R., et  al. (2001). A frameshift mutation in NOD2 associated with susceptibility to Crohn’s disease. Nature, 411, 603–606. Ohtsuki, T., Inada, T., & Arinami, T. (2004). Failure to confirm association between AKT1 haplotype and schizophrenia in a Japanese case-control population. Molecular Psychiatry, 9, 981–983. Pae, C. U., Serretti, A., Mandelli, L., Yu, H. S., Patkar, A. A., Lee, C. U., et  al. (2007). Effect of 5-haplotype of dysbindin gene (DTNBP1) polymorphisms for the susceptibility to bipolar I disorder. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 144B, 701–703. Pae, C. U., Yu, H. S., Amann, D., Kim, J. J., Lee, C. U., Lee, S. J., et al. (2008). Association of the trace amine associated receptor 6 (TAAR6) gene with schizophrenia and bipolar disorder in a Korean case control sample. Journal of Psychiatric Research, 42, 35–40. Palo, O. M., Antila, M., Silander, K., Hennah, W., Kilpinen, H., Soronen, P., et  al. (2007). Association of distinct allelic haplotypes of DISC1 with psychotic and bipolar spectrum disorders and with underlying cognitive impairments. Human Molecular Genetics, 16, 2517–2528. Pedrosa, E., Ye, K., Nolan, K. A., Morrell, L., Okun, J. M., Persky, A. D., et al. (2007). Positive association of schizophrenia to JARID2 gene. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 144B, 45–51. Perlis, R. H., Purcell, S., Fagerness, J., Kirby, A., Petryshen, T. L., Fan, J., et al. (2008). Familybased association study of lithium-related and other candidate genes in bipolar disorder. Archives of General Psychiatry, 65, 53–61. Perry, G. H., Dominy, N. J., Claw, K. G., Lee, A. S., Fiegler, H., Redon, R., et al. (2007). Diet and the evolution of human amylase gene copy number variation. Nature Genetics, 39, 1256–1260. Peters, K., Wiltshire, S., Henders, A. K., Dragovic, M., Badcock, J. C., Chandler, D., et al. (2008). Comprehensive analysis of tagging sequence variants in DTNBP1 shows no association with schizophrenia or with its composite neurocognitive endophenotypes. American Journal of Medical Genetics. Part B, Neuropsychiatric Genetics, 147B, 1159–1166. Petretto, E., Liu, E. T., & Aitman, T. J. (2007). A gene harvest revealing the archeology and complexity of human disease. Nature Genetics, 39, 1299–1301. Petryshen, T. L., Middleton, F. A., Kirby, A., Aldinger, K. A., Purcell, S., Tahl, A. R., et al. (2005). Support for involvement of neuregulin 1 in schizophrenia pathophysiology. Molecular Psychiatry, 10(366–374), 328. Polymeropoulos, M. H., Coon, H., Byerley, W., Gershon, E. S., Goldin, L., Crow, T. J., et  al. (1994). Search for a schizophrenia susceptibility locus on human chromosome 22. American Journal of Medical Genetics, 54, 93–99.

Genetic Studies of Schizophrenia

375

Prata, D., Breen, G., Osborne, S., Munro, J., St Clair, D., & Collier, D. (2008). Association of DAO and G72(DAOA)/G30 genes with bipolar affective disorder. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 147, 914–917. Prescott, C. A., & Gottesman, I. I. (1993). Genetically mediated vulnerability to schizophrenia. The Psychiatric Clinics of North America, 16, 245–267. Pulver, A. E., Karayiorgou, M., Lasseter, V. K., Wolyniec, P., Kasch, L., Antonarakis, S., et al. (1994). Follow-up of a report of a potential linkage for schizophrenia on chromosome 22q12q13.1: Part 2. American Journal of Medical Genetics, 54, 44–50. Pulver, A. E., Karayiorgou, M., Wolyniec, P. S., Lasseter, V. K., Kasch, L., Nestadt, G., et  al. (1994). Sequential strategy to identify a susceptibility gene for schizophrenia: Report of potential linkage on chromosome 22q12-q13.1: Part 1. American Journal of Medical Genetics, 54, 36–43. Pulver, A. E., Lasseter, V. K., Kasch, L., Wolyniec, P., Nestadt, G., Blouin, J. L., et al. (1995). Schizophrenia: A genome scan targets chromosomes 3p and 8p as potential sites of susceptibility genes. American Journal of Medical Genetics, 60, 252–260. Pulver, A. E., Nestadt, G., Goldberg, R., Shprintzen, R. J., Lamacz, M., Wolyniec, P. S., et  al. (1994). Psychotic illness in patients diagnosed with velo-cardio-facial syndrome and their relatives. The Journal of Nervous and Mental Disease, 182, 476–478. Purcell, S. M., Wray, N. R., Stone, J. L., Visscher, P. M., O’Donovan, M. C., Sullivan, P. F., et al. (2009). Common polygenic variation contributes to risk of schizophrenia and bipolar disorder. Nature, 460, 748–752. Qu, M., Tang, F., Yue, W., Ruan, Y., Lu, T., Liu, Z., et al. (2007). Positive association of the disrupted-in-Schizophrenia-1 gene (DISC1) with schizophrenia in the Chinese Han population. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 144, 266–270. Ray, P. N., Belfall, B., Duff, C., Logan, C., Kean, V., Thompson, M. W., et al. (1985). Cloning of the breakpoint of an X;21 translocation associated with Duchenne muscular dystrophy. Nature, 318, 672–675. Redon, R., Ishikawa, S., Fitch, K. R., Feuk, L., Perry, G. H., Andrews, T. D., et al. (2006). Global variation in copy number in the human genome. Nature, 444, 444–454. Riley, B. P., Makoff, A., Mogudi-Carter, M., Jenkins, T., Williamson, R., Collier, D., et al. (2000). Haplotype transmission disequilibrium and evidence for linkage of the CHRNA7 gene region to schizophrenia in southern African Bantu families. American Journal of Medical Genetics, 96, 196–201. Riley, B. P., & McGuffin, P. (2000). Linkage and associated studies of schizophrenia. American Journal of Medical Genetics, 97, 23–44. Risch, N. (1990). Linkage strategies for genetically complex traits. I. Multilocus models. American Journal of Human Genetics, 46, 222–228. Risch, N., & Merikangas, K. (1996). The future of genetic studies of complex human diseases. Science, 273, 1516–1517. Rizig, M. A., McQuillin, A., Puri, V., Choudhury, K., Datta, S., Thirumalai, S., et  al. (2006). Failure to confirm genetic association between schizophrenia and markers on chromosome 1q23.3 in the region of the gene encoding the regulator of G-protein signaling 4 protein (RGS4). American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 141, 296–300. Rosa, A., Gardner, M., Cuesta, M. J., Peralta, V., Fatjo-Vilas, M., Miret, S., et al. (2007). Familybased association study of neuregulin-1 gene and psychosis in a Spanish sample. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 144B, 954–957. Rudin, E. (1916). Zur Vererbung und Neuentstehung der Dementia Praecox. Berlin: Springer. Rujescu, D., Ingason, A., Cichon, S., Pietilainen, O. P., Barnes, M. R., Toulopoulou, T., et  al. (2009). Disruption of the neurexin 1 gene is associated with schizophrenia. Human Molecular Genetics, 18, 988–996. Sachs, N. A., Sawa, A., Holmes, S. E., Ross, C. A., DeLisi, L. E., & Margolis, R. L. (2005). A frameshift mutation in disrupted in Schizophrenia 1 in an American family with schizophrenia and schizoaffective disorder. Molecular Psychiatry, 10, 758–764.

376

B. Riley

Saggers-Gray, L., Heriani, H., Handoko, H. Y., Irmansyah, I., Kusumawardhani, A. A., Widyawati, I., et al. (2008). Association of PIP5K2A with schizophrenia: A study in an indonesian family sample. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 147B, 1310–1313. Sanders, A. R., Duan, J., Levinson, D. F., Shi, J., He, D., Hou, C., et al. (2008). No significant association of 14 candidate genes with schizophrenia in a large European ancestry sample: Implications for psychiatric genetics. The American Journal of Psychiatry, 165, 497–506. Sawcer, S., Jones, H. B., Feakes, R., Gray, J., Smaldon, N., Chataway, J., et al. (1996). A genome screen in multiple sclerosis reveals susceptibility loci on chromosome 6p21 and 17q22. Nature Genetics, 13, 464–468. Saxena, R., Voight, B. F., Lyssenko, V., Burtt, N. P., de Bakker, P. I., Chen, H., et  al. (2007). Genome-wide association analysis identifies loci for type 2 diabetes and triglyceride levels. Science, 316, 1331–1336. Schumacher, J., Jamra, R. A., Freudenberg, J., Becker, T., Ohlraun, S., Otte, A. C., et al. (2004). Examination of G72 and D-amino-acid oxidase as genetic risk factors for schizophrenia and bipolar affective disorder. Molecular Psychiatry, 9, 203–207. Schwab, S. G., Albus, M., Hallmayer, J., Honig, S., Borrmann, M., Lichtermann, D., et al. (1995). Evaluation of a susceptibility gene for schizophrenia on chromosome 6p by multipoint affected sib-pair linkage analysis. Nature Genetics, 11, 325–327. Schwab, S. G., Hoefgen, B., Hanses, C., Hassenbach, M. B., Albus, M., Lerer, B., et al. (2005). Further evidence for association of variants in the AKT1 gene with schizophrenia in a sample of European sib-pair families. Biological Psychiatry, 58, 446–450. Schwab, S. G., Knapp, M., Mondabon, S., Hallmayer, J., Borrmann-Hassenbach, M., Albus, M., et  al. (2003). Support for association of schizophrenia with genetic variation in the 6p22.3 gene, dysbindin, in sib-pair families with linkage and in an additional sample of triad families. American Journal of Human Genetics, 72, 185–190. Schwab, S. G., Knapp, M., Sklar, P., Eckstein, G. N., Sewekow, C., Borrmann-Hassenbach, M., et al. (2006). Evidence for association of DNA sequence variants in the phosphatidylinositol4-phosphate 5-kinase IIalpha gene (PIP5K2A) with schizophrenia. Molecular Psychiatry, 11, 837–846. Scott, L. J., Mohlke, K. L., Bonnycastle, L. L., Willer, C. J., Li, Y., Duren, W. L., et al. (2007). A genome-wide association study of type 2 diabetes in Finns detects multiple susceptibility variants. Science, 316, 1341–1345. Sebat, J., Lakshmi, B., Malhotra, D., Troge, J., Lese-Martin, C., Walsh, T., et al. (2007). Strong association of de novo copy number mutations with autism. Science, 316, 445–449. Sebat, J., Lakshmi, B., Troge, J., Alexander, J., Young, J., Lundin, P., et al. (2004). Large-scale copy number polymorphism in the human genome. Science, 305, 525–528. Shaw, S. H., Kelly, M., Smith, A. B., Shields, G., Hopkins, P. J., Loftus, J., et  al. (1998). A genome-wide search for schizophrenia susceptibility genes. American Journal of Medical Genetics, 81, 364–376. Shi, J., Badner, J. A., Gershon, E. S., & Liu, C. (2008). Allelic association of G72/G30 with schizophrenia and bipolar disorder: A comprehensive meta-analysis. Schizophrenia Research, 98, 89–97. Shi, J., Levinson, D. F., Duan, J., Sanders, A. R., Zheng, Y., Pe’er, I., et al. (2009). Common variants on chromosome 6p22.1 are associated with schizophrenia. Nature, 460, 753–757. Shifman, S., Bronstein, M., Sternfeld, M., Pisante-Shalom, A., Lev-Lehman, E., Weizman, A., et al. (2002). A highly significant association between a COMT haplotype and schizophrenia. American Journal of Human Genetics, 71, 1296–1302. Shin, H. D., Park, B. L., Kim, E. M., Lee, S. O., Cheong, H. S., Lee, C. H., et  al. (2007). Association analysis of G72/G30 polymorphisms with schizophrenia in the Korean population. Schizophrenia Research, 96, 119–124. Shinkai, T., De Luca, V., Hwang, R., Muller, D. J., Lanktree, M., Zai, G., et al. (2007). Association analyses of the DAOA/G30 and D-amino-acid oxidase genes in schizophrenia: Further evidence for a role in schizophrenia. Neuromolecular Medicine, 9, 169–177.

Genetic Studies of Schizophrenia

377

Silberberg, G., Darvasi, A., Pinkas-Kramarski, R., & Navon, R. (2006). The involvement of ErbB4 with schizophrenia: Association and expression studies. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 141B, 142–148. Sladek, R., Rocheleau, G., Rung, J., Dina, C., Shen, L., Serre, D., et al. (2007). A genome-wide association study identifies novel risk loci for type 2 diabetes. Nature, 445, 881–885. So, H. C., Chen, R. Y., Chen, E. Y., Cheung, E. F., Li, T., & Sham, P. C. (2008). An association study of RGS4 polymorphisms with clinical phenotypes of schizophrenia in a Chinese population. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 147B, 77–85. Sobell, J. L., Richard, C., Wirshing, D. A., & Heston, L. L. (2005). Failure to confirm association between RGS4 haplotypes and schizophrenia in Caucasians. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 139, 23–27. Song, W., Li, W., Feng, J., Heston, L. L., Scaringe, W. A., & Sommer, S. S. (2008). Identification of high risk DISC1 structural variants with a 2% attributable risk for schizophrenia. Biochemical and Biophysical Research Communications, 367, 700–706. Spielman, R. S., & Ewens, W. J. (1996). The TDT and other family-based tests for linkage disequilibrium and association. American Journal of Human Genetics, 59, 983–989. St Clair, D., Blackwood, D., Muir, W., Carothers, A., Walker, M., Spowart, G., et  al. (1990). Association within a family of a balanced autosomal translocation with major mental illness. Lancet, 336, 13–16. St Clair, D., Xu, M., Wang, P., Yu, Y., Fang, Y., Zhang, F., et al. (2005). Rates of adult schizophrenia following prenatal exposure to the Chinese famine of 1959-1961. JAMA., 294, 557–562. Stefansson, H., Ophoff, R. A., Steinberg, S., Andreassen, O. A., Cichon, S., Rujescu, D., et al. (2009). Common variants conferring risk of schizophrenia. Nature, 460, 744–747. Stefansson, H., Rujescu, D., Cichon, S., Pietilainen, O. P., Ingason, A., Steinberg, S., et al. (2008). Large recurrent microdeletions associated with schizophrenia. Nature, 455, 232–236. Stefansson, H., Sarginson, J., Kong, A., Yates, P., Steinthorsdottir, V., Gudfinnsson, E., et  al. (2003). Association of neuregulin 1 with schizophrenia confirmed in a Scottish population. American Journal of Human Genetics, 72, 83–87. Stefansson, H., Sigurdsson, E., Steinthorsdottir, V., Bjornsdottir, S., Sigmundsson, T., Ghosh, S., et al. (2002). Neuregulin 1 and susceptibility to schizophrenia. American Journal of Human Genetics, 71, 877–892. Steinthorsdottir, V., Thorleifsson, G., Reynisdottir, I., Benediktsson, R., Jonsdottir, T., Walters, G. B., et al. (2007). A variant in CDKAL1 influences insulin response and risk of type 2 diabetes. Nature Genetics, 39, 770–775. Straub, R. E., Jiang, Y., MacLean, C. J., Ma, Y., Webb, B. T., Myakishev, M. V., et al. (2002). Genetic variation in the 6p22.3 gene DTNBP1, the human ortholog of mouse dysbindin, is associated with schizophrenia. American Journal of Human Genetics, 71, 337–348. Straub, R. E., MacLean, C. J., O’Neill, F. A., Burke, J., Murphy, B., Duke, F., et  al. (1995). A potential vulnerability locus for schizophrenia on chromosome 6p24-22: Evidence for genetic heterogeneity. Nature Genetics, 11, 287–293. Sullivan, P. F. (2005). The genetics of schizophrenia. PLoS Medicine, 2, e212. Sullivan, P. F., Kendler, K. S., & Neale, M. C. (2003). Schizophrenia as a complex trait: Evidence from a meta-analysis of twin studies. Archives of General Psychiatry, 60, 1187–1192. Sullivan, P. F., Lin, D., Tzeng, J. Y., van den Oord, E., Perkins, D., Stroup, T. S., et al. (2008). Genomewide association for schizophrenia in the CATIE study: Results of stage 1. Molecular Psychiatry, 13, 570–584. Sun, S., Wang, F., Wei, J., Cao, L. Y., Wu, G. Y., Lu, L., et al. (2008). Association between interleukin-3 receptor alpha polymorphism and schizophrenia in the Chinese population. Neuroscience Letters, 440, 35–37. Susser, E. S., & Lin, S. P. (1992). Schizophrenia after prenatal exposure to the Dutch Hunger Winter of 1944-1945. Archives of General Psychiatry, 49, 983–988. Szeszko, P. R., Hodgkinson, C. A., Robinson, D. G., DeRosse, P., Bilder, R. M., Lencz, T., et al. (2008). DISC1 is associated with prefrontal cortical gray matter and positive symptoms in schizophrenia. Biological Psychology, 79, 103–110.

378

B. Riley

Takeshita, M., Yamada, K., Hattori, E., Iwayama, Y., Toyota, T., Iwata, Y., et al. (2008). Genetic examination of the PLXNA2 gene in Japanese and Chinese people with schizophrenia. Schizophrenia Research, 99, 359–364. Talbot, K., Eidem, W. L., Tinsley, C. L., Benson, M. A., Thompson, E. W., Smith, R. J., et  al. (2004). Dysbindin-1 is reduced in intrinsic, glutamatergic terminals of the hippocampal formation in schizophrenia. The Journal of Clinical Investigation, 113, 1353–1363. Talkowski, M. E., Seltman, H., Bassett, A. S., Brzustowicz, L. M., Chen, X., Chowdari, K. V., et al. (2006). Evaluation of a susceptibility gene for schizophrenia: Genotype based meta-analysis of RGS4 polymorphisms from thirteen independent samples. Biological Psychiatry, 60, 152–162. Tang, J. X., Chen, W. Y., He, G., Zhou, J., Gu, N. F., Feng, G. Y., et al. (2004). Polymorphisms within 5¢ end of the Neuregulin 1 gene are genetically associated with schizophrenia in the Chinese population. Molecular Psychiatry, 9, 11–12. Tang, J. X., Zhou, J., Fan, J. B., Li, X. W., Shi, Y. Y., Gu, N. F., et al. (2003). Family-based association study of DTNBP1 in 6p22.3 and schizophrenia. Molecular Psychiatry, 8, 717–718. Thiselton, D. L., Vladimirov, V. I., Kuo, P. H., McClay, J., Wormley, B., Fanous, A., et al. (2008). AKT1 is associated with schizophrenia across multiple symptom dimensions in the Irish study of high density schizophrenia families. Biological Psychiatry, 63, 449–457. Thiselton, D. L., Webb, B. T., Neale, B. M., Ribble, R. C., O’Neill, F. A., Walsh, D., et al. (2004). No evidence for linkage or association of neuregulin-1 (NRG1) with disease in the Irish study of high-density schizophrenia families (ISHDSF). Molecular Psychiatry, 9, 777–783. Thomson, P. A., Christoforou, A., Morris, S. W., Adie, E., Pickard, B. S., Porteous, D. J., et al. (2007). Association of neuregulin 1 with schizophrenia and bipolar disorder in a second cohort from the Scottish population. Molecular Psychiatry, 12, 94–104. Thomson, P. A., Wray, N. R., Millar, J. K., Evans, K. L., Hellard, S. L., Condie, A., et al. (2005). Association between the TRAX/DISC locus and both bipolar disorder and schizophrenia in the Scottish population. Molecular Psychiatry, 10(657–668), 616. Tienari, P. (1991). Interaction between genetic vulnerability and family environment: The Finnish adoptive family study of schizophrenia. Acta Psychiatrica Scandinavica, 84, 460–465. Tienari, P., Wynne, L. C., Sorri, A., Lahti, I., Laksy, K., Moring, J., et  al. (2004). Genotypeenvironment interaction in schizophrenia-spectrum disorder. Long-term follow-up study of Finnish adoptees. British Journal of Psychiatry, 184, 216–222. Tochigi, M., Zhang, X., Ohashi, J., Hibino, H., Otowa, T., Rogers, M., et al. (2006). Association study of the dysbindin (DTNBP1) gene in schizophrenia from the Japanese population. Neuroscience Research, 56, 154–158. Tosato, S., Dazzan, P., & Collier, D. (2005). Association between the neuregulin 1 gene and schizophrenia: A systematic review. Schizophrenia Bulletin, 31, 613–617. Tosato, S., Ruggeri, M., Bonetto, C., Bertani, M., Marrella, G., Lasalvia, A., et  al. (2007). Association study of dysbindin gene with clinical and outcome measures in a representative cohort of Italian schizophrenic patients. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 144B, 647–659. Tsuang, D. W., Skol, A. D., Faraone, S. V., Bingham, S., Young, K. A., Prabhudesai, S., et al. (2001). Examination of genetic linkage of chromosome 15 to schizophrenia in a large Veterans Affairs Cooperative Study sample. American Journal of Medical Genetics, 105, 662–668. Turunen, J. A., Peltonen, J. O., Pietilainen, O. P., Hennah, W., Loukola, A., Paunio, T., et  al. (2007). The role of DTNBP1, NRG1, and AKT1 in the genetics of schizophrenia in Finland. Schizophrenia Research, 91, 27–36. Tuzun, E., Sharp, A. J., Bailey, J. A., Kaul, R., Morrison, V. A., Pertz, L. M., et al. (2005). Finescale structural variation of the human genome. Nature Genetics, 37, 727–732. Van Den Bogaert, A., Schumacher, J., Schulze, T. G., Otte, A. C., Ohlraun, S., Kovalenko, S., et al. (2003). The DTNBP1 (dysbindin) gene contributes to schizophrenia, depending on family history of the disease. American Journal of Human Genetics, 73, 1438–1443. van den Oord, E., Sullivan, P. F., Chen, X., Kendler, K. S., & Riley, B. (2003). Identification of a high risk haplotype for the dystrobrevin binding protein 1 (DTNBP1) gene in the Irish study of high density schizophrenia families. Molecular Psychiatry, 8, 499–510.

Genetic Studies of Schizophrenia

379

Vilella, E., Costas, J., Sanjuan, J., Guitart, M., De, D. Y., Carracedo, A., et al. (2007). Association of schizophrenia with DTNBP1 but not with DAO, DAOA, NRG1 and RGS4 nor their genetic interaction. Journal of Psychiatric Research, 42, 278–288. doi:10.1016/j. jpsychires.2007.02.005. Vionnet, N., Hani, E., Dupont, S., Gallina, S., Francke, S., Dotte, S., et al. (2000). Genomewide search for type 2 diabetes-susceptibility genes in French whites: Evidence for a novel susceptibility locus for early-onset diabetes on chromosome 3q27-qter and independent replication of a type 2-diabetes locus on chromosome 1q21-q24. American Journal of Human Genetics, 67, 1470–1480. Vladimirov, V., Thiselton, D. L., Kuo, P. H., McClay, J., Fanous, A., Wormley, B., et al. (2007). A region of 35 kb containing the trace amine associate receptor 6 (TAAR6) gene is associated with schizophrenia in the Irish study of high-density schizophrenia families. Molecular Psychiatry, 12, 842–853. Walsh, T., McClellan, J. M., McCarthy, S. E., Addington, A. M., Pierce, S. B., Cooper, G. M., et al. (2008). Rare structural variants disrupt multiple genes in neurodevelopmental pathways in schizophrenia. Science, 320, 539–543. Walss-Bass, C., Raventos, H., Montero, A. P., Armas, R., Dassori, A., Contreras, S., et al. (2006). Association analyses of the neuregulin 1 gene with schizophrenia and manic psychosis in a Hispanic population. Acta Psychiatrica Scandinavica, 113, 314–321. Wang, X., He, G., Gu, N., Yang, J., Tang, J., Chen, Q., et al. (2004). Association of G72/G30 with schizophrenia in the Chinese population. Biochemical and Biophysical Research Communications, 319, 1281–1286. Weickert, C. S., Straub, R. E., McClintock, B. W., Matsumoto, M., Hashimoto, R., Hyde, T. M., et  al. (2004). Human dysbindin (DTNBP1) gene expression in normal brain and in schizophrenic prefrontal cortex and midbrain. Archives of General Psychiatry, 61, 544–555. Williams, H. J., Owen, M. J., & O’Donovan, M. C. (2007). Is COMT a susceptibility gene for schizophrenia? Schizophrenia Bulletin, 33, 635–641. Williams, N. M., Preece, A., Morris, D. W., Spurlock, G., Bray, N. J., Stephens, M., et al. (2004). Identification in 2 independent samples of a novel schizophrenia risk haplotype of the dystrobrevin binding protein gene (DTNBP1). Archives of General Psychiatry, 61, 336–344. Williams, N. M., Preece, A., Spurlock, G., Norton, N., Williams, H. J., McCreadie, R. G., et al. (2004). Support for RGS4 as a susceptibility gene for schizophrenia. Biological Psychiatry, 55, 192–195. Williams, N. M., Preece, A., Spurlock, G., Norton, N., Williams, H. J., Zammit, S., et al. (2003). Support for genetic variation in neuregulin 1 and susceptibility to schizophrenia. Molecular Psychiatry, 8, 485–487. Wiltshire, S., Hattersley, A. T., Hitman, G. A., Walker, M., Levy, J. C., Sampson, M., et al. (2001). A genomewide scan for loci predisposing to type 2 diabetes in a U.K. population (the Diabetes UK Warren 2 Repository): Analysis of 573 pedigrees provides independent replication of a susceptibility locus on chromosome 1q. American Journal of Human Genetics, 69, 553–569. Wood, L. S., Pickering, E. H., & Dechairo, B. M. (2007). Significant support for DAO as a schizophrenia susceptibility locus: Examination of five genes putatively associated with schizophrenia. Biological Psychiatry, 61, 1195–1199. Xu, J., Pato, M. T., Torre, C. D., Medeiros, H., Carvalho, C., Basile, V. S., et al. (2001). Evidence for linkage disequilibrium between the alpha 7-nicotinic receptor gene (CHRNA7) locus and schizophrenia in Azorean families. American Journal of Medical Genetics, 105, 669–674. Xu, B., Roos, J. L., Levy, S., van Rensburg, E. J., Gogos, J. A., & Karayiorgou, M. (2008). Strong association of de novo copy number mutations with sporadic schizophrenia. Nature Genetics, 40, 880–885. Xu, M. Q., Xing, Q. H., Zheng, Y. L., Li, S., Gao, J. J., He, G., et al. (2007). Association of AKT1 gene polymorphisms with risk of schizophrenia and with response to antipsychotics in the Chinese population. The Journal of Clinical Psychiatry, 68, 1358–1367. Yang, J. Z., Si, T. M., Ruan, Y., Ling, Y. S., Han, Y. H., Wang, X. L., et al. (2003). Association study of neuregulin 1 gene with schizophrenia. Molecular Psychiatry, 8, 706–709.

380

B. Riley

Yue, W., Liu, Z., Kang, G., Yan, J., Tang, F., Ruan, Y., et  al. (2006). Association of G72/G30 polymorphisms with early-onset and male schizophrenia. NeuroReport, 17, 1899–1902. Zeggini, E., Weedon, M. N., Lindgren, C. M., Frayling, T. M., Elliott, K. S., Lango, H., et  al. (2007). Replication of genome-wide association signals in U.K. samples reveals risk loci for type 2 diabetes. Science, 316, 1336–1341. Zhang, F., Sarginson, J., Crombie, C., Walker, N., St Clair, D., & Shaw, D. (2006). Genetic association between schizophrenia and the DISC1 gene in the Scottish population. American Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 141, 155–159. Zhang, F., St Clair, D., Liu, X., Sun, X., Sham, P. C., Crombie, C., et al. (2005). Association analysis of the RGS4 gene in Han Chinese and Scottish populations with schizophrenia. Genes, Brain, and Behavior, 4, 444–448. Zhang, X., Tochigi, M., Ohashi, J., Maeda, K., Kato, T., Okazaki, Y., et al. (2005). Association study of the DISC1/TRAX locus with schizophrenia in a Japanese population. Schizophrenia Research, 79, 175–180. Zhao, X., Leotta, A., Kustanovich, V., Lajonchere, C., Geschwind, D. H., Law, K., et al. (2007). A unified genetic theory for sporadic and inherited autism. Proceedings of the National Academy of Sciences of the United States of America, 104, 12831–12836. Zhao, X., Shi, Y., Tang, J., Tang, R., Yu, L., Gu, N., et al. (2004). A case control and family based association study of the neuregulin1 gene and schizophrenia. Journal of Medical Genetics, 41, 31–34. Zou, F., Li, C., Duan, S., Zheng, Y., Gu, N., Feng, G., et al. (2005). A family-based study of the association between the G72/G30 genes and schizophrenia in the Chinese population. Schizophrenia Research, 73, 257–261.

Proteomics of the Anterior Cingulate Cortex in Schizophrenia Danielle Clark, Irina Dedova, and Izuru Matsumoto

Abstract  Proteomics provides the opportunity to visualise, quantify and ­hypothesise differences in protein expression, functions and interactions with the possibility of elucidating the molecular mechanisms underlying the ­neuropathology of schizophrenia. The anterior cingulate cortex (ACC, Brodmann Area 24) is implicated in the pathogenesis of schizophrenia due to its normal physiological functions and connectivity, and from reports of structural, morphological and neurotransmitter aberrations within this brain area in the disease state. The current proteomic analysis, following subtraction of proteins similar to the alcoholic PFC proteome, and the risperidone-treated rat dorsal striatum, ­identified 22 proteins specifically altered in the ACC grey matter (GM) ­proteome, and 19 aberrant proteins in the ACC white matter (WM) proteome in schizophrenia. Functional classification of these proteins suggests that normal cellular ­metabolism is deficient, with an increase in oxidative stress and the utilisation of other energy pathways in both the GM and WM regions of the ACC in the disease state. Furthermore, this study reiterates the well-documented hypothesis of aberrant neurotransmission within this region in schizophrenia, and outlines the importance of both the WM tracts as well as the GM in aberrant synaptic activity. This study therefore provides some focussed avenues in the analysis of the molecular mechanisms of the ACC in schizophrenia. Keywords  Schizophrenia • Anterior cingulate cortex • Human brain • Proteomics • Metabolism • Cytoskeleton • Synaptic transmission

I. Dedova (*) Schizophrenia Research Institute, 384 Victoria Street, Darlinghurst, NSW 2010, Australia and School of Biomedical and Health Sciences, University of Western Sydney, NSW 1797, Australia e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_14, © Springer Science+Business Media, LLC 2011

381

382

D. Clark et al.

1 Introduction Schizophrenia is a chronic, debilitating disorder, characterised by delusions, ­hallucinations, disorganised speech and behaviour and changes in emotion (American Psychiatric Association, 1994). This incurable psychiatric illness affects approximately 1% of the population and according to the World Health Organisation, schizophrenia is the worlds fourth leading cause for disability (WHO, 1994). Despite the wealth of research into the manifestations and neuropathology of schizophrenia, it remains a clinical diagnosis without a definitive cause. There is, however, evidence to suggest genetics, together with ­environmental risk factors, play a role in triggering schizophrenia (Harrison & Owen, 2003; Tsuang, 2000; Weinberger, 1995). The literature also shows that schizophrenia affects the normal functioning of multiple brain areas, such as the anterior cingulate cortex (ACC) (Carter et al., 2001; Cleghorn et al., 1990; Hempel et al., 2003; Quintana et al., 2004), the dorsolateral prefrontal cortex (DLPFC) (Callicott et al., 2000; MacDonald & Carter, 2003; Weinberger et al., 1986) and the hippocampus (Heckers, 2001). The ACC, a component of the limbic system is believed to have an integral role in mediating aspects of both cognition and emotional behaviours (Allman et  al., 2001; MacLeod & MacDonald, 2000). Indeed, various functions such as emotions, executive control and attention are reported to be abnormal in schizophrenia patients (Carter et al., 2001; Hempel et al., 2003; Salgado-Pineda et al., 2003). Furthermore, abnormal activation of the ACC during hallucinations (Cleghorn et al., 1990) and task performances (Carter et  al., 1997; Quintana et  al., 2004) has been shown in affected patients. In addition to these functional changes within the ACC in schizophrenia, structural and morphological pathologies have been found (Benes et  al., 2001; Bouras et al., 2001; Chana et al., 2003; Salgado-Pineda et al., 2003). A reduction in the ACC grey matter (GM) and white matter (WM) volume (Salgado-Pineda et al., 2003) reduced laminar thickness (Bouras et al., 2001), diminished number of pyramidal neurons in the deeper lamina (Benes et al., 2001; Chana et al., 2003) and smaller size of pyramidal neurons (Bouras et al., 2001; Chana et al., 2003) have all been observed in schizophrenia patients relative to healthy individuals. In addition, functional imaging studies of schizophrenia patients have revealed altered blood flow and/or glucose metabolism compared to healthy controls (Haznedar et  al., 1997, 2004; Liddle et al., 1992). The manifestations of schizophrenia are thought to be due to abnormal neural circuitry between and within brain regions (Benes, 2000). Within the ACC, there is depleted dopamine activity (Suhara et  al., 2002), an increased density of ­glutamatergic neurons (Benes et al., 1992) and a reduced number of GABA cells co-expressing N-methyl-D-aspartate (NMDA) glutamate receptors (Woo et  al., 2004). The GM and WM of the ACC are therefore strong candidates involved in the neurological mechanisms causing the manifestations of schizophrenia. There is little understanding of the molecular changes that accompany the ­cellular pathology in schizophrenia, due in part to limitations of existing study techniques and lack of knowledge regarding normal human brain physiology

Proteomics of the Anterior Cingulate Cortex in Schizophrenia

383

(Thompson et  al., 1998). Recent advances in proteomic methodology offers new possibilities to investigate the expression of multi-protein systems simultaneously and to determine protein profiles specific for different diseases. Proteomics has been successfully applied to post-mortem human brain to study the global changes in protein expression caused by neurological and psychiatric conditions such as schizophrenia, Down syndrome, Alzheimer’s disease and alcoholism. This chapter will analyse results of our investigations into the alterations in the proteomic profiles of the ACC in schizophrenia. The detection of key proteins changed in schizophrenia may lead to the identification of the specific molecular pathways underlying the neuropathology of this brain disorder. Two dimensional gel electrophoresis (2DE) and mass spectrometry (MS) offers a large-scale starting point for further analysis and studies to focus on those molecular pathways specific to the region in schizophrenia. Due to the differences in cell composition and ­physiological specialisations of the GM and WM, the proteomic methodology was applied to each matter separately. By comparing proteins altered in each of the GM and WM of the ACC, region-specific proteins may be elucidated. By comparing the altered proteins of these regions to proteins identified as altered in another ­neuropathological disease, alcoholism, the specificity of identified changes will be further verified as disease-specific changes. All the schizophrenia patients in this study underwent extensive neuroleptic treatment during their lives. Thus, the ­possible effects of pre-mortem medication will be considered by cross-correlating changes identified in this study with that of chronic neuroleptic-treated animal model. Proteins changed in both studies will be ruled out in the current discussion, to focus on proteins altered due to the disease itself. The resultant lists of proteins altered due to schizophrenia in the ACC GM and WM is the basis for hypotheses introduced in the discussion of this chapter. These hypotheses may then be further investigated using immunohistological, western blot, genomic and other proteomic techniques to ascertain the specific molecular changes within the ACC in the disease state. This proteomic analysis, therefore, provides a broad platform for the elucidation of the role and underlying mechanisms of the ACC in schizophrenia.

2 Cross-Correlation of Altered Proteins in Different Brain Region, Disease and Risperidone-Treated Rats 2.1 Human ACC Samples of Schizophrenia Subjects and 2DE-Based Proteomic Analysis Fresh frozen post-mortem human brain tissue from the ACC of ten non-psychiatric control individuals and ten schizophrenia patients was obtained from the NSW Tissue Resource Centre (University of Sydney, NSW, Australia). The summary of individual demographics is shown in Table  1. Sample preparation and proteomic  methodology were according to the method established in our laboratory

384

D. Clark et al.

Table 1  Patient demographics Variable Control (mean ± SD) Schizophrenia (mean ± SD) Age (years) 48.6 ± 9.68 48 ± 10.73 Gender (F:M) 2:8 2:8 PMI (h) 23.4 ± 10.37 25.85 ± 10.9 Brain hemisphere (R:L) 5:5 5:5 pH 6.45 ± 0.36 6.39 ± 0.25 Duration illness (years) – 24.166 Cpze (mg) – 130–1,800 PMI Post-mortem interval, Cpze an estimated range of a donor’s lifetime daily neuroleptic ­medication use calculated in milligrams of chlorpromazine equivalent units

­(Alexander-Kaufman et al., 2006, 2007; Clark et al., 2006; Clark et al., 2007). The ACC from schizophrenia and control individuals was macroscopically divided into GM and WM (ten samples in each group, run in duplicate) and separately analysed using 2D gel electrophoresis (Figs. 2 and 3). Proteins differing in abundance between ­schizophrenia and control groups underwent mass spectrometry (MALDI-TOF MS) for ­identification (Alexander-Kaufman et al., 2006, 2007; Clark et al., 2006, 2007).

2.2 Changes in the ACC GM and WM Proteomes in Schizophrenia Relative to Healthy Controls Applying 2DE to the ACC GM of a schizophrenia cohort revealed 42 proteins with altered levels relative to the ACC GM proteome of the healthy cohort; 39 of these altered proteins were subsequently identified using mass spectrometry (see Clark et al., 2006 for a detailed table of results). Comparison between the proteome of the schizophrenia ACC WM and that of healthy ACC WM found the volume/density of 32 proteins to be altered in the schizophrenia cohort; 30 of these were then identified using MALDI-TOF MS, representing 27 unique genes (Clark et al., 2007). All identified proteins were subject to post hoc statistical Pearsons correlation with pre- and post-mortem factors, e.g. age, post-mortem interval (PMI), and duration of illness (see Table  1). As noted above, all patients were medicated with a wide range of doses. This made it difficult to apply post hoc statistical analysis with neuroleptic dose. In the GM proteome, three protein spots correlated with one of these factors: the abundance of both dihydropyrimidinase-related protein 2 (DRP2) and dimethylarginine dimethylaminohydrolase 1 (DDAH-1) correlated with brain pH, and the expression of acetyl CoA acetyltransferase correlated with PMI of the samples. Following statistical analysis, the WM schizophrenia proteome also found DRP2 and succinate dehydrogenase to have a positive correlation to brain pH. In addition, two proteins, phosphoglycerate mutase and DJ protein 1, correlated to PMI, whilst inositol monophosphate had a negative correlation with duration of illness in the schizophrenia cohort.

Proteomics of the Anterior Cingulate Cortex in Schizophrenia

385

2.3 Cross-Correlation of Identified Proteins in the ACC WM and GM Proteome in Schizophrenia To ascertain region specificity, proteins identified as being altered in the schizophrenia ACC WM were compared to those altered in the ACC GM proteome (Fig. 1a) (Clark et al., 2007). There were four proteins identified in both GM and WM: heat shock 70-kDa protein 1 (HSP70), DRP2, citrate synthase and fructose bisphosphate aldolase C (FBA-C). FBA-C was the only protein to change in the same direction in both the ACC GM and WM of schizophrenia average gel

a

b

Schizophrenia GM

36

4

Schizophrenia WM

Schizophrenia ACC GM

27

36

c

Alcoholic BA9 GM 10

42

d Schizophrenia ACC WM

27

3

Alcoholic BA9 WM

Schizophrenia ACC GM

24

36

RIS-Striatum

4

23

e Schizophrenia ACC WM

27

RIS-Striatum

4

23

Fig. 1  Cross correlation of altered proteins of the ACC grey and white matter in schizophrenia. (a) ACC GM compared with ACC WM schizophrenia; (b) schizophrenia ACC GM compared with alcoholic BA9 GM; (c) Schizophrenia ACC WM altered proteins and alcoholic BA9 WM altered proteins. (d) Schizophrenia ACC GM altered proteins and risperidone-treated rat dorsal striatum and (e). Schizophrenia ACC WM compared to the risperidone-treated rat dorsal striatum. The number of altered proteins identified in each study relative to controls is shown within each respective circle. The number of altered proteins identified in both studies are noted above the arrows

386

D. Clark et al.

Fig. 2  Schizophrenia ACC grey matter specific proteome. Identification of proteins altered in the schizophrenia ACC GM relative to control cases, after correlations with confounding factors, alcoholic BA9 GM, and risperidone-treated rat dorsal striatum. Solid outline is increased ­expression, dashed is decreased abundance and dotted outline denotes proteins not visualised in the schizophrenia proteome (p < 0.05, ANOVA)

Fig. 3  Schizophrenia ACC white matter specific proteome. Identification of proteins altered in the schizophrenia ACC WM relative to control cases, after correlations with confounding factors, alcoholic BA9 WM, and risperidone-treated rat dorsal striatum. Solid outline is increased ­expression, dashed is decreased abundance and dotted outline denotes proteins not visualised in the schizophrenia proteome (p < 0.05, ANOVA)

Proteomics of the Anterior Cingulate Cortex in Schizophrenia

387

(Table 2). However, none of these four proteins correspond to the same spot on 2D gels and as such may be specific isoforms of these proteins differentially affected in each of the GM and WM of ACC in schizophrenia (Clark et al., 2006, 2007).

2.4 Cross-Correlation of Altered Proteins in the ACC in Schizophrenia and Prefrontal Cortex Brodmann Area 9 (PFC BA9) in Alcoholism Comparing the proteins altered in the GM and WM proteomes in schizophrenia with those altered in other psychiatric and/or neurological diseases can suggest non-specific changes, or false positives, in the analysis, thus delineating proteins which are particular to schizophrenia neuropathology. The above ACC data were contrasted with the results of a proteomic investigation of the GM and WM of the PFC (BA9) in alcoholism (Alexander-Kaufman et al., 2006, 2007). The abundance of individual proteins can differ depending on the brain region due to factors such as the physiological specialisation of each brain region, region-specific gene expression and post-translational modifications. Thus, when comparing the abundance of proteins of the ACC and PFC, some of the differences in the direction of observed changes may represent differences between these brain regions rather than having a specific pathological significance. However, these studies share a majority of control cases (7 out of 10), as well as the experimental and analytical methodologies, providing a solid basis for comparison of the results and limiting intra-individual variations within control samples. Used as a preliminary tool, this comparison also aids in narrowing the affected protein list in the ACC GM and WM schizophrenia proteins to begin downstream analyses. Ten proteins identified in both the schizophrenic GM and alcoholic GM ­proteome analysis had altered expression in both diseases when compared to ­control samples: prohibitin, DRP2, dynamin 1 brain, N-ethylmaleimide sensitive fusion protein (NSF), DDAH 1, EH-domain-containing protein 3, phosphatidylethanolamine-binding protein (PEBP), guanine nucleotide-binding protein ­polypeptide beta 1, FBA-C, and peptidyl-prolyl cis-trans isomerise A (Fig.  1b) (Alexander-Kaufman et  al., 2006; Clark et  al., 2006). Interestingly, DDAH-1 and PEBP differed in the direction of change between the two diseases (Table 2). PEBP was increased in schizophrenia and had reduced abundance in the alcohol study, whilst DDAH-1 was reduced in schizophrenia and increased in alcoholism. However, DDAH-1 was found to correlate with brain pH in the schizophrenia study. This may suggest different molecular processes occurring in the diseases or may represent changes in the two brain regions (BA24 and BA9), or may be ­considered an artefact. Schizophrenia ACC WM alterations compared to the alcoholic PFC WM ­alterations revealed three proteins overlapping (alpha-internexin, ubiquitin ­carboxy-terminal hydrolase L1, phosphoglycerate mutase 1; 1C) between the two  diseases. All three proteins differed in the direction of change (Table 2) ­(Alexander-Kaufman et al., 2007;

388 Table  2  Proteins similar to, but differing in direction studies Protein Sz GM Alcohol GM DRP 2 – � Citrate synthase ¯ – HSP 70 � – Alpha internexin – – Ubiquitin Phosphoglycerate mutase – – DDAH-1 ¯ � PEBP � ¯ ATP synthase – –

D. Clark et al. of change relative to control between Ris

Sz WM

Alc WM – – – ¯ ¯ � – – –

� ¯ – � – ¯ – � – � ¯ ¯ – – – – ¯ � Arrows up and down indicate protein changes in the diseases relative to control cases. This table includes proteins showing opposite direction of a change in schizophrenia compared to the alcohol and neuroleptic-treated model Ris Risperidone-treated rat striatum proteome, Alc alcoholic PFC proteome, Sz schizophrenia, GM grey matter, WM white matter

Clark et  al., 2007). Alpha internexin was similar to controls and reduced in the ­uncomplicated and hepatic complicated cohorts of the alcoholism proteomic analysis, respectively, while it showed increased abundance in the schizophrenia WM ­proteome. Ubiquitin followed a similar trend to alpha internexin, where it was found to be reduced in alcoholism relative to control cases, and had a greater abundance in the schizophrenia cases relative to controls (not visualised in the control average gel). Phosphoglycerate mutase 1, however, was increased in alcoholism and reduced in the schizophrenia proteome relative to control cases. As in the GM and WM correlations, these proteins did not correspond to the same location on the 2D gel, and as such may also represent differing post-translational changes occurring in schizophrenia and alcoholism. Other possible causes of these findings are the unique molecular ­mechanisms underlying the disease, or simply the differences in each brain region. Whilst these are different brain regions between the two diseases, this ­comparison allows us to focus on proteins which may be specifically altered in schizophrenia (disease-specific changes), thereby enabling further research on the resultant refined hypotheses.

2.5 Altered Proteins of the Schizophrenia ACC vs. Altered Proteins Due to Neuroleptics in Animal Brain As all the schizophrenia patients in our studies were taking antipsychotics it is ­difficult to separate the effect of the medication on protein abundance from the effect of the disease itself. Risperidone is an atypical antipsychotic agent ­commonly used to treat schizophrenia symptoms (O’Brien et al., 2006). Findings

Proteomics of the Anterior Cingulate Cortex in Schizophrenia

389

in the ACC in schizophrenia were therefore compared with changes in the rat brain proteome due to chronic subcutaneous administration of ­risperidone. In this study, Sprague-Dawley rats were treated with therapeutically relevant doses (2.1 mg/kg/day) of risperidone for 4 weeks using surgically implanted subcutaneous osmotic minipumps, which provided continuous controlled drug administration (Karl et al, 2006). Changes in the proteome of the dorsal ­striatum, a brain area especially affected by antipsychotics, were found when a risperidonetreated group of animals was compared with a non-treated control group (O’Brien et al., 2006). The comparison of altered proteins of the ­risperidone-treated rat striatum with those observed in the schizophrenia ACC may uncover proteins altered due to antipsychotic agent and not the ­disease itself. Four proteins were altered in both the schizophrenia ACC GM and risperidonetreated rat striatum and were all changed in the same direction (Fig. 1d). The proteins were identified as annexin A5, creatinine kinase ubiquitous mitochondrial 1, DRP2 and FBA-C (Clark et al., 2006; O’Brien et al., 2006). This suggests that the levels of these proteins detected in the schizophrenia ACC GM may be due to the effect of neuroleptic medication. The four proteins found to be differentially expressed in both the risperidone striatum proteome and the schizophrenia ACC WM proteome (Fig.  1e) are as ­follows: ATP synthase, DRP2, FBA-C and phosphoglycerate mutase 1 (Clark et al., 2007; O’Brien et al., 2006). DRP 2 and ATP synthase were found to be altered in different directions between the disease and treatment states (Table 2): DRP2 was increased and ATP synthase reduced in the risperidone cohort, whereas DRP2 was reduced and ATP synthase elevated in the schizophrenia WM relative to control cases. The drug effects observed in the animal model striatum can be subtracted from the ACC results of schizophrenia to eliminate the possible effects of medication in the human samples. These proteins identified in both the ­risperidone study and the schizophrenia proteomic study can therefore be ­considered the result of the drug treatment.

3 Schizophrenia ACC-Specific Protein Changes The following discussion and tables are focussed on the proteins found to be altered in the schizophrenia ACC relative to controls, following subtraction of those found to be changed in alcoholism, risperidone treatment and confounding factors and both the ACC GM (Fig. 2) and WM (Fig. 3). These proteins are therefore considered region- and disease- specific without the effects from pre- and/or post-mortem factors. Twenty-two proteins are therefore considered to be diseaseand region-specifically changed within the ACC GM in schizophrenia (Table 3), whilst 19 proteins appear to be distinctively altered in the schizophrenia ACC WM ­proteome (Table  4). These proteins have been functionally classified to aid the discussion of the molecular mechanisms which may be occurring in the ACC GM and WM in schizophrenia.

Table 3  Schizophrenia specific regionally altered proteins of the ACC grey matter Functional classification Spot no. Direction of change Protein identification Metabolic 2591 � Acetyl CoA acetyltransferase cytosolic 2333 ¯ Isocitrate dehydrogenase [NAD] subunit a   717 Absent Isocitrate dehydrogenase [NAD] subunit a 1187 ¯ NADH oxidoreductase (ubiquinone) 1 a subcomplex, 5, 13kDa   740 � Creatine kinase B chain   991 Absent Biliverdin reductase B Oxidative stress 2637 � Superoxide dismutase [Cu-Zn]   916 ¯ Hydroxyacylglutathione hyhdrolase (Glx II)   772 � Annexin A1 Cytoarchitectural   762 ¯ Tubulin a-6 chain 2179 � Stathmin 2026 � Dynactin subunit 2 2077 � Actin-like protein 3 2078 Absent Actin-like protein 3 Synaptic 2507 Absent Serine/threonine protein phosphatase PP1 a catalytic subunit   953 � Synaptosomal associated protein 25 (SNAP 25) Signalling 476 � Dihydropyrimidinase related protein 1 (DRP 1) 4p16.1-p15

17q21

11q13.2

12q13.12 1p36.11 12q13.3 2q14.1 2q14.1

21q22.11 16p13.3 9q21.13

14q32.32 19q13.2

6q25.3 15q25.1 15q25.1 7q32.1

Chromosomal loci

390 D. Clark et al.

2617   881 2573

2231

2013   342   924

Absent � �

�

� Absent �

Mitochondrial inner membrane protein Novel protein (RP11.4615.3) human Chromosome 2 open reading frame 32

ADP-ribosylation factor 1

Glial fibrillary acidic protein Serotransferrin precursor (transferrin) Carbonic anhydrase

Direction of change refers to the abundance of protein in the schizophrenia proteome relative to control cases

Unknown function

Trafficking

Glial specific

2p11.2 1p36.32 2p14

1q42.13

17q21.31 3q22.1 8q21.2

Proteomics of the Anterior Cingulate Cortex in Schizophrenia 391

Table 4  Schizophrenia specific regionally altered proteins of the ACC white matter Functional classification Spot no. Direction of change Protein identification Metabolic   804 ¯ Carbonyl reductase [NADPH] 1103 ¯ Glycerol-3-phosphate dehydrogenase [NAD+] fragment   248 � Glycogen phosphorylase brain form   751 ¯ L-lactate dehydrogenase chain B   268 � N-acetyl glutamate synthase   370 ¯ Succinate dehydrogenase [ubiquinone] flavoprotein subunit Oxidative Stress   979 � DJ-1 protein Cytoarchitectural   636 � Tubulin 5 b   818 � Tubulin specific chaperone B   705 ¯ NAD-dependant deacetylase sirtuin 2   707 ¯ NAD-dependant deacetylase sirtuin 2   377 ¯ Neurofilament triple L protein Synaptic 2654 � Choline acetyltransferase fragment 1129 ¯ Syntaxin 1C   496 ¯ Aldehyde dehydrogenase family member 7 A1   836 � Vacuolar ATPase Signalling   883 � Inositol monophosphate 8q21.13

8p21.3

10q11.23 7q11.23 5q23.2

19q13.3 19q13.12 19q13.2 19q13.2 8p21.2

1p36.23

20p11.2 12p12.1 17q21.31 12q13.12

21q22.12 12q13.12

Chromosomal loci

392 D. Clark et al.

  938

  244   323 �

� Absent Proteasome subunit b type

Endoplasmin precursor Endoplasmin precursor

  638 Absent Adenosylhomocysteinase Direction of change refers to the abundance of protein in the schizophrenia proteome relative to control cases

Unknown function

Proteolysis

Trafficking

20q11.22

1q21.3

12q23.1 12q23.1

Proteomics of the Anterior Cingulate Cortex in Schizophrenia 393

394

D. Clark et al.

Interestingly, the majority of altered proteins specific to both the ACC GM and WM in schizophrenia were classified as metabolic and oxidative stress proteins (nine and seven for GM and WM, respectively), followed by cytoarchitectural (five for each of the GM and WM). Other classifications identified in the ACC GM in schizophrenia were synaptic (n = 2), glial specific (n = 3), signalling (n = 1) and trafficking (n = 1) and three proteins with unknown function. As in the schizophrenia GM, other functions of the altered proteins found in the schizophrenia WM were four synaptic proteins, two trafficking, one signalling, one proteolytic and one protein with unknown function. Based on these classifications, the underlying mechanisms of schizophrenia seem to be related to, or caused by, changes in the cytoarchitecture and metabolism of the ACC GM and WM. Functional imaging studies have revealed altered blood flow and/or glucose metabolism in the ACC in schizophrenia (Haznedar et al., 1997, 2004; Liddle et al., 1992). In both the GM and WM of the ACC, there may be a reduction in glucose metabolism with a compensatory up-regulation of proteins involved in alternative metabolic pathways and the consequent production of oxidative by-products and increased molecular response to oxidative stress. This is indicated by the reduced abundance in the schizophrenia proteome of isocitrate dehydrogenase (GM), NADH oxidoreductase (GM), and succinate dehydrogenase (WM), proteins all involved in cellular respiration. In addition, proteins involved in alternate energy pathways were increased, such as acetyl CoA acetyl transferase (GM) which is involved in the metabolism of ketones in times of oxidative stress. However, the indication of alternative energy pathways seems most apparent in the WM of the ACC, where cytosolic glycerol-3 phosphate dehydrogenase (GPDH-C), a protein which regenerates the supply of NAD+ for continued glycolysis (Hwang et al., 1999), and lactate dehydrogenase, which catalyses the conversion of pyruvate to lactate in times of energy need, were reduced in the schizophrenia ACC WM ­proteome. Furthermore, glycogen phosphorylase, involved in the production of lactate from glycogen under anaerobic conditions (Brown et al., 2005), is increased in the schizophrenia cohort. These findings suggest that, in the WM of the ACC in schizophrenia, the rate of normal metabolism is reduced, possibly leading to a ­subsequent reduction in GPDH-C, thereby reducing the amount of available ­metabolic substrate. This may result in increased production of lactate from ­glycogen, facilitated by glycogen phosphorylase. This study also suggests that, in the WM, lactate may be increased in schizophrenia, as a protein spot corresponding to lactate dehydrogenase is reduced in the disease group, a finding consistent with the literature (Altar et al., 2005). Perhaps consequently, there were proteins found differing in abundance in the ACC in schizophrenia involved with oxidative stress (in the GM: superoxide ­dismutase (CuZn) (SOD), annexin A1 and hydroxyacylglutathione hydrolase (Glx II) and in the WM DJ 1 protein). It is thought that the primary causative agents of oxidative stress in the brain are glutamate activation of ionotropic receptors (Tsai et al., 1998) and dopamine metabolism (Rabinovic et al., 2000). The effect of enhanced oxidative activity is varied, including inactivating enzymes, inhibiting presynaptic glutamate uptake and causing irreversible modifications to proteins (Tsai et  al., 1998). Some previous studies have questioned the effects of medication on the activity of oxyradicals and the cellular

Proteomics of the Anterior Cingulate Cortex in Schizophrenia

395

defence mechanisms against these products (Zhang et  al., 2003); however, these proteins involved in oxidative stress found in the schizophrenia ACC studies did not correlate with proteins found to be altered in the antipsychotic agent-treated rats, possibly suggesting that oxidative stress effects are due to the disease process itself. Somewhat surprisingly, this study revealed a higher proportion of synaptic ­proteins to be affected in the WM of the ACC, relative to the GM proteome in schizophrenia. One reason for this may be the limitations of the current proteomic technique to analyse only the soluble fraction of proteins from all possible proteins of each region. Despite this limitation, there were two protein products identified as altered in the ACC GM and four from the ACC WM in schizophrenia, and which have a role in synaptic transmission, adding further support for the well-documented hypothesis of neurotransmitter system aberrations involving the ACC (Benes, 1999; Benes et al., 2001; Okubo et al., 1997). Synaptosomal-associated protein 25 (SNAP-25), involved in aspects of neurotransmitter release from the presynaptic neurons, has previously been implicated in this hypothesis (Thompson et al., 1998, 2003) and is found to be increased in the ACC GM proteome of the schizophrenia cohort. Serine/threonine protein phosphatase (PP1) is located in both neurons and glia and regulates neurotransmission. The finding of a reduction of PP1 alpha subunit in this study, strengthens previous findings of altered pathways involving PP1, such as DARPP-32 (Svenningsson et al., 2003), and yatio (Wang et al., 1994; Westphal et al., 1999). DARPP-32 is phosphorylated by the activation of dopamine, serotonin and glutamate receptors, whilst yatio is an NMDA receptor-associated protein. The phosphorylation of DARPP-32, caused by the activity of neurotransmitter ­receptors, regulates PP1 (Svenningsson et  al., 2003). If, as this study suggests, PP1 has a reduced activity, it may be caused by the upstream alterations of neurotransmitters binding to DARPP-32. A downstream effect of a reduction of PP1 is on the interaction of PP1 with yatio, which acts to limit the current through NMDA glutamate channels (Westphal et al., 1999). Therefore, decreased binding of PP1 with yatio would result in enhanced currents through the NMDA receptor channels (Wang et al., 1994), indicating increased glutamate activity. This hypothesis is consistent with studies which have postulated increased glutamate afferent modulation on GABA cell population within ACC (Benes, 2000), which may then cause ­consequent alterations in areas receiving projections from the ACC. Aldehyde dehydrogenase, reduced in the ACC WM, is involved in many ­functions including the degradation of dopamine (Maring et  al., 1985), further implicating a role of dopamine in the neuropathology of schizophrenia. In addition, a reduction in vacuolar ATPase has been correlated with glutamate uptake ­inhibition in bovine brain synaptic vesicles (Wang & Floor, 1998). In the current study, this enzyme has an increased abundance in the schizophrenia ACC WM proteome, possibly indicating an enhanced rate of glutamate activity (Benes, 1999), which is also consistent with our findings above. Within the ACC WM proteome in schizophrenia, another transmitter system, the cholinergic system, is further implicated in the disease process (Holt et al., 1999; Karson et al., 1996), as the enzyme choline acetyltransferase (CHAT) was found to be increased in the disease cohort. In the literature, this enzyme has been found to

396

D. Clark et al.

be reduced in neuroleptic naïve schizophrenia patients (Holt et al., 1999), and it has been suggested that the enhanced level of this protein may be a medication effect. In this preliminary analysis, there was no similarity found in the risperidone study; however, further investigations are warranted to determine the effect of all other antipsychotic agents on protein levels and function in the ACC. Taken together, the changes in proteins involved in synaptic transmission, may be the key to the changes in circuitry within the ACC in schizophrenia.

4 Conclusions Proteomics provides the opportunity to visualise, quantify and hypothesise protein structure, function and interactions with the possibility of elucidating the molecular mechanisms underlying the neuropathology of schizophrenia. The current proteomic analysis, following subtraction of proteins similar to the alcoholic PFC proteome, and the risperidone-treated rat dorsal striatum, identified 22 proteins specifically altered in the ACC GM proteome, and 19 aberrant proteins in the ACC WM proteome in schizophrenia. Functional classification of these proteins suggests that normal cellular metabolism is deficient, with an increase in oxidative stress and the utilisation of other energy pathways in both the GM and WM regions of the ACC in the disease state. Furthermore, this study reiterates the well-documented hypothesis of aberrant neurotransmission within this region in schizophrenia, and outlines the importance of both the WM tracts as well as the GM in aberrant ­synaptic activity. This study therefore provides some focussed avenues in the analysis of the molecular mechanisms of the ACC in schizophrenia. Acknowledgements  This work was supported by grants from the NSW Government BioFirst Award and the Cecilia Kilkeary Foundation provided to I.M. D.C. was awarded a summer scholarship from NISAD. Tissue was received from the NSW Tissue Resource Centre, which is supported by the University of Sydney, Schizophrenia Research Institute, National Institute of Alcohol Abuse and Alcoholism and NSW Department of Health and NHMRC. We would like to thank Kimberly Alexander-Kaufman for her advice, guidance and technical assistance and Dr. Arifumi Hasegawa for his technical assistance.

References Alexander-Kaufman, K., Cordwell, S., et  al. (2007). A proteome analysis of the dorsolateral ­prefrontal cortex in human alcholic patients. Proteomics Clinical Applications, 1, 62–72. Alexander-Kaufman, K., James, G., et al. (2006). Differential protein expression in the prefrontal white matter of human alcoholics: a proteomics study. Molecular Psychiatry, 11(1), 56–65. Allman, J. M., Hakeem, A., et al. (2001). The anterior cingulate cortex. The evolution of an interface between emotion and cognition. Annals of the New York Academy of Sciences, 935, 107–117. Altar, C. A., Jurata, L. W., et al. (2005). Deficient hippocampal neuron expression of proteasome, ubiquitin, and mitochondrial genes in multiple schizophrenia cohorts. Biological Psychiatry, 58(2), 85–96.

Proteomics of the Anterior Cingulate Cortex in Schizophrenia

397

American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders. Washington DC Benes, F. M. (1999). Alterations of neural circuitry within layer II of anterior cingulate cortex in schizophrenia. Journal of Psychiatric Research, 33(6), 511–512. Benes, F. M. (2000). Emerging principles of altered neural circuitry in schizophrenia. Brain Research. Brain Research Reviews, 31(2–3), 251–269. Benes, F. M., Vincent, S. L., et al. (1992). Increased GABAA receptor binding in superficial layers of cingulate cortex in schizophrenics. The Journal of Neuroscience, 12(3), 924–929. Benes, F. M., Vincent, S. L., et al. (2001). The density of pyramidal and nonpyramidal neurons in anterior cingulate cortex of schizophrenic and bipolar subjects. Biological Psychiatry, 50(6), 395–406. Bouras, C., Kovari, E., et  al. (2001). Anterior cingulate cortex pathology in schizophrenia and bipolar disorder. Acta Neuropathologica (Berlin), 102(4), 373–379. Brown, A. M., Sickmann, H. M., et  al. (2005). Astrocyte glycogen metabolism is required for neural activity during aglycemia or intense stimulation in mouse white matter. Journal of Neuroscience Research, 79(1–2), 74–80. Callicott, J. H., Bertolino, A., et al. (2000). Physiological dysfunction of the dorsolateral ­prefrontal cortex in schizophrenia revisited. Cerebral Cortex, 10(11), 1078–1092. Carter, C. S., MacDonald, A. W., 3rd, et al. (2001). Anterior cingulate cortex activity and impaired self-monitoring of performance in patients with schizophrenia: an event-related fMRI study. The American Journal of Psychiatry, 158(9), 1423–1428. Carter, C. S., Mintun, M., et  al. (1997). Anterior cingulate gyrus dysfunction and selective ­attention deficits in schizophrenia: [15O]H2O PET study during single-trial Stroop task ­performance. The American Journal of Psychiatry, 154(12), 1670–1675. Chana, G., Landau, S., et  al. (2003). Two-dimensional assessment of cytoarchitecture in the ­anterior cingulate cortex in major depressive disorder, bipolar disorder, and schizophrenia: evidence for decreased neuronal somal size and increased neuronal density. Biological Psychiatry, 53(12), 1086–1098. Clark, D., Dedova, I., et  al. (2006). A proteome analysis of the anterior cingulate cortex gray ­matter in schizophrenia. Molecular Psychiatry, 11(5), 459–470, 423. Clark, D., Dedova, I., et al. (2007). Altered proteins of the anterior cingulate cortex white matter proteome in schizophrenia. Proteomics Clinical Applications, 1, 157–166. Cleghorn, J. M., Garnett, E. S., et al. (1990). Regional brain metabolism during auditory hallucinations in chronic schizophrenia. The British Journal of Psychiatry, 157, 562–570. Harrison, P. J., & Owen, M. J. (2003). Genes for schizophrenia? Recent findings and their pathophysiological implications. Lancet, 361(9355), 417–419. Haznedar, M. M., Buchsbaum, M. S., et al. (2004). Cingulate gyrus volume and metabolism in the schizophrenia spectrum. Schizophrenia Research, 71(2–3), 249–262. Haznedar, M. M., Buchsbaum, M. S., et al. (1997). Decreased anterior cingulate gyrus metabolic rate in schizophrenia. The American Journal of Psychiatry, 154(5), 682–684. Heckers, S. (2001). Neuroimaging studies of the hippocampus in schizophrenia. Hippocampus, 11(5), 520–528. Hempel, A., Hempel, E., et  al. (2003). Impairment in basal limbic function in schizophrenia ­during affect recognition. Psychiatry Research, 122(2), 115–124. Holt, D. J., Herman, M. M., et al. (1999). Evidence for a deficit in cholinergic interneurons in the striatum in schizophrenia. Neuroscience, 94(1), 21–31. Hwang, K., Jeong, D. W., et  al. (1999). Alteration of the NAD+/NADH ratio in CHO cells by stable transfection with human cytosolic glycerol-3-phosphate dehydrogenase: resistance to oxidative stress. Molecules and Cells, 9(4), 429–435. Karl, T., Duffy, L., et  al. (2006). Behavioural effects of chronic haloperidol and risperidon ­treatments in rats. Behavioural Brain Research, 171, 286–294. Karson, C. N., Mrak, R. E., et al. (1996). Decreased mesopontine choline acetyltransferase levels in schizophrenia. Correlations with cognitive functions. Molecular and Chemical Neuropathology, 29(2–3), 181–191.

398

D. Clark et al.

Liddle, P. F., Friston, K. J., et al. (1992). Cerebral blood flow and mental processes in schizophrenia. Journal of the Royal Society of Medicine, 85(4), 224–227. MacDonald, A. W., 3rd, & Carter, C. S. (2003). Event-related FMRI study of context processing in dorsolateral prefrontal cortex of patients with schizophrenia. Journal of Abnormal Psychology, 112(4), 689–697. MacLeod, C. M., & MacDonald, P. A. (2000). Interdimensional interference in the Stroop effect: uncovering the cognitive and neural anatomy of attention. Trends in Cognitive Sciences, 4(10), 383–391. Maring, J. A., Deitrich, R. A., et  al. (1985). Partial purification and properties of human brain aldehyde dehydrogenases. Journal of Neurochemistry, 45(6), 1903–1910. O’Brien, E., Dedova, I., et  al. (2006). Effects of chronic risperidone treatment on the striatal ­protein profiles in rats. Brain Research, 1113(1), 24–32. Okubo, Y., Suhara, T., et al. (1997). Decreased prefrontal dopamine D1 receptors in schizophrenia revealed by PET. Nature, 385(6617), 634–636. Quintana, J., Wong, T., et al. (2004). Anterior cingulate dysfunction during choice anticipation in schizophrenia. Psychiatry Research, 132(2), 117–130. Rabinovic, A. D., Lewis, D. A., et al. (2000). Role of oxidative changes in the degeneration of dopamine terminals after injection of neurotoxic levels of dopamine. Neuroscience, 101(1), 67–76. Salgado-Pineda, P., Baeza, I., et  al. (2003). Sustained attention impairment correlates to gray ­matter decreases in first episode neuroleptic-naive schizophrenic patients. Neuroimage, 19(2 Pt 1), 365–375. Suhara, T., Okubo, Y., et  al. (2002). Decreased dopamine D2 receptor binding in the anterior cingulate cortex in schizophrenia. Archives of General Psychiatry, 59(1), 25–30. Svenningsson, P., Tzavara, E. T., et al. (2003). Diverse psychotomimetics act through a common signaling pathway. Science, 302(5649), 1412–1415. Thompson, P. M., Egbufoama, S., et al. (2003). SNAP-25 reduction in the hippocampus of patients with schizophrenia. Progress in Neuropsychopharmacology & Biological Psychiatry, 27(3), 411–417. Thompson, P. M., Sower, A. C., et  al. (1998). Altered levels of the synaptosomal associated ­protein SNAP-25 in schizophrenia. Biological Psychiatry, 43(4), 239–243. Tsai, G., Goff, D. C., et  al. (1998). Markers of glutamatergic neurotransmission and oxidative stress associated with tardive dyskinesia. The American Journal of Psychiatry, 155(9), 1207–1213. Tsuang, M. (2000). Schizophrenia: genes and environment. Biological Psychiatry, 47(3), 210–220. Wang, L. Y., Orser, B. A., et al. (1994). Regulation of NMDA receptors in cultured hippocampal neurons by protein phosphatases 1 and 2A. Nature, 369(6477), 230–232. Wang, Y., & Floor, E. (1998). Hydrogen peroxide inhibits the vacuolar H+-ATPase in brain ­synaptic vesicles at micromolar concentrations. Journal of Neurochemistry, 70(2), 646–652. Weinberger, D. R. (1995). From neuropathology to neurodevelopment. Lancet, 346(8974), 552–557. Weinberger, D. R., Berman, K. F., et al. (1986). Physiologic dysfunction of dorsolateral prefrontal cortex in schizophrenia. I. Regional cerebral blood flow evidence. Archives of General Psychiatry, 43(2), 114–124. Westphal, R. S., Tavalin, S. J., et  al. (1999). Regulation of NMDA receptors by an associated phosphatase-kinase signaling complex. Science, 285(5424), 93–96. WHO. (1994). International classification of diseases. Geneva: World Health Organisation. Woo, T. U., Walsh, J. P., et al. (2004). Density of glutamic acid decarboxylase 67 messenger RNAcontaining neurons that express the N-methyl-D-aspartate receptor subunit NR2A in the ­anterior cingulate cortex in schizophrenia and bipolar disorder. Archives of General Psychiatry, 617, 649–657. Zhang, X. Y., Zhou, D. F., et  al. (2003). The effect of risperidone treatment on superoxide ­dismutase in schizophrenia. Journal of Clinical Psychopharmacology, 23(2), 128–131.

Proteome Effects of Antidepressant Medications Lucia Carboni, Chiara Piubelli, and Enrico Domenici

Abstract  In this review, we give an overview of the proteomic investigations aimed at elucidating the impact of antidepressants treatment on brain at the molecular level. We discuss the animal models in which currently available treatments of mood disorders have been investigated by proteomic approaches, with a critical analysis of the main studies and a comparison of the methodologies used. The results are reviewed in the light of the putative mechanism of action of antidepressant therapies, and overall suggest an activation of long-term events downstream of the neurotransmitter signalling. Keywords  2D electrophoresis • Antidepressive agents • Clorgyline • CRF1 antagonists • Desipramine • Differential in-gel electrophoresis • Escitalopram • Exercise • Fluoxetine • NK1 antagonists • Paroxetine • Proteomics • Sleep deprivation • Venlafaxine Abbreviations 2D 5HT CHAPS CRF Cy DIGE

Two-dimensional Serotonin 3-[(3-Cholamido propyl) dimethyl ammonio]-1-propanesulphonate Corticotropin releasing factor Cyanine Differential in-gel electrophoresis

L. Carboni (*) Neurosciences Centre of Excellence for Drug Discovery, GlaxoSmithKline, 37135, Verona, Italy e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_15, © Springer Science+Business Media, LLC 2011

399

400

DRP-2 ESC FRL FSL GAP43 GENDEP GFAP GO IEF iMAO IPA IPG MALDI-TOF MS MS/MS NE NK1 NOR pI PLS SAM SNAP SSRI TBP TCA WB

L. Carboni et al.

Dihydropyrimidinase-related protein 2 Escitalopram Flinders Resistant Line Flinders Sensitive Line Growth-associated protein 43 or neuromodulin Genome-based therapeutic drugs for depression Glial fibrillary acidic protein Gene ontology Isoelectric focusing Monoamine oxidase inhibitors Ingenuity Pathway Analysis Immobilised pH gradients Matrix-assisted laser desorption/ionisation – time of flight Maternal separation Tandem mass spectrometry Norepinephrine Neurokinine 1 receptor Nortriptyline Isoelectric point Partial least squares Significance Analysis of Microarrays Soluble NSF attachment protein Selective serotonin re-uptake inhibitors Tributylphosphine Tricyclic antidepressants Western blotting

1 Introduction Patients suffering from major depressive disorder experience episodes characterised by depressed mood and loss of interest or pleasure in nearly all activities. Additional symptoms, including changes in appetite, sleep, psychomotor activity, feelings of worthlessness or guilt, difficulty concentrating and suicidal ­ideations are also associated with the disorder (American Psychiatric Association, 2000). The diagnosis for major depressive disorder is based on criteria that evaluate the presence of the above-mentioned symptoms by an interview with the patient, since no laboratory findings diagnostic of a major depressive episode have been identified (American Psychiatric Association, 2000). The reported prevalence estimates for major depressive disorder in the United States population are 16.2% for lifetime, with mean episode duration of 16 weeks (Kessler et  al., 2003). Major depression is a heterogeneous disease with no established mechanism; a number of different hypothesis have been formulated to understand its pathophysiological bases (Belmaker & Agam, 2008; Manji, Drevets, & Charney, 2001; Nestler et al., 2002, Wong & Licinio, 2001).

Proteome Effects of Antidepressant Medications

401

2 Antidepressant Treatments 2.1 Pharmacotherapy Several effective and specific treatments are available for the therapy of mood ­disorders (Baldessarini, 2005; Holtzheimer & Nemeroff, 2006; Spigset & Mårtensson, 1999; Wong & Licinio, 2004). Nearly all pharmacological classes of antidepressant medications share the common mechanism of interfering with monoaminergic neurotransmission. The first discovered class of antidepressive agents consisted of monoamine oxidase inhibitors (MAOI) (Table 1). These enzymes mediate the inactivation by degradation of aminergic neurotransmitters, such as norepinephrine (NE), serotonin (5HT) and dopamine. MAO inhibition produces the immediate effect of increasing monoamine levels in the synaptic cleft. The second class of antidepressant compounds to be discovered were named tricyclic antidepressants (TCA), making reference to their chemical structure (Table 1). TCAs share with MAOIs the ability to increase the synaptic concentration of amine neurotransmitters, although with a different molecular mechanism. TCAs bind to the neurotransmitter transporters placed in the pre-synaptic membrane and inhibit the neurotransmitter re-uptake into the presynaptic terminal, thus prolonging its activity. In addition to the inhibition of 5HT and NE, members of this class also block histamine and muscarinic receptors, thus bringing about a number of unwanted effects. The more common adverse effects of TCAs include dry mouth and a sour or metallic taste, epigastric distress, constipation, dizziness, tachycardia, palpitations, postural hypotension, blurred vision, urinary retention, weight gain, weakness and fatigue. Although TCA are very efficacious, the emergence of frequent and serious side effects hampered the use of these medications. The attempt to overcome these shortcomings led to the discovery of a newer generation of antidepressants which selectively inhibited the 5HT transporters and were thus dubbed selective serotonin re-uptake inhibitors (SSRI) (Table 1). In addition to MAOI, TCA and SSRI, a number of other efficacious antidepressants that do not fall specifically into these classes are also reported in Table 1. The blockade of the transport-mediated removal of amine neurotransmitters is considered a crucial initiating event that induces a series of important secondary adaptations. Beyond their actions as reuptake inhibitors, the molecular changes induced by chronic antidepressant treatment remain incompletely understood. Nevertheless, a cascade of adaptive processes triggered by repeated administration of antidepressive agents are probably involved in mediating their therapeutic response (D’Sa & Duman, 2002; Malberg & Blendy, 2005; Schloss & Henn, 2004).

2.2 Other Antidepressant Treatments The treatment options for major depression include other alternatives in addition to pharmacotherapy. Different kinds of psychotherapies, such as cognitive, interpersonal and behaviour therapies, alone or in combination with pharmacotherapy, can

Pirlindole Trazodone

Nortriptyline

Protriptyline Trimipramine

Tranylcypromine

Fluoxetine Fluvoxamine Paroxetine Sertraline

Milnacipran Reboxetine Venlafaxine Viloxazine

Clomipramine Desipramine Doxepin Imipramine Maprotiline

Isocarboxazid Moclobemide Nialamide Phenelzine Selegiline

Bupropion Minaprine Mirtazapine Nefazodone Oxitriptan

Atomoxetine

Duloxetine

Escitalopram

Amoxapine

Iproniazid

Other Agomelatine

SNRI Desvenlafaxine

Table 1  Treatments for mood disorders therapy Antidepressants iMAO TCA SSRI Iproclozide Amitriptyline Citalopram

Other treatments Deep brain stimulation Electroconvulsive treatment Exercise Light therapy Psychotherapy Sleep deprivation Transcranial magnetic stimulation Vagus nerve stimulation

402 L. Carboni et al.

Proteome Effects of Antidepressant Medications

403

be effective for treating major depression (Beck, 2005). Other non-pharmacological approaches include physical treatments reported in Table  1 and discussed in the following references (Benedetti, Barbini, Colombo, & Smeraldi, 2007; Giedke & Schwärzler, 2002; Lawlor & Hopker, 2001; Padberg & Moller, 2003; Park, Goldman, Carpenter, Price, & Friehs, 2007; UK ECT Review Group, 2003).

3 Proteomic Approaches for Aiding the Discovery of Novel Antidepressants Although various treatment options are available, treatment-resistant depression is common: depending upon the definition used, prevalence is estimated to be around 19–34% (Keller, 2005). Moreover, even though newer antidepressants show a lower incidence of side effects, they nonetheless may be the cause for reduced compliance. It would be thus desirable if the next generation of antidepressants produced fewer side effects. In addition, although existing therapies can control the symptoms and remission in depression can be achieved, it is debatable whether they can effectively cure the disease. Thus, the need to discover new targets for therapeutic intervention is still significant (Nemeroff & Owens, 2002), and a proteomic approach can be adopted as a hypothesis-free method for the identification of pathways that are activated by antidepressant treatments which could represent a source of novel targets. In addition, the clinical development of antidepressant drugs still relies on the quantification of disease severity through symptom-based scales which are inevitably influenced by subjective factors and lack of a proper definition. Due to the lack of objective criteria to predict and monitor drug response based on biological markers, clinical trials in psychiatry are usually long-term investigations, involving large number of patients across multiple sites in the attempt to obtain sufficient statistical power and to compensate for the relatively low sensitivity of the psychopathologic rating scales. The investigation of the impact of antidepressants at the proteome level, both in the brain and in peripheral tissues, might help to elucidate their mode of action at the molecular level and in parallel to uncover potential markers for antidepressant efficacy. One of the major problems in a global analysis approach of human samples is the high variability across individuals, especially when it is not possible to monitor patient disease and treatment history (Miklos & Maleszka, 2001). The use of animal models can be advantageous since more standardised protocols can be applied, but it has some limitations due to the fact that animal models can only mimic some aspects of human psychiatric disease. The study of antidepressants in animal models can, however, contribute to the understanding of the protein networks involved in their mechanism of action. Proteins whose expression is significantly modulated by antidepressant treatment can be highlighted by proteomic analysis, and then selectively tested for their peripheral modulation in animals and in humans to assess their potential clinical utility.

404

L. Carboni et al.

4 Proteomic Analyses After Antidepressant Treatments 4.1 Antidepressive Agents 4.1.1 Proteomics in the Early Days An overview of the published studies involving proteomic analyses after antidepressant treatment is shown in Table  2. The studies reported before 2000 are rather rare, reflecting the inherent complexity of proteomic techniques and the relatively low interest by the scientific community in understanding the effect of antidepressants, and in general psychotropic drugs, beyond their primary impact on monoaminergic tone. One of the first studies aiming at identifying proteomic changes after antidepressant treatment in rat brain was carried out back in the 1980s (Heydorn, Creed, & Jacobowitz, 1984; study no. 1 in Tables  2 and 3). In this study, rats were treated acutely or chronically with 10 mg/kg desipramine or reserpine with the objective of determining whether proteins previously unknown to be associated with NE ­neurotransmission could be detected. A statistical analysis was performed to compare the optical densities of the proteins revealed in the 2D maps. In both parietal cortex and hippocampus, three proteins were found to be significantly modified after chronic but not acute desipramine treatment. Chronic reserpine treatment exerted a modification in the concentration of the same three proteins but in the opposite direction. The molecular mass and isoelectric point (pI) of the three proteins were determined, although the identification was not possible. Two of these proteins have approximately the same molecular mass, but a slightly different pI, suggesting that they may be two post-translational modifications of the same protein. Later, the same group studied proteome modifications in rat hippocampus and parietal cortex induced by chronic administration of clorgyline, an antidepressant belonging to a different class, i.e. a MAOI (Sills, Heydorn, Cohen, Creed, & Jacobowitz, 1986; study no. 2 in Tables 2 and 3). The concentration of one protein was augmented in the hippocampus, while the levels of two proteins were increased and three decreased in the parietal cortex. One of the two proteins decreased by the treatment was identified as neuron-specific enolase (Table 3). Proteins affected by antidepressant treatments according to proteomic studies are listed in Table 3; the most relevant biological processes in which the modulated proteins exert a function are also described. Although clorgyline treatment induced changes in the concentration of hippocampal and parietal cortex proteins, these modifications did not overlap with those induced by desipramine treatment. The different patterns of proteomic changes induced by the two different antidepressants suggest that any common changes related to their antidepressant action may be masked by mechanism-specific changes. It is also possible that the different proteomic changes observed are still related to common upstream or downstream antidepressant mechanisms (Sills et  al., 1986). Although preliminary in their nature, the above studies hint at the potential offered by proteomic approaches to investigate the impact of antidepressants at the intracellular level.

Table 2  Published studies involving proteomic analyses after antidepressant treatment No Study Species Brain region Antidepressant Dose, route 1 Heydorn et al. Male SpragueParietal cortex, Desipramine 10 mg/kg (1984) Dawley rat hippocampus twice/day, i.p. 2 Sills et al. (1986) Male SpragueParietal cortex, Clorgyline 1 mg/kg, pellet Dawley rat hippocampus Male SpragueHippocampus Fluoxetine 10 mg/kg, i.p. 3 Khawaja, Xu, Dawley rat Liang, & Barrett (2004) 4 Khawaja et al. Male SpragueHippocampus Venlafaxine 10 mg/kg, i.p. (2004) Dawley rat 5 Guest et al. Male Guinea Cerebral cortex Fluoxetine 10 mg/kg, s.c. (2004) pigsa 6 Guest et al. Male Guinea Cerebral cortex L-000760735 3 mg/kg, s.c. (2004) pigs 7 Carboni, Vighini, Male SpragueHippocampus, pre- Fluoxetine 5 mg/kg, p.o. et al. (2006) Dawley rat frontal cortex 8 Carboni, Vighini, Male SpragueHippocampus, pre- GR205171 5 mg/kg twice/ et al. (2006) Dawley rat frontal cortex day, p.o. 9 Carboni, Vighini, Male SpragueHippocampus, pre- DMP696 5 mg/kg, p.o. et al. (2006) Dawley rat frontal cortex 10 Bisgaard et al. Male Wistar Ventral Escitalopram 5 mg/kg i.p. (2007) rat hippocampus 11 Bisgaard et al. Male Wistar Ventral Vehicle (2007) rat hippocampus

2D PAGE, fingerprinting DIGE, fingerprinting DIGE, fingerprinting 2D PAGE, fingerprinting 2D PAGE, fingerprinting 2D PAGE, fingerprinting DIGE, MS/MS

2 Weeks

4 Weeks

4 Weeks

3 Weeks

3 Weeks

3 Weeks

4 Weeks

(continued)

DIGE, MS/MS

2D PAGE, fingerprinting

2 Weeks

Chronic mild stress Chronic mild stress

2D PAGE

3 Weeks

4 Weeks

Method 2D PAGE

Treatment Model 2 Weeks

Proteome Effects of Antidepressant Medications 405

Species

Brain region

Antidepressant

Dose, route

Cecconi et al. Male SpragueEmbryonic Fluoxetine 1 mM (2007) Dawley rat cortical neurons 13 McHugh et al. Mouse Embryonic Paroxetine 1 mM (2008) stem cells Male SpragueCholinergic basal 14 Basheer, Brown, Dawley rat forebrain Ramesh, Begum, & McCarley (2005) Male SpragueHippocampus 15 Ding, Vaynman, Dawley rat Souda, Whitelegge, & Gomez-Pinilla (2006) Administration routes: i.p. intraperitoneally; s.c. sub-cutaneously; p.o. orally a In these studies, the sex of the experimental animals is not reported, thus they are assumed to be male

12

Table 2  (continued) No Study

2 Weeks

3 Days

2D PAGE, fingerprinting 2D PAGE, MS/MS

Method

Exercise

2D PAGE, fingerprinting

Sleep 2D PAGE, deprivation fingerprinting

Treatment Model

406 L. Carboni et al.

(continued)

Table 3  Modulated proteins found in the analysed studies with associated biological process GO terms. Where GO term for biological processes where not available, GO terms for molecular function where added and indicated in parentheses Accession Protein UniProt entry number n Study (Table 2) Gene ontology (biological process) Actin filament polymerization; Actin ACTB_RAT P60711 6 Bisgaard et al. (2007) (10, 11); axonogenesis Carboni, Vighini, et al. (2006) (7–9); McHugh et al. (2008) (13) Cell differentiation; nervous system DihydropyrimidinaseDPYL2_RAT P47942 6 Bisgaard et al. (2007) (10); Carboni, development; regulation of axon related protein 2 Vighini, et al. (2006) (7–9); Khawaja extension; signal transduction et al. (2004) (3, 4) Anaerobic glycolysis; lactate LDHB_RAT P42123 5 Carboni, Vighini, et al. (2006) (7, 9); l-Lactate metabolic process; tricarboxylic Khawaja et al. (2004) (3, 4); Ding dehydrogenase acid cycle intermediate metabolic et al. (2006) (15) B chain process 60 kDa heat CH60_RAT P63039 4 Bisgaard et al. (2007) (11); Guest et al. (2004) Apoptosis, cell organisation and shock protein (5, 6); Ding et al. (2006) (15) biogenesis, cellular localization, cellular macromolecule metabolism, intracellular protein transport, protein folding, protein targeting, response to stress Glycolysis Alpha-enolase ENOA_RAT P04764 4 Carboni, Vighini, et al. (2006) (9); Cecconi et al. (2007) (12); Khawaja et al. (2004) (3, 4) Intermediate filament-based process; Glial fibrillary acidic GFAP_RAT P47819 4 Carboni, Vighini, et al. (2006) (7, 8); response to wounding protein Ding et al. (2006) (15); McHugh et al. (2008) (13)

Proteome Effects of Antidepressant Medications 407

P19527 P04691

NFL_RAT

TBB2B_RAT

1433E_RAT

1433Z_RAT

ACON_RAT

AINX_RAT

CH10_RAT

Neurofilament light polypeptide Tubulin beta-2B chain

14-3-3 Protein epsilon

14-3-3 Protein zeta delta

Aconitate hydratase

Alpha-internexin

10 kDa heat shock protein

P26772

P23565

Q9ER34

P63102

P62260

P02688

MBP_RAT

Myelin basic protein S

Accession number

UniProt entry

Protein

Table 3  (continued)

3

3

3

3

3

4

4

4

n Carboni, Vighini, et al. (2006) (9); Khawaja et al. (2004) (3, 4); Basheer et al. (2005) (14)

Gene ontology (biological process)

Central nervous system development; myelination; negative regulation of axonogenesis; synaptic transmission Guest et al. (2004) (5, 6); Basheer et al. (2005) Intermediate filament cytoskeleton (14); Ding et al. (2006) (15) organisation and biogenesis Microtubule-based movement; protein Bisgaard et al. (2007) (10, 11); polymerization Carboni, Vighini, et al. (2006) (9); Ding et al. (2006) (15) Carboni, Vighini, et al. (2006) (7, 8); Cell communication, cellular Cecconi et al. (2007) (12) localization, intracellular protein transport, intracellular signalling cascade, protein targeting Cell organisation and biogenesis, Carboni, Vighini, et al. (2006) (7, 9); Cecconi et al. (2007) (12) cellular localization, intracellular protein transport, protein targeting Carboni, Vighini, et al. (2006) (7–9) Cellular carbohydrate metabolism, organic acid metabolism, tricarboxylic acid cycle Carboni, Vighini, et al. (2006) (7, 9); Cell differentiation; intermediate Ding et al. (2006) (15) filament cytoskeleton organisation and biogenesis; nervous system development Carboni, Vighini, et al. (2006) (7); Khawaja Protein folding et al. (2004) (3, 4)

Study (Table 2)

408 L. Carboni et al.

P11598 Q63537

O08838

P36972

Protein disulfide-isomerase A3 PDIA3_RAT

SYN2_RAT

TBA1B_ MOUSE

AATC_RAT

AMPH_RAT

APT_RAT

ATPA_RAT

CALR_RAT

COX5B_RAT

Synapsin 2

Tubulin alpha-1B chain

Aspartate aminotransferase

Amphiphysin

Adenine phosphoribosyl transferase

ATP synthase alpha chain

Calreticulin

Cytochrome c oxidase subunit 5B

P12075

P18418

P15999

P13221

P05213

P08009

GSTM4_RAT

GSH S transferase Yb3

2

2

2

2

2

2

3

3

3

3

(continued)

Carboni, Vighini, et al. (2006) (7); Khawaja Metabolic process et al. (2004) (3, 4) Carboni, Vighini, et al. (2006) (7, 8); McHugh Cell redox homeostasis; positive et al. (2008) (13) regulation of apoptosis Carboni, Vighini, et al. (2006) (7–9) Regulation of neurotransmitter secretion; synaptic transmission Carboni, Vighini, et al. (2006) (7, 9); Basheer Microtubule-based movement; et al. (2005) (14) microtubule-based process; protein polymerization Carboni, Vighini, et al. (2006) (7, 9) Amino acid metabolic process; aspartate catabolic process; biosynthetic process Carboni, Vighini, et al. (2006) (7); Basheer Learning; regulation of GTPase et al. (2005) (14) activity; synaptic vesicle endocytosis Khawaja et al. (2004) (3, 4) Adenine metabolic process; lactation; nucleoside metabolic process; purine ribonucleoside salvage ATP synthesis coupled proton Carboni, Vighini, et al. (2006) (9); Ding et al. (2006) (15) transport; ion transport Carboni, Vighini, et al. (2006) (7, 8) Cellular calcium ion homeostasis; cortical actin cytoskeleton organisation and biogenesis; protein export from nucleus; protein folding Khawaja et al. (2004) (3, 4) Electron transport; respiratory gaseous exchange

  Proteome Effects of Antidepressant Medications 409

UniProt entry

P21575

DYN1_RAT

ENOG_RAT

FABPH_RAT

GMFB_RAT

GRP78_RAT

Gamma-enolase

Fatty acid-binding protein

Glia maturation factor beta

78 kDa glucose-regulated protein

P06761

Q63228

P07483

P07323

P80254

2

2

2

2

2

2

2

P10860

DOPD_RAT

2

n

O08557

Accession number

d-dopachrome decarboxylase Dynamin-1

DDAH1_RAT N(G),N(G)dimethylarginine dimethylaminohydrolase 1 Glutamate DHE3_RAT dehydrogenase 1

Protein

Table 3  (continued)

Carboni, Vighini, et al. (2006) (8); Cecconi et al. (2007) (12)

Khawaja et al. (2004) (3, 4)

Sills et al. (1986) (2); Carboni, Vighini, et al. (2006) (9) Khawaja et al. (2004) (3, 4)

Carboni, Vighini, et al. (2006) (8, 9)

Khawaja et al. (2004) (3, 4)

Carboni, Vighini, et al. (2006) (9); Ding et al. (2006) (15)

Khawaja et al. (2004) (3, 4)

Study (Table 2)

Fatty acid metabolic process; longchain fatty acid transport; negative regulation of cell proliferation; phosphatidylcholine biosynthetic process Protein binding, protein dimerization activity (GO, molecular function) ER overload response; anti-apoptosis; negative regulation of caspase activity

Arginine metabolic process; nitric oxide mediated signal transduction; positive regulation of angiogenesis Amino acid metabolic process; glutamate catabolic process; longterm memory; transmembrane receptor protein tyrosine kinase signalling pathway Inflammatory response; melanin biosynthetic process Receptor internalisation; receptormediated endocytosis; synaptic transmission Glycolysis

Gene ontology (biological process)

410 L. Carboni et al.

Q01713

P11980 P04636

IGF1_RAT

KCRB_RAT

KLF9_RAT

KPYM_RAT

MDHM_RAT

Insulin-like growth factor I

Creatine kinase, B chain

Krueppel-like factor 9

Pyruvate kinase isozymes M1/M2 Malate dehydrogenase

P07335

P08025

P63018

HSP7C_RAT

Heat shock cognate 71 kDa protein

P04906

GSTP1_RAT

GSH S transferase P

2

2

2

2

2

2

2

 

Carboni, Vighini, et al. (2006) (8, 9)

Carboni, Vighini, et al. (2006) (8, 9)

Khawaja et al. (2004) (3, 4)

Carboni, Vighini, et al. (2006) (9); McHugh et al. (2008) (13)

Khawaja et al. (2004) (3, 4)

Carboni, Vighini, et al. (2006) (7); Ding et al. (2006) (15)

Carboni, Vighini, et al. (2006) (7, 8)

(continued)

Glycolysis; malate metabolic process; tricarboxylic acid cycle

Anti-apoptosis; central nervous system development; metabolic process; xenobiotic metabolic process Cell cycle, chaperone cofactor dependent protein folding, response to stress Ras protein signal transduction; antiapoptosis; cell development; glial cell differentiation; insulin-like growth factor receptor signalling pathway; memory; nervous system development; positive regulation of cell proliferation Brain development; cellular chloride ion homeostasis; phosphocreatine metabolic process Progesterone receptor signalling pathway; regulation of transcription Glycolysis

Proteome Effects of Antidepressant Medications 411

P49432 P15399

O35832 P31044 P10111 O35244

ODPB_RAT

PBAS_RAT

PCTK3_RAT

PEBP1_RAT

PPIA_RAT

PRDX6_RAT

Serine/threonine-protein kinase PCTAIRE-3 Phosphatidylethanolaminebinding protein 1 Peptidyl-prolyl cis-trans isomerase A Peroxiredoxin-6

P12839

NFM_RAT

Neurofilament medium polypeptide Pyruvate dehydrogenase E1 component subunit beta Probasin

P63086

MK01_RAT

Mitogen-activated protein kinase 1

Accession number

UniProt entry

Protein

Table 3  (continued)

2

2

2

2

2

2

2

2

n

Carboni, Vighini, et al. (2006) (9); Cecconi et al. (2007) (12) Khawaja et al. (2004) (3, 4)

Khawaja et al. (2004) (3, 4)

Khawaja et al. (2004) (3, 4)

Khawaja et al. (2004) (3, 4)

Khawaja et al. (2004) (3, 4)

Guest et al. (2004) (5, 6)

Carboni, Vighini, et al. (2006) (7, 9)

Study (Table 2)

Hydrogen peroxide catabolic process; lipid catabolic process; response to oxidative stress

Negative regulation of MAPKKK cascade Protein folding

Cell cycle, cell–cell signalling, cellular localization, cellular metabolism, chemotaxis, development, induction of apoptosis, intracellular protein transport, intracellular signalling cascade, morphogenesis, nuclear import, positive regulation of transcription, response to DNA damage stimulus, response to stress, synaptic transmission, transmission of nerve impulse Intermediate filament cytoskeleton organisation and biogenesis Acetyl-CoA biosynthetic process from pyruvate; glycolysis Cellular physiological process, establishment of localization, transport Protein amino acid phosphorylation

Gene ontology (biological process)

412 L. Carboni et al.

P10536

RAB1B_RAT

RAB4A_RAT

RL18A_RAT

RL28_RAT RL35A_RAT RS19_RAT

Ras-related protein Rab-4A

60S Ribosomal protein L18a

60S Ribosomal protein L28 60S Ribosomal protein L35a 40S Ribosomal protein S19

2

P09005 P22789 O08950

ST2A2_RAT

Alcohol sulfotransferase A

Transcription initiation factor T2AG_RAT IIA gamma chain

2

2

2

P28663

2 2 2

2

2

2

2

Beta-soluble NSF attachment SNAB_MOUSE protein Serine protease inhibitor 2.1 SPI21_RAT

P17702 P04646 P17074

P62718

P05714

P17220

PSA2_RAT

Proteasome subunit alpha type-2 Ras-related protein Rab-1B

Khawaja et al. (2004) (3, 4)

Khawaja et al. (2004) (3, 4)

Bisgaard et al. (2007) (11); Basheer et al. (2005) (14) Khawaja et al. (2004) (3, 4)

Khawaja et al. (2004) (3, 4) Khawaja et al. (2004) (3, 4) Khawaja et al. (2004) (3, 4)

Khawaja et al. (2004) (3, 4)

Khawaja et al. (2004) (3, 4)

Khawaja et al. (2004) (3, 4)

Khawaja et al. (2004) (3, 4)

 

(continued)

Ubiquitin-dependent protein catabolic process Intracellular protein transport; regulation of transcription; small GTPase mediated signal transduction Protein transport; regulation of endocytosis; small GTPase mediated signal transduction Biosynthesis, macromolecule metabolism Translation Translation Anatomical structure development; gas transport; hemocyte development; positive regulation of cell motility; response to extracellular stimulus; translation Intracellular protein transport; vesiclemediated transport Protease inhibitor activity (GO, molecular function) Lipid metabolic process; steroid metabolic process Regulation of transcription, DNAdependent; transcription initiation from RNA polymerase II promoter

Proteome Effects of Antidepressant Medications 413

P51672 P05065 P09117 P07943

XCL1_RAT ALDOA_RAT

ALDOC_RAT

ALDR_RAT

ANXA5_RAT

ATP5H_RAT

ATP5I_RAT

Annexin A5

ATP synthase D chain

ATP synthase e chain

P29419

P31399

P14668

P51863

Q00981

VA0D1_MOUSE

UCHL1_RAT

Ubiquitin carboxyl-terminal hydrolase isozyme L1

P37805

Accession number

Vacuolar ATP synthase subunit d 1 Lymphotactin Fructose-bisphosphate aldolase A Fructose-bisphosphate aldolase C Aldose reductase

TAGL3_RAT

UniProt entry

Transgelin

Protein

Table 3  (continued)

1

1

1

1

1

2 1

2

2

2

n

Carboni, Vighini, et al. (2006) (7)

Carboni, Vighini, et al. (2006) (8)

Carboni, Vighini, et al. (2006) (7)

Carboni, Vighini, et al. (2006) (8)

Ding et al. (2006) (15)

Khawaja et al. (2004) (3, 4) Carboni, Vighini, et al. (2006) (9)

Carboni, Vighini, et al. (2006) (7, 9)

Carboni, Vighini, et al. (2006) (7); Ding et al. (2006) (15) Carboni, Vighini, et al. (2006) (9); Cecconi et al. (2007) (12)

Study (Table 2)

Central nervous system development; muscle development Axon target recognition; axon transport of mitochondrion; axonogenesis; cell proliferation; response to stress; sensory perception of pain; ubiquitin-dependent protein catabolic process ATP biosynthetic process; ion transport Chemotaxis; immune response Glycolysis; response to hypoxia; response to nicotine Fructose metabolic process; glycolysis Carbohydrate metabolic process; positive regulation of JAK-STAT cascade; stress-activated protein kinase signalling pathway Anti-apoptosis; negative regulation of coagulation; protein homooligomerization; signal transduction ATP synthesis coupled proton transport; ion transport ATP synthesis coupled proton transport; ion transport

Gene ontology (biological process)

414 L. Carboni et al.

DEF5_MOUSE

EF1A1_RAT

Defensin-related cryptdin-5

Elongation factor 1-alpha 1

Glyceraldehyde 3-phosphate dehydrogenase Guanidine binding protein

GBB2_RAT

G3P_RAT

ENPL_MOUSE EZRI_MOUSE

P08082 P84087

CLCB_RAT CPLX2_RAT

Endoplasmin Ezrin

Q9d8b3

CHM4B_MOUSE

P54313

P04797

P08113 P26040

P62630

P28312

P63100

CANB1_RAT

Calcineurin subunit B isoform 1 Charged multivesicular body protein 4b Clathrin light chain B Complexin 2

P10719

ATPB_RAT

ATP synthase subunit beta

1

1

1 1

1

1

1 1

1

1

1

 

Carboni, Vighini, et al. (2006) (7)

Carboni, Vighini, et al. (2006) (9)

Carboni, Vighini, et al. (2006) (9) Carboni, Vighini, et al. (2006) (9)

Carboni, Vighini, et al. (2006) (9)

McHugh et al. (2008) (13)

Carboni, Vighini, et al. (2006) (9) Carboni, Vighini, et al. (2006) (9)

Carboni, Vighini, et al. (2006) (9)

Carboni, Vighini, et al. (2006) (9)

Ding et al. (2006) (15)

(continued)

Intracellular protein transport Mast cell degranulation; membrane fusion; neurotransmitter transport; synaptic vesicle exocytosis; vacuole organisation and biogenesis Defense response; defense response to bacterium Biosynthesis, cellular macromolecule metabolism, translation Protein folding; response to stress Establishment and/or maintenance of apical/basal cell polarity; regulation of cell shape Apoptosis; glucose metabolic process; glycolysis G-protein coupled receptor protein signalling pathway; signal transduction

ATP synthesis coupled proton transport; ion transport; receptormediated endocytosis Cell development, cell differentiation, morphogenesis Protein transport

Proteome Effects of Antidepressant Medications 415

Q99PT1 P09606

P30033

P38647

P07901

P11499 O88600

GDIR_MOUSE

GLNA_RAT

GNAO2_RAT

GRP75_MOUSE

H2B1_RAT

HS90A_MOUSE

HS90B_MOUSE

HSP74_RAT

Rho GDP-dissociation inhibitor 1 Glutamine synthetase

Guanine nucleotidebinding protein G(o) subunit alpha 2 Stress-70 protein

Histone H2B

Heat shock protein HSP 90-alpha

Heat shock protein HSP 90-beta Heat shock 70 kDa protein 4

Q00715

Q61598

GDIB_MOUSE

Rab GDP dissociation inhibitor beta

Accession number

UniProt entry

Protein

Table 3  (continued)

1

1

1

1

1

1

1

1

1

n

Bisgaard et al. (2007) (10)

Carboni, Vighini, et al. (2006) (9)

Carboni, Vighini, et al. (2006) (9)

Cecconi et al. (2007) (12)

McHugh et al. (2008) (13)

Carboni, Vighini, et al. (2006) (8)

Carboni, Vighini, et al. (2006) (9)

Carboni, Vighini, et al. (2006) (7)

Carboni, Vighini, et al. (2006) (8)

Study (Table 2)

Protein export from nucleus; protein folding; response to heat; response to stress Defense response to bacterium; nucleosome assembly Nitric oxide biosynthetic process; positive regulation of cytotoxic T cell differentiation; protein folding; response to heat; response to stress Protein folding; response to heat; response to stress Response to stimulus, response to stress, response to temperature stimulus, response to unfolded protein

Ammonia assimilation cycle; regulation of neurotransmitter levels Cell communication, signal transduction

Protein transport; regulation of GTPase activity; small GTPase mediated signal transduction Rho protein signal transduction

Gene ontology (biological process)

416 L. Carboni et al.

P08461

P16617

ODP2_RAT

PGK1_RAT

PHB_MOUSE PROF2_RAT

Dihydrolipoyllysineresidue acetyltransferase component of pyruvate dehydrogenase complex Phosphoglycerate kinase 1

Prohibitin Profilin-2

P67778 Q9EPC6

P46460 O35619

P07936

NEUM_RAT

NSF_MOUSE O35619_MOUSE

P25809 P19804

KCRU_RAT NDKB_RAT

Vesicle-fusing ATPase Vesicle associated membrane protein 2

Creatine kinase Nucleoside diphosphate kinase B Neuromodulin

1 1

1

1

1 1

1

1 1

McHugh et al. (2008) (13) Carboni, Vighini, et al. (2006) (7)

Ding et al. (2006) (15)

Carboni, Vighini, et al. (2006) (9)

Carboni, Vighini, et al. (2006) (9) Basheer et al. (2005) (14)

Basheer et al. (2005) (14)

Ding et al. (2006) (15) Carboni, Vighini, et al. (2006) (7)

 

(continued)

Carbohydrate metabolic process; glycolysis; phosphorylation DNA replication Cytoskeleton organisation and biogenesis; regulation of actin polymerization and/or depolymerization

Phosphocreatine biosynthetic process Negative regulation of apoptosis; nucleotide metabolic process Cell differentiation; glial cell differentiation; nervous system development; protein kinase C activation; regulation of cell growth; response to wounding; tissue regeneration Protein transport Calcium ion-dependent exocytosis; membrane fusion; synaptic vesicle exocytosis; vesicle-mediated transport Acetyl-CoA biosynthetic process; glycolysis; tRNA aminoacylation for protein translation

Proteome Effects of Antidepressant Medications 417

Q5RI75

O89086

RASEF_MOUSE

RBM3_MOUSE

Synaptosomal-associated protein 25

40S Ribosomal protein SA Splicing factor, arginine/serine-rich 1 Alpha-soluble NSF attachment protein

Q9cz13

QCR1_MOUSE

Cytochrome b-c1 complex subunit 1 RAS and EF-hand domain-containing protein homolog Putative RNA-binding protein 3

P60881

1

1

SNAA_MOUSE Q9DB05

SNP25_RAT

1 1

1

1

1

1

n

RSSA_RAT P38983 SFRS1_MOUSE Q6PDM2

P42669

PURA_MOUSE

Transcriptional activator protein Pur-alpha

Accession number

UniProt entry

Protein

Table 3  (continued)

Basheer et al. (2005) (14)

Bisgaard et al. (2007) (11)

Carboni, Vighini, et al. (2006) (7) Cecconi et al. (2007) (12)

Cecconi et al. (2007) (12)

McHugh et al. (2008) (13)

Carboni, Vighini, et al. (2006) (8)

Ding et al. (2006) (15)

Study (Table 2)

Apical protein localization; brain development; intracellular protein transport; neuron differentiation; vesicle-mediated transport Axonogenesis; long-term memory; neurotransmitter secretion; neurotransmitter uptake; regulation of synaptogenesis; sleep; synaptic transmission; synaptic vesicle docking during exocytosis

miRNA-mediated gene silencing, production of miRNAs; response to cold; translation Translation mRNA processing; protein targeting

Protein transport; small GTPase mediated signal transduction

Apoptosis; cell differentiation; cell proliferation; nervous system development; regulation of transcription Electron transport; proteolysis

Gene ontology (biological process)

418 L. Carboni et al.

Q9Z2I9 P68370

STMN1_RAT

STXB1_RAT

SUCB1_MOUSE

TBA1A_RAT

Stathmin

Syntaxin binding protein 1

Succinyl-CoA ligase [ADPforming] beta-chain Tubulin alpha-1A chain

P61765

P13668

Q64105

SPRE_MOUSE

Sepiapterin reductase

P07632

SODC_RAT

Superoxide dismutase [Cu-Zn]

1

1

1

1

1

1

Carboni, Vighini, et al. (2006) (7)

Carboni, Vighini, et al. (2006) (7)

Carboni, Vighini, et al. (2006) (9)

Carboni, Vighini, et al. (2006) (7)

McHugh et al. (2008) (13)

Carboni, Vighini, et al. (2006) (9)

 

(continued)

DNA fragmentation during apoptosis; activation of MAPK activity; anti-apoptosis; nervous system development; removal of superoxide radicals; response to oxidative stress Metabolic process; tetrahydrobiopterin biosynthetic process Axonogenesis; cell differentiation; intracellular signalling cascade; microtubule depolymerization; negative regulation of microtubule polymerization; nervous system development Axon target recognition, axonogenesis, membrane organisation and biogenesis, nervous system development, neurogenesis, neuron development, neuron differentiation, neuron morphogenesis during differentiation, synaptic vesicle transport Metabolic process; tricarboxylic acid cycle Cell organisation and biogenesis, cytoskeleton organisation and biogenesis, cytoskeleton-dependent intracellular transport

Proteome Effects of Antidepressant Medications 419

VIME_ MOUSE

P20152

P50516

TRAP1_MOUSE Q9CQN1 UBE2N_RAT Q9EQX9

VATA_MOUSE

1 1

P63029

TCTP_RAT

Vacuolar ATP synthase catalytic subunit A Vimentin

1

P42932

1

1

1

1

TCPQ_MOUSE

n

P01849

TCA_MOUSE

T-cell receptor alpha chain C region T-complex protein 1 subunit theta Translationally controlled tumour protein Heat shock protein 75 kDa Ubiquitin-conjugating enzyme E2 N

Accession number

UniProt entry

Protein

Table 3  (continued)

McHugh et al. (2008) (13)

Carboni, Vighini, et al. (2006) (7)

Carboni, Vighini, et al. (2006) (9) Basheer et al. (2005) (14)

Carboni, Vighini, et al. (2006) (9)

Ding et al. (2006) (15)

McHugh et al. (2008) (13)

Study (Table 2)

Protein folding; response to stress DNA repair; protein modification process; ubiquitin-dependent protein catabolic process ATP biosynthetic process; ion transport; Intermediate filament-based process

Cellular protein metabolic process; protein folding Anti-apoptosis, cell proliferation

Gene ontology (biological process)

420 L. Carboni et al.

Proteome Effects of Antidepressant Medications

421

4.1.2 Proteomics in the New Millennium Proteomics has witnessed a rapid evolution during the 1990s due to a series of ­technological improvements, resulting in an explosion of the number and types of studies in all research areas. Improvements in 2D gel approaches (such as the consolidation of immobilised pH strip methods and developments in imaging analysis software) as well as fundamental advances in mass spectrometry protein detection and identification have revolutionised the field (Patterson & Aebersold, 2003). The emerging literature is also testifying a rapid growth in the field of proteomic approaches to psychiatry and to the study of antidepressant actions, prompted by the perceived limited knowledge of the molecular mechanisms underlying their action. One of the first attempts to characterise the impact of different antidepressants on a proteome-wide scale through modern 2D gel approaches is represented by the study reported by Khawaja et al. (2004). Their ultimate objective was finding new potential targets for pharmacological intervention looking at brain changes induced by different classes of antidepressants. The study was carried out in rat hippocampus after chronic treatment with two different antidepressants, venlafaxine (a 5HT and NE re-uptake inhibitor; study no. 4 in Tables 2 and 3) and fluoxetine (a SSRI; study no. 3 in Tables 2 and 3), using a differential proteomics approach to uncover a subset of proteins which were regulated by both antidepressants. The reason for searching only for overlapping proteins was to increase the probability of detecting molecular changes related to therapeutic efficacy. A quantitative analysis was not feasible with the technical method used in the study, thus a qualitative comparison was documented. By establishing a cut-off of 1.5-fold intensity changes, 31 protein spots showed increased staining intensity after both antidepressant treatments, whereas two were down-regulated. Modulated proteins belonged mainly to pathways involved in neurogenesis, cell proliferation and nervous system development as well as in intracellular signal transduction, energy metabolism and macromolecule biosynthesis (Table 3). Since all available antidepressants share the ability to modify the monoaminergic tone at the synaptic cleft, changes induced in brain function by chronic administration may result either from their therapeutic properties or from targeting specific neurotransmitter systems. To explore this, Guest et al. (2004) carried out a comparative analysis in guinea pig cerebral cortex proteomes after pharmacological treatment with a SSRI antidepressant and a NK1 receptor antagonist. The latter treatment represents a potential novel non-aminergic therapeutic approach in development for depression (Stout, Owens, & Nemeroff, 2001). Since a differential in-gel electrophoresis (DIGE) method was adopted, both groups of samples were run together on the same gels. Proteins modulated by treatment with fluoxetine (study no. 5 in Tables 2 and 3) and L-000760735 (NK1 antagonist; study no. 6 in Tables 2 and 3) belonged to the chaperone class and to neurofilament proteins. Several Heat shock protein 60 isoforms were modulated in opposite directions, suggesting that the treatment was influencing the ratio between post-translationally modified forms more than the absolute levels. In contrast, neurofilament protein levels were reduced by both pharmacological treatments. The authors were prompted by these findings to further investigate the

422

L. Carboni et al.

modifications in neurofilament protein abundance after antidepressant treatment with other techniques. Since the down-regulation suggested that an antidepressant action can be exerted by influencing synaptic remodelling, a comparison was performed between fluoxetine-treated and control animals by evaluating potential differences in postsynaptic densities in the somatosensory cortex. The results gave further support to the hypothesis that antidepressant treatments increase cortical synaptic activity (Guest et al., 2004). Following the same line of reasoning, another proteomic study was performed with the objective of discovering overlapping or distinct molecular changes induced by treatment with established or putative antidepressants acting through different mechanisms (Carboni, Vighini, et al., 2006). The comparison of changes induced by the different compounds can help assessing whether new classes of putative antidepressants could induce similar or unique changes when compared to fluoxetine. Hippocampal and frontal-prefrontal cortical proteomes were compared in rats chronically treated with fluoxetine (study no. 7 in Tables 2 and 3), with a NK1 receptor antagonist, GR205171 (study no. 8 in Tables 2 and 3), or with a CRF receptor 1 antagonist, DMP696 (study no. 9 in Tables  2 and 3). Protein modulations shared by all three compounds involved changes in levels of actin isoforms, of synapsin 2, aconitate hydratase and dihydropyrimidinase-related ­protein 2 (DRP-2). However, caution should be used in interpreting the latter finding, because DRP-2 belongs to a group of proteins that often show changes after several kinds of perturbations (Fountoulakis, 2004). In addition to overlapping modifications, each pharmacologically active compound induced a specific pattern of protein modulation (Table 3). The common theme that was detected in changes induced by each treatment supports the hypothesis that efficacious antidepressants share the ability of modulating neural plasticity. Another approach was undertaken by Bisgaard et al. (2007), who adopted a proteomic perspective in the attempt of identifying molecular correlates of resistance to antidepressant treatment. In addition, molecular mechanisms of the liability to environmental stressors were also investigated, since there is a well-known correlation between stress exposure and major depression (Kendler, Karkowski, & Prescott, 1999). Modified proteins can be relevant in the liability to depressive disorder, since stress exposure can trigger the onset of major depression episodes in predisposed individuals. The proteomic study was carried out in the ventral hippocampus of rats exposed to a rodent model of depression (chronic unpredictable stress) and antidepressant treatment with an SSRI, escitalopram (ESC). The design included six experimental groups: the rats exposed to the chronic mild stress were split into ESC responders, ESC non-responders, vehicle-treated and stress-resilient animals. The latter group included rats that did not show a reduction in sucrose consumption after stress exposure, a behaviour frequently detected in stressed animals and considered to be related to anhedonia. Two additional groups were not exposed to stress and only received the antidepressant or vehicle. The aim of the study was to identify markers of anhedonia, of resistance to ESC, and of stress resilience. No putative anhedonia marker could be identified. Markers of ESC resistance were defined as those proteins showing a significant difference of expression between ESC responders and non-responders (study no. 10 in Tables 2 and 3), which were further subcategorised

Method for protein References solubilisation 5.6% 2-Mercap­ Heydorn toethanol, et al. 20% (1984) glycerol, 0.4% ampholines pH 3–10, 2% NP-40 Sills et al. 5% 2-Mercap­ (1986) toethanol, 20% glycerol, 0.4% ampholines pH 3–10, 1.6% ampholines pH 5–7, 2% NP-40 6 M urea, 2 M Khawaja thiourea, et al. 40 mM tris, (2004) 65 mM DTT, 4% CHAPS Guest et al. 30 mM Tris pH 8, (2004) 7 M urea, 2 M thiourea and 4% CHAPS Not indicated

3 Individuals /group

Silver staining

–

Silver 4.5% staining Iodoace­ tamide

260 mM Cy3 and Cy5 5 Individuals Iodoace­ /group for quantif­ tamide ication. (after Sypro Ruby IEF) for spot excision

Conventional tubes gels/ ampholines

IPG

IPG

IEF method Conventional tubes gels/ ampholines

Number of replicates Not indicated

Alkylation method Staining – Silver staining

DeCyder

HT analyzer

Computerised scanning densitometry (Goldman et al.19821)

Software for spot a nalysis Computerised scanning densitometry (Goldman et al.1982)

Table 4  Summary of the technologic approaches adopted by the different papers listed in references

Student’s t test (p < 0.05)

Fold change cut-off ³1.5

Student’s t test (p < 0.01)

Determination of spot differences Student’s t test (p < 0.05)

(continued)

Immunoblot for 4 proteins

–

MS-Fit

MS-Fit

–

Results confirm­ ation –

–

Software for identif­ ication –

Proteome Effects of Antidepressant Medications 423

10 mM HEPES, 137 mM NaCl, 4.6 mM KCl, 1.1 mM KH2PO4, 0.6 mM MgSO4 8 M urea, 100 mM DTT, 4% CHAPS, 0.8% ampholines

7 M urea, 2 M Carboni, thiourea, Vighini 20 mM tris, et al. 5 mM TBP, (2006) 3% CHAPS, 1% ampholines pH 3.5–10

Ding et al. (2006)

Basheer et al. (2005)

Method for protein References solubilisation

Table 4  (continued)

–

–

10 mM Iodoace­ tamide (before IEF)

IPG

IPG

Sypro Ruby and Pro-Q Diamond phosphop­ rotein staining Sypro Ruby

Silver staining or Coomassie blue

Alkylation method Staining

Conventional tubes gels/ ampholines

IEF method

Fold change cut-off ³1.6

PDQuest

PDQuest

5 Individuals/ group

8 Individuals/ group

Profound Immunoblot for 2 proteins

Profound Immunoblot for 4 or MSproteins Fit

Student’s t test Mascot, – Pro­ (p < 0.05) found ,PLS and (p < 0.05) Protein and fold Probe change cut-off ³1.15

Fold change cut-off ³2

Software Determination for Results identif­ of spot confirm­ ication differences ation

Progenesis Pool from 7 Individuals/ group, 2 replicate gels

Number of replicates

Software for spot a nalysis

424 L. Carboni et al.

McHugh et al. (2008)

Silver 240 mM Staining Iodoace­ for tamide quantifi­ (after cation; IEF) Sypro Ruby for spot excision

4 cell prepara­ tions/ group

PDQuest

PDQuest 4 cell prepara­­tions/ group, 4 replicate gels each

IPG 7 M urea, 2 M thiourea, 20 mM tris, 5 mM TBP, 3% CHAPS, 1% ampholines pH 3.5–10 IPG 8 M urea, 2 M thiourea, 20 mM tris, 130 mM DTT, 4% CHAPS, ampholines pH 3–10

Cecconi et al. (2007)

Sypro 20 mM Ruby Iodoace­ tamide (before IEF)

Cy3 and Cy5 Pools from 2 or DeCyder 19 mM 3 indivi­duals; for quantifi­ Iodoace­ cation. Silver 6 pooled tamide samples/ staining for (after group spot IEF) excision

IPG 7 M urea, 2 M thiourea, 30 mM tris, 4 % CHAPS

Bisgaard et al. (2007)

Student’s t test Mascot – (p < 0.05), (for fold change MS/ cut-off ³1.3, MS) SAM2 and hierarchical cluster analysis Student’s t test Mascot, Immunoblot for 8 Pro­ (p < 0.05) proteins found and and hierarchical Protein cluster Probe analysis – Mascot Student’s (for t test MS/ (p < 0.02) MS)

Proteome Effects of Antidepressant Medications 425

426

L. Carboni et al.

based on their expression profiles in the different groups. Modulated proteins were found to be involved in cytoskeleton reorganisation, protein folding and cell differentiation (Table 3). The chronic mild stress marker proteins were revealed by comparing animals resilient to chronic mild stress vs. healthy and anhedonic rats (study no. 11 in Tables 2 and 3). Actin, tubulin, Heat shock protein 60 and a and b SNAPs were thus included among putative stress resilience markers (Table  3) (Bisgaard et al., 2007). 4.1.3 Proteomic Investigations in Cellular Models A number of studies have been reported which apply proteomic approaches to the characterisation of antidepressant activity on central nervous system model systems. The search for molecular changes induced by fluoxetine on neuronal cells beyond the serotonin re-uptake inhibition was addressed with a proteomic approach by Cecconi et al. (2007). The experiments were carried out on a rat primary culture enriched in cortical neurons which was exposed to a sub-chronic treatment with fluoxetine (study no. 12 in Tables 2 and 3). Six proteins were differently expressed by more than 100% and seven proteins were differently expressed by more than 50% (Table 3). Among the identified proteins, cyclophilin A, 14-3-3 protein z/d and GRP78 are involved in neuroprotection, in serotonin biosynthesis and in axonal transport, respectively. Trends of changes in the same direction as those detected in the 2D gel analyses were also detected in western blotting (WB) for most proteins. Nonetheless, the quantitative difference between the results obtained by 2D electrophoresis and by WB suggested again that most changes detected by the former technique specifically involve post-translationally modified forms, which can only be separated in 2D maps (Cecconi et al., 2007). In recent years, evidence is accumulating that efficacious antidepressants are associated with the ability to stimulate neurogenesis (Malberg, Eisch, Nestler, & Duman, 2000) and this mechanism has been implicated in the therapeutic action. McHugh et al. (2008) used a proteomic approach to improve the understanding of the molecular and cellular processes affected by the exposure to an antidepressant in mouse neural cells (study no. 13 in Tables  2 and 3). This model system is obtained by stem cell differentiation and consists of dopaminergic and NEergic neurons together with astrocytes. A 2-week exposure to the SSRI antidepressant paroxetine induced modifications in the proteome that included a marked reduction of GFAP levels and a modification in actin, creatine kinase, B  chain disulphide isomerase, prohibitin, RAS and EF-hand domain-containing protein homolog, sepiapterin reductase, T-cell receptor alpha chain C region, vimentin, defensin-related cryptdin-5 and stress-70 protein (Table 3) (McHugh et al., 2008).

4.2 Non-Pharmacologic Interventions Sleep deprivation is reported to be an effective, although short-lasting, treatment for major depression, eliciting rapid improvement in many signs and symptoms in a high

Proteome Effects of Antidepressant Medications

427

percentage of patients (Giedke & Schwärzler, 2002). It is thus conceivable that ­proteomic changes observed after sleep deprivation may be involved in its antidepressant efficacy. In rats, a qualitative proteomic analysis of cholinergic basal forebrain (­including horizontal and diagonal band of Broca, substantia innominata and magnocellular preoptic area) after exposure to 6 h of sleep deprivation was carried out (Basheer et al., 2005; study no. 14 in Tables 2 and 3). WB analysis of modulated proteins showed a modification in the ratios of different post-translationally modified isoforms of alpha tubulin, GAP43, and amphiphysin I (Table 3). It is thus conceivable that these proteins may be involved in molecular changes associated with antidepressant efficacy. On the other hand, since the antidepressant mechanism of action of sleep deprivation is unknown, it is possible that the brain regions chosen in this study may not have singled out the most relevant changes relevant to its antidepressant effects. Several lines of evidence suggest that exercise can be an effective intervention for treating depression, although its long-term effectiveness is not completely established (Lawlor & Hopker, 2001). A proteomic study evaluated changes in rat hippocampus after exposure to a light, voluntary exercise protocol (Ding et al., 2006; study no. 15 in Tables 2 and 3). The analysis was performed by 2D electrophoresis with qualitative comparisons. Phosphorylation specific staining was also performed on maps. Among several spots representing different post-translational modifications of the same protein, changes in respective ratios were found in energy metabolism proteins (fructosebisphosphate aldolase C, Phosphoglycerate kinase 1, l-lactate dehydrogenase B chain, alpha and beta subunits of ATP synthase, Creatine kinase, and Glutamate dehydrogenase); cytoskeletal proteins (beta tubulin, glial fibrillary acidic protein, and alpha internexin); chaperones (heat shock protein 8, heat shock protein 60, Heat shock cognate 71 kDa, T-complex protein 1 subunit theta, and Transgelin-3) (Table 3). In addition, a differential phosphorylation pattern was observed for tubulin, neurofilament light polypeptide, GFAP, Heat shock cognate 71 kDa, and Transcriptional activator protein Pur-alpha (Table  3). The authors suggest that exercise modulates synaptic plasticity and energy metabolism in hippocampus, thus possibly contributing to an enhancement of cognitive function observed in humans and animals (Ding et al., 2006). It is possible that these same molecular changes are also involved in the putative antidepressant benefits caused by exercise.

5 Critical Evaluations of Proteomic Techniques in Studies with Antidepressants 5.1 2D Gel Electrophoresis Protein expression profiling for the study of the antidepressants effect relied mainly on the use of 2D gel electrophoresis combined with mass spectrometry. Except for some cases in which conventional isoelectric focussing was still used (Basheer et al., 2005; Heydorn et al., 1984) according to the method of O’Farrell (1975), the most commonly used approaches for the 2D separation of proteins employed

428

L. Carboni et al.

immobilised pH gradients (IPG) in the first dimension (Bisgaard et  al., 2007; Carboni, Vighini, et al., 2006; Cecconi et al., 2007; Ding et al., 2006; Guest et al., 2004; Khawaja et al., 2004; McHugh et al., 2008), as it is widely recognised that this method has remarkably improved the resolution, the reproducibility and the handling of 2D maps (Bjellqvist et al., 1982; Görg, Weiss, & Dunn, 2004; Hanash, 2003; Righetti, Campostrini, Pascali, Hamdan, & Astner, 2004; Tannu & Hemby, 2006). The most common standard solubilisation solution included urea and thiourea as chaotropic agents and CHAPS as zwitterionic detergent (Table 4). A reduction and alkylation step was often applied to avoid the formation of mixed disulphide bridges among polypeptide chains (Herbert et al., 2001). After the run, protein spots were frequently detected by silver staining. A fluorescent staining (Sypro Ruby) has been adopted by some groups for 2D gel staining because of its enhanced sensitivity compared to the Coomassie’s blue, its larger dynamic range for quantification compared to the silver and its compatibility with downstream identification analysis. Another common technique, the 2D DIGE, has been adopted in some studies (Bisgaard et  al., 2007; Guest et  al., 2004) in which the proteins are labelled by cyanine fluorescent dyes (Cy2, Cy3 and Cy5) before the first dimension run. After the labelling with different Cy tags, control and experimental samples can be mixed together and separated in the same gel, thus preventing gel-to-gel variability. However, DIGE has some disadvantages including possible artefacts in optical density ratios due to differences in dye-labelling of some proteins making dye reversal necessary to prevent dye bias.

5.2 Gel Matching and Differential Analysis The qualitative and quantitative spot analysis is now commonly performed by the use of dedicated software (in the papers considered in this chapter: PDQuest from Bio-Rad, Progenesis from Nonlinear Dynamics, Investigator HT analyzer from Genomic Solutions and DeCyder from Amersham) by which the user is supported in spot detection and matching by automatic tools. Despite recent improvements, the automatic spot analysis requires accurate manual checking, which still remains one of the most time-consuming steps in the process of 2D map analysis. The variability across gels can be reduced by running and processing gels in parallel, but still the same spot can be shift in position or can be missing in some gels. Missing spots occur when a protein spot, detected in a reference gel, is not present in the sample gel. This can be due to both a real absence of a protein or to detection failure due to mismatching, running, or staining variability. The last question can be addressed by analysing multiple replicate gels for each experimental condition and checking gel matching by an expert user. The minimal number of replicates for a good quality analysis is still under discussion (Grove, Hollung, Uhlen, Martens, & Faergestad, 2006). In the papers examined in this chapter, three to eight biological replicates were considered for the 2D analysis. In some cases samples were pooled and run in duplicate (Basheer et al., 2005) or, when a larger number of animals were investigated (Bisgaard et  al., 2007; n = 12), samples were pooled in pairs yielding six

Proteome Effects of Antidepressant Medications

429

b­ iological replicates. Proteomic studies produce large multivariate datasets causing issues in data analysis that have not yet yielded consensus solutions across the proteomic community. In some of the papers considered here, the evaluation of treatment effect was performed setting an arbitrary cut-off to the fold change of the spot volume among the analysed conditions (Basheer et  al., 2005; Ding et  al., 2006; Khawaja et al., 2004), but to test for significant effects of the treatments, the most commonly used analyses were statistical univariate methods, such as Student’s t-test. In some papers (Bisgaard et al., 2007; Carboni, Vighini, et al., 2006; Cecconi et al., 2007), a multivariate analysis was also performed, such as partial least squares (PLS) or hierarchical cluster analysis. In a multivariate analysis, all protein variables can be evaluated at the same time, thus preventing the false positives that can be produced by repeated univariate tests applied to a large data set. Some studies indicate that both univariate and multivariate analysis should be used in proteomic data evaluation for maximising data recovery (Karp, Griffin, & Lilley, 2005). Both analysis methods are thus complementary, so changes detected in common are thought to have a higher probability of being significant (Karp et al., 2005).

5.3 Protein Identification Spot excision from the gels before identification analysis was often performed in these studies by hand, or by the use of a spot-cutter robot. This last option could be preferable for improving cutting precision, in particular when a fluorescent staining is used and the spots are excised from several replicate gels. Most of the papers describing proteomic profiling of the antidepressants applied MALDI-TOF peptide fingerprinting mass spectrometry analysis that provides the most rapid route for protein identification. Two papers (Bisgaard et al., 2007 and McHugh et al., 2008) reported on protein identification by the use of tandem mass spectrometry (MS/ MS) by which amino acid sequences of individual peptides can be obtained, thus improving the effectiveness of the identification. Among the dedicated software used for protein identification from sequence databases (mainly Swiss-Prot and NCBInr) the most used were: MS-Fit (Protein Prospector), Mascot (Matrix Science) and Profound (PROWL, The Rockefeller University). These tools use different algorithms for assigning a significance score to the protein hits. It is reported that each algorithm produced best results in a specific range of molecular masses (Govorun & Archakov, 2002); across the different algorithms, ProFound, which uses Bayesian statistics for protein ranking, gives the most reliable identification results over the whole range of protein masses. In some papers (Carboni, Vighini, et al., 2006; Cecconi et al., 2007; Basheer et al., 2006), different software packages were used in parallel to confirm the identification. Different criteria were adopted for the acceptance of protein identifications by the different authors: some considered the scores generated by the software; some evaluated the number of matched peptides and the sequence coverage. The data quality evaluation is not easy due to the fact that often the criteria for the identification acceptance are not comparable, particularly comparing work done in different years. The Proteomic community is

430

L. Carboni et al.

still evolving (Stead, Preece, & Brown, 2006) and efforts are being made to ­standardise the way in which proteomics results should be reported (for example by the Human Proteome Organisation–Proteomics Standard Initiative).

5.4 Result Confirmation In some papers, confirmation of results was obtained by WB of a number of selected proteins. The low number of confirmations could be justified by the fact that one-dimensional WB is not sufficiently powerful for detecting the changes revealed in 2D maps. In fact, since 2D polyacrylamide gel electrophoresis can separate different isoforms of a protein (e.g. post-translational modified proteins that differ only in pI), changes detected in 2D gels can involve only a small percentage of the total protein quantity. In these cases, WB could not have the suitable sensitivity for detecting small variations in the total amount of the protein and is therefore not the best method for confirmation of results. An overview of the technologic approaches adopted in the different papers described in this chapter is summarised in Table 4.

6 Ongoing Studies A number of other studies using proteomic approaches for investigating the molecular changes induced by antidepressant treatments are under development, as can be ascertained by an analysis of the literature available in form of meeting presentations and posters. An overview of the animal models under investigation will be given in this section, even though the complete lists of modulated protein are not yet available. Proteomic investigations are ongoing within an integrated project, financed by the Europen Union, named Genome-based therapeutic drugs for depression (GENDEP). In this project, large-scale clinical pharmacogenomic studies on depressed patients are combined with preclinical investigations on animal models of disease, focusing on treatment with antidepressants. Simultaneous and wide-scale analyses of gene and protein expression are being adopted, since they provide powerful strategies for the exploration at a molecular level of complex pathophysiological mechanisms including the response to treatment with psychotropic agents. Within GENDP proteomic studies, the changes in protein profiles induced by antidepressant treatments were addressed in a rat model of depression that included both genetic and environmental components. The Flinders Sensitive Line (FSL), a genetically selected rat model of depression displaying good face, predictive and construct validity (Overstreet, Friedman, Mathé, & Yadid, 2005), was investigated via a 2D proteomic approach. As a control, analyses were carried out on the ­corresponding Flinders Resistant Line (FRL), which does not show the depressivelike behaviour. To evaluate gene-environment interactions, the FSL and FRL

Proteome Effects of Antidepressant Medications

431

a­ nimals were subjected to maternal separation (MS), since early life trauma is ­considered to be an important antecedent of major depression (Heim & Nemeroff, 2001). Both stressed and control FSL and FRL rats were treated with the SSRI ESC admixed to food pellets (25 mg/kg/day) or vehicle for 31 days. Image analysis was carried out on maps prepared from prefrontal/frontal cortex and hippocampus, and protein levels were evaluated with statistical tests. Comparisons were carried out between groups to detect changes induced by genetic background, stress exposure and pharmacological treatment in each brain region. In hippocampus, after MS, ESC treatment regulated 16 and 30 proteins in FSL and FRL, respectively; without MS, 18 or 29 proteins were modified in the two strains. In prefrontal/frontal cortex, after MS, ESC treatment regulated 13 and 14 proteins in FSL and FRL, respectively; without MS, 10 or 11 proteins were modified in the two strains. Modulated proteins were identified by peptide fingerprinting mass spectrometry. In the gene– environment interaction model, proteins modulated by ESC treatment belonged mainly to energy metabolism and cellular morphogenesis pathways. ESC treatment exerted a different interaction on brain proteome in the FRL genetic background, since a higher number of proteins was differentially expressed and, although with some overlap, a distinct expression profile was induced. Specifically, regulation of proteins related to vesicular trafficking and signal transduction was revealed. A different response was also induced by ESC treatment in FSL animals not subjected to MS: only four proteins were affected in common (Carboni, Piubelli, et al., 2006; Carboni et al., 2007; Piubelli et al., 2006). A parallel study in a different animal cohort was also carried out following the same experimental design. In this investigation, the proteomic analysis was carried out on purified synaptic terminals in order to collect information about specific changes induced in this sub-cellular compartment. Statistical analysis of maps prepared from prefrontal/frontal cortex synaptosomes revealed 37 proteins differently regulated in FSL vs. FRL rats. The stressful experience of MS significantly dysregulated 48 proteins in FSL, and 24 proteins in FRL synaptosomes. Chronic ESC treatment differently regulated 33 protein spots in FSL and 7 protein spots in FSL subjected to MS. Interestingly, in FSL rats, 3 of the proteins down-regulated by MS were up-regulated by ESC (Mallei, Giambelli, Barbiero, El Khoury et  al. 2006; Mallei, Giambelli, Barbiero, Mathé et al. 2006). The effect of nortriptyline (NOR) treatment was also examined in another GENDEP proteomic study, which utilised the same treatment schedule already described for ESC. Comparisons were carried out among groups with different genetic backgrounds in order to detect changes induced by the exposure to MS stress and by NOR treatment. The results showed that different proteins were modulated in FSL and FRL groups by MS and NOR treatment. Modulated proteins were mainly involved in cytoskeleton reorganisation, protein folding, energy metabolism, vesicle trafficking, regulation of neurogenesis and central nervous system development, suggesting new molecular correlates of vulnerability to depression, and of response to NOR treatment (Piubelli et al., 2007). In an additional study included in the GENDEP project, another animal model was investigated with a proteomic approach. The Learned Helplessness rat model

432

L. Carboni et al.

is obtained by exposing rats to aversive stimuli under circumstances in which they cannot control or predict these stimuli (Henn & Vollmayr, 2005). The model is well characterised and has good face, construct and predictive validity. This model employs a strictly elaborated protocol, defining rats as being “learned helpless” or “non-learned helpless”. An expression proteomic study was carried out to identify the molecular changes induced in hippocampus and prefrontal/frontal cortex of stressed animals with or without sub-chronic treatment with ESC. Changes in protein levels between basal “learned helpless” and “non-learned helpless” were identified, as well as after ESC treatment. Most alterations affected proteins involved in cellular metabolism, signal transduction and neurotransmission (e.g. Munc-18, 14-3-3 proteins and Wnt2b protein) (Giambelli et al., 2007). Since antidepressants and exercise are reported to increase neurogenesis in the dentate gyrus (Sahay & Hen, 2007; Uda, Ishido, Kami, & Masuhara, 2006), a proteomic study was undertaken to gain more understanding of neurogenesis in rat hippocampus. The overall aim of the study was to identify key proteins that mediate the proliferative effects of fluoxetine-treatment (10 mg/kg for 21 days), voluntary running and social housing on hippocampal progenitor cells using a proteomic approach. Hippocampal proteins from animals exposed to different treatments were extracted and fractionated into four fractions: cytosolic, membrane/organelle, nucleic and cytoskeletal fractions. Protein spots were detected, quantified identified with MALDI-TOF mass spectrometry (Paulson, Klint, Brange, Sihlbom, & Erikssom, 2007; Paulson, Klint, Sihlbom, & Eriksson, 2006). A major problem in the treatment of major depression is selecting the best medication for each individual patient, especially as most antidepressants have an unexplained lack of therapeutic effect in up to 40% of patients. It is possible that sequence variation in certain genes may account, at least in part, for individual differences in antidepressant response. With the objective of discovering genes relevant for the antidepressant efficacy, microarray analysis of gene expression and 2D electrophoretic analysis of brain proteins were run in parallel in a rat model of chronic paroxetine treatment. Proteins were separated from the hippocampus and frontal cortex and 30 protein spots that significantly differed in quantity between control and paroxetine-exposed rats were observed (Rogers, Allington, McHugh, Joyce, & Kennedy, 2004).

7 An Integrative Overview of Proteomic Findings An overall analysis of the changes in the proteome profile of brain regions exerted by different antidepressants is suggestive of a broad effect of treatment at different functional levels. Notwithstanding the inherent bias of the 2D gel approach towards abundant proteins and soluble intracellular components, a closer look of the list of altered proteins across the different studies suggests a number of biological processes that are recurrently modulated by antidepressant treatment. Taking the interpretation of data from studies with different design and different areas with a number of ­experimental caveats, it appears that antidepressant treatments are able to modulate the expression of proteins involved in energy production, cellular assembly and

Proteome Effects of Antidepressant Medications

433

organisation, cell-to-cell signalling and interactions, cell cycle, and many other ­biological processes as defined by the Gene Ontology terms (see Table 3). By making use of the Ingenuity Pathway Analysis (IPA) tool, we have further conducted a functional and pathway analysis of the proteomic effect of antidepressant treatments. The IPA tool (Ingenuity Systems, Mountain View, CA, USA; https:// analysis.ingenuity.com/pa) is a web-based system that provides computational ­algorithms to functionally analyse large datasets and identify or generate gene/protein networks that are formed by the proteins of interest. The analysis is supported by the company’s knowledge base, which contains gene–gene and ­protein–protein interactions collated from extensive annotation of literature findings. An overall analysis of the functional categories captured by the Ingenuity knowledge base confirms a substantial over-representation of proteins belonging to neurological disease among the disease category, which is perhaps not unexpected given that these proteomic studies have been carried out in neuronal tissues. When looking at the proteomic data and their associated biological processes, cellular assembly and organisation, cellular function and maintenance, nervous system development and function appear to be the mechanisms mostly impacted by antidepressant treatments (Fig. 1). Starting from the hypothesis that changes that are in common between different antidepressant treatments might be more likely to be involved in their therapeutic efficacy, we have then limited our analysis to proteins that have been found by at least three studies and examined the pathways generated by the system. The top scoring network identified by IPA with the above criteria is characterised by proteins associated with neurological disease and belonging to the cellular assembly and organisation process. Twelve out of 33 proteins in the network depicted in Fig. 2 are part of the experimental dataset constituted by proteins highlighted at least in three studies. Interestingly, the study which displays the highest degree of overlaps is the proteomic dataset obtained by chronic treatment with a prototypical SSRI antidepressant (fluoxetine) in the rat hippocampus (Carboni, Vighini, et al., 2006). The identified network suggests an involvement of isoforms of 14-3-3, a protein enriched in brain which plays important roles in a wide range of vital regulatory processes, including signal transduction, cell survival, cell cycle regulation (Kjarland, Keen, & Kleppe, 2006), and possibly in neuronal plasticity events, since it can interact with cytoskeleton elements (Sun, Bittner, & Holz, 2003). An up-regulation of 14-3-3 zeta isotype at mRNA and protein level has been also described in RBL-2H3 cells upon incubation with fluoxetine (Baik et al., 2005), suggesting its activation by antidepressant treatment. In parallel, several proteins in the network suggest that modifications occur at the level of cytoskeleton proteins. Interestingly, recent literature indicates that microtubule dynamics are altered by stress in rodents and suggests the existence of a link between stress-induced cytoskeletal ­microtubular changes and depression. Indeed, both acute and chronic treatment with ­antidepressant drugs has been shown to affect the expression of microtubular proteins (Bianchi, Hagan, & Heidbreder, 2005). Finally, even though not detected in at least three of the proteomic experiments, pathway analysis identified IFN-gamma as a putative network member with a number of significant interactions. The results is consistent with the postulated role of proinflammatory cytokines in the aetiology and

434

L. Carboni et al.

-log(p-value) 0

1

2

Threshold

3

4

5

6

7

8

9

10

11

12

13

Neurological Disease Hematological Disease Immunological Disease Respiratory Disease Cellular Assembly and Organization Cellular Function and Maintenance Nervous System Development and Function Cell Death Cell Morphology Cell Signaling Molecular Transport Nucleic Acid Metabolism Small Molecule Biochemistry Genetic Disorder Psychological Disorders

Fig. 1  Functional classification of the proteins identified in the full dataset (all published material on antidepressants) using ingenuity pathway analysis. The most significant biological functions and diseases associated to the experimental results are represented by solid bars. The significance value, represented in the X axis by −log(p value), is a measure of the likelihood that the association between the experimental set of molecules and a given biological process is not due to random chance. The p value is calculated using the right-tailed Fisher Exact Test. In this method, the p value for a given function is calculated by considering the number of functional analysis molecules that participate in that function and the total number of molecules that are known to be associated with that function in Ingenuity’s knowledge base. p Values less than 0.05 (indicated in the figure by x values higher than the threshold line) indicate a statistically significant, non-random association

p­ athophysiology of major depression (Schiepers, Wichers, & Maes, 2005) and in the known ability of different antidepressant treatments to exert a common, negative immunoregulatory effect by suppressing the IFN-gamma/IL-10 production ratio possibly contributing to their therapeutic efficacy (Kubera et al., 2001).

Proteome Effects of Antidepressant Medications

435

Extracellular Space

Unknown

4

2 IL1F9

1 MBP

DSPP

3

Jnk 5

IFNG

8

Plasma Membrane KIR2DS2

7

6 Cytoplasm

Calmodulin

BASP1

9 ACO2

NEFH 10

13 11

ACTB

12

NEFL

14

15

INA

PHGDH

RPL23A

17

16

PSME3

18

SSBP1

GFAP

GSTM2

YWHAZ

19

21

20

GSTA2

22 HSPD1

23 TPI1 26

BCKDHB

YWHAE

24 RPL19

SLC8A2

27

29 28

30

RPS3 TUBA1B

32

35

Nucleus

33

25

HSPE1

SYNPO2

31

HIST2H2BE

Cof lin

MYCN

14-3-3 (β,ε,ζ)

34

Fig. 2  Resulting pathways from Ingenuity analysis obtained with the proteins appearing in at least three of the published studies; proteins in grey are part of the experimental data set. Protein abbreviations: 1 MBP myelin basic protein; 2 IL1F9 interleukin 1 family, member 9; 3 IFNG interferon gamma; 4 DSPP dentin phosphophoryn; 5 Jnk Jnk protein; 6 KIR2DS2 killer cell immunoglobulin-like receptor; 7 BASP1 brain abundant membrane attached signal protein 1; 8 Calmodulin Calmodulin protein; 9 ACO2 aconitate Hydratase 2; 10 NEFH neurofilament h; 11 NEFL Neurofilament Triplet L; 12 INA alpha internexin; 13 ACTB actin beta; 14 PHGDH phosphoglycerate dehydrogenase; 15 RPL19 ribosomal protein L19; 16 PSME3 proteasome activator complex subunit 3; 17 SSBP1 single-stranded DNA binding protein 1; 18 GFAP glial fibrillary acidic protein; 19 GSTM2 glutathione S-transferase M2; 20 YWHAZ 14-3-3-zeta; 21 GSTA2 glutathione S-transferase A2; 22 HSPD1 heat shock 60kDa protein 1; 23 TPI1 triosephosphate isomerase 1; 24 BCKDHB branched chain keto acid dehydrogenase E1; 25 Cofilin cofilin protein; 26 YWHAE 14-3-3 epsilon; 27 SLC8A2 solute carrier family 8 (sodium-calcium exchanger); 28 SYNPO2 synaptopodin 2; 29 RPL19 ribosomal protein L19; 30 HSPE1 heat shock 10kDa protein 1 (chaperonin 10); 31 RPS3 ribosomal protein S3; 32 TUBA1B tubulin alpha 1b; 33 HIST2H2BE histone cluster 2; 34 MYCN N-myc proto-oncogene protein; 35 14-33 (b, e, z) 14-3-3 (beta, epsilon, zeta)

436

L. Carboni et al.

By relaxing the inclusion criteria to proteins appearing in at least two studies, the most significant network observed is characterised by proteins belonging to the following functional categories: Cellular Assembly and Organisation, Cellular Function and Maintenance, and Nervous System Development and Function, with 21 out of 35 proteins derived from the experimental dataset (Fig. 3). Interestingly, Extracellular Space

Unknown

3

LOC299282

PP2A4

MBP 2

1

5 Calcineurin protein(s)

IGF 1 Plasma Membrane

7

SYN2

AMPH

6 11

Cytoplasm

Rock 8

DPYSL2 Actin

9

10 Calmodulin Pkc(s) 12

YWHAE 14 DNM1 13

Ras

15 GFAP

16

18

17 Cofilin

YWHAZ

INA

19 NEFL

20 22

21 HSPD1

MeK1/2 23

PKM2

PCTK3

Rsk

24

25 26

NEFM 27

HSPE1

HSP 28 GOT 29

GMFB 30

31 CKB

PEBP1 32

NFkB

33

GOT1 34 Nucleus

Creatine Kinase 35

Fig. 3  Resulting pathways from Ingenuity analysis obtained with the proteins appearing in at least two of the published studies; proteins in grey are part of the experimental data set. Protein abbreviation: 1 LOC299282 serine protease inhibitor; 2 IGF1 insulin-like growth factor 1; 3 MBP myelin basic protein; 4 PP2A protein phosphatase 2a; 5 Calcineurin proteins calcineurin proteins; 6 SYN2 synapsin II; 7 AMPH amphiphysin; 8 Actin actin; 9 DPYSL2 dihydropyrimidinase-like 2; 10 calmodulin calmodulin protein; 11 Rock Rho associated kinase; 12 Pkc(s) protein kinase c; 13 DNM1 dynamin 1; 14 YWHAE 14-3-3 epsilon; 15 GFAP glial fibrillary acidic protein; 16 Ras p21 Ras; 17 Cofilin cofilin protein; 18 INA alpha internexin; 19 YWHAZ 14-3-3-zeta; 20 NEFL neurofilament triplet L; 21 PCTK3 PCTAIRE protein kinase 3; 22 PKM2 pyruvate kinase M; 23 Mek1/2 Mkk½; 24 Rsk P90-RSK; 25 HSPD1 heat shock 60 kDa protein 1; 26 HSPE1 heat shock 10 kDa protein 1 (chaperonin 10); 27 NEFM neurofilament triplet m; 28 HSP heat shock protein; 29 GOT 2-oxoglutarate-glutamate aminotransferase; 30 GMFB glia maturation factor; 31 CKB creatine kinase b chain; 32 PEBP1 phosphatidylethanolamine binding protein 1; 33 NFkB nuclear factor NF-kappa-B; 34 GOT1 glutamate oxaloacetate transaminase 1; 35 Creatine kinase creatin kinase protein

Proteome Effects of Antidepressant Medications

437

the network is still characterised by the presence of 14-3-3 protein and ­cytoskeleton/ neurofilament proteins, including IGF-1 as an additional key member of the network. IGF-1, here displaying the highest number of direct or functional interactions with the network members, has been suggested as one of the modulators of the long-term effects associated to antidepressant treatment, in particular neurogenesis (Manev & Manev, 2001). Synthesis and release of IGF-1 has been shown to be stimulated by serotonergic mechanisms, and recently IGF-1 itself has been shown to display anxiolytic and antidepressant-like effects in rodent models either upon direct administration or indirect modulation of its levels (Hoshaw, Malberg, & Lucki, 2005; Malberg et al., 2007). In conclusion, the available data confirm that through analysis of large datasets from unbiased proteomic approaches it is possible to identify and confirm a series of putative modulators of the therapeutic effects of antidepressants. Proteomic profiling of antidepressants appears a promising approach to generate valid hypotheses and candidate markers to be assessed in follow-up investigations for their utility as biomarkers or as starting points for new antidepressant therapies.

References American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders, text revision (4th ed.). Washington: American Psychiatric Association. Baik, S. Y., Jung, K. H., Choi, M. R., Yang, B. H., Kim, S. H., Lee, J. S., et al. (2005). Fluoxetineinduced up-regulation of 14-3-3zeta and tryptophan hydroxylase levels in RBL-2H3 cells. Neuroscience Letters, 374, 53–57. Baldessarini, R. J. (2005). Drug therapy of depression and anxiety disorders. In: Laurence L. Brunton, John S. Lazo & Keith L. Parker (Eds.), Goodman & Gilman’s the pharmacological basis of therapeutics (11th ed) McGraw-Hill Professional Publishing. Basheer, R., Brown, R., Ramesh, V., Begum, S., & McCarley, R. W. (2005). Sleep deprivationinduced protein changes in basal forebrain: implications for synaptic plasticity. Journal of Neuroscience Research, 82, 650–658. Beck, A. T. (2005). The current state of cognitive therapy: A 40-year retrospective. Archives of General Psychiatry, 62, 953–959. Belmaker, R. H., & Agam, G. (2008). Major depressive disorder. The New England Journal of Medicine, 358, 55–68. Benedetti, F., Barbini, B., Colombo, C., & Smeraldi, E. (2007). Chronotherapeutics in a psychiatric ward. Sleep Medicine Reviews, 11, 509–522. Bianchi, M., Hagan, J. J., & Heidbreder, C. A. (2005). Neuronal plasticity, stress and depression: involvement of the cytoskeletal microtubular system? Current Drug Targets. CNS and Neurological Disorders, 4, 597–611. Bisgaard, C. F., Jayatissa, M. N., Enghild, J. J., Sanchez, C., Artemychyn, R., & Wiborg, O. (2007). Proteomic investigation of the ventral rat hippocampus links DRP-2 to escitalopram treatment resistance and SNAP to stress resilience in the chronic mild stress model of depression. Journal of Molecular Neuroscience, 32, 132–144. Bjellqvist, B., Ek, K., Righetti, P. G., Gianazza, E., Görg, A., Westermeier, R., et  al. (1982). Isoelectric focusing in immobilized pH gradients: principle, methodology and some applications. Journal of Biochemical and Biophysical Methods, 6, 317–339. Carboni, L., Piubelli, C., El Khoury, A., Gruber, S., Andersson, W., Mathé, A. A., & Domenici, E. (2007). Escitalopram treatment in the FSL rat model of depression with gene-environment interaction: Proteomic analysis of hippocampus and frontal/prefrontal cortex. Program No. 907.14 Neuroscience Meeting Planner. San Diego, CA: Society for Neuroscience.

438

L. Carboni et al.

Carboni, L., Piubelli, C., Vighini, M., El Khoury, A., Gruber, S., Anderssen, W., et  al. (2006). Proteomic analysis of hippocampus and frontal cortex in a rat model of depression with geneenvironment interaction and antidepressant treatment. European Neuropsychopharmacology, 16S4, S336–S337. Carboni, L., Vighini, M., Piubelli, C., Castelletti, L., Milli, A., & Domenici, E. (2006). Proteomic analysis of rat hippocampus and frontal cortex after chronic treatment with fluoxetine or ­putative novel antidepressants: CRF1 and NK1 receptor antagonists. European Neuropsychopharmacology, 16, 521–537. Cecconi, D., Mion, S., Astner, H., Domenici, E., Righetti, P.G., & Carboni, L. (2007). Proteomic analysis of rat cortical neurons after fluoxetine treatment. Brain Research, 1135, 41–51. D’Sa, C., & Duman, R. S. (2002). Antidepressants and neuroplasticity. Bipolar Disorders, 4, 183–194. Ding, Q., Vaynman, S., Souda, P., Whitelegge, J. P., & Gomez-Pinilla, F. (2006). Exercise affects energy metabolism and neural plasticity-related proteins in the hippocampus as revealed by proteomic analysis. The European Journal of Neuroscience, 24, 1265–1276. Fountoulakis, M. (2004). Application of proteomics technologies in the investigation of the brain. Mass Spectrometry Reviews, 23, 231–258. Giambelli, R., Vollmayr, B., Barbiero, V. S., Mallei, A., Dormann, C., Gass, P., Racagni, G., & Popoli, M. (2007). Proteomic analysis of a behavioral animal model of depression after antidepressant treatment. National Congress of the Italian Society of Neuroscience, Verona, Italy, 27–30 September 2007. Giedke, H., & Schwärzler, F. (2002). Therapeutic use of sleep deprivation in depression. Sleep Medicine Reviews, 6, 361–377. Goldman, D., Merril, C.R., Polinsky, R.J., & Ebert, M.H. (1982). Lymphocyte proteins in Huntington’s disease: quantitative analysis by use of two-dimensional electrophoresis and computerized densitometry. Clinical Chemistry, 28, 1021–1025. Görg, A., Weiss, W., & Dunn, M. J. (2004). Current two-dimensional electrophoresis technology for proteomics. Proteomics, 4, 3665–3685. Govorun, V. M., & Archakov, A. I. (2002). Proteomic technologies in modern biomedical science. Biochemistry, 67, 1109–1123. Grove, H., Hollung, K., Uhlen, A. K., Martens, H., & Faergestad, E. M. (2006). Challenges related to analysis of protein spot volumes from two-dimensional gel electrophoresis as revealed by replicate gels. Journal of Proteome Research, 5, 3399–3410. Guest, P. C., Knowles, M. R., Molon-Noblot, S., Salim, K., Smith, D., Murray, F., et al. (2004). Mechanisms of action of the antidepressants fluoxetine and the substance P antagonist L-000760735 are associated with altered neurofilaments and synaptic remodeling. Brain Research, 1002, 1–10. Hanash, S. (2003). Disease proteomics. Nature, 422, 226–232. Heim, C., & Nemeroff, C. B. (2001). The role of childhood trauma in the neurobiology of mood and anxiety disorders: preclinical and clinical studies. Biological Psychiatry, 49, 1023–1039. Henn, F. A., & Vollmayr, B. (2005). Stress models of depression: Forming genetically vulnerable strains. Neuroscience and Biobehavioral Reviews, 29, 799–804. Herbert, B., Galvani, M., Hamdan, M., Olivieri, E., MacCarthy, J., Pedersen, S., et  al. (2001). Reduction and alkylation of proteins in preparation of two-dimensional map analysis: Why, when, and how? Electrophoresis, 22, 2046–2057. Heydorn, W. E., Creed, G. J., & Jacobowitz, D. M. (1984). Effect of desmethylimipramine and reserpine on the concentration of specific proteins in the parietal cortex and the hippocampus of rats as analyzed by two-dimensional gel electrophoresis. The Journal of Pharmacology and Experimental Therapeutics, 229, 622–628. Holtzheimer, P. E., III, & Nemeroff, C. B. (2006). Emerging treatments for depression. Expert Opinion on Pharmacotherapy, 7, 2323–2339. Hoshaw, B. A., Malberg, J. E., & Lucki, I. (2005). Central administration of IGF-I and BDNF leads to long-lasting antidepressant-like effects. Brain Research, 1037, 204–208.

Proteome Effects of Antidepressant Medications

439

Karp, N. A., Griffin, J. L., & Lilley, K. S. (2005). Application of partial least squares discriminant analysis to two-dimensional difference gel studies in expression proteomics. Proteomics, 5, 81–90. Keller, M. B. (2005). Issues in treatment-resistant depression. The Journal of Clinical Psychiatry, 66(suppl 8), 5–12. Kendler, K. S., Karkowski, L. M., & Prescott, C. A. (1999). Causal relationship between stressful life events and the onset of major depression. The American Journal of Psychiatry, 156, 837–841. Kessler, R. C., Berglund, P., Demler, O., Jin, R., Koretz, D., Merikangas, K. R., et al. (2003). The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R). JAMA, 289, 3095–3105. Khawaja, X., Xu, J., Liang, J. J., & Barrett, J. E. (2004). Proteomic analysis of protein changes developing in rat hippocampus after chronic antidepressant treatment: Implications for depressive disorders and future therapies. Journal of Neuroscience Research, 75, 451–460. Kjarland, E., Keen, T. J., & Kleppe, R. (2006). Does isoform diversity explain functional differences in the 14-3-3 protein family? Current Pharmaceutical Biotechnology, 7, 217–223. Kubera, M., Kenis, G., Budziszewska, B., Bosmans, E., Scharpe, S., Basta-Kaim, A., et al. (2001). Lack of a modulatory effect of imipramine on glucocorticoid-induced suppression of interferon-gamma and interleukin-10 production in  vitro. Polish Journal of Pharmacology, 53, 289–294. Lawlor, D. A., & Hopker, S. W. (2001). The effectiveness of exercise as an intervention in the management of depression: Systematic review and meta-regression analysis of randomised controlled trials. BMJ, 322, 763–767. Malberg, J. E., & Blendy, J. A. (2005). Antidepressant action: to the nucleus and beyond. Trends in Pharmacological Sciences, 26, 631–638. Malberg, J. E., Eisch, A. J., Nestler, E. J., & Duman, R. S. (2000). Chronic antidepressant ­treatment increases neurogenesis in adult rat hippocampus. The Journal of Neuroscience, 20, 9104–9110. Malberg, J. E., Platt, B., Rizzo, S. J., Ring, R. H., Lucki, I., Schechter, L. E., et  al. (2007). Increasing the levels of insulin-like growth factor-I by an IGF binding protein inhibitor produces anxiolytic and antidepressant-like effects. Neuropsychopharmacology, 32, 2360–2368. Mallei, A., Giambelli, R., Barbiero, V. S., El Khoury, A., Gruber, S., Andersson, W., Mathé, A. A., Aitchison, K., McGuffin, P., Racagni, G., & Popoli, M. (2006). Global analysis of the synaptoproteome in a rat model of depression with gene-environment interaction. Program No. 291.13 Neuroscience Meeting Planner. Atlanta, GA: Society for Neuroscience. Mallei, A., Giambelli, R., Barbiero, V. S., Mathé, A. A., El Khoury, A., Carboni, L., et al. (2006). Functional proteomic analysis of an animal model of depression combining genetic vulnerability and environmental stress. European Neuropsychopharmacology, 16S4, S169–S170. Manev, H., & Manev, R. (2001). S100B: an old neurotrophic factor with putative new roles in psychiatric illnesses. Journal of Psychiatric Research, 35, 347–350. Manji, H. K., Drevets, W. C., & Charney, D. S. (2001). The cellular neurobiology of depression. Natural Medicines, 7, 541–547. McHugh, P. C., Rogers, G. R., Loudon, B., Glubb, D. M., Joyce, P. R., & Kennedy, M. A. (2008). Proteomic analysis of embryonic stem cell-derived neural cells exposed to the antidepressant paroxetine. Journal of Neuroscience Research, 86, 306–316. Miklos, G. L., & Maleszka, R. (2001). Integrating molecular medicine with functional proteomics: realities and expectations. Proteomics, 1, 30–41. Nemeroff, C. B., & Owens, M. J. (2002). Treatment of mood disorders. Nature Neuroscience, 5(suppl), 1068–1070. Nestler, E. J., Barrot, M., DiLeone, R. J., Eisch, A. J., Gold, S. J., & Monteggia, L. M. (2002). Neurobiology of depression. Neuron, 34, 13–25. O’Farrell, P. H. (1975). High resolution two-dimensional electrophoresis of proteins. The Journal of Biological Chemistry, 250, 4007–4021.

440

L. Carboni et al.

Overstreet, D. H., Friedman, E., Mathé, A. A., & Yadid, G. (2005). The Flinders Sensitive Line rat: a selectively bred putative animal model of depression. Neuroscience and Biobehavioral Reviews, 29, 739–759. Padberg, F., & Moller, H. J. (2003). Repetitive transcranial magnetic stimulation: Does it have potential in the treatment of depression? CNS Drugs, 17, 383–403. Park, M. C., Goldman, M. A., Carpenter, L. L., Price, L. H., & Friehs, G. M. (2007). Vagus nerve stimulation for depression: rationale, anatomical and physiological basis of efficacy and future prospects. Acta Neurochirurgica. Supplementum, 97, 407–416. Patterson, S. D., & Aebersold, R. H. (2003). Proteomics: the first decade and beyond. Nature Genetics, 33, 311–323. Paulson, L. M., Klint, T., Brange, S., Sihlbom, C., & Erikssom, P.M. (2007). Proteomic studies of induced adult neurogenesis. 27th Annual Meeting of the Australian Neuroscience Society held jointly with the International Brain Research Organisation (IBRO) 7th World Conference of Neuroscience, Melbourne, Australia, 12–17 July 2007. Paulson, L., Klint, T., Sihlbom, C., & Eriksson, P. (2006). Proteomic studies of fluoxetine induced adult neurogenesis in hippocampus compared with cerebellum. Fifth Annual World Congress of the Human Proteome Organization (HUPO), Long Beach, CA, 28 October–1 November 2006. Piubelli, C., Carboni, L., Vighini, M., Gruber, S., Andersson, W., El Koury, A., et  al. (2006).  The Flinders rat model of depression: Gene-environment interaction and antidepressant  effects by proteomic analysis. American Journal of Medical Genetics, 141B, 782. Piubelli C., De Luca M., Gruber S., Andersson W., El Khoury A., Mathé A. A., Domenici E., & Carboni L. (2007). The Flinders rat model of depression: gene-environment interaction and nortriptyline effects by proteomic analysis. Italian Society of Neuroscience, Verona, Italy, 27–30 September 2007. Righetti, P. G., Campostrini, N., Pascali, J., Hamdan, M., & Astner, H. (2004). Quantitative proteomics: a review of different methodologies. European Journal of Mass Spectrometry, 10, 335–348. Rogers, G. R., Allington, M. D., McHugh, P. C., Joyce, P. R., & Kennedy, M. A. (2004). Combined proteomic/microarray approach to identify novel candidate genes for individual variation in response to antidepressants. International Congress of the World Federation of Biological Psychiatry, Sydney, Australia, 9–13 February 2004. Sahay, A., & Hen, R. (2007). Adult hippocampal neurogenesis in depression. Nature Neuroscience, 10, 1110–1115. Schiepers, O. J., Wichers, M. C., & Maes, M. (2005). Cytokines and major depression. Progress in Neuropsychopharmacology & Biological Psychiatry, 29, 201–217. Schloss, P., & Henn, F. A. (2004). New insights into the mechanisms of antidepressant therapy. Pharmacology & Therapeutics, 102, 47–60. Sills, M. A., Heydorn, W. E., Cohen, R. M., Creed, G. J., & Jacobowitz, D. M. (1986). Effect of chronic treatment with clorgyline on the relative concentration of specific proteins in the ­hippocampus and parietal cortex of the rat. Neuropharmacology, 25, 143–150. Spigset, O., & Mårtensson, B. (1999). Fortnightly review: drug treatment of depression. BMJ, 318, 1188–1191. Stead, D. A., Preece, A., & Brown, A. J. (2006). Universal metrics for quality assessment of protein identifications by mass spectrometry. Molecular & Cellular Proteomics, 5, 1205–1211. Stout, S. C., Owens, M. J., & Nemeroff, C. B. (2001). Neurokinin 1 receptor antagonists as potential antidepressants. Annual Review of Pharmacology and Toxicology, 41, 877–906. Sun, L., Bittner, M. A., & Holz, R. W. (2003). Rim, a component of the presynaptic active zone and modulator of exocytosis, binds 14-3-3 through its N terminus. The Journal of Biological Chemistry, 278, 38301–38309. Tannu, N. S., & Hemby, S. E. (2006). Methods for proteomics in neuroscience. Progress in Brain Research, 158, 41–82.

Proteome Effects of Antidepressant Medications

441

Uda, M., Ishido, M., Kami, K., & Masuhara, M. (2006). Effects of chronic treadmill running on neurogenesis in the dentate gyrus of the hippocampus of adult rat. Brain Research, 1104, 64–72. UK ECT Review Group. (2003). Efficacy and safety of electroconvulsive therapy in depressive disorders: a systematic review and meta-analysis. Lancet, 361, 799–808. Wong, M. L., & Licinio, J. (2001). Research and treatment approaches to depression. Nature Reviews. Neuroscience, 2, 343–351. Wong, M. L., & Licinio, J. (2004). From monoamines to genomic targets: a paradigm shift for drug discovery in depression. Nature Reviews. Drug Discovery, 3, 136–151.

Part V

Neurological Disorders

MicroRNAs in Neurodegenerative Disorders Catherine L. Clelland and James D. Clelland

Abstract  MicroRNAs (miRNAs) are a class of small RNA molecules that are involved in cellular posttranscriptional regulation. In this review, we describe some key features of miRNA-mediated gene regulation in the central nervous system, and the recent evidence suggesting a role for miRNA dysregulation in the pathogenesis of neurodegenerative disorders such as Alzheimer’s disease, Parkinson’s disease and Huntington’s disease. Finally, we comment on the potential for miRNA-based therapeutic strategies for treatment of these disorders. Keywords  miRNA • Posttranscriptional regulation • Neurodegenerative disorders • Synapse • Molecular genetics

1 Introduction Our knowledge of mammalian gene expression mechanisms has grown enormously over the past decade. From a basic DNA to RNA to protein model, with regulation at DNA promoter or enhancer sites, we now recognize a large and growing family of interconnecting and overlapping miRNA molecules, that influence the ­expression of genes and the translation of mRNA transcripts into proteins (Zamore & Haley, 2005). The discovery of miRNA species and their roles in biological processes has also opened up possibilities for the elucidation of new disease mechanisms, as every biological control system has the potential to malfunction or to function inappropriately. Therefore, there are now large-scale efforts to determine the involvement of the new

C.L. Clelland (*) Department of Pathology and Cell Biology and Taub Institute for Research on Alzheimer’s Disease and the Aging Brain, Columbia University Medical Center, 630 West 168th Street, New York, NY 10032, USA e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_16, © Springer Science+Business Media, LLC 2011

445

446

C.L. Clelland and J.D. Clelland

“small RNA” regulatory mechanisms, in diseases as diverse as cancers and psychiatric and neurological disorders (Hammond, 2006; Perkins, Jeffries, & Sullivan, 2005).

2 MicroRNAs (miRNAs): A Widespread Class of Small RNA Molecules that Regulate Protein Expression miRNAs are noncoding ~19–25-nt single-stranded RNAs that are enzymatically cleaved from ~70-nt pre-miRNAs, and that have been shown to interact with target mRNAs via partial hybridization to the 3¢ noncoding regions of target mRNAs, and to regulate cellular gene expression via mRNA degradation and/or transcription and translation repression (Bartel, 2004; Makeyev & Maniatis, 2008; Vasudevan, Tong, & Steitz, 2007; Zamore & Haley, 2005). By 2004, ~200 mammalian miRNAs had been discovered. The estimated number of mammalian miRNA genes has now risen to tens of thousands (reviewed in Landgraf et al., 2007), highlighting the extensive research that is ongoing in this field. For example, one group recently reported data from a large-scale sequencing approach of 250 small RNA libraries from 26 different organs/cell types from both humans and rodents. Their mammalian miRNA “expression atlas” consisted of 340, 303, and 205 distinct mature miRNA species from human, mouse, and rat, respectively (Landgraf et al., 2007). Homology between mammalian miRNAs has also been observed. For example, the sequences of over 140 miRNAs are identical between mice and humans (Bartel, 2004; Weber, 2005; Zamore & Haley, 2005).

2.1 Overview of the Mammalian miRNA Expression Pathway (see Fig. 1 and review in Winter, Jung, Keller, Gregory, & Diederichs (2009)) 1. Pri-miRNA transcription and cleavage to pre-miRNAs – Pri-miRNAs, consisting on average of a 33-base pair (bp) hairpin stem, a terminal loop, and two singlestranded regions flanking the hairpin stem, are transcribed from genomic DNA via RNA Polymerase II/III-directed transcription. Also taking place inside the nucleus, Pri-miRNAs are cleaved by the Drosha/DGCR8 microprocessor complex, generating the “precursor miRNA” (pre-miRNA): ~70 nt in length, maintaining approximately 25–30 bp of the Pri-miRNA stem and the terminal loop. 2. Pre-miRNA export and processing, and mature miRNA strand selection – An Exportin-5 protein complex recognizes the pre-miRNAs, binds to them and exports them from the nucleus to the cytoplasm, where the RNase Dicer/Tar RNA binding protein (TARB) complex cleaves away the pre-miRNA terminal loop. This leaves an approximately 22-nt, double-stranded RNA molecule. The functional strand is designated the mature miRNA and is loaded, along with the Argonaute (Ago2) protein, into the RNA induced silencing complex (RISC, see below). The other, nonfunctional miRNA* strand is degraded.

MicroRNAs in Neurodegenerative Disorders

3

Cytoplasm

Pri-miRNA

Transcription 5

447

,

Nucleus

,

Cleavage Drosha DGCR8

5 3

,

,

Pre-miRNA 5

Exportin-5 Induced Nuclear Export

3

,

,

Cleavage Dicer 5

3

,

,

3

∗

5

TRBP

,

,

Mature miRNA 5

,

3

, 3

,

∗

5

,

Degradation

Mature miRNA within RISC 3

,

RISC

5

,

Fig. 1  Overview of the mammalian miRNA expression and translation repression pathway. See text for details. Adapted from Winter et al. (2009)

2.2 Mechanisms of miRNA Posttranscriptional Regulation miRNAs often hybridize imperfectly to the 3¢UTR of target mRNA molecules. Binding of the miRNA appears to initiate via the “seed sequence,” that consists of ~2–8 nt in the miRNA molecule that pairs with the 3¢UTR of the target mRNA (Bartel, 2004; Lim et al., 2005). Depending upon the degree of complementarity of the miRNA with its target, it is thought that the miRNA RISC complex silences target mRNAs via triggering of mRNA degradation, inhibiting translation elongation, or deadenylation. However, the mechanisms of miRNA-mediated translational repression are complex and not yet well understood (Kosik & Krichevsky, 2005; Zamore & Haley, 2005). For example, studies have recently shown that mRNA targets are also

448

C.L. Clelland and J.D. Clelland

efficiently repressed via 5¢UTR binding sites (Lytle, Yario, & Steitz, 2007), and adding to the complexity of miRNA regulation, it has also been proposed that some miRNAs may function via transcription up-regulation (Vasudevan et al., 2007). Of interest, two landmark 2008 studies, both of which employed a joint mass spectrometry/mRNA microarray profiling approach, have shown that the widespread changes in protein synthesis observed in response to miRNA transfection in vitro were strongly correlated with mRNA levels (Baek et al., 2008; Selbach et al., 2008). For example, Baek et al. 2008 reported a significant correlation between regulated protein and mRNA levels (r2 = 0.63, p < 10−12), and in addition that all proteins with changes of 50% or more had a detectable change in message level. This work suggests that the effects of miRNA regulation can be detected at multple levels in the cell.

3 Regulatory Roles of miRNAs in Synapse Formation miRNAs appear to play roles in regulating cell division, differentiation and homeostasis, which require close regulation of protein expression (Kosik & Krichevsky, 2005; Miska, 2005; Zamore & Haley, 2005). In model systems, disruption of miRNA function can induce multisystem abnormalities, indicating the biological importance of miRNA regulation, and data from recent studies has begun to illustrate the very specific pattern of spatial and temporal miRNA regulation during neuronal differentiation and development (Kim et al., 2004; Kosik & Krichevsky, 2005). In the developing neuron, axons and dendrites extend from the neuronal cell body. Dendrites receive their inputs from neuronal axon terminals, and as the dendritic spines form contacts with axon terminals from other neurons, synapses are created. Recently, and using a combination of knockout murine models and cell culture assays, an essential role for miRNAs in regulating synapse formation was identified, suggesting that miRNAs may also be involved in learning and memory (Schratt et al., 2006; Siegel et al., 2009; Wayman et al., 2008). Specifically, Schratt et al. over-expressed a brain-specific miRNA (miR-134) in cultured rat neurons and observed significantly decreased dendritic spine size. The researchers searched for possible targets for miR-134 and identified one potential target, Limk1, whose protein product is involved in building dendritic spines. Of significance, Limk1 knockout mice showed abnormalities in dendritic spine structure that were very similar to the effects of miR-134 over-expression (Schratt et al., 2006). Conversely, introduction of the miRNA miR-132 into hippocampal neurons was shown to enhance dendrite morphogenesis, via inhibited translation of the p250GAP protein (Wayman et al., 2008). p250GAP regulates several Rho family GTPases including Rac, and it has been suggested that the miRNAs miR-134 and miR-132 antagonistically regulate spine growth by activation of the Rac signaling pathway (reviewed in Schratt, 2009a, 2009b). miR-124, a miRNA that is abundantly expressed in the CNS and has a specific pattern of expression in the differentiating neuron (Kosik & Krichevsky, 2005), was also shown to positively regulate neurite outgrowth via Rac activity (Yu, Chung, Deo, Thompson, & Turner, 2008). More recently, miR-138

MicroRNAs in Neurodegenerative Disorders

449

miR-138 5 ,

,

3 ,

5

5

,

,

3

3

,

APT1

RhoA activity

Spine Shrinkage

miR-132

p250

Rac activity

5

,

5

,

5

Spine Growth

LimK1 5

miR-134 ,

, 5

,

3

,5

3

3

3 ,

,

3

,

3

,

,

,

Fig. 2  Dynamic regulation of dendritic spine morphogenesis. In the neuron, spine size is controlled by the dynamic relationship between spine shrinkage and spine growth. Left miR-138 negatively regulates expression of the APT1 protein, which increases the palmitoylation state of its G-protein target, induces RhoA activity and results in spine shrinkage. Spine morphogeneis is achieved (in part) via homoestatic spine growth antagonistically regulated by the miRNAs-132 and 134 (right). Specifically, miR-132 inhibits translation of its target the Rac inactivating protein p250GAP(p250). Rac activation results in spine growth, possibly through a Limk1 mediated ­pathway, which itself is inhibited via miR-134 (as Limk1 is a direct target of miR-134 repression). Activated/up-regulated targets (arrows pointing upwards), inactivated/down-regulated targets (arrows pointing downwards). Adapted from Schratt (2009a)

was identified in an enrichment microarray screen as a synaptodendritic miRNA (Siegel et al., 2009). Functional studies using miR-138 and anti-miR-138 oligonucleotides showed that this miRNA was involved in the negative regulation of ­hippocampal neuron dendritic spine size, and that loss of acyl-protein thioesterase 1 (APT1), the mRNA target of miR-138, resulted in a suppression of spine ­enlargement (Siegel et  al., 2009). The dynamic balance between dendritic spine shrinkage and growth that is regulated by miRNAs, is summarized in Fig. 2.

4 A Role for miRNA Dysregulation in Neurodegenerative Disorders 4.1 Neurodegeneration as a Consequence of Deficits in miRNA Processing A number of important studies have begun to suggest an etiological role of ­miRNAs in neurodegeneration. The first class of studies have investigated CNS phenotypes that arise due to miRNA processing disruption. For instance, in Drosophila, Bilen et  al., demonstrated that loss of global miRNA function due to loss of the Drosophila dicer-1 gene in the eye, enhanced the toxicity of Ataxin-3, the pathogenic protein that causes the human polyQ disease spinocerebellar ataxia type 3 (Bilen, Liu, Burnett, Pittman, & Bonini, 2006). Of particular significance, this study also showed that the severe ­neurodegeneration observed in flies expressing human tau [the microtubule-associated protein found in the pathogenic neurofibrillary tangles observed in the brains of patients

450

C.L. Clelland and J.D. Clelland

with tauopathies such as Alzheimer’s disease (AD) and frontotemporal dementia (FTD)], was dramatically increased in Drosophila with reduced RD3-1 activity (a dsRNA-binding protein needed for Dicer-regulated processing of pre-miRNAs to mature forms), and that up-regulation of the Drosophila miRNA ban suppresses this neurodegeneration (Bilen, Liu, & Bonini, 2006; Bilen, Liu, Burnett, et  al., 2006) ­suggesting a protective role for this miRNA. Similarly, inactivation of Dicer in a murine Purkinje cell model system lead to progressive loss of miRNAs and a concurrent ­cerebellar degeneration and development of ataxia (Schaefer et al., 2007). A noteworthy study from 2008 documented multiple behavioral and cognitive ­deficits in a mammalian model of altered miRNA biogenesis (Stark et al., 2008). DGCR8, which as described above plays an important role in miRNA processing (Fig. 1), maps to the DiGeorge Critical region on human chromosome 22q11 (hemizygocity for which has been associated with an increased risk of schizophrenia), and the syntenic region on murine chromosome 16 (Sutherland, Kim, & Scambler, 1998). Stark et al., created a mouse model hemizygous for 1.5 kb of chromosome 16 (which included DGCR8). Transgenic mice had altered pre-pulse inhibition and spatial-working memory (murine behaviors considered to model human schizophrenia), plus exhibited abnormal spine morphogenesis. Moreover, these animals had a significant loss of mature miRNAs that included miR-134, coupled with upregulation of their corresponding pri-forms (Stark et al., 2008), suggesting that loss of miRNA processing contributes to the cognitive phenotype in the 22q11-deletion mouse model. Following analysis of an additional murine model, a conditional Dicer knockout, Kim et  al. 2007 showed that loss of dicer exclusively in dopamine neurons (and thus mature miRNAs in these neurons) resulted in a phenotype resembling Parkinson’s disease (PD). Kim et al. 2007 then identified a candidate miRNA, miR133b, that they consider contributes to the Parkinson’s phenotype in these mice, and they confirmed that miR-133b, which is specifically expressed in midbrain ­dopamine neurons, was found to be significantly down-regulated in human PD brains when compared to control brain tissue.

4.2 miRNA Dysregulation in Human Neurodegenerative Disorders A second class of studies have compared the brains of patients with neurodegenerative disease to control brains, and these studies have identified miRNA that are ­dysregulated and thus may have a role in neurodegeneration. For example, miRNA loss was ­identified in both a mouse model of Huntington’s disease (HD), as well as in the brains from deceased Huntington’s patients (Johnson et al., 2008), which the authors suggest may reflect increased transcriptional repression of the Pri-miRNAs by REST, the transcriptional repressor found at high levels in the nucleus of HD neurons. Dysregulation of miRNA in the brains of patients with AD has also been reported (Hebert et al., 2008; Lukiw, 2007; Wang, Rajeev, et al., 2008), and an etiological role

MicroRNAs in Neurodegenerative Disorders

451

for miRNA dysregulation in the AD pathology has been suggested (Almeida, Bolanos, & Moreno, 2005; Hebert et  al., 2008). In the first study to investigate miRNAs in AD dementia, 13 individual miRNAs were spotted onto nylon membranes and hybridized to RNA extracted from the hippocampal CA1 brain region from five AD patients and adult controls, and their levels measured (Lukiw, 2007). Three ­miRNAs were found to be significantly up-regulated in the AD brains, miR-9, miR125b, and miR-128, and Carrettiero, Hernandez, Neveu, Papagiannakopoulos, & Kosik (2009) have suggested that upregulation of miR-128 results in decreased levels of BAG2, a cochaperone protein that helps target ubiquitin-independent degradation of pathogenic tau. Studies by Wang, Rajeev, et al. (2008) and Hebert et al. (2008) both used microarray analysis to identify two miRNAs that were down-regulated in AD brains, were computationally predicted to regulate BACE1 mRNA, and had inverse correlations with levels of BACE1 protein in the AD brain tissue. Validation by dual luciferase assay confirmed that miR-107 and mir-29a/b-1 both regulate BACE1 (the endopeptidase that cleaves the b-amyloid precursor protein, generating the neurotoxic b-amyloid peptide (Ab)), and may suggest that loss of miRNAs can contribute to increased BACE1 protein and Ab levels in AD (Hebert et al., 2008).

4.3 Loss of miRNA Target Gene Binding Is Associated with Neurodegeneration Two interesting papers have also reported that the presence of genomic variations in the 3¢UTR miRNA binding sites within their target genes, confers risk for developing a neurodegenerative disorder. Specifically, Wang, van der Walt, et al. (2008) showed that a polymorphism in the 3¢UTR of the fibroblast growth factor 20 (FGF20), which is targeted by miR-433, significantly increased the risk of developing PD, and the authors suggested that this may result from increased a-synuclein expression due to loss of miR-433 binding and FGF20 repression. Likewise, a 3¢UTR variant in ­progranulin (GRN) located in the miR-659 binding-site was identified as a major susceptibility factor for developing TDP43 positive FTD (Rademakers et al., 2008).

5 Conclusions and Future Perspectives Since their initial discovery, many landmark papers have been published showing the crucial role of miRNAs in regulating many biological systems, and now ­important studies have provided data to illustrate the critical balance of miRNAs and their regulated mRNAs in synaptic development and plasticity (Schratt et al., 2006; Siegel et  al., 2009; Wayman et  al., 2008). Furthermore, the enhanced neuronal degeneration observed in Dicer knockout models in both Drosophila and mouse (Bilen, Liu, & Bonini, 2006; Bilen, Liu, Burnett, et al., 2006; Kim et al., 2007), the dysregulation of miRNA observed in studies of human diseased brain tissue, (Hebert

452

C.L. Clelland and J.D. Clelland

et  al., 2008; Johnson et  al., 2008; Lukiw, 2007; Wang, Rajeev, et  al., 2008), and in vitro studies showing that some miRNAs can directly regulate the gene expression of proteins implicated in neurodegenerative disorders (Hebert et al., 2008), all suggest that miRNAs play an important role in the neurodegenerative etiology of Spinocerebellar ataxia type 3, PD, HD, and AD. It is hoped that future expansion of this work may directly link the dysregulation of miRNA regulated pathways to specific aspects of the in vivo pathology seen in human neurodegenerative diseases. Finally, elucidating a role for miRNAs in neurodegeneration may allow for the development of medications designed to prevent or mitigate miRNA dysregulation, or alternatively therapies that act via up- or down-regulation of the targeted mRNA. Current research in this area holds some promise. For example, research into miRNA delivery via viral vectors and depletion of miRNAs via “sponges,” has shown some positive in vitro results (Brown & Naldini, 2009; Rodriguez-Lebron & Paulson, 2006). However, the issues of achieving sufficient target knockdown, the presence of off-target effects, and the possibility of escape mutants still remain to be addressed (Brown & Naldini, 2009). Nonetheless, recent and ongoing ­developments in this field provide grounds for great optimism and for the ­prediction that miRNA based therapeutics will become some of the most exciting biological medications of the upcoming decades. Acknowledgments  This publication was made possible by Grant Number KL2 RR024157 (KL2 award to C Clelland) from the National Center for Research Resources (NCRR), a component of the National Institutes of Health (NIH), and NIH Roadmap for Medical Research. Its contents are solely the responsibility of the authors and do not necessarily represent the official view of NCRR or NIH. Information on NCRR is available at the NCRR Website. Information on Re-engineering the Clinical Research Enterprise can be obtained from the NIH Roadmap website.

References Almeida, A., Bolanos, J. P., & Moreno, S. (2005). Cdh1/Hct1-APC is essential for the survival of postmitotic neurons. The Journal of Neuroscience, 25, 8115–8121. Baek, D., Villen, J., Shin, C., Camargo, F. D., Gygi, S. P., & Bartel, D. P. (2008). The impact of microRNAs on protein output. Nature, 455, 64–71. Bartel, D. P. (2004). MicroRNAs: Genomics, biogenesis, mechanism, and function. Cell, 116, 281–297. Bilen, J., Liu, N., & Bonini, N. M. (2006). A new role for microRNA pathways: Modulation of degeneration induced by pathogenic human disease proteins. Cell Cycle, 5, 2835–2838. Bilen, J., Liu, N., Burnett, B. G., Pittman, R. N., & Bonini, N. M. (2006). MicroRNA pathways modulate polyglutamine-induced neurodegeneration. Molecular Cell, 24, 157–163. Brown, B. D., & Naldini, L. (2009). Exploiting and antagonizing microRNA regulation for therapeutic and experimental applications. Nature Reviews. Genetics, 10, 578–585. Carrettiero, D. C., Hernandez, I., Neveu, P., Papagiannakopoulos, T., & Kosik, K. S. (2009). The cochaperone BAG2 sweeps paired helical filament- insoluble tau from the microtubule. The Journal of Neuroscience, 29, 2151–2161. Hammond, S. M. (2006). MicroRNAs as oncogenes. Currert Opinion in Genetics and Development, 16, 4–9.

MicroRNAs in Neurodegenerative Disorders

453

Hebert, S. S., Horre, K., Nicolai, L., Papadopoulou, A. S., Mandemakers, W., Silahtaroglu, A. N., et al. (2008). Loss of microRNA cluster miR-29a/b-1 in sporadic Alzheimer’s disease correlates with increased BACE1/beta-secretase expression. Proceedings of the National Academy of Sciences of the United States of America, 105, 6415–6420. Johnson, R., Zuccato, C., Belyaev, N. D., Guest, D. J., Cattaneo, E., & Buckley, N. J. (2008). A microRNA-based gene dysregulation pathway in Huntington’s disease. Neurobiology of Disease, 29, 438–445. Kim, J., Inoue, K., Ishii, J., Vanti, W. B., Voronov, S. V., Murchison, E., et al. (2007). A microRNA feedback circuit in midbrain dopamine neurons. Science, 317, 1220–1224. Kim, J., Krichevsky, A., Grad, Y., Hayes, G. D., Kosik, K. S., Church, G. M., et  al. (2004). Identification of many microRNAs that copurify with polyribosomes in mammalian neurons. Proceedings of the National Academy of Sciences of the United States of America, 101, 360–365. Kosik, K. S., & Krichevsky, A. M. (2005). The elegance of the microRNAs: A neuronal ­perspective. Neuron, 47, 779–782. Landgraf, P., Rusu, M., Sheridan, R., Sewer, A., Iovino, N., Aravin, A., et al. (2007). A mammalian microRNA expression atlas based on small RNA library sequencing. Cell, 129, 1401–1414. Lim, L. P., Lau, N. C., Garrett-Engele, P., Grimson, A., Schelter, J. M., Castle, J., et al. (2005). Microarray analysis shows that some microRNAs downregulate large numbers of target mRNAs. Nature, 433, 769–773. Lukiw, W. J. (2007). Micro-RNA speciation in fetal, adult and Alzheimer’s disease hippocampus. Neuroreport, 18, 297–300. Lytle, J. R., Yario, T. A., & Steitz, J. A. (2007). Target mRNAs are repressed as efficiently by microRNA-binding sites in the 5¢ UTR as in the 3¢ UTR. Proceedings of the National Academy of Sciences of the United States of America, 104, 9667–9672. Makeyev, E. V., & Maniatis, T. (2008). Multilevel regulation of gene expression by microRNAs. Science, 319, 1789–1790. Miska, E. A. (2005). How microRNAs control cell division, differentiation and death. Current Opinion in Genetics & Development, 15, 563–568. Perkins, D. O., Jeffries, C., & Sullivan, P. (2005). Expanding the ‘central dogma’: The regulatory role of nonprotein coding genes and implications for the genetic liability to schizophrenia. Molecular Psychiatry, 10, 69–78. Rademakers, R., Eriksen, J. L., Baker, M., Robinson, T., Ahmed, Z., Lincoln, S. J., et al. (2008). Common variation in the miR-659 binding-site of GRN is a major risk factor for TDP43positive frontotemporal dementia. Human Molecular Genetics, 17, 3631–3642. Rodriguez-Lebron, E., & Paulson, H. L. (2006). Allele-specific RNA interference for neurological disease. Gene Therapy, 13, 576–581. Schaefer, A., O’Carroll, D., Tan, C. L., Hillman, D., Sugimori, M., Llinas, R., et  al. (2007). Cerebellar neurodegeneration in the absence of microRNAs. The Journal of Experimental Medicine, 204, 1553–1558. Schratt, G. (2009a). Fine-tuning neural gene expression with microRNAs. Current Opinion in Neurobiology, 19, 213–219. Schratt, G. (2009b). microRNAs at the synapse. Nature Reviews. Neuroscience, 10, 842–849. Schratt, G. M., Tuebing, F., Nigh, E. A., Kane, C. G., Sabatini, M. E., Kiebler, M., et al. (2006). A brain-specific microRNA regulates dendritic spine development. Nature, 439, 283–289. Selbach, M., Schwanhausser, B., Thierfelder, N., Fang, Z., Khanin, R., & Rajewsky, N. (2008). Widespread changes in protein synthesis induced by microRNAs. Nature, 455, 58–63. Siegel, G., Obernosterer, G., Fiore, R., Oehmen, M., Bicker, S., Christensen, M., et al. (2009). A functional screen implicates microRNA-138-dependent regulation of the depalmitoylation enzyme APT1 in dendritic spine morphogenesis. Nature Cell Biology, 11, 705–716. Stark, K. L., Xu, B., Bagchi, A., Lai, W. S., Liu, H., Hsu, R., et al. (2008). Altered brain microRNA biogenesis contributes to phenotypic deficits in a 22q11-deletion mouse model. Nature Genetics, 40, 751–760.

454

C.L. Clelland and J.D. Clelland

Sutherland, H. F., Kim, U. J., & Scambler, P. J. (1998). Cloning and comparative mapping of the DiGeorge syndrome critical region in the mouse. Genomics, 52, 37–43. Vasudevan, S., Tong, Y., & Steitz, J. A. (2007). Switching from repression to activation: MicroRNAs can up-regulate translation. Science, 318, 1931–1934. Wang, G., van der Walt, J. M., Mayhew, G., Li, Y. J., Zuchner, S., Scott, W. K., et  al. (2008). Variation in the miRNA-433 binding site of FGF20 confers risk for Parkinson disease by overexpression of alpha-synuclein. American Journal of Human Genetics, 82, 283–289. Wang, W. X., Rajeev, B. W., Stromberg, A. J., Ren, N., Tang, G., Huang, Q., et al. (2008). The expression of microRNA miR-107 decreases early in Alzheimer’s disease and may accelerate disease progression through regulation of beta-site amyloid precursor protein-cleaving enzyme 1. The Journal of Neuroscience, 28, 1213–1223. Wayman, G. A., Davare, M., Ando, H., Fortin, D., Varlamova, O., Cheng, H. Y., et al. (2008). An activity-regulated microRNA controls dendritic plasticity by down-regulating p250GAP. Proceedings of the National Academy of Sciences of the United States of America, 105, 9093–9098. Weber, M. J. (2005). New human and mouse microRNA genes found by homology search. The FEBS Journal, 272, 59–73. Winter, J., Jung, S., Keller, S., Gregory, R. I., & Diederichs, S. (2009). Many roads to maturity: MicroRNA biogenesis pathways and their regulation. Nature Cell Biology, 11, 228–234. Yu, J. Y., Chung, K. H., Deo, M., Thompson, R. C., & Turner, D. L. (2008). MicroRNA miR-124 regulates neurite outgrowth during neuronal differentiation. Experimental Cell Research, 314, 2618–2633. Zamore, P. D., & Haley, B. (2005). Ribo-gnome: The big world of small RNAs. Science, 309, 1519–1524.

Specific and Surrogate Cerebrospinal Fluid Markers in Creutzfeldt–Jakob Disease Gianluigi Zanusso, Michele Fiorini, Pier Giorgio Righetti, and Salvatore Monaco

Abstract  Detection of cerebrospinal fluid (CSF) biomarkers is a major challenge for laboratories involved in neurological disorder diagnostics and the long list of putative markers now available reflects the enormous efforts and the relevance of CSF in neurology. The result of these intensive studies on CSF is that specific biomarkers are included as supportive criteria for neurodegenerative disorder diagnosis. The diagnostic strategies for biomarkers detection include the detection of surrogate markers which act as bystanders of an ongoing pathogenic process or of those distinct proteins involved in the pathogenic mechanisms. Creutzfeldt–Jakob represents an example in which a surrogate marker such as 14-3-3 had been included among the diagnostic criteria. However, 14-3-3 protein is a very sensitive marker which might indicate positive in several neurological conditions thus reducing its specificity, and it is now coupled with elevated Tau protein levels (>1,300 pg/ml) reaching a specificity of around 100%. This indicates that the aim of CSF diagnostics, not only in sCJD but more widely in different neurodegenerative disorders, is focused on the combination of different CSF biomarkers, resulting in up- or down-regulation in a distinct neurological disorder, in order to increase the specificity and sensibility. Here, we show a series of biomarkers modified in sCJD and the biochemical methods applicable for the detection of novel biomarkers in the CSF, as well as future perspectives for novel biomarkers detection in the CSF. Keywords  Prion diseases • Creutzfeldt–Jakob Disease • Cerebrospinal fluid • Two-dimensional gel • Biomarkers • Proteomics • Intravital diagnostics

G. Zanusso (*) Department of Neurological and Visual Sciences, Section of Clinical Neurology, University of Verona, P.le L.A. Scuro, 10, 37134 Verona, Italy e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_17, © Springer Science+Business Media, LLC 2011

455

456

G. Zanusso et al.

1 Introduction Prion diseases are a group of human and animal neurodegenerative disorders ­characterized by neuronal cell loss, spongiform changes, gliosis and deposition of abnormal amyloid protein. In humans, prion diseases encompass sporadic, genetic and acquired forms which segregate with Creutzfeldt–Jakob disease (CJD), Fatal insomnia and Gerstmann–Straussler–Scheinker syndrome disease phenotypes. The most common human form of prion disorder is represented by sporadic CJD which accounts for about 80% of all human cases. It occurs worldwide with an annual incidence of about one case per million people. Affected patients usually present with dementia accompanied by visual hallucinations, cerebellar ­dysfunction, extrapyramidal signs and myoclonus which invariably elapse into akinetic mutism. In sCJD, the variability of clinical and pathological phenotypes is highly influenced by the combination of a host-dependent methionine/valine (M/V) polymorphism at the codon 129 in the prion protein gene and by two major PrP Sc conformers (Zanusso & Monaco, 2005). Definite diagnosis of sCJD relies on the detection of pathological PrP (PrPSc) by western blot and immunocytochemistry in autoptic brain tissues (Zerr et al., 1998) (Table 1). Thus, the quest for intra-vitam biomarkers is a major issue to ­investigators, and in this context, cerebrospinal fluid (CSF) is a major candidate.

2 Search for Pathological Values in Cerebrospinal Fluid of sCJD Patients CSF is the mechanical liquid jacket for the spinal cord and brain. It is a component of the extracellular space, collecting and exchanging the various substances formed during the metabolic activity of the nervous system. Thus, abnormalities in its ­composition indicate an ongoing “non-physiological” process. While in a wide variety of inflammatory, infectious, neoplastic and demyelinating diseases, the CSF composition is altered, in neurodegenerative disorders it is normal or only slightly modified. Table 1  Diagnostic criteria for sporadic Creutzfeldt– Jakob disease

I II

III

Rapidly progressive dementia A – Myoclonus B – Visual or cerebellar problem C – Pyramidal or extrapiramidal features D – Akinetic mutism A – Typical EEG B – Positive 14-3-3

Possible CJD: I and 2 of II and duration less than 2 years Probable CJD: I and 2 of II and IIIA, or possible CJD and IIIB Definite CJD: Neuropathologically confirmed diagnosis

Specific and Surrogate Cerebrospinal Fluid Markers in Creutzfeldt–Jakob Disease

457

In the CSF of sCJD patients, laboratory standard analyses, which include cell count, blood–brain barrier markers and oligoclonal bands, are not substantially informative. Nevertheless, inflammatory biomarkers might not be unexpected, given increasing evidence for the role of inflammation in this disease (Janssen, Godbolt, Ioannidis, Thompson, & Rossor, 2004). The potential involvement of the immune system was suggested more than 20 years ago, and different studies have reported changes in IgG concentration in the serum and in the CSF of sheep with natural and experimental scrapie (Collis & Kimberlin, 1983; Strain et al., 1984). Experimental transmission studies indicate that the immune-system plays a major role during the peripheral prion neuroinvasion, and that lymphoreticular organs are necessary for accumulating and replicating prions (Aguzzi & Heikenwalder, 2006; Lasmezas et  al., 1996). However, in humans, this search is restricted to variant CJD characterized by the presence of lymphotrophic prions, but also in this disorder, lack of evidence of immune-system activation is likely insufficient to trigger detectable biomarkers in body fluids, including plasma and CSF.

2.1 The Search for PrPSc Several studies have been performed to demonstrate PrPSc in CSF and in other body fluids. There are several lines of evidence supporting CSF as a major candidate for testing: first, transmission studies indicate the presence in CSF of doses of infectivity (Brown et al., 1994); second, PrPSc is deposited in a widespread manner in the brain parenchyma (Budka et al., 1995). In the CSF of sCJD patients, only the soluble cellular form of PrP (PrPC) has been detected (Tagliavini et  al., 1992). A detailed 2-D map characterization of human PrPC in CSF was provided in our previous study, showing the high ­microheterogeneity of glycosylated PrPC isoforms, including PrPC truncated ­fragments (Castagna et al., 2002). However, the pathological insoluble and partial protease resistant PrP (PrPSc) was never detected (Wong et al., 2001). A possible explanation is that aggregated PrPSc undergoes an impaired drainage from the brain to the CSF of sCJD-affected patients. By employing “scanning for intensely ­fluorescent targets” (SIFT), which is an ultra-sensitive quantitative detection method capable of detecting femtomolar concentrations of PrPSc, aggregated PrPSc was found in only 20% of the CSF samples taken from CJD-diagnosed patients (Bieschke et al., 2000; Giese et al., 2000). The kinetics of aggregated PrPSc in the CSF might reflect what has been found in Alzheimer’s disease, in which an inverse ratio is observed between the aggregation state of A-beta in the brain tissue and its levels in the CSF. Increased levels of Amyloid beta protein (A-beta) are observed only during the onset of episodic memory impairment, when soluble A-beta has not been fully entrapped into ­insoluble aggregates or amyloid deposits; after this disease phase, A-beta levels are normal or decreased (Andreasen & Blennow, 2005; Carrette et al., 2003; Sunderland et al., 2003).

458

G. Zanusso et al.

In order to produce a diagnostic test based on PrPSc recognition, different ­laboratories generated monoclonal antibodies which specifically recognize and immunoprecipitate PrPSc and are therefore able to retrieve a few PrPSc molecules from large volumes (Korth et  al., 1997; Paramithiotis et  al., 2003), but the ­inconsistent amount of antigen (PrPSc) might be the reason of why a diagnostic test is not yet available. A promising novel approach seems to be offered by an automated protein ­misfolding cyclic amplification (PMCA). This technology amplifies PrPSc by ­autocatalytic replication to reach a sufficient threshold to detect PrPSc by western blot analysis in the blood of experimentally infected animals (Saa, Castilla, & Soto, 2006). Still, results on in vivo detection of PrPSc in body fluids from sCJD affected patients are not yet available and a large-scale validation is needed.

2.2 Search for Surrogate Markers The lack of PrPSc detection in the CSF has led to the search for surrogate markers helpful in CJD diagnosis. Essentially, biomarker identification has followed two different approaches: by searching for surrogate markers related to neuronal damage and/or to glial activation or by the identification of novel protein markers. In the former, candidate biomarkers are selected and measured based on their presence in the neuropathological lesions of sCJD and/or their involvement in disease. In this context, the most investigated brain-derived proteins were neuron-related markers, which include neuron-specific enolase (NSE), S-100b, 14-3-3 protein, Tau protein, Apolipoprotein E (APOE), and amyloid beta 1–42 (Sanchez-Juan et al., 2006). A diagnostic limitation of these markers is that they increase when neuronal damage occurs, regardless of the cause. In addition, the alterations of levels of brain-derived proteins in the CSF do not always reflect the degree of neuronal damage and the disease severity, but, if markers are used in an appropriate clinical setting or in combination with other markers, their sensitivity and specificity increase consistently (Boesenberg-Grosse et al., 2006). For instance, the 14-3-3 protein is a biomarker indicative of a rapid neuronal destruction and thus useful for differentiating sCJD from other more slowly progressive dementias. However, increased levels of 14-3-3 protein might be induced by epileptic seizures and also by “silent conditions” such as neuroleptic medication intake, indicating that it is a biomarker that is highly susceptible to false positive results. In this context, the combination of 14-3-3 protein and levels of Tau (a cytoskeleton protein), leaking in the CSF only in the presence of significant ­damage, eliminates 14-3-3 false positives and leads to a specificity around 100% for sCJD diagnosis (Zanusso et al., 2005). An additional pathological hallmark of sCJD is the activation of microglia. These cells represent resident macrophages of the CNS, and in their activated state release oxidative tissue-toxic-mediators such as pro-inflammatory and anti-­ inflammatory cytokines, and a broad range of mediators such as prostaglandins and F2-isoprostane. CSF levels of these mediators have been extensively studied,

Specific and Surrogate Cerebrospinal Fluid Markers in Creutzfeldt–Jakob Disease

459

including cytokine levels, PGE2, and 8-epi-PGF2a, markers of lipid peroxidation, but a specific biomarker was not identified (Arlt et al., 2002; Cartier et al., 2004; Kettlun et al., 2003; Minghetti et al., 2000; Stoeck et al., 2005); (Table 2).

2.3 2D-PAGE Approach Analysis of the CSF proteome in different neurological conditions represents a useful broad-based approach for identifying novel markers and for measuring a large number of molecules in an unbiased manner. In addition, statistical methods can be applied to develop combinations or panels that distinguish disease states from controls. The fundamental step in a typical discovery-based proteomic experiment consists of protein separation that is usually achieved by one of three methods: twodimensional gel electrophoresis (2D-PAGE), liquid chromatography, or, more recently, “protein chips” on activated surfaces that bind proteins based on chemical characteristics. 2D-PAGE analysis is a technique which, in the first dimension, separates ­proteins based on their surface charges and, in the second dimension, on their molecular masses. After protein separation, a comparative analysis of CSF samples is carried out between age-matched control subjects and pathological samples, in this case sCJD. From this matching analysis, two main results may be obtained: identification of novel proteins found in sCJD but absent in normal controls as “on/off markers,” or altered expression of CSF proteins found to be differently upor down-regulated in sCJD samples versus controls. Of course, this second result requires a well-controlled normal baseline, with a definite range of normal levels, to consider a biomarker as abnormal. A crucial step consists in the method of revealing proteins entrapped in the ­polyacrylamide gels, based on stains with variable degrees of sensitivity. The most diffuse protocol of gel staining used in most laboratories is silver stain, which is easily applicable, highly sensitive, inexpensive, and reproducible with a detection limit nearly 100 times lower than that obtained with Coomassie staining. Silver stain is the elective method when an on/off result is required, but for quantitative expression, the gel-to-gel variations confound the analysis process. Recently, a technique of gel staining based on fluorescent dyes with different excitation and emission wavelengths allows the labeling of protein before 2D-PAGE. This technique, termed differential in-gel electrophoresis (DIGE), exploits two mass- and charge-matched cyanine dyes (Cy3 and Cy5) for minimally labeling Lys residues in proteins of two different samples, which are subsequently mixed and run on the same gel. In later studies (Knowles et al., 2003), a third dye (Cy2) was used as an internal standard for providing a link for inter-gel comparison and facilitating a more robust statistical analysis. Thus, multiple CSF samples from normal controls and patients can be co-separated and visualized with different fluorescent dyes on a single 2D gel. The images obtained are then evaluated separately and differences

Table 2  CSF surrogate markers with utility for sCJD diagnosis Protein Method of detection References 14-3-3 pan WB Beaudry et al. (1999); Blennow, Johansson, and Zetterberg (2005); Collins et al. (2000); Huang, Marie, Livramento, Chammas, and Nitrini (2003); Jimi et al. (1992); Lemstra et al. (2000); Rosenmann et al. (1997); Satoh et al. (2006); Shiga, Wakabayashi, Miyazawa, Kido, and Itoyama (2006); Van Everbroeck, Boons, and Cras (2005); Van Everbroeck, Green, Pals, Martin, and Cras (1999); Van Everbroeck, Quoilin, Boons, Martin, and Cras (2003); Weber, Otto, Bodemer, and Zerr (1997); Zerr et al. (1996, 1998) 14-3-3 isoforms WB Otto et al. (2000); Shiga et al. (2006); Wiltfang et al. (1999); Zerr et al. (1996) 14-3-3 ELISA Kenney et al. (2000); Peoc’h, Schroder, Laplanche, Ramljak, and Muller (2001) NSE ELISA Beaudry et al. (1999); Jimi et al. (1992); Kohira et al. (2000); Kropp et al. (1999); Van Everbroeck et al. (2005); Weber et al. (1997); Zerr et al. (1996) S-100 ELISA Beaudry et al. (1999); Cepek et al. (2005); Otto et al. (1997a); Van Everbroeck et al. (2005); Weber et al. (1997) TAU ELISA Blennow et al. (2005); Buerger et al. (2006); Cepek et al. (2005); Otto et al. (1997b); Riemenschneider et al. (2003); Van Everbroeck et al. (1999, 2003, 2005); Weber et al. (1997) P-TAU ELISA Blennow et al. (2005); Buerger et al. (2006); Goodall et al. (2006); Riemenschneider et al. (2003); Satoh et al. (2006) Ab 1-42 ELISA Otto et al. (2000); Van Everbroeck et al. (2006); Wiltfang et al. (2003) H-FABP ELISA Guillaume, Zimmermann, Burkhard, Hochstrasser, and Sanchez (2003); Steinacker et al. (2004) Oxidative stress markers Ubiquitin Radioimmunoassay Manaka et al. (1992) Malondialdehide HPLC Bleich et al. (2000) Minghetti et al. (2002) PG E2/F2 Chemiluminescent and colorimetric enzyme immunoassay Lipid peroxidation/ Capillary gas chromatography, Arlt et al. (2002); Janssen et al. (2004) composition hplc, enzymatic kit Lactatedehydrogenase ? Schmidt et al. (2004) Cytochines ELISA Stoeck, Bodemer, and Zerr (2006); Stoeck et al. (2005)

460 G. Zanusso et al.

Specific and Surrogate Cerebrospinal Fluid Markers in Creutzfeldt–Jakob Disease

461

between samples determined quantitatively, by using a dedicated analysis software. Moreover, these dyes have a linear response to variation in protein concentration over five orders of magnitude with a mass sensitivity in the order of picograms. The protein spots may be excised from the gel and subjected to proteolytic digestion for identification by MS (Fig. 1). A key objective is to identify a reference map which is obtained by analyzing normal CSF samples collected from subjects, after informed consent. After 2D separation and gel staining, images are captured in digital format and then ­subjected to computer-assisted image analysis. This includes image normalization, cropping and background subtraction, and matching of the spots between the reference gel and samples.

3 Future Perspectives The focus of clinical proteomics research, in searching for biomarkers, is not just to find a single polypeptide chain as an indicator of a disease, as in the past, but rather to find a panel of biomarkers, whose total expression will tend towards 100% sensitivity coupled to 100% specificity. In sCJD, an important step forward has been made with the identification of the 14-3-3 protein as a biomarker, but still the specificity is not 100% and both false positives and false negatives occur; however, the rates of “false” results may be reduced by adding one or more additional markers, such as, e.g., Tau protein. Almost all neurodegenerative diseases, including sCJD, are “conformational disorders” linked to proteins which undergo an aberrant processing and aggregate in neural tissues. This abnormal protein metabolism also involves, together with the “aggregated protein” marker of the disease, a cascade of additional proteins which co-participate as bystanders or with a specific diseaseactivity (Monaco et  al., 2006). In this context, the proteomic approach, which couples 2D-PAGE with MS analysis, has been consistently improved and the panel of potential diagnostic biomarkers enlarged. Among these novel proteins, there are several reliable candidates, namely Cystatin C, Apo-J and the family of FABP ­proteins that, coupled with 14-3-3 supported pre-mortem diagnosis, might be included for an unequivocal assessment of sCJD (Table 3). There is also an additional path today, as a most interesting prospective panel for biomarker discovery is emerging, notably in the low Mr region of the serum ­polypeptides that contains a large population of fragments, derived mainly from two sources, high-abundance circulating proteins and cell and tissue proteins. While some researchers have dismissed those serum peptides as biological trash, recent work suggests that such peptides could indeed reflect biological events and contain diagnostic biomarkers, due to the fact that, in many diseases, there is an abnormal enzymatic breakdown of proteins, thus rendering the presence of these circulating peptides disease-specific and, in consequence, of diagnostic value. Geho, Liotta, Petricoin, Zhao, and Araujo (2006) have given this intriguing title to one of their recent publications: “The amplified peptidome: the new treasure chest

462

G. Zanusso et al. I. CSF Collection from Normal and sCJD Affected Subjects

II. Sample Preparation

III. 2-D Gel Run IV. Gel Matching Analysis

V. Comparative Analysis, Statistical Evaluation

VI. Reference Map

VII. Spot Excision and Tryptic Digestion

Relatlve Intenslty

VIII. Mass Spectrometry Analysis

100 T9 80 T21 22 60 T19 T20 40 T 10 T20 21 20 T10 0 800 1000 1500

T13

T15 2000

2500

3000

3500

4000

Mass/charge

IX. Bioinformatics Data Base Search and Protein Identification

Fig. 1  2D-gel electrophoresis and mass spectrometry as biomarker discovery tools

Specific and Surrogate Cerebrospinal Fluid Markers in Creutzfeldt–Jakob Disease

463

Table 3  Proteins identified in the CSF from sCJD patients by means of 2D-PAGE analysis and differential proteomics Protein Reference APO-E Choe, Green, Knight, Thompson, and Lee (2002) Electrophoresis Cystatin-C Sanchez et al. (2004) Proteomics 14-3-3 Zerr et al. (1996) The Lancet 14-3-3 Harrington, Merril, Asher, and Gajdusek (1986) NEJM 14-3-3 Hsich, Kenney, Gibbs, Lee, and Harrington (1996) NEJM H-FABP Guillaume et al. (2003) Proteomics Antioxidants Krapfenbauer et al. (2002) Electrophoresis Several markersa Piubelli et al. (2006) Proteomics Prostaglandine d syntase Harrington, Fonteh, Biringer, R Huhmer, and Cowan (2006) Disease Markers a Upregulated: a1 antichymotrypsin, apolipoprotein A1, apolipoprotein A4, HP2-a haptoglobin, ubiquitin, transferrin, cystatin-C. Downregulated: apolipoprotein J, fibrinogen g-chain, prostaglandin D2 synthase, gelsolin, fibrin b, complement factor B/3a

of candidate biomarkers.” Although this approach has not been tried as yet on CSF, it is worth exploring: CSF can be easily de-proteinized, the free peptides captured with the “Equalizer Bead” technology (Righetti & Boschetti, 2007; Righetti et al., 2006) and then profiled via SELDI-MS or sequenced by LC-MS (work in progress). Acknowledgments  This work was supported in part by Fondazione Cariverona, and from the Italian Ministry of Health in collaboration with Istituto Superiore di Sanità, Grant # 4AN/F10. P.G.R. is supported by grants from PRIN 2006 (MIUR, Rome) and by Fondazione Cariplo.

References Aguzzi, A., & Heikenwalder, M. (2006). Pathogenesis of prion diseases: current status and future outlook. Nature Reviews. Microbiology, 4, 765–775. Review. Andreasen, N., & Blennow, K. (2005). CSF biomarkers for mild cognitive impairment and early Alzheimer’s disease. Clinical Neurology and Neurosurgery, 107, 165–173. Review. Arlt, S., Kontush, A., Zerr, I., Buhmann, C., Jacobi, C., Schroter, A., et al. (2002). Increased lipid peroxidation in cerebrospinal fluid and plasma from patients with Creutzfeldt–Jakob disease. Neurobiology of Disease, 10, 150–156. Beaudry, P., Cohen, P., Brandel, J. P., Delasnerie-Laupretre, N., Richard, S., Launay, J. M., et al. (1999). 14-3-3 protein, neuron-specific enolase, and S-100 protein in cerebrospinal fluid of patients with Creutzfeldt–Jakob disease. Dementia and Geriatric Cognitive Disorders, 10, 40–46. Bieschke, J., Giese, A., Schulz-Schaeffer, W., Zerr, I., Poser, S., Eigen, M., et  al. (2000). Ultrasensitive detection of pathological prion protein aggregates by dual-color scanning for intensely fluorescent targets. Proceedings of the National Academy of Sciences of the United States of America, 97, 5468–5473. Bleich, S., Kropp, S., Degner, D., Zerr, I., Pilz, J., Gleiter, C. H., et al. (2000). Creutzfeldt–Jakob disease and oxidative stress. Acta Neurologica Scandinavica, 101, 332–334.

464

G. Zanusso et al.

Blennow, K., Johansson, A., & Zetterberg, H. (2005). Diagnostic value of 14-3-3 beta immunoblot and T-tau/P-tau ratio in clinically suspected Creutzfeldt–Jakob disease. International Journal of Molecular Medicine, 16, 1147–1149. Boesenberg-Grosse, C., Schulz-Schaeffer, W. J., Bodemer, M., Ciesielczyk, B., Meissner, B., Krasnianski, A., et al. (2006). Brain-derived proteins in the CSF: do they correlate with brain pathology in CJD? BMC Neurology, 21, 6–35. Brown, P., Gibbs, C. J., Jr., Rodgers-Johnson, P., Asher, D. M., Sulima, M. P., Bacote, A., et al. (1994). Human spongiform encephalopathy: the National Institutes of Health series of 300 cases of experimentally transmitted disease. Annals of Neurology, 35, 513–529. Budka, H., Aguzzi, A., Brown, P., Brucher, J. M., Bugiani, O., Gullotta, F., et  al. (1995). Neuropathological diagnostic criteria for Creutzfeldt–Jakob disease (CJD) and other human spongiform encephalopathies (prion diseases). Brain Pathology, 5, 459–466. Buerger, K., Otto, M., Teipel, S. J., Zinkowski, R., Blennow, K., DeBernardis, J., et al. (2006). Dissociation between CSF total tau and tau protein phosphorylated at threonine 231 in Creutzfeldt–Jakob disease. Neurobiology of Aging, 27, 10–15. Cartier, L., Garcia, L., Kettlun, A. M., Castaneda, P., Collados, L., Vasquez, F., et  al. (2004). Extracellular matrix protein expression in cerebrospinal fluid from patients with tropical ­spastic paraparesis associated with HTLV-I and Creutzfeldt–Jakob disease. Scandinavian Journal of Clinical and Laboratory Investigation, 64, 101–107. Castagna, A., Campostrini, N., Farinazzo, A., Zanusso, G., Monaco, S., & Righetti, P. G. (2002). Comparative two-dimensional mapping of prion protein isoforms in human cerebrospinal fluid and central nervous system. Electrophoresis, 23, 2242–2246. Cepek, L., Steinacker, P., Mollenhauer, B., Wiese, B., Ciesielczyk, B., Bibl, M., et  al. (2005). Follow-up investigations of tau protein and S-100B levels in cerebrospinal fluid of patients with Creutzfeldt–Jakob disease. Dementia and Geriatric Cognitive Disorders, 19, 376–382. Choe, L. H., Green, A., Knight, R. S., Thompson, E. J., & Lee, K. H. (2002). Apolipoprotein E and other cerebrospinal fluid proteins differentiate ante mortem variant Creutzfeldt–Jakob disease from ante mortem sporadic Creutzfeldt–Jakob disease. Electrophoresis, 23, 2242–2246. Collis, S.C., & Kimberlin, R.H. (1983). Further studies on changes in immunoglobulin G in the sea and CSF of Herdwick sheep with natural and experimental scrapie. Journal of Comparative Pathology, 93, 331–338. Collins, S., Boyd, A., Fletcher, A., Gonzales, M., McLean, C. A., Byron, K., et  al. (2000). Creutzfeldt–Jakob disease: diagnostic utility of 14-3-3 protein immunodetection in cerebrospinal fluid. Journal of Clinical Neuroscience, 7, 203–208. Carrette, O., Demalte, I., Scherl, A., Yalkinoglu, O., Corthals, G., Burkhard, P., et  al. (2003). A panel of cerebrospinal fluid potential biomarkers for the diagnosis of Alzheimer’s disease. Proteomics, 3, 1486–1494. Geho, D. H., Liotta, L. A., Petricoin, E. F., Zhao, W., & Araujo, R. P. (2006). The amplified ­peptidome: the new treasure chest of candidate biomarkers. Current Opinion in Chemical Biology, 10, 50–55. Giese, A., Bieschke, J., Eigen, M., & Kretzschmar, H. A. (2000). Putting prions into focus: ­application of single molecule detection to the diagnosis of prion diseases. Archives of Virology. Supplementum, 16, 161–171. Goodall, C. A., Head, M. W., Everington, D., Ironside, J. W., Knight, R. S., & Green, A. J. (2006). Raised CSF phospho-tau concentrations in variant Creutzfeldt–Jakob disease: diagnostic and pathological implications. Journal of Neurology, Neurosurgery and Psychiatry, 77, 89–91. Guillaume, E., Zimmermann, C., Burkhard, P. R., Hochstrasser, D. F., & Sanchez, J. C. (2003). A potential cerebrospinal fluid and plasmatic marker for the diagnosis of Creutzfeldt–Jakob disease. Proteomics, 3, 1495–1499. Harrington, M. G., Fonteh, A. N., Biringer, R. G., R Huhmer, A. F., & Cowan, R. P. (2006). Prostaglandin D synthase isoforms from cerebrospinal fluid vary with brain pathology. Disease Markers, 22, 73–81. Harrington, M. G., Merril, C. R., Asher, D. M., & Gajdusek, D. C. (1986). Abnormal proteins in the cerebrospinal fluid of patients with Creutzfeldt–Jakob disease. The New England Journal of Medicine, 315, 279–283.

Specific and Surrogate Cerebrospinal Fluid Markers in Creutzfeldt–Jakob Disease

465

Hsich, G., Kenney, K., Gibbs, C. J., Jr., Lee, K. H., & Harrington, M. G. (1996). The 14-3-3 brain protein in cerebrospinal fluid as a marker for transmissible spongiform encephalopathies. The New England Journal of Medicine, 335, 924–930. Huang, N., Marie, S. K., Livramento, J. A., Chammas, R., & Nitrini, R. (2003). 14-3-3 protein in the CSF of patients with rapidly progressive dementia. Neurology, 61, 354–357. Janssen, J. C., Godbolt, A. K., Ioannidis, P., Thompson, E. J., & Rossor, M. N. (2004). The prevalence of oligoclonal bands in the CSF of patients with primary neurodegenerative dementia. Journal of Neurology, 251, 184–188. Jimi, T., Wakayama, Y., Shibuya, S., Nakata, H., Tomaru, T., Takahashi, Y., et al. (1992). High levels of nervous system-specific proteins in cerebrospinal fluid in patients with early stage Creutzfeldt–Jakob disease. Clinica Chimica Acta, 211, 37–46. Kenney, K., Brechtel, C., Takahashi, H., Kurohara, K., Anderson, P., & Gibbs, C. J., Jr. (2000). An enzyme-linked immunosorbent assay to quantify 14-3-3 proteins in the cerebrospinal fluid of suspected Creutzfeldt–Jakob disease patients. Annals of Neurology, 48, 395–398. Kettlun, A., Collados, L., Garcia, L., Cartier, L. A., Wolf, M. E., Mosnaim, A. D., et al. (2003). Matrix metalloproteinase profile in patients with Creuztfeldt-Jakob disease. International Journal of Clinical Practice, 57, 475–478. Knowles, M. R., Cervino, S., Skynner, H. A., Hunt, S. P., de Felipe, C., Salim, K., et al. (2003). Multiplex proteomic analysis by two-dimensional differential in-gel electrophoresis. Proteomics, 3, 1162–1171. Kohira, I., Tsuji, T., Ishizu, H., Takao, Y., Wake, A., Abe, K., et al. (2000). Elevation of neuronspecific enolase in serum and cerebrospinal fluid of early stage Creutzfeldt–Jakob disease. Acta Neurologica Scandinavica, 102, 385–387. Korth, C., Stierli, B., Streit, P., Moser, M., Schaller, O., Fischer, R., et al. (1997). Prion (PrPSc)-specific epitope defined by a monoclonal antibody. Nature, 390, 74–77. Kropp, S., Zerr, I., Schulz-Schaeffer, W. J., Riedemann, C., Bodemer, M., Laske, C., et al. (1999). Increase of neuron-specific enolase in patients with Creutzfeldt–Jakob disease. Neuroscience Letters, 261, 124–126. Krapfenbauer, K., Yoo, B. C., Fountoulakis, M., Mitrova, E., & Lubec, G. (2002). Expression patterns of antioxidant proteins in brains of patients with sporadic Creutzfeldt–Jacob disease. Electrophoresis, 23, 2541–2547. Lasmezas, C. I., Cesbron, J. Y., Deslys, J. P., Demaimay, R., Adjou, K. T., Rioux, R., et al. (1996). Immune system-dependent and -independent replication of the scrapie agent. Journal of Virology, 70, 1292–1295. Lemstra, A. W., van Meegen, M. T., Vreyling, J. P., Meijerink, P. H., Jansen, G. H., Bulk, S., et al. (2000). 14-3-3 testing in diagnosing Creutzfeldt–Jakob disease: a prospective study in 112 patients. Neurology, 55, 514–516. Manaka, H., Kato, T., Kurita, K., Katagiri, T., Shikama, Y., Kujirai, K., et  al. (1992). Marked increase in cerebrospinal fluid ubiquitin in Creutzfeldt–Jakob disease. Neuroscience Letters, 139, 47–49. Minghetti, L., Greco, A., Cardone, F., Puopolo, M., Ladogana, A., Almonti, S., et  al. (2000). Increased brain synthesis of prostaglandin E2 and F2-isoprostane in human and experimental transmissible spongiform encephalopathies. Journal of Neuropathology and Experimental Neurology, 59, 866–871. Minghetti, L., Cardone, F., Greco, A., Puopolo, M., Levi, G., Green, A. J., et al. (2002). Increased CSF levels of prostaglandin E(2) in variant Creutzfeldt–Jakob disease. Neurology, 58, 127–129. Monaco, S., Zanusso, G., Mazzucco, S., & Rizzuto, N. (2006). Cerebral amyloidoses: molecular pathways and therapeutic challenges. Current Medicinal Chemistry, 13, 1903–1913. Otto, M., Esselmann, H., Schulz-Shaeffer, W., Neumann, M., Schroter, A., Ratzka, P., et  al. (2000). Decreased beta-amyloid1-42 in cerebrospinal fluid of patients with Creutzfeldt–Jakob disease. Neurology, 54, 1099–1102. Otto, M., Stein, H., Szudra, A., Zerr, I., Bodemer, M., Gefeller, O., et al. (1997). S-100 protein concentration in the cerebrospinal fluid of patients with Creutzfeldt–Jakob disease. Journal of Neurology, 244, 566–570.

466

G. Zanusso et al.

Otto, M., Wiltfang, J., Tumani, H., Zerr, I., Lantsch, M., Kornhuber, J., et  al. (1997). Elevated levels of tau-protein in cerebrospinal fluid of patients with Creutzfeldt–Jakob disease. Neuroscience Letters, 225, 210–212. Paramithiotis, E., Pinard, M., Lawton, T., LaBoissiere, S., Leathers, V. L., Zou, W. Q., et  al. (2003). A prion protein epitope selective for the pathologically misfolded conformation. Natural Medicines, 9, 893–899. Peoc’h, K., Schroder, H. C., Laplanche, J., Ramljak, S., & Muller, W. E. (2001). Determination of 14-3-3 protein levels in cerebrospinal fluid from Creutzfeldt–Jakob patients by a highly sensitive capture assay. Neuroscience Letters, 301, 167–170. Piubelli, C., Fiorini, M., Zanusso, G., Milli, A., Fasoli, E., Monaco, S., et al. (2006). Searching for markers of Creutzfeldt–Jakob disease in cerebrospinal fluid by two-dimensional mapping. Proteomics, 6(Suppl 1), S256–S261. Riemenschneider, M., Wagenpfeil, S., Vanderstichele, H., Otto, M., Wiltfang, J., Kretzschmar, H., et al. (2003). Phospho-tau/total tau ratio in cerebrospinal fluid discriminates Creutzfeldt–Jakob disease from other dementias. Molecular Psychiatry, 8, 343–347. Righetti, P. G., Boschetti, E., Lomas, L., & Citterio, A. (2006). Protein Equalizer technology: the quest for a “democratic proteome”. Proteomics, 6, 3980–3992. Righetti, P. G., & Boschetti, E. (2007). Sherlock Holmes and the proteome - a detective story. The FEBS Journal, 274, 897–905. Rosenmann, H., Meiner, Z., Kahana, E., Halimi, M., Lenetsky, E., Abramsky, O., et al. (1997). Detection of 14-3-3 protein in the CSF of genetic Creutzfeldt–Jakob disease. Neurology, 49, 593–595. Saa, P., Castilla, J., & Soto, C. (2006). Ultra-efficient replication of infectious prions by automated protein misfolding cyclic amplification. The Journal of Biological Chemistry, 281, 35245–35252. Sanchez, J. C., Guillaume, E., Lescuyer, P., Allard, L., Carrette, O., Scherl, A., et  al. (2004). Cystatin C as a potential cerebrospinal fluid marker for the diagnosis of Creutzfeldt–Jakob disease. Proteomics, 4, 2229–2233. Sanchez-Juan, P., Green, A., Ladogana, A., Cuadrado-Corrales, N., Saanchez-Valle, R., Mitrovaa, E., et  al. (2006). CSF tests in the differential diagnosis of Creutzfeldt–Jakob disease. Neurology, 67, 637–643. Satoh, K., Shirabe, S., Eguchi, H., Tsujino, A., Eguchi, K., Satoh, A., et al. (2006). 14-3-3 protein, total tau and phosphorylated tau in cerebrospinal fluid of patients with Creutzfeldt–Jakob disease and neurodegenerative disease in Japan. Cellular and Molecular Neurobiology, 26, 45–52. Schmidt, H., Otto, M., Niedmann, P., Cepek, L., Schroter, A., Kretzschmar, H. A., et al. (2004). CSF lactate dehydrogenase activity in patients with Creutzfeldt–Jakob disease exceeds that in other dementias. Dementia and Geriatric Cognitive Disorders, 17, 204–206. Shiga, Y., Wakabayashi, H., Miyazawa, K., Kido, H., & Itoyama, Y. (2006). 14-3-3 protein levels and isoform patterns in the cerebrospinal fluid of Creutzfeldt–Jakob disease patients in the progressive and terminal stages. Journal of Clinical Neuroscience, 13, 661–665. Steinacker, P., Mollenhauer, B., Bibl, M., Cepek, L., Esselmann, H., Brechlin, P., et al. (2004). Heart fatty acid binding protein as a potential diagnostic marker for neurodegenerative ­diseases. Neuroscience Letters, 370, 36–39. Stoeck, K., Bodemer, M., & Zerr, I. (2006). Pro- and anti-inflammatory cytokines in the CSF of patients with Creutzfeldt–Jakob disease. Journal of Neuroimmunology, 172, 175–181. Stoeck, K., Bodemer, M., Ciesielczyk, B., Meissner, B., Bartl, M., Heinemann, U., et al. (2005). Interleukin 4 and interleukin 10 levels are elevated in the cerebrospinal fluid of patients with Creutzfeldt–Jakob disease. Archives of Neurology, 62, 1591–1594. Strain, G. M., Barta, O., Olcott, B. M., & Braun, W. F., Jr. (1984). Serum and cerebrospinal fluid concentrations of immunoglobulin G in Suffolk sheep with scrapie. American Journal of Veterinary Research, 45, 1812–1813. Sunderland, T., Linker, G., Mirza, N., Putnam, K. T., Friedman, D. L., Kimmel, L. H., et  al. (2003). Decreased beta-amyloid 1-42 and increased tau levels in cerebrospinal fluid of patients with Alzheimer disease. JAMA, 289, 2094–2103.

Specific and Surrogate Cerebrospinal Fluid Markers in Creutzfeldt–Jakob Disease

467

Tagliavini, F., Prelli, F., Porro, M., Salmona, M., Bugiani, O., & Frangione, B. (1992). A soluble form of prion protein in human cerebrospinal fluid: implications for prion-related encephalopathies. Biochemical and Biophysical Research Communications, 184, 1398–1404. Van Everbroeck, B., Boons, J., & Cras, P. (2005). Cerebrospinal fluid biomarkers in Creutzfeldt–Jakob disease. Clinical Neurology and Neurosurgery, 107, 355–360. Van Everbroeck, B., Green, A. J., Pals, P., Martin, J. J., & Cras, P. (1999). Decreased Levels of Amyloid-beta 1-42 in Cerebrospinal Fluid of Creutzfeldt–Jakob Disease Patients. Journal of Alzheimer’s Disease, 1, 419–424. Van Everbroeck, B., Quoilin, S., Boons, J., Martin, J. J., & Cras, P. (2003). A prospective study of CSF markers in 250 patients with possible Creutzfeldt–Jakob disease. Journal of Neurology, Neurosurgery and Psychiatry, 74, 1210–1214. Van Everbroeck, B., Michotte, A., Sciot, R., Godfraind, C., Deprez, M., Quoilin, S., et al. (2006). Increased incidence of sporadic Creutzfeldt–Jakob disease in the age groups between 70 and 90 years in Belgium. European Journal of Epidemiology, 21, 443–447. Weber, T., Otto, M., Bodemer, M., & Zerr, I. (1997). Diagnosis of Creutzfeldt–Jakob disease and related human spongiform encephalopathies. Biomedicine & Pharmacotherapy, 51, 381–387. Wiltfang, J., Esselmann, H., Smirnov, A., Bibl, M., Cepek, L., Steinacker, P., et al. (2003). Betaamyloid peptides in cerebrospinal fluid of patients with Creutzfeldt–Jakob disease. Annals of Neurology, 54, 263–267. Wiltfang, J., Otto, M., Baxter, H. C., Bodemer, M., Steinacker, P., Bahn, E., et al. (1999). Isoform pattern of 14-3-3 proteins in the cerebrospinal fluid of patients with Creutzfeldt–Jakob disease. Journal of Neurochemistry, 73, 2485–2490. Wong, B. S., Green, A. J., Li, R., Xie, Z., Pan, T., Liu, T., et  al. (2001). Absence of proteaseresistant prion protein in the cerebrospinal fluid of Creutzfeldt–Jakob disease. The Journal of Pathology, 194, 9–14. Zanusso, G., Fiorini, M., Farinazzo, A., Gelati, M., Benedetti, M. D., Ferrari, S., et  al. (2005). Phosphorylated 14-3-3zeta protein in the CSF of neuroleptic-treated patients. Neurology, 64, 1618–1620. Zanusso, G., & Monaco, S. (2005). Molecular mechanisms of human prion diseases. Drug Discovery Today: Disease Mechanisms, 2, 511–518. Zerr, I., Bodemer, M., Gefeller, O., Otto, M., Poser, S., Wiltfang, J., et al. (1998). Detection of 14-3-3 protein in the cerebrospinal fluid supports the diagnosis of Creutzfeldt–Jakob disease. Annals of Neurology, 43, 32–40. Zerr, I., Bodemer, M., Otto, M., Poser, S., Windl, O., Kretzschmar, H. A., et al. (1996). Diagnosis of Creutzfeldt–Jakob disease by two-dimensional gel electrophoresis of cerebrospinal fluid. Lancet, 348, 846–849.

Genome-Wide Expression Studies in Autism-Spectrum Disorders: Moving from Neurodevelopment to Neuroimmunology Roberto Sacco, Antonio M. Persico, Krassimira A. Garbett, and Károly Mirnics

Abstract  Autism has long been thought to stem from abnormal ­neurodevelopment. Surprisingly, microarray-based genome-wide expression studies, involving either postmortem brain tissue or lymphoblastoid cell lines, provide converging ­evidence supporting prominent roles for the immune system in the pathogenesis of autismspectrum disorders (ASDs). In particular, bioinformatic analyses, employing biological databases and gene network prediction software, point toward the involvement of multiple genes interconnected in immune-related pathways. Taken together, these findings suggest that a dysreactive immune process could derange neurodevelopment during critical periods in a large subset of children with autism. These conclusions are also supported by neuropathological and immunological studies, which are briefly summarized. Genome-wide expression studies can thus lead to a better understanding of autism pathogenesis and facilitate the identification of subgroups of patients with a similar underlying pathophysiology (“endophenotypes”), eventually leading to more effective therapeutic strategies. The characterization of peripheral gene-expression patterns and immunological abnormalities can also contribute to design laboratory-based diagnostic tools for the early detection of ASDs. Keywords  Autism • Autistic disorder • Immune system • Gene expression • Microarrays • Neurodevelopment • Neuroinflammation • Pervasive developmental disorders

A.M. Persico (*) Laboratory of Molecular Psychiatry and Neurogenetics, University “Campus Bio-Medico”, Via Alvaro del Portillo 21, 00128 Rome, Italy e-mail: [email protected] K. Mirnics (*) Department of Psychiatry, Vanderbilt University, 8130A MRB III, 465 21st Avenue South, Nashville, TN 37203, USA e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_18, © Springer Science+Business Media, LLC 2011

469

470

R. Sacco et al.

1 Introduction Autism is a complex Pervasive Developmental Disorder (PDD) characterized by impaired language, communication and social skills, as well as by repetitive and stereotypic patterns of behavior, appearing before 3 years of age (APA, 1994). Our understanding of autism pathogenesis is still far from satisfactory, despite major efforts in the past two decades aimed at elucidating the genetic and neurobiological underpinnings of this disease (DiCicco-Bloom et al., 2006; Persico & Bourgeron, 2006). On the one hand, categorical diagnoses represent only an easy-to-use approximation to the clinical complexity of “autism-spectrum disorders” (ASDs), a dimensional continuum ranging from minimal autistic traits to full-blown autism (Piven, Palmer, Jacobi, Childress, & Arndt, 1997). Co-morbidity with epilepsy and mental retardation, present in up to 30 and 80% of autistic patients, respectively, contributes an additional level of clinical complexity (Fombonne, 1999; Tuchman & Rapin, 2002). On the other hand, despite heritability estimates greater than 90%, an unexpected degree of complexity is also encountered at the genetic level. In fact, these prominent genetic contributions encompass interindividual heterogeneity, rare disease-causing de novo mutations, common variants at numerous loci conferring vulnerability or protection, phenocopies, incomplete penetrance, genomic instability often accompanied by parental germline mosaicism, and gene–gene and gene–environment interactions (Lintas & Persico, 2009; Persico & Bourgeron, 2006). A third level of complexity is represented by the neuroanatomical underpinnings of the disease. In general, developmental anomalies later leading to autism begin prenatally, with reduced apoptosis and/or excessive cell proliferation, ­abnormal cell migration, differentiation and synaptogenesis (Bauman & Kemper, 2005; DiCicco-Bloom et al., 2006; Miller et al., 2005; Persico & Bourgeron, 2006); postnatally, these anomalies yield excessive head and body growth in a large subset of patients (Courchesne et  al., 2007; Lainhart et  al., 2006; Sacco et  al., 2007). However, postmortem neuroanatomical studies have unveiled an extreme degree of heterogeneity, with no two brains displaying entirely superimposable abnormalities at the microscopic level. In addition, the involvement of the central nervous system (CNS) is accompanied by systemic signs and symptoms such as macrosomy (Lainhart et al., 2006; Sacco et al., 2007), immune dysreactivity (see below), abnormal intestinal permeability with gastrointestinal symptoms (Campbell et al., 2009; D’Eufemia et  al., 1996; Jyonouchi, Geng, Ruby, & Zimmerman-Bier, 2005) and renal peptiduria (Reichelt, Knivsberg, Nodland, Stensrud, & Reichelt, 1997). These systemic abnormalities are not simple to articulate with neuropathological findings, genetic underpinnings and behavioral symptoms. Genome-wide “omic” approaches, namely the comprehensive study of all genes and their function through genomics, epigenomics, transcriptomics, ­proteomics, and metabolomics, represent an alternative, hypothesis-free, “carpet-bombing” ­strategy which can provide useful information especially in complex, polygenic disorders such as ASDs. The advent of microarray technologies has significantly contributed to the progress of omic approaches. The number of ­published studies focussed on neuropsychiatric disorders is still relatively limited, but is quickly rising.

Genome-Wide Expression Studies in Autism-Spectrum Disorders

471

2 Genome-Wide Expression Studies in Autism Spectrum Disorders In general, two different biomaterials have been employed in array-based genomewide expression studies, either postmortem brain tissue or blood cells, either in the form of peripheral mononuclear blood cells (PMBCs) or as immortalized ­lymphoblastoid cell lines (LCLs). Postmortem brain tissue offers the advantage of exploring genome-wide expression in the tissue most directly implicated in abnormal behaviors. At the same time, the availability of postmortem tissue is limited and RNA quality further reduces the number of brain specimens which can be employed in genome-wide expression profiling. On the other hand, RNA extracted from PMBCs or LCLs is conceivably more distant from the pathophysiology of these disorders. However, it is more manageable and easier to obtain, ensuring that larger sample sizes can be collected and analyzed. The latter approach has been successfully applied to the study of multiple neurological, neurogenetic and psychiatric disorders, such as multiple sclerosis (Iglesias et al., 2004), Tourette syndrome (Du et  al., 2006), neurofibromatosis type I, Down syndrome, tuberous sclerosis complex II (Tang et al., 2004), and bipolar disorder (Kakiuchi et al., 2003). To date, seven published studies have used array-based genome-wide approaches to identify gene expression differences between autistic patients and controls, as summarized in Table 1 and discussed below. 1. Purcell et  al. (2001) examined postmortem cerebellar cortical tissue samples from ten autistic individuals and 23 controls. Two different microarray platforms were used, the Atlas Human Neurobiology array (Clontech), containing 588 human cDNAs (n = 4 AUT/CON pairs), and the UniGEM V2 array (Incyte Genomics), encompassing 9,374 cDNA probes (n = 2 pools of 9 AUT vs. 4 CON individuals, respectively). Cerebellar tissues from the remaining patients and controls were used for confirmatory experiments, performed using reverse-­ transcription PCR (RT-PCR), western blotting and receptor autoradiography. The latter experiments also involved prefrontal cortex and basal ganglia. Mean age (±SD) was 19 ± 14 years, and 22 ± 15 years for patients and controls, ­respectively. The authors chose to investigate the cerebellar cortex because ­neuroanatomical studies had previously documented cytoarchitectonic ­abnormalities in this brain region, especially decreased Purkinje cell numbers (Bauman & Kemper, 2005). At least 30 differentially expressed genes were ­identified, including three ionotropic AMPA glutamate receptor subunits, GluR1, GluR2 and GluR3, as well as the astroglial excitatory amino acid transporter EAAT2 (also known as SLC1A2), all overexpressed in the cerebellar cortex of autistic individuals. No significant difference was found with RNAs encoding subunits belonging to other glutamate receptor types or proteins involved in other neurotransmissions. Contrary to this increase found at the transcript level, glutamate receptor autoradiography revealed decreased AMPA receptors in both granule cell and molecular layers of the autistic cerebellar cortex. No significant difference in NMDA receptor density was found in the cerebellum, nor were

Whole human genome array G4112A (Agilent): 41,000 probes (31,044 expressed in LCLs)

LCLs

LCLs

Five male MZ twins discordant for autism severity and/or degree of language impairment

8 autistics with fragile X, 7 autistics with 15q11-q13 dup, 15 controls (all males)

6 autistics and 6 controls, matched for age, sex and PMI

Hu, Frank, Heine, Lee, and Quackenbush (2006)

Nishimura et al. (2007)

Garbett et al. (2008)

Superior temporal U133 Plus 2.0 GeneChip (Affymetrix): 54,000 gyrus (BA probes 41/42)

Microarray platform Atlas human neurobiology array (Clontech): 588 cDNA probesUniGEM V2 array (Incyte Genomics): 9,374 cDNA probes TIGR 40 K human set array: 39,936 human cDNA probes

Tissue sources Cerebellum (also prefrontal cortex and caudateputamen for RA)

Sample size 10 autistics and 10 controls

Reference Purcell, Jeon, Zimmerman, Blue, and Pevsner (2001)

Table 1  Genome-wide expression studies in autism spectrum disorders

qRT-PCR

qRT-PCR

qRT-PCR

Confirmation qRT-PCR WB RA

� Pro-inflammatory cytokines ¯ Genes involved in neuronal differentiation, brain development and axon guidance FX and dup (15q): �/¯ cell communication and signal transduction + immunerelated genes FX only: �/¯ chaperone and protein folding + RNA binding and mRNA metabolism � Immune-related genes ¯ Genes involved in neuronal development and neurite outgrowth

Main expression changes in ASD � AMPA-type glutamate receptors and transporters

472 R. Sacco et al.

17 autistics (early onset), 18 autistics (late onset), 12 controls

LCLs

U133 Plus 2.0 GeneChip (Affymetrix): 54,000 probes.

qRT-PCR

� NK and CD8+ cellrelated genes, neurodevelopmental and glutamate-related genes (SEMA4c and GLUD1) qRT-PCR � NK cell receptors and Enstrom et al. (2009b) 52 ASD, 27 controls LCLs U133 Plus 2.0 GeneChip cytotoxic function (Affymetrix): 54,000 ¯ Cellular proliferation, probes. cellular biosynthesis, cellular metabolism, ribosomal processing, intracellular organelles qRT-PCR Hu et al. (2009) 86 autistics (severe, mild, LCLs TIGR 40 K human set All autistics: �/¯ androgen savant) and 30 controls array: 39,936 human sensitivity cDNA probes Severe and mild only: �/¯ inflammation, apoptosis, neurological diseases, neurodevelopment Severe only: �/¯ circadian rhythm genes aut autistics; con controls; LCLs lymphoblastoid cell lines; MZ monozygotic; qRT-PCR quantitative reverse transcription – polymerase chain reaction; WB western blotting; RA receptor autoradiography

Gregg et al. (2008)

Genome-Wide Expression Studies in Autism-Spectrum Disorders 473

474

R. Sacco et al.

there changes in either AMPA or NMDA glutamate receptor density in the ­prefrontal cortex and the caudate-putamen. Differences in RNA or protein levels were not correlated with age, PMI or history of seizures. The authors suggest that the discrepancy between increased RNA levels and decreased protein levels of AMPA receptor subunits might be a reflection of excitotoxicity in the cerebellar cortex, which is enriched in glutamatergic neurons and dependent upon glial glutamate transport to maintain extracellular glutamate concentrations within physiological limits. 2. Hu et al. (2006) performed a gene expression profiling using LCLs from (1) three monozygotic (MZ) twin pairs discordant for severity of autistic symptoms (i.e., one twin was autistic and the other twin only had autism-spectrum traits), and (2) from two MZ twin pairs concordant for an autism diagnosis, but ­discordant for severity of language impairment. Cy5- and Cy3-labeled RNAs from each twin pair were co-hybridized to TIGR 40  K Human Set arrays ­containing 39,936 human cDNA probes. Microarray data were analyzed using various approaches, including Significant Analysis of Microarray (SAM) within the MeV package and false discovery rate (FDR). The more severely affected twin in the three discordant MZ twin pairs displayed 25 up-regulated and 19 down-regulated transcripts, relative to the other twin. The two pairs of MZ twins, concordant for an autism diagnosis but discordant for severity of ­language impairment, showed differences in gene expression profile largely overlapping with those found in the three discordant MZ twin pairs. Interestingly, ­independent global functional analyses performed using Ingenuity Pathway Analysis (IPA) on the pooled microarray data obtained from discordant and from concordant MZ twins, analyzed separately, revealed an extended gene ­network showing inter-relationships between differentially expressed genes, centered in both cases around TNFa and encompassing ­several proinflammatory cytokines, such as IL1B, IL4 and IL6. Only seven genes involved in ­neurodevelopment, neural function or in other diseases of the nervous system were found to be differentially expressed, namely NAGLU, ASS, FLAP/ALOX5AP, DAPK1, IL6ST, CHL1, and ROBO1. Moreover, in two pairs where an unaffected sibling was also available, three of these genes (ASS, FLAP and CHL1) showed a higher mean log2 ratio for the more severely affected twin than for the less severely affected twin, when each was compared to its normal sibling. Altogether, these results should be viewed as exploratory, due to the very small sample size assessed in this study (n = 3 discordant and 2 concordant MZ twin pairs). Nonetheless, they do point towards epigenetic mechanisms ­triggered by inflammatory ­processes conceivably underlying the correlation between ­differential gene expression at these loci and the severity of autistic symptoms in MZ twin pairs. 3. Nishimura et al. (2007) used LCLs to profile the transcriptome of eight males with autism due to a fragile X mutation (FMR1-FM), seven males with autism due to a 15q11–q13 duplication [dup(15q)] and 15 nonautistic controls. Samples were hybridized to Whole Human Genome Arrays (Agilent) and microarray data were analyzed using (1) Analysis of Variance (ANOVA), SAM and Rank Product

Genome-Wide Expression Studies in Autism-Spectrum Disorders

475

Analysis to identify differentially expressed genes; (2) principal component analysis and hierarchical clustering to classify gene expression patterns into separate groups based on genetic diagnosis; and (3) Functional Annotation Clustering and IPA to classify differentially-expressed genes into ontology groups and to generate gene networks based upon known protein and gene ­interactions. A set of 68 differentially expressed genes was found to distinguish controls from autistics, regardless of genetic etiology. In addition, 52 transcripts were found specifically altered in the FMR1-FM and 12 genes in the dup(15q) sample. Among the ten genes most differentially expressed between FMR1-FM and dup(15q) were FMR1, CYFIP1, UBE3A, HERC2 and NIPA2. The authors thus speculate about a hypothetical molecular link between autism in FMR1-FM and dup(15q), as possibly mediated by the CYFIP1 protein, which was up-­ regulated in the dup(15q) group and is able to antagonize FMRP. In particular, the authors speculate that a reduction in FMRP or the induction of CYFIP1, ­possibly acting through JAKMIP1 and GPR155, could play relevant roles in autism by affecting GABRB receptors and consequently also mGluR1 receptors, thereby increasing glutamate sensitivity. This argument appears rather speculative, since additional experiments performed in cell culture and on Fmr-1 knockout (ko) mice, while confirming the existence of abnormal expression patterns for these genes, show an opposite change of JAKMIP1 and GPR155 transcript/ protein levels in neural SH-SY5Y cells, as compared to LCLs. Further research will thus be necessary to conclusively establish the exact role of these genes in the pathogenesis of autistic symptoms. More interestingly, classifying all differentially-expressed genes into ontology groups, four gene clusters were identified, including: (1) genes related to cell ­communication and signal transduction; (2) genes related to immune response and defence response (such as TNFRSF8, IGSF3, IL4R, IL21R, CCL17, CCL22); (3) genes related to chaperone and protein folding; and (4) genes related to RNA ­binding and mRNA metabolism. Importantly, cell communication and immune genes were both most enriched in the 68 genes shared by FMR1-FM and dup(15q) patients, regardless of genetic etiology. In contrast, genes encoding chaperones and RNA binding proteins were selectively enriched in the 52 genes dysregulated in FMR1-FM, a finding consistent with the roles of FMRP in RNA binding and in the regulation of translation. The authors thus propose that genome-wide expression can contribute to the identification of etiologically homogenous subsets of patients and can potentially provide useful diagnostic biomarkers, such as blood levels of JAKMIP1 and GPR155 RNA or protein. 4. Garbett et al. (2008) assessed the transcriptome of postmortem brains belonging to six autistic/control pairs matched for age, sex and postmortem interval (PMI). RNA was extracted from the superior temporal gyrus (Brodmann Area 41/42), a neocortical area known to host well-described abnormalities in autism and ­playing a critical role in the perception and comprehension of spoken language (Zilbovicius et al., 2006). Three different statistical approaches were employed to provide probability estimates for differences in expression levels: (1) pairwise

476

R. Sacco et al.

analysis; (2) groupwise analysis; and (3) relative rank analysis. Genes were defined as “differentially expressed” when displaying a difference in average logarithmic ratio (ALR) >1 (corresponding to >2-fold change) and at least two of the three P-values below 0.05. This strategy allowed the authors to maximize the true discovery rate and to minimize the possible confounding effect arising from limited sample size and cohort diversity. Applying the statistical strategy outlined above, 130 up-regulated and 22 down-regulated genes yielded P < 0.05 in all three statistical approaches; 69 additional genes reached P < 0.05 in two of the three strategies; finally, 66 additional genes reached statistical significance in only one of three analytical approaches. By definition, all these genes displayed greater than twofold expression differences between patients and controls. A two-way hierarchical cluster analysis of the expression data was also carried out in order to assess whether samples would separate into two discrete groups based on gene expression patterns and to what extent these groups correspond to clinical diagnoses. The clustering of expression patterns successfully identified five of the six autistic patients as homogeneous and different from the remaining subjects (six controls and one patient). Interindividual transcriptomic variability was also significantly greater among autistic individuals compared to controls, confirming the higher degree of neurobiological heterogeneity present among autistic patients. Finally, using an automated enrichment detection tool, the authors classified differentially-expressed genes into common biological functions. Expression profiling revealed the up-regulation of many immune-related genes and reduced transcript levels for several genes involved in neuronal development and outgrowth. Innate immune genes appear especially upregulated in ASD brain samples, although the involvement of the IL2RB, TH1TH2 and FAS pathways suggests the co-activation of T cell-mediated acquired immune mechanisms. According to functional pathway analysis, the annotated and differentially expressed transcripts belong to 31 gene sets, with 19 of these 31 gene sets related to antigen-specific immune responses, inflammation, cell death, ­autoimmune disorders, migration and targeting of the immune response to specific cells (NKT pathway). An interesting parallel could be drawn between these transcriptome changes and the pattern characterizing the late recovery phase of experimental autoimmune encephalomyelitis (Baranzini, Bernard, & Oksenberg, 2005). Overall, these results point toward a potential viral trigger leading to an early-onset, chronic autoimmune process, ultimately causing altered neurodevelopment and/or abnormal brain function in autistic individuals. 5. Gregg et al. (2008) assessed gene expression in LCLs derived from 35 patients with autistic disorder, 14 patients with ASD (i.e., not fulfilling diagnostic criteria for autistic disorder) and 12 controls. The autistic group was further subdivided into 17 children with early onset and no regression vs. 18 children with late onset and/ or history of regression. Altogether, 55 differentially expressed genes ­distinguished the autistic group from controls, 140 genes were differentially expressed between early onset autism and normal controls, only 20 genes distinguished autistics with regression from controls, while no differences were detected between ASD patients and controls. A total of 11 genes shared abnormal expression levels in

Genome-Wide Expression Studies in Autism-Spectrum Disorders

477

all three autistic groups (autism, early onset autism and autism with regression). Functional pathway analysis found this gene set primarily related to the KEGG pathway, involved in natural killer cell-mediated cytotoxicity. The NK cell signaling pathway was also identified by Ingenuity Pathways Analysis (IPA). These 11 autism-related genes are all highly expressed in NK cells and/or CD8+ cytotoxic T cells. NK cells represent a primary innate defense against viral and bacterial infections, as well as malignant transformation. NK cells bind to infected cells and, if the target cells are MHC-I negative, they secrete perforins which permeabilize the cell membrane and allow the entry of granzymes, thereby initiating caspase-mediated apoptosis. Also, cytotoxic CD8+ T cells play a critical role against viruses, intracellular microrganisms and neoplastic cells by recognizing specific antigens on target cells, initiating perforin-induced membrane permeabilization and accomplishing granzyme-mediated apoptosis. In addition, the IL-2 pathway was also identified by IPA as enriched in all ASD patients and in those with an early-onset contrasted with typically developing controls. IL-2 is known to act on CD8+ T cells, where it stimulates proliferation, while also affecting perforin and granzyme gene expression. Several up-regulated genes, such as SH21B/EAT2, RUNX3, GiMAP6 and DNMT1, profoundly modulate NK and/or CD8+ T cell function or proliferation. Other genes involved in neural development or in glutamatergic neurotransmission were also up-regulated in autism, including semaphorin 4C (SEMA4C) and glutamate dehydrogenase (GLUD1). 6. In another study, the same research group (Enstrom et al., 2009b) carried out a genome-wide expression analysis using LCLs from 52 children with autistic spectrum disorders and 27 typically developing control children. They found 82 probes, corresponding to 59 known genes up-regulated and 544 probes downregulated in the autistic group compared to controls. The majority of up-­regulated transcripts were related to NK cell function, while most down-regulated genes were implicated in cellular proliferation and differentiation. Overexpressed genes include both killer cell immunoglobulin receptors (KIRs) and inhibitory KIRs, as well as genes encoding other NK effector molecules. This transcript profile appears somewhat paradoxical, because there is increased expression of both inhibitory and cytolitic NK-related genes. Down-regulated genes belong to several functional groups, including cellular biosynthesis, ribosomal processing, cellular metabolism and intracellular organelles. Given the previous and the present results, cellular and functional analyses of NK cells in ASD patients were carried out. These analyses showed increased NK cell counts and cytolytic activity under unstimulated conditions, but significantly decreased cytolytic activity and IFNg production by NK cells under stimulation in autistic patients compared to controls. These results suggest that NK cells are chronically-­activated in autistic children and that, if stimulated, they may either be unable to respond as much as NK cells in controls or they could even undergo a paradoxical down-regulation. Interestingly, a similar pattern is also observed in autoimmune disorders, such as multiple sclerosis, whereby circulating lymphocytes are activated in the periphery and are not able to react any further when exposed to mitogen stimulation in vitro

478

R. Sacco et al.

(Gershwin, Haselwood, Dorshkind, & Castles, 1979; Vervliet & Schandene, 1985). 7. Hu et al. (2009) recently used a “phenotypic approach” to guide the ­identification of autism genes and pathways contributing to specific clinical phenotypes. Genome-wide expression analysis was performed on 116 LCLs from 86 male autistic patients and 30 age-matched controls, using the TIGR 40  K human arrays as in Hu et  al. (2006). Autistic individuals were divided into three ­phenotypic subgroups, according to ADI-R scores: (1) patients with severe ­language impairment and high overall severity scores, (2) patients with mild severity scores, including many individuals clinically diagnosed with Asperger’s Syndrome or PDD-NOS, and (3) patients displaying high scores in savant skill categories, regardless of language impairment severity. A two-class SAM ­analysis with a FDR <5% revealed a set of 530 differentially expressed genes distinguishing autistic patients from controls, regardless of subgrouping. Interestingly, a four-class SAM analysis with FDR £0.1% unveiled an ­expression matrix of the top 123 most significant genes supporting quantitative as well as qualitative differences in gene expression profiles that distinguish the three ASD phenotypes from one another, as well as from controls. The same outcome was obtained applying a principal component analysis on these 123 ­differentially expressed genes, although there appears to be some overlap between the savant subgroup and controls. Twenty transcripts separate all three subgroups of ­autistic patients from controls. Interestingly, all of them are novel and 19 of them map to putative intronic or intergenic noncoding regions. The majority of these transcripts are linked to androgen sensitivity and steroid hormone biosynthetic pathway, according to gene expression studies of androgen insensitivity (Holterhus, Hiort, Demeter, Brown, & Brooks, 2003). Other functions ­associated with genes shared by two of the three autistic subgroups include apoptosis, inflammation and oxidative stress (ITGAM and NFKB1 were also checked by qRT-PCR), synaptogenesis and axon guidance (RHOA, SLIT2), epigenetic regulation and protein ubiquitination (MBD2). Among the differentially expressed genes unique to each subgroup, as many as 15 circadian rhythm ­regulatory or responsive genes were identified as differentially expressed in the most severe subgroup. This gene cluster comprises PER3, NPAS2, AANAT, CRY1, DPYD, and BHLHB2/DEC1. Finally, the top multigene network ­identified by IPA in the severely and mildly affected subgroups includes many genes involved in neurite outgrowth, neurogenesis, axon pathfinding, brain morphology, cell death, cell-mediated immune responses, neurological and neuroinflammatory diseases. The authors conclude that shared differentially expressed genes may underlie the basic deficits present in autistic individuals, while subgroup-­specific genes may confer specific clinical features and ­contribute to determine autism severity. In conclusion, with the possible exception of the study by Purcell et al. (2001), based on microarrays carrying a relatively limited number of probes compared to those employed in subsequent studies, the other six published studies provide ­converging evidence supporting an involvement of the immune system as playing a significant

Genome-Wide Expression Studies in Autism-Spectrum Disorders

479

role in the pathogenesis of autistic disorder, especially with regard to NK and CD8+ T cells. It remains to be established whether and to what extent immune anomalies, such as defective NK cell function, represent causes, ­consequences or susceptibility factors, exerting a permissive effect upon neural inflammation or deranging normal neuro-immune interactions. Yet, the ­convergence of CNS and LCL data on an abnormal immune activation, especially ­concerning innate immune response mechanisms, spurs further interest into ­investigations aimed at elucidating the intimate nature of immune contributions to altered ­behavior in autism.

3 Abnormal Immunity and Neuroinflammation in Autism Genome-wide expression studies represent one of several approaches now supporting a dysfunctional activation of the immune system as a relevant player in autism pathogenesis (Ashwood, Wills, & Van de Water, 2006; Cohly & Panja, 2005; Pardo, Vargas, & Zimmerman, 2005; Patterson, 2009; Sperner-Unterweger, Winkler, & Fuchs, 2006). Postmortem tissues similar or identical to those used for some genome-wide expression studies were also examined in neuropathological and neurochemical studies. Vargas, Nascimbene, Krishnan, Zimmerman, and Pardo (2005) convincingly documented the existence of chronic neuroinflammation in the brains of autistic individuals, possibly located in the frontal and temporal lobes, the same brain regions characterized by excessive growth (Courchesne et al., 2007). Interestingly, a positive history of allergies or autoimmune disorders in the patient or in his/her first-degree relatives was associated with larger head circumference and body growth in a sample of 241 autistic children (Sacco et al., 2007). Neuroinflammation in autistic brains is characterized by hyperproliferation of astroglial cells and by activated microglia, in the absence of lymphocyte infiltration or immunoglobulin deposition in the CNS. Consequently, there is an increased production of cytokines generally exerting pro-inflammatory effects, such as IL-1, IL-6, IL-12, MCP-1, IFN-g, and TNF-a, both in postmortem cerebral and cerebellar cortex, and in the CSF of autistic children. Increased levels of some ­anti-­inflammatory cytokines, such as transforming growth factor-b1 (TGF-b1), were also documented (Vargas et  al., 2005), while both RNA and protein levels of PKC-beta, which drives the expression of many pro-inflammatory cytokines, are down-regulated (Lintas et  al., 2009). These findings could reflect a compensatory adjustment attempting to attenuate neuroinflammation and/or repair damaged ­tissue. In another study, similar increases in MCP-1 and TGF-b1 were also recorded in the cerebrospinal fluid, cerebellum and anterior cingulate gyrus, while increased levels of eotaxine, IL-6, IL-10 and MCP-3 were found in the anteriore cingulate gyrus of postmortem brains (Zimmerman et al., 2005). Several pro-inflammatory cytokines, including TNF-a, IFN-g, IL-6, IL-8 and GM-CSF, were also significantly elevated in postmortem frontal cortical specimens of eight autistic patients compared to sexmatched controls (Li et al., 2009)

480

R. Sacco et al.

In general, cytokines and chemokines produced by activated immune cells can negatively influence neurodevelopment, but can also modulate cognitive and emotional processing ultimately affecting behavior “here and now,” as commonly occurs with infectious diseases, injury and inflammation. The sustained and prolonged overexpression of this cytokine network could thus play relevant roles both in early pathogenetic events leading to autism, and in determining the specific pattern of autistic symptoms later displayed by a given patient (e.g., sleep-wake cycle abnormalities, sensory disturbances, etc.). Not only can the immune system affect the nervous system, but conversely neuropeptides produced by the central and peripheral nervous system can also affect the immune system, for example by ­influencing the recruitment and chemotaxis of innate immune cells. In addition to postmortem findings, several immune abnormalities have been reported in autistic patients assessed in vivo: –– Decreased peripheral lymphocyte counts and reduced frequency of naïve CD4+ T cells (Ferrante et al., 2003). –– Decreased response to T cell mitogens (Engstrom et al., 2003; Plioplys, Greaves, Kazemi, & Silverman, 1994). –– Incomplete or partial T cell activation, compatible with an increased numbers of DR+ T cells, in the absence of IL-2 receptor expression. This activation is inversely correlated with decreased amounts of the C4B complement component (Warren et al., 1990). –– Dysregulation of apoptosis in CD4+ T lymphocytes (Engstrom et al., 2003). –– Impaired humoral immunity, with increased IgG, IgG4, IgM and IgE levels, decreased levels of IgA and IgG1, and reduced humoral response to protein antigens (Enstrom et al., 2009a; Heuer et al., 2008; Warren et al., 1997). –– Altered NK cell frequency and reduced NK cell lytic activity in autistic patients compared to controls (Enstrom et  al., 2009b; Warren, Foster, & Margaretten, 1987). –– Elevated numbers of circulating monocytes, correlated with increased plasma levels of the Th1-type cytokine interferon-g (IFN-g) and neopterin (Sweeten, Posey, & McDougle, 2003; Sweeten, Posey, Shankar, & McDougle, 2004). Also, urinary levels of neopterin and biopterin are elevated in some (Messahel et al., 1998), though not all studies (Eto, Bandy, & Butterworth, 1992), while studies assessing the CSF have most frequently reported decreased neopterin levels (Tani, Fernell, Watanabe, Kanai, & Långström, 1994; Zimmerman et al., 2005). In inflammatory conditions, such as autoimmune disorders and cancer, increased neopterin concentrations and accelerated tryptophan catabolism stem from the overproduction of IFN-g (Wirleitner, Neurauter, Schröcksnadel, Frick, & Fuchs, 2003). In vitro, IFN-g induces neopterin formation and tryptophan degradation via indoleamine (2,3)-dioxygenase, and both events are effectively suppressed by the Th2-type cytokines IL-4 and IL-10 (Wirleitner et al., 2003). –– Activated Th2-type cytokine profile over Th1-type, due to an imbalance between Th1/Th2 subsets of CD4+/CD8+ T cells, in favor of Th2 (Korvatska, Van de Water,

Genome-Wide Expression Studies in Autism-Spectrum Disorders

–– ––

–– ––

481

Anders, & Gershwin, 2002). However, this likely represents an oversimplification, as cytokine studies have yielded conflicting results pointing toward more complex profiling than the traditional Th1/Th2 distinction (Ashwood et al., 2006). Several studies have indeed reported pro-inflammatory cytokine patterns in CSF and blood (Chez, Dowling, Patel, Khanna, & Kominsky, 2007; Croonenberghs, Bosmans, Deboutte, Kenis, & Maes, 2002; Jyonouchi, Sun, & Le, 2001; Molloy et al., 2006; Singh, 1996; Zimmerman et al., 2005). Presence of auto-antibodies directed to self-antigens located in the CNS (Cabanlit, Wills, Goines, Ashwood, & Van de Water, 2007; Connolly et  al., 2006; Vojdani et al., 2002; Wills et al., 2009). Presence of maternal antibodies reactive against fetal antigens from brain extracts (Braunschweig et  al., 2008; Croen et  al., 2008; Dalton et  al., 2003; Singer et al., 2008; Zimmerman et al., 2007). These antibodies, injected in pregnant animals, yield stereotypic behaviors in monkeys (Martin et al., 2008), and decreased orientation and exploratory behavior in rodents (Dalton et al., 2003). Fever-induced reductions in aberrant behaviors, including irritability, hyperactivity, stereotypies, and inappropriate speech (Curran et al., 2007). Familiality for autoimmune diseases, asthma and allergies (Comi, Zimmermann, Frye, Law, & Peeden, 1999; Croen, Grether, Yoshida, Odouli, & Van de Water, 2005; Mouridsen, Rich, Isager, & Nedergaard, 2007).

With reference to studies published to date in this field, readers should be cautioned that their quality and reliability is highly heterogeneous. Some inconsistencies could thus at least partly stem from methodological issues. However, “true” inconsistencies between studies could also conceivably stem not only from interindividual heterogeneity in basic pathophysiological mechanisms but also from samples either differing by sex and age (especially in reference to the influence on immune function of sex hormones at or after puberty) and/or passing through different phases of “activation” or “quiescence” in the disease, possibly both at the behavioral and at the immune level. Nonetheless, despite these limitations, the presence of immune abnormalities in autism appears strongly supported by the results summarized above. Further evidence of a possible involvement of the immune system in ASD comes from several genetic studies, describing an association between autistic disorder and immune gene variants, such as HLA and complement alleles (Odell et al., 2005; Torres, Maciulis, & Odell, 2001; Warren et al., 1996). MHC class I plays a crucial role in brain development, neuronal differentiation and activitydependent synaptic plasticity (Boulanger, 2004; Huh et  al., 2000). Moreover, animal models show that the maternal immune response to infection or to experimental immunogens such as LPS or poly(I:C) can influence fetal brain development through increased levels of circulating cytokines, especially IL-6 (Meyer et al., 2006; Patterson, 2002; Patterson, 2009; Ponzio, Servatius, Beck, Marzouk, & Kreider, 2007; Shi et al, 2009; Smith, Li, Garbett, Mirnics, & Patterson, 2007; Yamashita, Fujimoto, Nakajima, Isagai, & Matsuishi, 2003). As an example,

482

R. Sacco et al.

maternal influenza virus infection during mid-gestation yields neuropathological and behavioral abnormalities in the mouse offspring, which are compatible with those seen in humans with autism (Fatemi et al., 2002; Shi, Fatemi, Sidwell, & Patterson, 2003, Shi et al, 2009). These abnormalities can also be evoked in the absence of viral infection, by activating the maternal immune response through the administration of poly I:C and are absent in IL-6 receptor knock-out mice (Shi et  al., 2009). These data suggest that cerebellar abnormalities may occur during embryonic development, following an activation of the maternal immune system. Also, an infection with Borna disease virus (BDV) at later stages in development (i.e., in neonatal rats) induces neuronal death in the cerebellum, hippocampus and neocortex, yielding behavioral abnormalities showing some overlap with autistic symptoms in humans (Hornig, Solbrig, Horscroft, Weissenböck, & Lipkin, 2001).

4 Conclusions and Future Perspectives Neuropathological, immunological, and genome-wide expression studies performed to date both postmortem and in  vivo, collectively document two main immune dysfunctions in autism, namely (1) a basal overproduction of pro-inflammatory cytokines coupled with a reduction in stimulus-triggered innate and acquired immune responses, and (2) the presence of auto-antibodies against selfantigens located in the CNS (Cohly & Panja, 2005). The expression pattern present in most autistic brains examined to date resembles a late-stage autoimmune event rather than an acute autoimmune response or the nonspecific immune activation seen in neurodegenerative diseases (Garbett et al., 2008). The presence of both an innate and an acquired immune component, together with animal models demonstrating virally-triggered neurodevelopmental abnormalities mediated by maternal cytokines overproduced during pregnancy, conceivably points toward a potential viral trigger for an early-onset chronic autoimmune process leading to altered neurodevelopment and to a persistent activation of the immune system in the brain of autistic patients. Future research will have to elucidate how this abnormal immune state comes about and is maintained, unveil the mechanisms through which it influences neural development, behavior and cognition both acutely and long-term, and define viable therapeutic options capable of blunting the pathological processes underlying autism. Acknowledgments  We would like to thank the donor families, the Autism Tissue Program (Princeton, N.J.), the Harvard Brain Tissue Resource Center (Belmont, MA), and the NICHD Brain and Tissue Bank (Baltimore, MD) for providing the brain tissue samples. A.M.P. is supported by the Italian Ministry for University, Scientific Research and Technology (PRIN n.2006058195), the Italian Ministry of Health (RFPS-2007-5-640174) and Autism Speaks (Princeton, NJ) and K.M. is supported by VUKC Startup Fund, R01 MH079299, and K02 MH070786.

Genome-Wide Expression Studies in Autism-Spectrum Disorders

483

References American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington: American Psychiatric Press. Ashwood, P., Wills, S., & Van de Water, J. (2006). The immune response in autism: a new frontier for autism research. Journal of Leukocyte Biology, 80, 1–15. Baranzini, S. E., Bernard, C. C., & Oksenberg, J. R. (2005). Modular transcriptional activity characterizes the initiation and progression of autoimmune encephalomyelitis. Journal of Immunology, 174, 7412–7422. Bauman, M. L., & Kemper, T. L. (2005). Neuroanatomic observations of the brain in autism: a review and future directions. International Journal of Developmental Neuroscience, 23, 183–187. Boulanger, L. M. (2004). MHC class I in activity-dependent structural and functional plasticity. Neuron Glia Biology, 1, 283–289. Braunschweig, D., Ashwood, P., Krakowiak, P., Hertz-Picciotto, I., Hansen, R., Croen, L. A., et al. (2008). Autism: maternally derived antibodies specific for fetal brain proteins. Neurotoxicology, 29, 226–231. Cabanlit, M., Wills, S., Goines, P., Ashwood, P., & Van de Water, J. (2007). Brain-specific autoantibodies in the plasma of subjects with autistic spectrum disorder. Annals of the New York Academy of Sciences, 1107, 92–103. Campbell, D. B., Buie, T. M., Winter, H., Bauman, M., Sutcliffe, J. S., Perrin, J. M., et al. (2009). Distinct genetic risk based on association of MET in families with co-occurring autism and gastrointestinal conditions. Pediatrics, 123, 1018–1024. Chez, M. G., Dowling, T., Patel, P. B., Khanna, P., & Kominsky, M. (2007). Elevation of tumor necrosis factor-alpha in cerebrospinal fluid of autistic children. Pediatric Neurology, 36, 361–365. Cohly, H. H., & Panja, A. (2005). Immunological findings in autism. International Review of Neurobiology, 71, 317–341. Comi, A. M., Zimmermann, A. W., Frye, V. H., Law, P. A., & Peeden, J. N. (1999). Familial clustering of autoimmune disorders and evaluation of medical risk factors in autism. Journal of Autism and Developmental Disorders, 35, 103–116. Connolly, A. M., Chez, M., Streif, E. M., Keeling, R. M., Golumbek, P. T., Kwon, J. M., et al. (2006). Brain-derived neurotrophic factor and autoantibodies to neural antigens in sera of children with autistic spectrum disorders, Landau-Kleffner syndrome, and epilepsy. Biological Psychiatry, 59, 354–363. Courchesne, E., Pierce, K., Schumann, C. M., Redcay, E., Buckwalter, J. A., Kennedy, D. P., et al. (2007). Mapping early brain development in autism. Neuron, 56, 399–413. Croen, L. A., Braunschweig, D., Haapanen, L., Yoshida, C. K., Fireman, B., Grether, J. K., et al. (2008). Maternal mid-pregnancy autoantibodies to fetal brain protein: the early markers for autism study. Biological Psychiatry, 64, 583–588. Croen, L. A., Grether, J. K., Yoshida, C. K., Odouli, R., & Van de Water, J. (2005). Maternal autoimmune diseases, asthma and allergies, and childhood autism spectrumdisorders: a casecontrol study. Archives of Pediatrics & Adolescent Medicine, 159, 151–157. Croonenberghs, J., Bosmans, E., Deboutte, D., Kenis, G., & Maes, M. (2002). Activation of the inflammatory response system in autism. Neuropsychobiology, 45, 1–6. Curran, L. K., Newschaffer, C. J., Lee, L. C., Crawford, S. O., Johnston, M. V., & Zimmerman, A. W. (2007). Behaviors associated with fever in children with autism spectrum disorders. Pediatrics, 120, e1386–e1392. D’Eufemia, P., Celli, M., Finocchiaro, R., Pacifico, L., Viozzi, L., Zaccagnini, M., et al. (1996). Abnormal intestinal permeability in children with autism. Acta Paediatrica, 85, 1076–1079. Dalton, P., Deacon, R., Blamire, A., Pike, M., McKinlay, I., Stein, J., et al. (2003). Maternal neuronal antibodies associated with autism and a language disorder. Annals of Neurology, 53, 533–537.

484

R. Sacco et al.

DiCicco-Bloom, E., Lord, C., Zwaigenbaum, L., Courchesne, E., Dager, S. R., Schmitz, C., et al. (2006). The developmental neurobiology of autism spectrum disorder. The Journal of Neuroscience, 26, 6897–6906. Du, X., Tang, Y., Xu, H., Lit, L., Walker, W., Ashwood, P., et al. (2006). Genomic profiles for human peripheral blood T cells, B cells, natural killer cells, monocytes, and polymorphonuclear cells: comparisons to ischemic stroke, migraine, and Tourette syndrome. Genomics, 87, 693–703. Engstrom, A. H., Ohlson, S., Stubbs, E. G., Maciulus, A., Caldwell, V., Odell, J. D., et al. (2003). Decreased expression of CD95 (FAS/APO-1) on CD4+ T-lymphocytes from participants with autism. Journal of Developmental and Physical Disabilities, 15, 155–163. Enstrom, A., Krakowiak, P., Onore, C., Pessah, I. N., Hertz-Picciotto, I., Hansen, R. L., et  al. (2009a). Increased IgG4 levels in children with autism disorder. Brain, Behavior, and Immunity, 23, 389–395. Enstrom, A., Lit, L., Onore, C. E., Gregg, J. P., Hansen, R. L., Pessah, I. N., et al. (2009b). Altered gene expression and function of peripheral blood natural killer cells in children with autism. Brain, Behavior, and Immunity, 23, 124–133. Eto, I., Bandy, M. D., & Butterworth, C. E., Jr. (1992). Plasma and urinary levels of biopterin, neopterin, and related pterins and plasma levels of folate in infantile autism. Journal of Autism and Developmental Disorders, 22, 295–308. Fatemi, S. H., Earle, J., Kanodia, R., Kist, D., Emamian, E. S., Patterson, P. H., et al. (2002). Prenatal viral infection leads to pyramidal cell atrophy and macrocephaly in adulthood: implications for genesis of autism and schizophrenia. Cellular and Molecular Neurobiology, 22, 25–33. Ferrante, P., Saresella, M., Guerini, F. R., Marzorati, M., Musetti, M. C., & Cazzullo, A. G. (2003). Significant association of HLA A2-DR11 with CD4 naive decrease in autistic children. Biomedicine & Pharmacotherapy, 57, 372–374. Fombonne, E. (1999). The epidemiology of autism: a review. Psychological Medicine, 29, 769–786. Garbett, K., Ebert, P. J., Mitchell, A., Lintas, C., Manzi, B., Mirnics, K., et al. (2008). Immune transcriptome alterations in the temporal cortex of subjects with autism. Neurobiology of Disease, 30, 303–311. Gershwin, M. E., Haselwood, D., Dorshkind, K., & Castles, J. J. (1979). Altered responsiveness to mitogens in subgroups of patients with juvenile rheumatoid arthritis. Journal of Clinical & Laboratory Immunology, 1, 293–297. Gregg, J. P., Lit, L., Baron, C. A., Hertz-Picciotto, I., Walker, W., Davis, R. A., et al. (2008). Gene expression changes in children with autism. Genomics, 91, 22–29. Heuer, L., Ashwood, P., Schauer, J., Goines, P., Krakowiak, P., Hertz-Picciotto, I., et al. (2008). Reduced levels of immunoglobulin in children with autism correlates with behavioral symptoms. Autism Research, 1, 275–283. Holterhus, P. M., Hiort, O., Demeter, J., Brown, P. O., & Brooks, J. D. (2003). Differential geneexpression patterns in genital fibroblasts of normal males and 46, XY females with androgen insensitivity syndrome; evidence for early programming involving the androgen receptor. Genome Biology, 4, R37. Hornig, M., Solbrig, M., Horscroft, N., Weissenböck, H., & Lipkin, W. I. (2001). Borna disease virus infection of adult and neonatal rats: models for neuropsychiatric disease. Current Topics in Microbiology and Immunology, 253, 157–177. Hu, V. W., Frank, B. C., Heine, S., Lee, N. H., & Quackenbush, J. (2006). Gene expression profiling of lymphoblastoid cell lines from monozygotic twins discordant in severity of autism reveals differential regulation of neurologically relevant genes. BMC Genomics, 7, 118. Hu, V. W., Sarachana, T., Kim, K. S., Nguyen, A., Kulkarni, S., Steinberg, M. E., et al. (2009). Gene expression profiling differentiates autism case-controls and phenotypic variants of autism spectrum disorders: evidence for circadian rhythm dysfunction in severe autism. Autism Research, 2(2), 78–97. Huh, G. S., Boulanger, L. M., Du, H., Riquelme, P. A., Brotz, T. M., & Shatz, C. J. (2000). Functional requirement for class I MHC in CNS development and plasticity. Science, 290, 2155–2159.

Genome-Wide Expression Studies in Autism-Spectrum Disorders

485

Iglesias, A. H., Camelo, S., Hwang, D., Villanueva, R., Stephanopoulos, G., & Dangond, F. (2004). Microarray detection of E2F pathway activation and other targets in multiple sclerosis peripheral blood mononuclear cells. Journal of Neuroimmunology, 150, 163–177. Jyonouchi, H., Geng, L., Ruby, A., & Zimmerman-Bier, B. (2005). Dysregulated innate immune responses in young children with autism spectrum disorders: their relationship to gastrointestinal symptoms and dietary intervention. Neuropsychobiology, 51, 77–85. Jyonouchi, H., Sun, S., & Le, H. (2001). Proinflammatory and regulatory cytokine production associated with innate and adaptive immune responses in children with autism spectrum disorders and developmental regression. Journal of Neuroimmunology, 120, 170–179. Kakiuchi, C., Iwamoto, K., Ishiwata, M., Bundo, M., Kasahara, T., Kusumi, I., et  al. (2003). Impaired feedback regulation of XBP1 as a genetic risk factor for bipolar disorder. Nature Genetics, 35, 171–175. Korvatska, E., Van de Water, J., Anders, T. F., & Gershwin, M. E. (2002). Genetic and immunologic considerations in autism. Neurobiology of Disease, 9, 107–125. Lainhart, J. E., Bigler, E. D., Bocian, M., Coon, H., Dinh, E., Dawson, G., et al. (2006). Head circumference and height in autism: a study by the collaborative program of excellence in autism. American Journal of Medical Genetics, 140, 2257–2274. Li, X., Chauhan, A., Sheikh, A. M., Patil, S., Chauhan, V., Li, X. M., et  al. (2009). Elevated immune response in the brain of autistic patients. Journal of Neuroimmunology, 207, 111–116. Lintas, C., & Persico, A. M. (2009). Autistic phenotypes and genetic testing: state-of-the-art for the clinical geneticist. Journal of Medical Genetics, 46, 1–8. Lintas, C., Sacco, R., Garbett, K., Mirnics, K., Militerni, R., Bravaccio, C., et  al. (2009). Involvement of the PRKCB1 gene in autistic disorder: significant genetic association and reduced neocortical gene expression. Molecular Psychiatry, 14(7), 705–718. Martin, L. A., Ashwood, P., Braunschweig, D., Cabanlit, M., Van de Water, J., & Amaral, D. G. (2008). Stereotypies and hyperactivity in rhesus monkeys exposed to IgG from mothers of children with autism. Brain, Behavior, and Immunity, 22, 806–816. Messahel, S., Pheasant, A. E., Pall, H., Ahmed-Choudhury, J., Sungum-Paliwal, R. S., & Vostanis, P. (1998). Urinary levels of neopterin and biopterin in autism. Neuroscience Letters, 241, 17–20. Meyer, U., Nyffeler, M., Engler, A., Urwyler, A., Schedlowski, M., Knuesel, I., et al. (2006). The time of prenatal immune challenge determines the specificity of inflammation-mediated brain and behavioral pathology. The Journal of Neuroscience, 26, 4752–4762. Miller, M. T., Stromland, K., Ventura, L., Johansson, M., Bandim, J. M., & Gillberg, C. (2005). Autism associated with conditions characterized by developmental errors in early embryogenesis: a mini review. International Journal of Developmental Neuroscience, 23, 201–219. Molloy, C. A., Morrow, A. L., Meinzen-Derr, J., Schleifer, K., Dienger, K., Manning-Courtney, P., et al. (2006). Elevated cytokine levels in children with autism spectrum disorder. Journal of Neuroimmunology, 172, 198–205. Mouridsen, S. E., Rich, B., Isager, T., & Nedergaard, N. J. (2007). Autoimmune diseases in parents of children with infantile autism: a case-control study. Developmental Medicine and Child Neurology, 49, 429–432. Nishimura, Y., Martin, C. L., Vazquez-Lopez, A., Spence, S. J., Alvarez-Retuerto, A. I., Sigman, M., et al. (2007). Genome-wide expression profiling of lymphoblastoid cell lines distinguishes different forms of autism and reveals shared pathways. Human Molecular Genetics, 16, 1682–1698. Odell, D., Maciulis, A., Cutler, A., Warren, L., McMahon, W. M., Coon, H., et  al. (2005). Confirmation of the association of the C4B null allelle in autism. Human Immunology, 66, 140–145. Pardo, C. A., Vargas, D. L., & Zimmerman, A. W. (2005). Immunity, neuroglia and neuroinflammation in autism. International Review of Psychiatry, 17, 485–495. Patterson, P. H. (2002). Maternal infection: window on neuroimmune interactions in fetal brain development and mental illness. Current Opinion in Neurobiology, 12, 115–118.

486

R. Sacco et al.

Patterson, P. H. (2009). Immune involvement in schizophrenia and autism: etiology, pathology and animal models. Behavioural Brain Research, 204(2), 313–321. Persico, A. M., & Bourgeron, T. (2006). Searching for ways out of the autism maze: genetic, epigenetic and environmental clues. Trends in Neurosciences, 29, 349–358. Piven, J., Palmer, P., Jacobi, D., Childress, D., & Arndt, S. (1997). Broader autism phenotype: evidence from a family history study of multiple-incidence autism families. The American Journal of Psychiatry, 154, 185–190. Plioplys, A. V., Greaves, A., Kazemi, K., & Silverman, E. (1994). Lymphocyte function in autism and Rett syndrome. Neuropsychobiology, 29, 12–16. Ponzio, N. M., Servatius, R., Beck, K., Marzouk, A., & Kreider, T. (2007). Cytokine levels during pregnancy influence immunological profiles and neurobehavioral patterns of the offspring. Annals of the New York Academy of Sciences, 1107, 118–128. Purcell, A. E., Jeon, O. H., Zimmerman, A. W., Blue, M. E., & Pevsner, J. (2001). Postmortem brain abnormalities of the glutamate neurotransmitter system in autism. Neurology, 57, 1618–1628. Reichelt, W. H., Knivsberg, A. M., Nodland, M., Stensrud, M., & Reichelt, K. L. (1997). Urinary peptide levels and patterns in autistic children from seven countries, and the effect of dietary intervention after 4 years. Developemental Brain Dysfunctions, 10, 44–55. Sacco, R., Militerni, R., Frolli, A., Bravaccio, C., Gritti, A., Elia, M., et  al. (2007). Clinical, ­morphological, and biochemical correlates of head circumference in autism. Biological Psychiatry, 62, 1038–1047. Shi, L., Fatemi, S. H., Sidwell, R. W., & Patterson, P. H. (2003). Maternal influenza infection causes marked behavioral and pharmacological changes in the offspring. The Journal of Neuroscience, 23, 297–302. Shi, L., Smith, S. E., Malkova, N., Tse, D., Su, Y., & Patterson, P. H. (2009). Activation of the maternal immune system alters cerebellar development in the offspring. Brain, Behavior, and Immunity, 23, 116–123. Singer, H. S., Morris, C. M., Gause, C. D., Gillin, P. K., Crawford, S., & Zimmermann, A. W. (2008). Antibodies against fetal brain in sera of mothers with autistic children. Journal of Neuroimmunology, 194, 165–172. Singh, V. K. (1996). Plasma increase of interleukin-12 and interferon-gamma. Pathological ­significance in autism. Journal of Neuroimmunology, 66, 143–145. Smith, S. E., Li, J., Garbett, K., Mirnics, K., & Patterson, P. H. (2007). Maternal immune ­activation alters fetal brain development through interleukin-6. The Journal of Neuroscience, 27, 10695–10702. Sperner-Unterweger, B., Winkler, C., & Fuchs, D. (2006). Immune activation in autism. Pediatric Neurology, 34, 333. Sweeten, T. L., Posey, D. J., & McDougle, C. J. (2003). High blood monocyte counts and ­neopterin levels in children with autistic disorder. The American Journal of Psychiatry, 160, 1691–1693. Sweeten, T. L., Posey, D. J., Shankar, S., & McDougle, C. J. (2004). High nitric oxide production in autistic disorder: a possible role for interferon-gamma. Biological Psychiatry, 55, 434–437. Tang, Y., Schapiro, M. B., Franz, D. N., Patterson, B. J., Hickey, F. J., Schorry, E. K., et al. (2004). Blood expression profiles for tuberous sclerosis complex 2, neurofibromatosis type 1, and Down’s syndrome. Annals of Neurology, 56, 808–814. Tani, Y., Fernell, E., Watanabe, Y., Kanai, T., & Långström, B. (1994). Decrease in 6R–5, 6, 7, 8-tetrahydrobiopterin content in cerebrospinal fluid of autistic patients. Neuroscience Letters, 181, 169–172. Torres, A. R., Maciulis, A., & Odell, D. (2001). The association of MHC genes with autism. Frontiers in Bioscience, 6, D936–D943. Tuchman, R., & Rapin, I. (2002). Epilepsy in autism. Lancet Neurology, 1, 352–358. Vargas, D. L., Nascimbene, C., Krishnan, C., Zimmerman, A. W., & Pardo, C. A. (2005). Neuroglial activation and neuroinflammation in the brain of patients with autism. Annals of Neurology, 57, 67–81.

Genome-Wide Expression Studies in Autism-Spectrum Disorders

487

Vervliet, G., & Schandene, L. (1985). In vitro correction of the interleukin-2 and interferongamma defect in multiple sclerosis. Clinical and Experimental Immunology, 61, 556–561. Vojdani, A., Campbell, A. W., Anyanwu, E., Kashanian, A., Bock, K., & Vojdani, E. (2002). Antibodies to neuron-specific antigens in children with autism: possible cross-reaction with encephalitogenic proteins from milk, Chlamydia pneumoniae and Streptococcus group A. Journal of Neuroimmunology, 129, 168–177. Warren, R. P., Foster, A., & Margaretten, N. C. (1987). Reduced natural killer cell activity in autism. Journal of the American Academy of Child and Adolescent Psychiatry, 26, 333–335. Warren, R. P., Odell, J. D., Warren, W. L., Burger, R. A., Maciulis, A., Daniels, W. W., et  al. (1996). Strong association of the third hypervariable region of HLA-DR beta 1 with autism. Journal of Neuroimmunology, 67, 97–102. Warren, R. P., Odell, J. D., Warren, W. L., Burger, R. A., Maciulis, A., Daniels, W. W., et  al. (1997). Immunoglobulin A deficiency in a subset of autistic subjects. Journal of Autism and Developmental Disorders, 27, 187–192. Warren, R. P., Yonk, L. J., Burger, R. A., Cole, P., Odell, J. D., Warren, W. L., et  al. (1990). Deficiency of suppressor-inducer (CD4+CD45RA+) T cells in autism. Immunological ­investigations, 19, 245–251. Wills, S., Cabanlit, M., Bennett, J., Ashwood, P., Amaral, D. G., & Van de Water, J. (2009). Detection of autoantibodies to neural cells of the cerebellum in the plasma of subjects with autism spectrum disorders. Brain, Behavior, and Immunity, 23, 64–74. Wirleitner, B., Neurauter, G., Schröcksnadel, K., Frick, B., & Fuchs, D. (2003). Interferongamma-induced conversion of tryptophan: immunologic and neuropsychiatric aspects. Current Medicinal Chemistry, 10, 1581–1591. Yamashita, Y., Fujimoto, C., Nakajima, E., Isagai, T., & Matsuishi, T. (2003). Possible association between congenital cytomegalovirus infection and autistic disorder. Journal of Autism and Developmental Disorders, 33, 455–459. Zilbovicius, M., Meresse, I., Chabane, N., Brunelle, F., Samson, Y., & Boddaert, N. (2006). Autism, the superior temporal sulcus and social perception. Trends in Neurosciences, 29, 359–366. Zimmerman, A. W., Connors, S. L., Matteson, K. J., Lee, L. C., Singer, H. S., Castaneda, J. C., et  al. (2007). Maternal antibrain antibodies in autism. Brain, Behavior, and Immunity, 21, 351–357. Zimmerman, A. W., Jyonouchi, H., Comi, A. M., Connors, S. L., Milstien, S., Varsou, A., et al. (2005). Cerebrospinal fluid and serum markers of inflammation in autism. Pediatric Neurology, 33, 195–201.

Protein Expression Profile of Alzheimer’s Disease Mouse Model Generated by Difference Gel Electrophoresis (DIGE) Approach Daria Sizova

Abstract  Alzheimer’s disease (AD) is a well-described neurological disorder characterized by the presence of a number of biological processes. Deposition of b-amyloid plaques in brain, one of the main features of AD pathology, was ­successfully modeled in the middle of the 1990s by a generation of the b-amyloid precursor protein (APP) transgenic mouse lines. The Thy1-APP751SL transgenic line, one of this “new generation” AD animal models, reproduces several ­characteristic features of Alzheimer`s disease such as Ab peptide deposition, dystrophic neurite formation, and progressive neuronal death. A recently developed technique named difference gel electrophoresis (DIGE) made identification of qualitative and quantitative differences in protein expression between two biological samples considerably more sensitive, straightforward and reliable. Application of DIGE allows the study of the AD-related proteome to progress with a new level of confidence. Here, we describe application of a novel DIGE technique coupled with MS protein identification to the generation of a distinct AD-related protein expression profile using cortices of 14-monthold ­Thy1-APP751SL transgenic mice. Using this approach, we identified 15 ­different proteins which are significantly regulated in AD pathology. Resulting AD-related proteome comprises a number of proteins that were already known to be ­implicated in AD and neurodegeneration, as well as several proteins for which relationship with AD had not been shown before. Identified proteins were grouped according to their key biological pathways. Acquired data are discussed in the view of existing literature on the AD proteome. Keywords  Alzheimer’s disease • Proteomics • DIGE technique • Biomarkers

D. Sizova (*) Department of Genetics, University of Pennsylvania School of Medicine, Clinical Research Building, Room 755, 415 Curie Boulevard, Philadelphia, PA 19104-6149, USA e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_19, © Springer Science+Business Media, LLC 2011

489

490

D. Sizova

1 Introduction Alzheimer’s disease (AD) is the most common form of dementia. It gradually destroys a person’s memory and ability to organize visual information. It causes language deterioration, brings changes in behavior and personality and has ­ultimately a fatal progression. Two different types of AD are distinguished: (1) family (early onset), and (2) sporadic (late onset with no clear family history). The latter represents the overwhelming majority of cases. Although the early symptoms of AD can be easily confused with mental changes taking place in aging, this ­condition is a well-described neurological disorder characterized by the presence of a number of distinct biological processes. Deposition of b-amyloid plaques in brains of affected individuals is believed to be the “prime mover” of the pathogenesis. Ab peptide, the major constituent of neuritic plaques and vascular deposits in AD patients, is derived by proteolysis from the larger b-amyloid precursor protein (APP). Point mutations of the APP gene have been identified in early onset AD cases in some families. Development of neurofibrillary tangles (filamentous inclusions in neuronal cell bodies and ­proximal dendrites) are the second basic biological feature of AD. These tangles are predominantly composed of a filamentous, hyperphosphorylated form of the microtubule-associated protein Tau (Wolfe, 2002) Although AD symptoms were first described almost 100 years ago, there is neither early diagnostics nor effective treatment for this disease at the present time. This fact can be explained in part by the absence of good animal models for this neurological disorder for a long period of time. Only at the end of the 1990s, several transgenic mouse lines based on expression of different mutant human APPs have been created. These mice carry mutant APPs transgenes with an optimized translation initiation site under the control of brain specified promoters. As a result, they display several main features of AD, which makes them appropriate animal models for AD-related studies. Study of AD proteome is essential not only for a better understanding of the molecular basis of this disease but also for defining new targets and biological markers for early diagnostics and new treatment options. Since a single suitable biomarker is still eluding identification, a valuable test for distinguishing AD from other dementia cases could be based on a pathology specific pattern of changes in protein expression (Castano, Roher, Esh, Kokjohn, & Beach, 2006). The distinct progress that was made in the field of proteomics during the last decade brought several new techniques which allowed the generation of protein expression profiles on a new level of confidence. One of these new approaches is difference gel ­electrophoresis (DIGE). In the present chapter, we describe application of a novel DIGE technique coupled with MS protein identification to a generation of a distinct AD-related protein expression profile using cortices of 14-month-old APP ­transgenic mice. Acquired data are discussed in relation to existing literature on AD proteome.

Protein Expression Profile of Alzheimer’s Disease

491

2 Transgenic Mouse Model as a Tool for Study of Alzheimer’s Disease Proteome The lack of an accurate animal model of AD was a serious obstacle to the ­understanding of this neurodegenerative disorder in terms of development of the disease, identification of the main biological processes involved in Ab peptide deposition and neurofibrillary tangles formation, and, as a consequence, ­development of possible diagnostic and treatment options. Indeed, only model animals could provide scientists with a possibility to work with freshly removed and highly ­homogeneous samples and also to study this disease at different developmental stages. Initial attempts to develop transgenic mouse models of AD were unsuccessful (Price & Sisodia, 1994). At the end of the 1990s, animal lines based on the expression of different mutant forms of human APP were created. These transgenic animals carry the mutant APP genes under control of brain-specified promoters with optimized translation initiation sites. The Thy1-APP751SL transgenic mice that were used in the present study are members of this “new generation” AD models and reproduce several characteristic features of Alzheimer`s disease such as Ab peptide deposition, dystrophic neurites formation, and progressive neuronal death (Blanchard et al., 2003). The Thy1-APP751SL transgenic mice (Tg) were generated as described in Blanchard et  al. (2003) by introducing a transgene carrying the human APP751 cDNA with the London (V642I) and Swedish (KM595/596NL) mutations under control of the murine Thy1 promoter (Thy1-APP751). Both mutations are described in two well-studied familial forms of AD; a brain-specific Thy1 promoter provides tissue-targeted expression of the corresponding transgene. To improve the translation initiation site of APP, an optimized Kozack consensus sequence was introduced. As reported in Blanchard et al. (2003), the resulting transgenic mice produce high levels of human amyloid Ab peptide in the brain. At 6 months of age, they develop fibrillar amyloid deposits, associated with dystrophic neurites and neuronal stress markers, therefore representing an appropriate animal model for AD. However, it should be mentioned that such mice represent only a partial model for Alzheimer`s disease as they produce amyloid plaques but not neurofibrillary tangles, the second basic feature of AD. In this case, proteomic analysis provides a picture reflecting cellular response to the presence of amyloid deposition. Although the Thy1-APP751SL transgenic model does not reproduce all major signs of AD, use of transgenic mouse samples has a number of advantages compared to work with human samples, as done in some previous AD-related proteomic studies (Castegna, Aksenov, Aksenova, et al., 2002; Castegna, Aksenov, Thongboonkerd, et al., 2002; Schonberger, Edgar, Kydd, Faull, & Cooper, 2001; Tsuji, Shiozaki, Kohno, Yoshizato, & Shimohama, 2002). Human samples directly reflect AD pathology but they differ in at least four ways: (1) there is extreme heterogeneity of samples in terms of postmortem delay leading to different extent of protein degradation; (2)  the stage of disease development may differ a lot from individual to individual; (3) patients` ages

492

D. Sizova

may vary; and (4) other pathologies may coexist. All these issues are not actual when an appropriate transgenic model is available. Disease ­progression takes place at a similar rate in each mouse while use of wild-type (wt) littermates as a control allows work on the same genetic background. The use of freshly removed and perfused brain cortices for sample preparation makes resulting samples extremely homogeneous. Consequently, proteomic analysis of APP t­ransgenic mouse model provides a profile that is related directly to the amyloid pathology and is not affected by inter-individual differences. In the present study we used 14-month-old Thy1-APP751 transgenic animals (Tg) and wt littermates from a pure C57Bl6 background. It was demonstrated that these mice exhibit many amyloid deposits in various brain regions, in particular the ­cortex or the hippocampus, as well as memory deficits in behavioral tests (Piot-Grosjean, Cudennec, Dhenain, Delay-Goyet, & Delatour, 2004). Two groups of six males and six females (12 Tg animals and 12 wt littermates in total) were chosen for our study. Separate proteomic assays were performed for males and females since it has been observed that the kinetics of intracellular Ab peptide accumulation as well as plaque deposition slightly differed depending of mouse sex. It was noticed that females on average start developing histological signs of amyloid deposition several weeks earlier than males. However, no significant difference between male and female samples was detected at the proteomic level (Sizova et al., 2007). To prepare tissue samples, mice were first anesthetized and perfused transaortically with phosphate-buffered saline to eliminate serum protein contamination before removing the brain. Cortices were used for the present study because plaque deposits are most abundant in this brain region. The cortices were removed as described in Glowinski and Iversen (1966), and then rapidly frozen on dry ice and kept at −80°C.

3 Generation of AD Proteomic Profile by DIGE Analysis A novel approach named DIGE and developed by Amersham Biosciences was applied to generate a complete proteomic profile of the AD mouse model. This technique has been developed especially for identifying qualitative and quantitative differences in protein expression between two biological samples (Orange, Hawkins, Story, & Fowler, 2000). Several essential points make this approach ­considerably better than other proteomic techniques previously used for this ­purpose. The application of large 2D gels and fluorescent dyes allows a high level of resolution and sensitivity. In addition, running the transgenic and control ­samples on the same gel eliminates the usual pitfall of matching of 2D patterns from different gels and makes the comparison more straightforward and reliable. Finally, the efficient and automated computer image analysis allows a thorough and robust quantification and statistical analysis, in order to select only significant changes (Marouga, David, & Hawkins, 2005).

Protein Expression Profile of Alzheimer’s Disease

493

3.1 Sample Preparation and Labeling Protein samples were prepared from the whole individual cortices by lyophilization followed by dry tissue crushing in aluminum foil and subsequent protein extraction and solubilization in the urea/thiourea-containing solution. The protein concentration of the extracts was determined by the Bradford method, and final supernatants were aliquoted and stored at −80°C. Four pools of protein extracts were generated according to mouse type and sex (Tg male, wt male, Tg female, wt female) by ­mixing equal protein amount of cortex extract for each of the six corresponding animals. Protein extract pools were conjugated with fluorescent dyes through N-hydroxysuccinimidyl linkages. This procedure successfully labels about 10% of the protein molecules. Sample labeling procedure is schematically represented in Fig. 1a, step 1. For analytical gels, 50 mg of proteins from wt extract pool (white circle in Fig. 1a), 50 mg of proteins from Tg extract pool (black circle), or a mixture of both (25 mg wt + 25 mg Tg, dashed circle), were incubated with Cy2, Cy5 or Cy3 cyanine dyes, respectively. For calibrating gels, the mixture of 25  mg of proteins from wt extract pool and 25 mg of proteins from Tg extract pool (dashed circle) was incubated in parallel with each of the three Cy2, Cy5 and Cy3 dyes; resulting samples were pooled before gel loading.

3.2 Protein Separation and Determination of Differentiating Spots All samples were separated by two-dimensional gel electrophoresis (2D-E). First, dimension electrophoresis was carried out using narrow immobilized pH gradient gels (pH 4.5–6.0, 5.5–6.7, 6.0–9.0) and a horizontal electrophoresis apparatus. Then, 50 mg of labeled proteins (for analytical and calibrating gels) or 500 mg of unlabeled proteins (for preparative gels) were loaded and isoelectric focusing (IEF) was performed. After equilibrating in a solution containing 0.1 M Tris-HCl, 36% urea, 30% glycerol, and either 0.5% DTT or 4.5% iodoacetamide, pH gradient gel strips were ready to be applied to the second dimension gels (12.7% SDS-PAGE). For each condition (mouse gender and pH), three analytical gels, three calibration gels, and four preparative gels were run in parallel. Preparative gels containing unlabeled protein samples were stained with the SYPRO Ruby dye. Differentiating protein spots were determined by the DIGE approach. Experimental design of a single DIGE Assay is shown in Fig. 1a. The whole experimental procedure can be represented as a sequence of three consecutive steps: (1) experimental determination of a threshold value which allows the determination of the level of experimental noise (calibrating assay), (2) determination of regulated spots (analytical assay), and (3) physical matching of spots on preparative gels for extraction of the regulated proteins and identification by MS (preparative assay). To  make the result of calibrating and analytical assays statistically significant,

494

D. Sizova

a

Single DIGE Assay Calibrating gel × 3

Analytical gel × 3

Preparative gel × 4 Wt gel × 2 Tg gel × 2

Main steps

Cy5 Cy3 Cy2

Cy5 Cy3 Cy2

No labeling

1. Sample labeling

2. 2-D separation

All ten 2-DE gels are run in parallel

3. Gel staining

No staining

Staining with SYPRO Ruby

4. Gel scanning

At three wavelengths (for Cy2, Cy3, and Cy5)

At one wavelength (for SYPRO Ruby)

Calculation of a threshold value

5. Data treatment

Identification of regulated spots

Excision of regulated spots MS protein identification

b

4.5

6.0

6.0

First DIGE Assay

Second DIGE Assay 5.5

4.5

9.0

Third DIGE Assay

Complete DIGE Analysis

6.7

pH range

9.0

Fig. 1  Experimental design of difference gel electrophoresis (DIGE) analysis. (a) Three calibrating, three analytical, and four preparative (two with wt and two with Tg samples) gels are run in parallel in a single DIGE assay. A mixture of equivalent amounts of wt and Tg samples (dashed circle) is labeled separately with each of the three Cy dyes and resulting blend is loaded on calibrating gels. In the case of analytical gels wt samples (white circle) are labeled with Cy2, Tg samples (black circle) – with Cy5, and a mixture wt/Tg (dashed circle) – with Cy3. Ten-times-higher amount of non-labeled wt (large white circle) or Tg samples (large black circle) is loaded on preparative gels. All gels are read at appropriate wavelengths (preparative gels are previously stained with SYPRO Ruby) and corresponding pictures are treated to calculate a threshold value (calibration gels), identify regulated spots (analytical gels), and excise regulated spots for MS protein identification (preparative gels). (b) One complete DIGE analysis is composed of three independent DIGE assays (each one with its own pH range of the first direction of 2-DE) to cover the whole possible protein pIs range

Protein Expression Profile of Alzheimer’s Disease

495

c­ orresponding gels were run in triplicate. The preparative gels were run in duplicate to ensure enough material for identification of regulated spots. Ten gels (three ­calibrating, three analytical and four preparative) run altogether in parallel constitute one single DIGE assay. Moreover, to improve protein spot resolution, each comparison (Tg male vs. wt male, and Tg female vs. wt female) was carried out using three overlapping linear pH gradients for the first dimension electrophoresis: pH 4.5–6.0, pH 5.5–6.7, and pH 6.0–9.0 (see Fig. 1b). Thus, three complete DIGE assays (for three different pH gradients) constituted one complete DIGE analysis for each mouse sex group. In the calibrating assay (Fig. 1a, left panel), the mixture of equivalent amounts of Tg and wt pools was labeled with three different dyes (Cy2, Cy3, or Cy5) and the resulting protein mixture was separated by 2D-E. Three digital pictures were then collected using a ProXpress fluorescent gel scanner (Perkin Elmer, Norwalk, CT, USA). Gel images were normalized by adjusting the exposure times according to the average pixel values observed. The quantification of spot abundance ratios was ­carried out using the DeCyder software (Amersham Bioscience), specifically developed for 2D-DIGE in complex with the experimental design described above. First, for each gel, the Cy2, Cy5, and Cy3 images were merged, allowing the co-detection of spot boundaries on the three images. For each spot, the spot volume (sum of pixel intensities) was calculated in the Cy2 or Cy5 channels and then ­normalized according to the corresponding Cy3 spot volume. Comparison of ­normalized Cy2 and Cy5 protein expression intensities within each gel gave a ­standardized expression ratio. This value was compared across all gels for each matched spot, and an average value was calculated using the triplicate values from each experimental condition. The threshold of significance for differentially regulated spots was defined from the analysis of three calibration gels, loaded with a mixture of wt plus Tg proteins labeled with each of the three dyes (Cy2, Cy3, and Cy5). After the image analysis described above, the distribution of normalized expression ratios was approximated by a Gaussian distribution and the ratio value corresponding to 2 standard deviations (95% confidence limit) was determined. Since in the calibration assay all three samples were the same Tg/wt mixture, the differences in spot intensities detected after comparison of the corresponding gel pictures resulted from experimental noise only. These experimental noise variations were calculated for three independent calibrating gels run in parallel. In the present assay, the ratio corresponding to the 95% confidence interval was found to be very close to 1.5, thus this value was taken as a threshold level for the subsequent analytical assay. For the analytical assay (Fig. 1a, central panel), a set of triplicate analytical gels was run to identify spots differing in protein amounts between Tg and wt samples. They were run exactly in the same way as calibrating gels, except that the three different samples were labeled with the three different dyes: 50 mg of proteins from wt extract pool are labeled with Cy2, 50 mg of proteins from Tg extract pool are labeled with Cy5, and a reference sample corresponding to a mixture of both (25 mg wt + 25 mg Tg) is labeled with Cy3. For each spot, an expression ratio was calculated between the Tg and wt channels after normalization using the Cy3-labeled reference sample as described above. The Tg/wt ratio was calculated for the three

496

D. Sizova

analytical gels, and the final value taken into account was an average of these three independent measurements. According to the threshold level detected in the ­calibration assay, only spots with expression ratios higher than 1.5 were considered as significantly regulated in the present study. By convention, when a calculated ratio was smaller than 1, the inverse ratio was indicated with a negative sign. A Student’s t test was performed and the corresponding p value was calculated for each spot ratio. These values reflect the probability of obtaining the observed data if the two groups (Tg and wt) had the same protein intensity and show the statistical significance of the results. Spots were defined as up-regulated when their ­expression was increased in Tg compared with wt, and down-regulated when their expression was decreased in Tg. Preparative gels (Fig.  1a, right panel) were loaded with non-labeled protein samples, using 10 times more proteins than analytical gels. Two gels were run with the wt samples, and two gels with the Tg samples, to ensure recovery of enough material for spot identification. After gel staining with the SYPRO Ruby fluorescent dye, spots found to be significantly regulated after analysis of the Cy-dyelabeled images were matched on the SYPRO Ruby protein patterns. Selected spots were then excised from preparative gels using an automated spot picker (Amersham Biosciences). Up-regulated spots were excised from preparative gels containing Tg protein samples, and down-regulated from preparative gels containing wt samples. Finally, excised spots were analyzed by mass spectrometry for identification of the corresponding protein(s). The whole set of spots found to be significantly regulated in the present study (35 in total) was distributed among all six DIGE experiments. However, a higher number of spots was found in the pH gradient 4.5–6.0 (14 spots for males, 5 spots for females) than in the two others (pH 5.5–6.7: 4 spots for males, 6 spots for females; pH 6.0–9.0: 3 spots for males, 2 spots for females). All 35 spots were excised and subjected to mass-spectrometry identification.

3.3 MS Identification of Regulated Proteins Excised regulated spots were prepared for protein identification as follows. In-gel digestion was performed with an automated protein digestion system. The gel slices were washed with NH4HCO3 and acetonitrile. The cysteine residues were reduced by dithiothreitol and alkylated by iodoacetamide. After dehydration with acetonitrile, the proteins were cleaved in the gel with modified porcine trypsin. Resulting peptides were extracted with acetonitrile in formic acid and then used for MALDI-TOF-MS and/or nanoLC-MS-MS. For MALDI-TOF-MS analysis, 0.5 ml of peptide extract was co-crystallized in the matrix a-cyano-4-hydroxycinnamic acid. MALDI mass measurements were carried out on an Ultraflex™ TOF/TOF (Bruker Daltonics, USA). The spectra were internally calibrated using two trypsin autolysis peaks. Proteins were identified by peptide mass fingerprinting using the program MASCOT (Matrix Science, UK)

Protein Expression Profile of Alzheimer’s Disease

497

against SWISS-PROT and TrEMBL databases. One missed cleavage per peptide was allowed, and a mass tolerance was set up at 35 ppm. Search was performed on all species available in the databases, and some variable modifications (carbamidomethylation for cysteine and oxidation for methionine) were taken into account. In a few cases, MALDI-TOF-MS gave ambiguous results or none at all. Then we selected corresponding samples for an additional round of tandem mass ­spectrometry coupled with chromatographic separation using a capillary LC-MS/ MS system. This methodology was essential for the identification of low ­expression and/or low molecular weight proteins for which MALDI-TOF-MS analysis ­generated poor signals. Such hierarchical approach was effective in reducing the number of samples for analysis by the time-consuming LC-MS/MS procedure. The two approaches (MALDI-TOF-MS and nanoLC-MS/MS) are complementary, and the use of two different ionization processes also contributed to the detection of a wider range of peptides. LC-MS/MS procedure was performed as follows. Chromatographic separations were conducted on a reversed-phase (RP) capillary column with a 200 nl/min flow. Nanoscale capillary liquid chromatography-tandem mass spectrometric (LC-MS-MS) analysis of the digested proteins was performed using a CapLC capillary LC system coupled to a hybrid quadrupole orthogonal acceleration time-offlight tandem mass spectrometer. Mass data acquisitions were piloted by MassLynx software. Mass data collected during a LC-MS/MS analysis were processed and converted into a PKL file to be submitted to the search software MASCOT. Searches were conducted with a tolerance on mass measurement of 150 ppm in MS mode and 0.25 Da in MS/MS mode. The hierarchical approach described above (MALDI-TOF-MS followed by nanoLCMS/MS) allowed us to identify proteins in all 35 regulated spots that were found to be significantly regulated in the present study. Twenty-two spots were ­successfully analyzed by MALDI-TOF-MS, and the corresponding proteins were identified by peptide mass fingerprinting using the Mascot search engine against SWISS-PROT and TrEMBL databases. The average sequence coverage of the identified proteins was around 45%. Eighteen spots were successfully analyzed by LC-MS/MS, and the corresponding proteins were identified by database search using generated collision-induced dissociation MS/MS spectra. In this case the sequence coverage was lower (26% in average), but this was compensated by ­protein sequence information. When the coverage was lower than 20%, a manual sequencing was performed on several peptides to confirm the presence of the ­protein in the mixture. Five spots were analyzed by both techniques, leading quite often to identification of the same proteins, thus confirming ambiguous MALDI-TOF-MS data. However, in several cases, additional proteins were identified by LC-MS/MS. Although no precise calibration of isoelectrical point or molecular weight was made on the 2D gels, we ensured that the theoretical pI and MW of identified proteins were consistent with the pH range where the spot had been detected and with its height on the gel. Finally, we summarized protein identification results in order to retrieve entries corresponding to mouse proteins, and to replace distinct database entries corresponding in fact to the same protein.

498

D. Sizova

3.4 Analysis of Resulting Protein Set It is important to emphasize that unambiguous protein identification was not possible for all 35 regulated spots. For 28 spots out of 35, a unique protein per spot was identified, and was then considered to be responsible for the observed variation in spot intensity. The remaining spots (7) were found to contain several proteins, and the regulated protein could thus not be unequivocally assigned. However, for six of them, we observed that one protein had already been unequivocally identified in another regulated spot, and that it was often identified with a better sequence coverage than the other proteins found in the same spot. Therefore, we considered that this protein was likely to account for the variation observed. The last spot contains two proteins unique to this spot and identified with similar MS sequence coverage. In this case, no experimental element allowed favoring one of the two proteins, and we further considered both as potentially regulated in the transgenic animals. A second important point to be noted is that several proteins were identified in more than one spot. Among the 35 spots identified, 11 spots contained the glial fibrillary acidic protein (GFAP), 4 spots contained the apolipoprotein E (ApoE), 3 spots contained the dihydropyrimidinase-related protein 2 (DRP-2, also known as CRMP2), and 3 spots contained the peroxiredoxin 6 (Prxd-6). Thus, these 21 spots account for only four distinct proteins. Multiple spots for one single protein result from three different situations. First, they can be found in gels run with different pH gradients, because of some overlap between the three pH ranges (e.g., DRP-2, Prxd-6) or, second, they can be identified in gels run with similar pH gradients but with different mouse sex samples. In these cases, they correspond to the identification of the same protein in independent experiments, and they are regulated in the same way and with comparable ratios. Alternatively, different spots corresponding to the same gene can be found in the same gel (the same gradient, the same mouse sex) suggesting the regulation of several isoforms from the same protein. This is the case for GFAP, which was found in 8 spots in one experiment (male samples, pH gradient 4.5–6.0) and in 3 spots in a separate experiment (female samples, pH gradient 4.5–6.0). The existence of several transcripts and numerous protein isoforms is well documented for this protein (Condorelli et al., 1999; Lubec, Krapfenbauer, & Fountoulakis, 2003; Lubec et al., 1999). Isoforms and post-translational events are not necessarily regulated in the same way or to the same extent under conditions where protein expression is modified. Interestingly, in our study, the various ­isoforms identified for the GFAP protein are all regulated in the same way (up-regulated), but with some variability in the intensity of the regulation depending on the spot/isoform considered. This was also observed for serum albumin, ApoE and CaMK II. As a result of this analysis, 15 proteins were identified as regulated in the cortex of Tg mice compared to wt littermates. They are all listed in Table 1. Eight of these proteins were identified in more than one spot, among which five were regulated in both male and female samples. The remaining proteins were identified more often only in the male group. A possible explanation for this discrepancy is that females, known

Olfactory marker protein Moesin N-ethylmaleimide sensitive fusion protein Others

Q64288 P26041 Q8C3R2

−2.0    1.5    1.5

Cottrell et al. (2005) and Schonberger et al. (2001)

Table 1  Proteomic profile of AD mouse model generated by DIGE analysis AD related studies SwissProt Average Protein name number ratio Proteomic Inflammation and oxidative stress Glial fibrillary acidic protein P03995    2.4 Cottrell et al. (2005), Kanninen, Goldsteins, Auriola, Alafuzoff, and Koistinaho (2004), Shin et al. (2004), and Tsuji et al. (2002) Complement C1q subcomponent, P14106    1.7 B chain precursor Peroxiredoxin 6 Q91WT2    1.8 Schonberger et al. (2001) Cholesterol metabolism Apolipoprotein E precursor P08226    2.2 Vercauteren et al. (2004) ACAT2 Q8CAY6    1.5 Neuronal and synaptic signaling Dihydropyrimidinase related O08553    1.7 Castegna, Aksenov, Thongboonkerd, et al. protein-2 (2002), Kanninen et al. (2004), Lubec et al. (1999), Schonberger et al. (2001), Shin et al. (2004), Tsuji et al. (2002), and Vercauteren et al. (2004) CaMK-II alpha subunit P46096 −1.6 Synaptotagmin I P46096    1.5 Schonberger et al. (2001)

(continued)

Ginsberg, Hemby, Lee, Eberwine, and Trojanowski (2000) and Yao et al. (2003)

Dickey et al. (2003)

Dickey et al. (2003)

Dickey et al. (2003)

Transcriptomic

Protein Expression Profile of Alzheimer’s Disease 499

P07724

Serum albumin precursor

   4.3

Average ratio (Schonberger et al. (2001) and Vercauteren et al. (2004)

AD related studies Proteomic

Transcriptomic

Serotransferrin precursor Q921I1    2.2 Dickey et al. (2003) Pyruvate kinase 3/M2 Q91YI8    1.6 Vercauteren et al. (2004) Dynamin 1-like Q8BNQ5    1.5 Regulated proteins identified by DIGE analysis are grouped according to the main known AD-related biological pathways. Protein names and SwissProt accession numbers are given together with their average ratio of regulation obtained by DIGE. References to AD-related studies where these proteins were reported previously (if any) are cited in the last two columns

SwissProt number

Protein name

Table 1  (continued)

500 D. Sizova

Protein Expression Profile of Alzheimer’s Disease

501

to have a more complex hormonal regulation, may have a greater inter-­individual variability which would mask less significant variations after pooling the individual samples. Most proteins were found to be up-regulated, with average ratios ranging from 1.5 to 4.3. Only two proteins (CaMK-II and olfactory marker protein [OMP]) were found to be down-regulated, with average ratios of −1.6 and −2.0, respectively.

4 Resulting Protein Expression Profile in Relation to the Current Knowledge of Alzheimer Disease Analysis of existing literature allowed us to group the resulting proteins in three major groups according to the biological pathways in which they are mainly involved (Table 1): inflammation and oxidative stress, cholesterol metabolism, and neuronal and synoptic signaling. Only 4 proteins out of 15 did not match with any of these three groups; they are listed separately. A literature survey revealed that eight of the identified proteins have been previously reported to be regulated in proteomic studies, either in AD human samples (Castegna, Aksenov, Thongboonkerd, et al., 2002; Cottrell et al., 2005; Kanninen et al., 2004; Lubec et al., 1999; Schonberger et al., 2001; Tsuji et al., 2002), or in transgenic animals (Shin et al., 2004; Vercauteren et al., 2004). We also confirmed the regulation of five proteins which had been identified as regulated only at the transcriptome level (Dickey et al., 2003; Ginsberg et al., 2000; Yao et al., 2003). Finally, in the present study, five proteins have been identified for the first time as regulated in an AD-related model. The inflammation and oxidative stress group of identified proteins comprises GFAP, complement protein C1q, and peroxiredoxin 6. The first two glial proteins are related to the inflammatory response, and both have been shown to be up-­ regulated in the present study. Chronic inflammation has been consistently observed in the brains of AD patients or transgenic mice which were developing amyloid plaques (Casserly & Topol, 2004; Matsuoka et al., 2001). Such response of damaged brain regions is believed to participate in the pathological process of AD, although its role is possibly both protective and harmful (Shen & Meri, 2003). Two characteristic features of inflammation are the presence of activated ­microglial cells and reactive astrocytes surrounding plaques, as well as expression of ­inflammatory mediators such as cytokines or complement factors (Casserly & Topol, 2004). GFAP is an intermediate filament protein that is specifically expressed in astrocytes. Moreover, it is known to be dramatically overexpressed during reactive astrogliosis (Messing & Brenner, 2003). C1q is the initial ­component of the classical complement pathway, and it can be secreted by both microglia and astrocytes. It was recently shown (van Beek, 2003, p. 18) that this protein binds Ab fibrillar aggregates, resulting in the activation of the complement cascade. In addition, C1q was observed to favor Ab fibrillogenesis, both in  vitro and in transgenic mice (Boyett et al., 2003; Webster, Glabe, & Rogers, 1995). Finally, in the absence of

502

D. Sizova

C1q in Tg2576 mice, the level of activated glia decreases without changing the number of amyloid plaques compared to regular Tg2576 animals (Fonseca, Zhou, Botto, & Tenner, 2004). These findings suggest a direct contribution of C1q to the pathology induced by amyloid plaque formation. Up-regulation of both GFAP and C1q has been described earlier in AD patient brain samples (Ross et  al., 2003; Yasojima, Schwab, McGeer, & McGeer, 1999), and in different amyloid transgenic models (Irizarry, Soriano, et al., 1997; Matsuoka et al., 2001), thus confirming the reliability of the present study. Moreover, in Kanninen et al. (2004)), it was shown that the level of glycosylation of GFAP is also increased in AD brains. Finally, immunochemistry studies demonstrated that GFAP and C1q are concentrated in the surrounding of plaques: GFAP was found in reactive astrocytes, and C1q was observed in microglia and amyloid plaques, both in AD patients brain samples and transgenic mouse brains (Afagh, Cummings, Cribbs, Cotman, & Tenner, 1996; Matsuoka et al., 2001). Peroxiredoxin 6 (Prxd-6), the third protein from the inflammation and oxidative stress group, also known as 1-Cys peroxiredoxin or antioxidant protein-2, is a bifunctional enzyme with glutathione peroxidase and phospholipase A2 activities (Chen, Dodia, Feinstein, Jain, & Fisher, 2000). Restoration of membrane integrity is believed to be its main function performed by both reducing and cleaving ­oxidized lipids. Overexpression and inactivation experiments showed a role of this protein in oxidative defense (Manevich et al., 2002; Wang et al., 2003). Up-regulation of Prxd-6 in the present model can be a reflection of changes in the oxidative status of the Tg mouse brains. This suggestion is supported by the fact that elevated oxidative stress was observed in human AD samples (Butterfield & Lauderback, 2002). Moreover, oxidative stress markers associated with plaque formation were found in other amyloid transgenic mouse models (Butterfield & Lauderback, 2002). Prxd-6 was also identified as up-regulated in AD human samples (Schonberger et  al., 2001), thus strengthening its potential role in pathology. Proteins involved in inflammation and oxidative damage are currently proposed as potential disease biomarkers in AD (Galasko, 2005). The other pathway known to be implicated in AD (Casserly & Topol, 2004) is cholesterol metabolism. An important role of this process in AD progression was demonstrated by retrospective studies showing a 70% lower prevalence of AD for individuals chronically treated with inhibitors of cholesterol synthesis (Wolozin, Brown, Theisler, & Silberman, 2004). Even more importantly, several cholesterolrelated gene polymorphisms are tightly associated with this disease. In particular, the presence of e4 form of the ApoE gene increases the risk of sporadic AD at least eightfold, making it the strongest identified risk factor for late-onset forms of AD (Buxbaum & Tagoe, 2000). In this respect, it is not surprising that two proteins that are significantly regulated in our model belong to the cholesterol metabolism group. These proteins are ApoE and acyl-coA: cholesterol acyltransferase 2 (ACAT-2). ApoE, the major apolipoprotein in the brain, is mainly synthesized by astrocytes. It is believed to play a role in cholesterol transport (Teunissen, Vente, Steinbusch, & Bruijn, 2002). APP transgenic mice have been shown to develop fewer plaques when they were deficient for ApoE (Bales et al., 1999), suggesting

Protein Expression Profile of Alzheimer’s Disease

503

that ApoE favors the formation of amyloid plaques. However, it is not clear whether this is due to the modulation of cholesterol metabolism (known to affect Ab ­processing) or to a direct effect on Ab aggregation or clearance (Bales, Dodart, DeMattos, Holtzman, & Paul, 2002). ApoE was previously shown to be up-­ regulated in AD human samples, and to be co-localized with amyloid plaques (Bales et al., 2002). ACAT-2, a second protein involved in the cholesterol metabolism pathway which is found to be up-regulated in our model, is one of two enzymes that catalyze the formation of cholesterol esters stored in cytosolic lipid droplets (Chang et al., 2001). Modulation of ACAT activity (either ACAT-1 or -2), and, consequently, of cholesterol ester production, is correlated with a parallel modulation of Ab generation (Puglielli et  al., 2001). This suggests a possible ­functional relationship of ACAT-2 with amyloid processing. Altogether, the increased level of expression of ApoE and ACAT-2 in our model could be an indication of altered cholesterol metabolism, possibly in relation with neuronal stress. Five other proteins found to be significantly regulated in the present AD model can be grouped according to their common affiliation with the neuronal and synaptic signaling pathways. These proteins are dihydropyrimidinase related protein-2 (DRP-2), CaMK II a subunit, synaptotagmin I, moesin, N-ethylmaleimide sensitive fusion protein (NSF), and OMP. Interestingly, changes in neuronal proteins are overall (with exception of OMP) of lower intensity compared to regulated proteins from the first two groups (oxidative stress and cholesterol metabolism). This is consistent with the lack of major neuronal loss described in these particular APP transgenic mice and in similar AD mouse models (Blanchard et al., 2003; Irizarry, McNamara, Fedorchak, Hsiao, & Hyman, 1997). DRP-2 is a protein known to be involved in axon pathfinding during ­development (Charrier et al., 2003). This protein was previously reported to have a relationship with AD pathology: changes in expression of this protein were observed in AD brain samples in several proteomic studies (Castegna, Aksenov, Thongboonkerd et  al., 2002; Lubec et al., 1999; Schonberger et al., 2001). However, depending on the study, the protein was either de-regulated (Lubec et al., 1999), down-regulated (Schonberger et  al., 2001), less glycosylated (Kanninen et  al., 2004), or oxidatively modified (Castegna, Aksenov, Thongboonkerd, et al., 2002). In addition, DRP-2 was shown to be altered (hyperphosphorylated) in neurofibrillary tangles (Gu, Hamajima, & Ihara, 2000). As stressed above, the major difference between our animal model and AD human samples lies in the lack of neurofibrillary tangles, a second hallmark of AD pathology. This leads to the absence of massive neuronal loss in the mouse cortices, a feature widely observed in human AD brains. The decreased level of DRP-2 reported in AD human samples is thus possibly related to the neuronal loss, while the increased level of DRP-2 observed in our model may reflect neuritic reorganization and formation of dystrophic neurites around amyloid plaques. CaMK II a subunit (down-regulated in the present model) is a very abundant kinase essential for synaptic signaling. Synaptotagmin I and NSF (both up-­regulated) are two proteins involved in vesicle trafficking. Regulation of these three proteins may reflect changes in synaptic properties. Both synaptotagmin I and NSF were previously shown to be overexpressed in AD patient brain (Schonberger et al., 2001),

504

D. Sizova

and NSF was identified in a functional proteomic study as a partner for APP (Cottrell et al., 2005). OMP is believed to act as a modulator of the olfactory signal-transduction cascade. It is almost exclusively expressed in the mature olfactive neurons, but was also found to be present in small groups of neurons in several areas of the mouse CNS including the cortex (Baker, Grillo, & Margolis, 1989; Buiakova et al., 1996). This protein was never shown to have a relationship with AD pathology. It is highly down-regulated in our model (ratio −2.0) which could indicate a corresponding alteration of neuronal signaling in the APP transgenic mouse brains. Finally, we found a set of proteins unambiguously up-regulated in the present mouse model that does not match with any pathway presently known to be involved in AD pathology. These proteins are albumin, serotransferrin, M-pyruvate kinase (ratio 1.5), and dynamin-1 like protein (ratio 1.6). Two serum proteins, albumin and transferrin, were found to be up-regulated with extremely high ratios of 4.3 and 2.2, respectively. It is important to note that albumin is highly present in serum but almost absent from adult brain (Tabernero, Medina, Sanchez-Abarca, Lavado, & Medina, 1999). From the first look, this makes its up-regulation in our model quite surprising. However, such a high level of this protein in AD cortexes can be explained by the leakage of the blood–brain-barrier (BBB) (Wisniewski, Vorbrodt, & Wegiel, 1997) that has been previously described in some areas of the cerebral cortex of 4-month-old APP transgenic mice (Ujiie, Dickstein, Carlow, & Jefferies, 2003). Moreover, two recent reports suggest that an increased level of albumin in the brain of APP transgenic mice is possible even in the absence of BBB leakage. Indeed, adsorption of albumin on vessel walls in APP/PS1 double transgenic mice was observed by Poduslo et al. (2001). Another study (Kumar-Singh et al., 2005) showed high association of dense-core amyloid plaques with blood vessels and infiltration of albumin around plaques in two APP transgenic mouse lines similar to our model. It is important to note that an increased expression of albumin has already been found previously in AD patient brain samples (Schonberger et  al., 2001). Up-regulation of transferrin can also be explained by the BBB breakdown, but this protein is also expressed in normal adult brain. Transferrin, whose major function is iron transport and metal cations, has been proposed to play a role in AD disease development (Farrar et  al., 1990; Todorich & Connor, 2004). Moreover, genetic association of the transferring C2 allele with AD has been described (Namekata et  al., 1997). M-pyruvate kinase is an ubiquitous enzyme involved in glycolysis. As has been suggested (Bigl, Bruckner, Arendt, Bigl, & Eschrich, 1999), AD pathology induces the increased metabolic activity of glial cells, which can explain the up-regulation of M-pyruvate kinase in the present model. Dynamin1-like protein (Dnm1l) is a recently characterized protein possibly playing a role in mitochondria and peroxisome fission (Frank et al., 2001; Koch et al., 2003). The finding that these four proteins are significantly regulated in the AD animal model could uncover new pathways involved in the disease progression. In general, proteomic analyses of pathological samples allow the identification of fewer regulated genes than transcriptomic analyses. This fact can be partially explained by the nature of the separation techniques generally used in proteomics which are based on the protein physico-chemical properties, and which can resolve efficiently only a selected subset of the whole proteoma. For example, it can be

Protein Expression Profile of Alzheimer’s Disease

505

expected that membrane and basic proteins, as well as very small polypeptides, will be considerably underrepresented when using 2D-E. As a consequence, it is likely that a subset of regulated proteins will not be found. In our case, a good example of that is the APP transgene, which is known to be strongly expressed only in ­transgenic animals, but was not identified in the present study (most likely because it is a membrane protein). On the other hand, proteomic studies have the strong advantage of directly measuring the protein expression, a parameter more ­physiologically relevant than the RNA level. Moreover, they enable the resolution of distinct isoforms of a protein and the identification of post-translational ­modifications which can be functionally important.

5 Conclusion and Perspectives Significant advancement in the field of neurodegenerative disorders can be expected in the coming years because of an exponential growth of molecular biology approaches and dramatic improvement in their quality. Indeed, in the last decade there has been considerable progress simultaneously in two domains, the contribution from each of them being extraordinary valuable for AD studies. First, AD transgenic mouse models representing a number of important features of this disease were finally created and, second, development of novel high-resolution proteomic methods allowed performing studies of AD related proteome on a new level of confidence. The use of a Thy1-APP751SL transgenic mouse model provided us with a way to avoid several major disadvantages of human samples, such as variations in postmortem delay, patient age, sex and genetic background. This allowed pooling of cortices from different animals of the same sex and made the samples extremely homogeneous. The fact that these mice display some of the major signs of AD such as Ab peptide deposition, dystrophic neurites formation, and progressive neuronal death gives more significance to the resulting proteomic profile. Moreover, this mouse model has a high potential for future studies related to AD. Generation of proteomic profiles from different regions of AD model mouse brains would yield new and important information. The use of brain samples from mouse groups of different ages with different stages of amyloid plaque deposition could also provide additional knowledge on disease development and progression. DIGE, a relatively novel approach developed by Amersham Biosciences in the middle of the 90th, allowed us to avoid general disadvantages of a classical 2D-E analysis followed by MS identification of proteins in selected gel spots. Use of large 2D gels and fluorescent dyes gave a higher level of protein spot resolution while running of transgenic and control samples on the same gel eliminated issues concerned to “gel-to-gel” differences. Moreover, computer image analysis including statistical treatment made resulting data much more reliable. Finally, automatic sample excision in a dust-free environment ensured the absence of contaminating keratin, the main problem reported by scientists performing MS-protein ­identification from gel slices.

506

D. Sizova

In order to get a complete proteomic profile from analyzed samples, we applied both MALDI-TOF-MS and nanoLC-MS/MS methods for identification of certain proteins when MALDI-TOF alone provided ambiguous data. As a result, we ­generated a proteomic profile consisting of 15 proteins that could be divided into three groups corresponding to the main known AD-related biological pathways: inflammation and oxidative stress, cholesterol metabolism, and neuronal and ­synaptic signaling. A literature survey showed that some of these proteins already had a documented link with AD; some had been found to be regulated in previous proteomic or transcriptomic studies, while others had never been reported in this context. Identification of previously reported proteins served to validate our original protocol, and strengthen the finding of new proteins related to AD pathology. The reported study may be considered as just one step of a long way to the exhaustive picture of the AD-related proteome. New developing methods in different biological and biochemical areas are expected to accelerate the process of data accumulation. Generation of new animal models of AD displaying both keystone features of AD – Ab deposition and Tau tangle formation – are opening dramatic new horizons for studies of AD. New techniques for the preparation of samples such as laser capture microdissection (Liao et al., 2004) could provide improved starting material from small brain areas or even Ab plaques alone. Alternative ­proteomic methods such as MudPIT (multidimensional protein identification technology, based on multidimensional high-pressure liquid chromatography coupled with tandem mass spectrometry; McDonald & Yates, 2002), or protein microarray technologies will enlarge the proteoma subset analyzed and shorten the time of analysis. Only combinations of different approaches to the study of both human and animal model AD proteomes will allow the generation of a detailed and reliable description of the cellular changes associated with this widespread disease. These data will provide an indispensable knowledge for the understanding of disease pathogenesis and, hopefully, will give clues to disease biomarkers and treatment. Acknowledgments  I would like to thank Dr. Anita Diu-Hercend for indispensable participation in all steps of this work, for fruitful discussions, for continuous help and encouragement, and for critical reading of the present manuscript. I am very grateful to Dr. Elodie Charbaut for her help in data treatment, evaluation, and representation. The author also sincerely acknowledges Dr. Veronique Blanchard and Dr. Philippe Delay-Goyet (AD mouse model), Emmanuelle Deretz, Sophie Bouvier, Dr. Fabienne Parker and Dr. Marc Duchesne (proteomics), and Dr. Francois Delalande, Dr. Anthony High, Dr. Florence Poirier, and Dr. Alain Van Dorsselaer (mass spectroscopy) for their contribution to the generation of experimental data.

References Afagh, A., Cummings, B. J., Cribbs, D. H., Cotman, C. W., & Tenner, A. J. (1996). Localization and cell association of C1q in Alzheimer’s disease brain. Experimental Neurology, 138, 22–32. Baker, H., Grillo, M., & Margolis, F. L. (1989). Biochemical and immunocytochemical ­characterization of olfactory marker protein in the rodent central nervous system. The Journal of Comparative Neurology, 285, 246–261.

Protein Expression Profile of Alzheimer’s Disease

507

Bales, K. R., Dodart, J. C., DeMattos, R. B., Holtzman, D. M., & Paul, S. M. (2002). Apolipoprotein E, amyloid, and Alzheimer disease. Molecular Interventions, 2, 363–375. Bales, K. R., Verina, T., Cummins, D. J., Du, Y., Dodel, R. C., Saura, J., et  al. (1999). Apolipoprotein E is essential for amyloid deposition in the APP(V717F) transgenic mouse model of Alzheimer’s disease. Proceedings of the National Academy of Sciences of the United States of America, 96, 15233–15238. Bigl, M., Bruckner, M. K., Arendt, T., Bigl, V., & Eschrich, K. (1999). Activities of key glycolytic enzymes in the brains of patients with Alzheimer’s disease. Journal of Neural Transmission, 106, 499–511. Blanchard, V., Moussaoui, S., Czech, C., Touchet, N., Bonici, B., Planche, M., et al. (2003). Time sequence of maturation of dystrophic neurites associated with Abeta deposits in APP/PS1 transgenic mice. Experimental Neurology, 184, 247–263. Boyett, K. W., DiCarlo, G., Jantzen, P. T., Jackson, J., O’Leary, C., Wilcock, D., et  al. (2003). Increased fibrillar beta-amyloid in response to human clq injections into hippocampus and cortex of APP+PS1 transgenic mice. Neurochemical Research, 28, 83–93. Buiakova, O. I., Baker, H., Scott, J. W., Farbman, A., Kream, R., Grillo, M., et  al. (1996). Olfactory marker protein (OMP) gene deletion causes altered physiological activity of ­olfactory sensory neurons. Proceedings of the National Academy of Sciences of the United States of America, 93, 9858–9863. Butterfield, D. A., & Lauderback, C. M. (2002). Lipid peroxidation and protein oxidation in Alzheimer’s disease brain: Potential causes and consequences involving amyloid beta-­peptideassociated free radical oxidative stress. Free Radical Biology & Medicine, 32, 1050–1060. Buxbaum, J. N., & Tagoe, C. E. (2000). The genetics of the amyloidoses. Annual Review of Medicine, 51, 543–569. Casserly, I., & Topol, E. (2004). Convergence of atherosclerosis and Alzheimer’s disease: Inflammation, cholesterol, and misfolded proteins. Lancet, 363, 1139–1146. Castano, E. M., Roher, A. E., Esh, C. L., Kokjohn, T. A., & Beach, T. (2006). Comparative ­proteomics of cerebrospinal fluid in neuropathologically-confirmed Alzheimer’s disease and non-demented elderly subjects. Neurological Research, 28, 155–163. Castegna, A., Aksenov, M., Aksenova, M., Thongboonkerd, V., Klein, J. B., Pierce, W. M., et al. (2002). Proteomic identification of oxidatively modified proteins in Alzheimer’s disease brain. Part I: Creatine kinase BB, glutamine synthase, and ubiquitin carboxy-terminal hydrolase L-1. Free Radical Biology & Medicine, 33, 562–571. Castegna, A., Aksenov, M., Thongboonkerd, V., Klein, J. B., Pierce, W. M., Booze, R., et al. (2002). Proteomic identification of oxidatively modified proteins in Alzheimer’s disease brain. Part II: Dihydropyrimidinase-related protein 2, alpha-enolase and heat shock cognate 71. Journal of Neurochemistry, 82, 1524–1532. Chang, T. Y., Chang, C. C., Lin, S., Yu, C., Li, B. L., & Miyazaki, A. (2001). Roles of acyl-­coenzyme A: Cholesterol acyltransferase-1 and -2. Current Opinion in Lipidology, 12, 289–296. Charrier, E., Reibel, S., Rogemond, V., Aguera, M., Thomasset, N., & Honnorat, J. (2003). Collapsin response mediator proteins (CRMPs): Involvement in nervous system development and adult neurodegenerative disorders. Molecular Neurobiology, 28, 51–64. Chen, J. W., Dodia, C., Feinstein, S. I., Jain, M. K., & Fisher, A. B. (2000). 1-Cys peroxiredoxin, a bifunctional enzyme with glutathione peroxidase and phospholipase A2 activities. The Journal of Biological Chemistry, 275, 28421–28427. Condorelli, D. F., Nicoletti, V. G., Barresi, V., Conticello, S. G., Caruso, A., Tendi, E. A., et al. (1999). Structural features of rat GFAP gene and identification of a novel alternative transcript. Journal of Neuroscience Research, 56, 219–228. Cottrell, B. A., Galvan, V., Banwait, S., Gorostiza, O., Lombardo, C. R., Williams, T., et al. (2005). A pilot proteomic study of amyloid precursor interactors in Alzheimer’s disease. Annals of Neurology, 58, 277–289. Dickey, C. A., Loring, J. F., Montgomery, J., Gordon, M. N., Eastman, P. S., & Morgan, D. (2003). Selectively reduced expression of synaptic plasticity-related genes in amyloid precursor protein + presenilin-1 transgenic mice. The Journal of Neuroscience, 23, 5219–5226.

508

D. Sizova

Farrar, G., Altmann, P., Welch, S., Wychrij, O., Ghose, B., Lejeune, J., et al. (1990). Defective gallium-transferrin binding in Alzheimer disease and Down syndrome: Possible mechanism for accumulation of aluminium in brain. Lancet, 335, 747–750. Fonseca, M. I., Zhou, J., Botto, M., & Tenner, A. J. (2004). Absence of C1q leads to less neuropathology in transgenic mouse models of Alzheimer’s disease. The Journal of Neuroscience, 24, 6457–6465. Frank, S., Gaume, B., Bergmann-Leitner, E. S., Leitner, W. W., Robert, E. G., Catez, F., et  al. (2001). The role of dynamin-related protein 1, a mediator of mitochondrial fission, in apoptosis. Developmental Cell, 1, 515–525. Galasko, D. (2005). Biomarkers for Alzheimer’s disease – clinical needs and application. Journal of Alzheimer’s Disease, 8, 339–346. Ginsberg, S. D., Hemby, S. E., Lee, V. M., Eberwine, J. H., & Trojanowski, J. Q. (2000). Expression profile of transcripts in Alzheimer’s disease tangle-bearing CA1 neurons. Annals of Neurology, 48, 77–87. Glowinski, J., & Iversen, L. L. (1966). Regional studies of catecholamines in the rat brain. I. The disposition of [3H]norepinephrine, [3H]dopamine and [3H]dopa in various regions of the brain. Journal of Neurochemistry, 13, 655–669. Gu, Y., Hamajima, N., & Ihara, Y. (2000). Neurofibrillary tangle-associated collapsin response mediator protein-2 (CRMP-2) is highly phosphorylated on Thr-509, Ser-518, and Ser-522. Biochemistry, 39, 4267–4275. Irizarry, M. C., McNamara, M., Fedorchak, K., Hsiao, K., & Hyman, B. T. (1997). APPSw ­transgenic mice develop age-related A beta deposits and neuropil abnormalities, but no ­neuronal loss in CA1. Journal of Neuropathology and Experimental Neurology, 56, 965–973. Irizarry, M. C., Soriano, F., McNamara, M., Page, K. J., Schenk, D., Games, D., et  al. (1997). Abeta deposition is associated with neuropil changes, but not with overt neuronal loss in the human amyloid precursor protein V717F (PDAPP) transgenic mouse. The Journal of Neuroscience, 17, 7053–7059. Kanninen, K., Goldsteins, G., Auriola, S., Alafuzoff, I., & Koistinaho, J. (2004). Glycosylation changes in Alzheimer’s disease as revealed by a proteomic approach. Neuroscience Letters, 367, 235–240. Koch, A., Thiemann, M., Grabenbauer, M., Yoon, Y., McNiven, M. A., & Schrader, M. (2003). Dynamin-like protein 1 is involved in peroxisomal fission. The Journal of Biological Chemistry, 278, 8597–8605. Kumar-Singh, S., Pirici, D., McGowan, E., Serneels, S., Ceuterick, C., Hardy, J., et  al. (2005). Dense-core plaques in Tg2576 and PSAPP mouse models of Alzheimer’s disease are centered on vessel walls. The American Journal of Pathology, 167, 527–543. Liao, L., Cheng, D., Wang, J., Duong, D. M., Losik, T. G., Gearing, M., et al. (2004). Proteomic characterization of postmortem amyloid plaques isolated by laser capture microdissection. The Journal of Biological Chemistry, 279, 37061–37068. Lubec, G., Krapfenbauer, K., & Fountoulakis, M. (2003). Proteomics in brain research: Potentials and limitations. Progress in Neurobiology, 69, 193. Lubec, G., Nonaka, M., Krapfenbauer, K., Gratzer, M., Cairns, N., & Fountoulakis, M. (1999). Expression of the dihydropyrimidinase related protein 2 (DRP-2) in Down syndrome and Alzheimer’s disease brain is downregulated at the mRNA and dysregulated at the protein level. Journal of Neural Transmission. Supplementum, 57, 161–177. Manevich, Y., Sweitzer, T., Pak, J. H., Feinstein, S. I., Muzykantov, V., & Fisher, A. B. (2002). 1-Cys peroxiredoxin overexpression protects cells against phospholipid peroxidation-mediated membrane damage. Proceedings of the National Academy of Sciences of the United States of America, 99, 11599–11604. Marouga, R., David, S., & Hawkins, E. (2005). The development of the DIGE system: 2D ­fluorescence difference gel analysis technology. Analytical and Bioanalytical Chemistry, 382, 669–678. Matsuoka, Y., Picciano, M., Malester, B., LaFrancois, J., Zehr, C., Daeschner, J. M., et al. (2001). Inflammatory responses to amyloidosis in a transgenic mouse model of Alzheimer’s disease. The American Journal of Pathology, 158, 1345–1354.

Protein Expression Profile of Alzheimer’s Disease

509

McDonald, W. H., & Yates, J. R. (2002). Shotgun proteomics and biomarker discovery. Disease Markers, 18, 99–105. Messing, A., & Brenner, M. (2003). GFAP: Functional implications gleaned from studies of genetically engineered mice. Glia, 43, 87–90. Namekata, K., Imagawa, M., Terashi, A., Ohta, S., Oyama, F., & Ihara, Y. (1997). Association of transferrin C2 allele with late-onset Alzheimer’s disease. Human Genetics, 101, 126–129. Orange, P., Hawkins, E., Story, C., & Fowler, S. (2000). Fluorescence 2-D difference gel electrophoresis. Life Science News, 5, 1–4. Piot-Grosjean, O., Cudennec, C. L., Dhenain, M., Delay-Goyet, P., & Delatour, B. (2004). 9th international conference on Alzheimer’s disease and related disorders. Philadelphia, USA, personal communication. Poduslo, J. F., Curran, G. L., Wengenack, T. M., Malester, B., Duff, K. (2001) Permeability of proteins at the blood-brain barrier in the normal adult mouse and double transgenic mouse model of Alzheimer’s disease. Neurobiology of Disease, 8, 555–567. Price, D. L., & Sisodia, S. S. (1994). Cellular and molecular biology of Alzheimer’s disease and animal models. Annual Review of Medicine, 45, 435–446. Puglielli, L., Konopka, G., Pack-Chung, E., Ingano, L. A., Berezovska, O., Hyman, B. T., et al. (2001). Acyl-coenzyme A: Cholesterol acyltransferase modulates the generation of the amyloid beta-peptide. Nature Cell Biology, 3, 905–912. Ross, G. W., O’Callaghan, J. P., Sharp, D. S., Petrovitch, H., Miller, D. B., Abbott, R. D., et al. (2003). Quantification of regional glial fibrillary acidic protein levels in Alzheimer’s disease. Acta Neurologica Scandinavica, 107, 318–323. Schonberger, S. J., Edgar, P. F., Kydd, R., Faull, R. L., & Cooper, G. J. (2001). Proteomic analysis of the brain in Alzheimer’s disease: Molecular phenotype of a complex disease process. Proteomics, 1, 1519–1528. Shen, Y., & Meri, S. (2003). Yin and Yang: Complement activation and regulation in Alzheimer’s disease. Progress in Neurobiology, 70, 463–472. Shin, S. J., Lee, S. E., Boo, J. H., Kim, M., Yoon, Y. D., Kim, S. I., et al. (2004). Profiling proteins related to amyloid deposited brain of Tg2576 mice. Proteomics, 4, 3359–3368. Sizova, D., Charbaut, E., Delalande, F., Poirier, F., High, A., Parker, F., et al. (2007). Proteomic analysis of brain tissue from an Alzheimer’s disease mouse model by two-dimensional difference gel electrophoresis. Neurobiology of Aging, 28, 357–570. Tabernero, A., Medina, A., Sanchez-Abarca, L. I., Lavado, E., & Medina, J. M. (1999). The effect of albumin on astrocyte energy metabolism is not brought about through the control of cytosolic Ca2+ concentrations but by free-fatty acid sequestration. Glia, 25, 1–9. Teunissen, C. E., Vente, J., Steinbusch, H. W., & Bruijn, C. D. (2002). Biochemical markers related to Alzheimer’s dementia in serum and cerebrospinal fluid. Neurobiology of Aging, 23, 485–508. Todorich, B. M., & Connor, J. R. (2004). Redox metals in Alzheimer’s disease. Annals of the New York Academy of Sciences, 1012, 171–178. Tsuji, T., Shiozaki, A., Kohno, R., Yoshizato, K., & Shimohama, S. (2002). Proteomic profiling and neurodegeneration in Alzheimer’s disease. Neurochemical Research, 27, 1245–1253. Ujiie, M., Dickstein, D. L., Carlow, D. A., & Jefferies, W. A. (2003). Blood-brain barrier permeability precedes senile plaque formation in an Alzheimer disease model. Microcirculation, 10, 463–470. van Beek, J., Elward, K., Gasque, P. (2003). Activation of complement in the central nervous system: roles in neurodegeneration and neuroprotection. Ann. NY Acad. Sci. 992, 56–71. Vercauteren, F. G., Clerens, S., Roy, L., Hamel, N., Arckens, L., Vandesande, F., et  al. (2004). Early dysregulation of hippocampal proteins in transgenic rats with Alzheimer’s diseaselinked mutations in amyloid precursor protein and presenilin 1. Molecular Brain Research, 132, 241–259. Wang, X., Phelan, S. A., Forsman-Semb, K., Taylor, E. F., Petros, C., Brown, A., et al. (2003). Mice with targeted mutation of peroxiredoxin 6 develop normally but are susceptible to oxidative stress. The Journal of Biological Chemistry, 278, 25179–25190.

510

D. Sizova

Webster, S., Glabe, C., & Rogers, J. (1995). Multivalent binding of complement protein C1Q to the amyloid beta-peptide (A beta) promotes the nucleation phase of A beta aggregation. Biochemical and Biophysical Research Communications, 217, 869–875. Wisniewski, H. M., Vorbrodt, A. W., & Wegiel, J. (1997). Amyloid angiopathy and blood-brain barrier changes in Alzheimer’s disease. Annals of the New York Academy of Sciences, 826, 161–172. Wolfe, M. S. (2002). Therapeutic strategies for Alzheimer’s disease. Nature Reviews. Drug Discovery, 1, 859–866. Wolozin, B., Brown, J., Theisler, C., & Silberman, S. (2004). The cellular biochemistry of ­cholesterol and statins: Insights into the pathophysiology and therapy of Alzheimer’s disease. CNS Drug Reviews, 10, 127–146. Yao, P. J., Zhu, M., Pyun, E. I., Brooks, A. I., Therianos, S., Meyers, V. E., et al. (2003). Defects in expression of genes related to synaptic vesicle trafficking in frontal cortex of Alzheimer’s disease. Neurobiology of Disease, 12, 97–109. Yasojima, K., Schwab, C., McGeer, E. G., & McGeer, P. L. (1999). Up-regulated production and activation of the complement system in Alzheimer’s disease brain. The American Journal of Pathology, 154, 927–936.

Proteomic Analysis of CNS Injury and Recovery Günther K.H. Zupanc and Marianne M. Zupanc

Abstract  Despite the enormous medical and economic consequences of traumatic injury to the central nervous system (CNS), little is known about the proteins involved in the resulting pathology. Major advances in the identification of such proteins have been made in recent years through application of differential proteome analysis. Such an approach has revealed a number of novel proteins as potential regulators of the degenerative and regenerative processes that take place in the mammalian brain and spinal cord after a traumatic insult. Some of these proteins may serve as diagnostic and prognostic markers to assess the severity of tissue damage. Comparative proteome analysis of regeneration-competent vs. regeneration-deficient systems are likely to provide new insights into the cellular signals that could be targeted for therapeutic intervention to increase the repair capacity of the human CNS. Keywords  Axotomy • Cerebrospinal fluid • Comparative proteomics • Degeneration • Differential proteomics • Plasticity • Regeneration • Spinal cord injury • Traumatic brain injury • Vestibular compensation Abbreviations 2D PAGE CNS GFAP SCI TBI

Two-dimensional polyacrylamide gel electrophoresis Central nervous system Glial fibrillary acidic protein Spinal cord injury Traumatic brain injury

G.K.H. Zupanc (*) Department of Biology, Northeastern University, 134 Mugar Life Sciences, 360 Huntington Avenue, Boston, MA 02115, USA e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_20, © Springer Science+Business Media, LLC 2011

511

512

G.K.H. Zupanc and M.M. Zupanc

1 Introduction Injury to the central nervous system (CNS) is one of the leading causes of death and disability. In the USA, an estimated 2 million incidents of traumatic brain injury (TBI) occur each year, resulting in the death of approximately 200,000 people and in the hospitalization of an additional half million (http://www.nidcd.gov/health/ voice/tbrain.htm). Between 250,000 and 400,000 Americans live with spinal cord injury (SCI), of which 50% are paraplegic; approximately 11,000–13,000 new ­spinal cord injuries occur each year (http://www.travisroyfoundation.org; http:// www.sci-info-pages.com/facts.html). The pathology of these injuries often includes severe neuronal damage, ­culminating in the loss of whole cells through necrosis (for review, see Rowland & Sciarra, 1989). Since the human CNS, like that of other mammals, lacks the ability to replace neurons lost to injury by newly generated ones, the structural damage is lasting, often resulting in permanent loss of certain functions. To develop effective therapies, it is essential to learn more about the cellular mechanisms underlying the processes of degeneration and regeneration that occur after the traumatic insult, and to identify the multitude of molecular signals involved in the individual steps of these processes. Such information is also important to define novel biochemical markers as diagnostic and prognostic indicators, since with the methods currently available it is often difficult to accurately assess the severity of the brain and spinal cord damage. A large-scale identification of proteins associated with CNS injury has become feasible in recent years with the advent of proteomics. In particular, the establishment of differential proteomics has enabled investigators to compare the proteomes of injured vs. intact cells (for a review of the various approaches available, see Monteoliva & Albar, 2004). Although the number of studies that have employed proteome analysis is still low, the first results are extremely promising and ­underscore the enormous potential of this approach. In the following, we will review the major achievements made thus far, discuss potential problems and ­limitations of the use of proteomics, and indicate possible directions for future research in the study of CNS injury.

2 Proteomics of Injury-Associated Proteins in the Mammalian CNS 2.1 Traumatic Brain Injury 2.1.1 Traumatic Brain Injury During Early Postnatal Development The first attempt to perform, by use of proteomics, a large-scale analysis of changes in protein abundance after TBI was made by Jenkins and associates (Jenkins et al., 2002).

Proteomic Analysis of CNS Injury and Recovery

513

The authors caused damage to the hippocampus of rats through moderate controlled cortical impact. This paradigm mimics part of the events that take place in closed-head injuries. At the age chosen – postnatal day 17 – the immature rat brain exhibits, like the adult brain, limited potential for structural and functional recovery (for review, see Payne & Lomber, 2002). Twenty-four hours after the injury, tissue samples from injured rats and sham-treated control rats, respectively, were processed for two-dimensional polyacrylamide gel electrophoresis (2D PAGE), and differences in protein spot intensities between the two conditions were analyzed. Out of 595 matching spots found under both conditions, the following six identified proteins showed at least twofold, statistically significant changes: b-actin, a-­ tubulin, b-tubulin, phosphatidylinositol transfer protein-a, and 78-kDa ­glucose-regulated protein precursor, all of which exhibited decreases in protein abundance in the injured rats, compared to sham-treated control animals; and Cu/ Zn superoxide dismutase, which displayed an increase in protein abundance. The decrease in abundance of b-actin, a-tubulin, and b-tubulin might reflect a ­degradation of these cytoskeletal proteins. Decrease in protein abundance of the 78-kDa glucose-regulated protein precursor is likely to also result in a lowered abundance of 78-kDa glucose-regulated protein. The latter has been shown to have a neuroprotective function by reducing apoptotic cell death (Tsuchiya et al., 2003). Thus, the overall effect of the reduced abundance of the 78-kDa glucose-regulated protein could be to increase the number of cells undergoing apoptotic cell death in the hippocampus. These findings are remarkable from a comparative point of view, because in the adult fish brain – a biological system that exhibits an enormous potential for both structural and functional recovery – the abundance of 78-kDa glucose-regulated protein is increased 3 days after cerebellar lesion (Zupanc, Wellbrock, & Zupanc, 2006; see also the Sect. 5). The increase in Cu/Zn superoxide dismutase in the immature rat brain, on the other hand, may reflect an attempt of the hippocampus to (partially) regenerate. This antioxidant enzyme appears to play an important role in stressful conditions after injury in that it protects tissue from cell death by reducing the production of free radicals (Reaume et al., 1996; Yin et al., 2001). In a follow-up study employing the same TBI-model as used by Jenkins et al. (2002), Kochanek et  al. (2006) performed a proteome analysis of global protein changes occurring 2 weeks following the controlled cortical impact in postnatal day 17 rats. At this time point, only three proteins showed alterations in protein spot intensity which were at least 50% higher or lower than controls and statistically significant. These proteins – the abundance of each was increased, compared to the sham-treated control animals – were 14-3-3-g, vimentin, and glial fibrillary acidic protein (GFAP). The members of the 14-3-3 protein family bind to phosphoserine/phosphothreonine motifs in a sequence-specific manner in a large number of target proteins, including proteins involved in cell cycle regulation, apoptosis, and stress responses (Darling, Yingling, & Wynshaw-Boris, 2005; Hermeking & Benzinger, 2006; Porter, Khuri, & Fu, 2006). Through coordination with survival kinases, 14-3-3

514

G.K.H. Zupanc and M.M. Zupanc

inhibits multiple pro-apoptotic molecules, thus suppressing apoptotic cell death. The increase in protein abundance 2 weeks after the controlled cortical impact might be related to this apoptosis-suppressing effect. The increase in the intensity of spots associated with vimentin and GFAP is in agreement with numerous studies that have described an activation of astrocytes in the CNS in response to various kinds of injury. The function of this response, which is commonly referred to as reactive gliosis and involves up-regulation of the glial intermediary filament proteins GFAP and vimentin, is rather enigmatic. Both detrimental effects, such as the formation of a scar barrier that prevents axons from growing into the area of the lesion, and beneficial functions, such as neurotrophic and protective activities, have been proposed (for reviews, see Pekny & Nilsson, 2005; Sofroniew, 2005). 2.1.2 Traumatic Brain Injury During Adulthood The controlled cortical impact model can also be used as a paradigm to study TBI in adult rats. Two investigations have demonstrated the power of ­combining this paradigm with proteome analysis to identify changes in protein abundance after TBI. In the first of these investigations, Kobeissy et al. (2006) collected samples of intact and injured cortical tissue 48  h post injury. These samples were analyzed by employing cation/anion exchange chromatography, followed by 1D PAGE. In total, 59 proteins exhibited statistically significant differences in abundance between intact and injured samples. Of these proteins, abundance was decreased in 21 ­proteins and increased in 38 proteins, compared to the naïve control animals. Among the proteins with decreased abundance in the post-injury samples were the cytoskeletal-associated proteins profilin, cofilin, and microtubule-associated ­protein 2; the neuron-specific proteins collapsin response mediator protein-2 and neuronal protein 22; and the glycolytic ­p roteins glyceraldehyde-3-phosphate dehydrogenease, enolase, aldehyde ­dehydrogenase, glutamate dehydrogenase, and hexokinase. Among the proteins with increased abundance in the post-injury samples were the glycolytic ­proteins lactate dehydrogenase, brain creatine kinase, and malate dehydrogenase; the ubiquitin-associated proteins ubiquitin carboxyl-­terminal hydrolase L1 and proteasome subunit a type 7; members of the 14-3-3 chaperone protein family; and C-reactive protein, transferrin, and ceruloplasmin, all members of the acute phase protein family. The latter are indicative of an inflammatory response, which is a well-known aspect of the pathology encountered after TBI in the mammalian brain. In addition to changes in the abundance of whole proteins, proteomic analysis also indicated the occurrence of proteolytic processes after cortical injury. For example, aII-spectrin, a protein with a molecular weight of 280 kDa, appeared as a 120-kDa spectrin breakdown product. This is in agreement with the well-­ characterized degradation of aII-spectrin to a 120-kDa fragment following ­caspase-3 proteolysis during apoptosis.

Proteomic Analysis of CNS Injury and Recovery

515

Another potential breakdown product was the 55-kDa fragment of collapsin response mediator protein-2 (also known as dihydropyrimidinase-related ­protein-2), which, as an intact protein, has a molecular weight of 62 kDa. Collapsin response mediator protein-2 is a mammalian homologue of the Caenorhabditis elegans Unc33, which is required for proper axon morphology and pathfinding in worms (Hedgecock, Culotti, Thomson, & Perkins, 1985; Siddiqui & Culotti, 1991). This observation is remarkable because, even under permissive conditions in which the generation of new cells is stimulated through injury, these cells commonly fail to further develop in the adult mammalian brain (Arvidsson, Collin, Kirik, Kokaia, & Lindvall, 2002). Thus, the inactivation of a protein mediating developmental ­processes, such as elongation of processes or axonal pathfinding, could explain the failure of the new cells to undergo further development. A third proteolyzed protein was synaptotagmin. This integral membrane protein is found on the surface of synaptic vesicles and involved in calcium-mediated release of neurotransmitters. The proteomics data indicate that the intact 65-kDa protein is degraded, resulting in a fragment of 37 kDa. As inactivation of proteins through proteolysis is likely to play a central role in many processes interfering with the ability of tissue to regenerate after TBI, the study of this aspect is particularly important for the definition of new therapeutic strategies to improve brain repair. Such work is likely to benefit from the recent development of degradomics – the specific application of genomic and proteomic approaches for the identification and characterization of proteases, including the substrates that are targeted by these proteases and their endogenous inhibitors, in an organism (for review, see López-Otín & Overall, 2002). Whereas the development of methods based on proteomics is still in its infancy, protease DNA microarray chips, employing complementary DNA or oligonucleotide-specific probes for the proteases found in a species, are already available (for review, see Doucet & Overall, 2008). The second study in which the cortical impact model was combined with ­proteome analysis in the adult rat brain has addressed a more specific aspect of the cellular processes associated with TBI. In this investigation, Opii et  al. (2007) ­identified mitochondrial proteins that undergo oxidative modifications after TBI. Such modifications are believed to be initiated by uptake of excessive amounts of Ca2+ by mitochondria. In the intact brain, one function of mitochondria – besides mediating ATP generation – is to act as a high-capacity Ca2+ repository to maintain Ca2+ homeostasis. After injuries, however, the massive release of Ca2+ into the intercellular space, and the eventual overload of mitochondria with Ca2+, leads to an increase in mitochondrial production of reactive oxygen species. This is thought to induce a number of other processes, including an opening of the mitochondrial permeability transition pore, a release of cytochrome C, and inhibition of ATP production (Lifshitz, Sullivan, Hovda, Wieloch, & McIntosh, 2004; Nicholls & Budd, 2000; Rego & Oliveira, 2003; Sullivan, Rabchevsky, Waldmeier, & Springer, 2005). Ultimately, cell death will occur. Thus, the oxidation of mitochondrial ­proteins appears to be a major factor contributing to the secondary wave of cell death that commonly occurs after the primary insult.

516

G.K.H. Zupanc and M.M. Zupanc

To identify oxidized mitochondrial proteins, young adult rats were subjected to controlled moderate unilateral cortical impact, and total mitochondria were isolated from cortical and hippocampal tissue (Opii et  al., 2007). Changes in oxidative ­levels of proteins, relative to contralateral (or sham-operated) controls, were ­measured through carbonylation of proteins using anti-dinitrophenylhydrazine immunochemical detection of 2D-oxyblots. The spots on these blots were matched with corresponding spots on 2D PAGE maps. Protein spots that showed a ­significant increase in protein carbonylation in the ipsilateral side, relative to the contralateral side, were excised from the gels and subjected to further analysis. This approach enabled the authors to identify a number of proteins that are oxidatively modified after TBI. These proteins include pyruvate dehydrogenase, voltage-dependent anion channel, fumarate hydratase 1, ATP synthase, and prohibitin from the cortex; and cytochrome C oxidase Va, isovaleryl coenzyme A dehydrogenase, enolase-1, and glyceraldehyde-3-phosphatase dehydrogenase from the hippocampus. Oxidative damage of these proteins, which are critically involved in energy metabolism, ­presumably leads to their inactivation and, thus, to a decrease in mitochondrial bioenergetics, as commonly observed after TBI.

2.2 Spinal Cord Injury Only three studies have been published thus far that have, by employing a proteomics approach, examined global changes in protein abundance after SCI (Ding et al., 2006; Kang, So, Moon, & Kim, 2006; Tsai, Shen, Kuo, Cheng, & Chak, 2008). Two of these investigations used transection models (Ding et al., 2006; Kang et al., 2006), and one a contusion model (Tsai et  al., 2008), of traumatic SCI. Injuries were applied in the region of spinal segments T9/T10 of rats, causing complete paraplegic symptoms of the hind limbs. Kang et  al. (2006) found, 24  h after the injury, that out of 239 protein spots identified, spot intensity was increased at least twofold in 39 spots in tissue from the injured spinal cord, compared to the normal tissue. Spot intensity was reduced at least twofold in 29 spots. As expected, proteins exhibited differential expression that are potentially involved in cellular transport, metabolism, and apoptosis. In addition, an increase in protein abundance was found in a number of proteins of the axonal cytoskeleton that might reflect the attempt of the spinal cord shortly after the injury to regenerate nerve fibers. Such proteins include a- and b-tubulin, ­neurofilament proteins, and peripherin. This is in line with the observation that in the mammalian spinal cord axonal sprouting may take place after injury – despite the lack of the ability to spontaneously regenerate a significant number of whole axons (for review, see Johansson, 2007). A major impediment for the regrowth of axons is the glial scar formed by reactive astrocytes (Busch & Silver, 2007). Correspondingly, the spot intensity associated with GFAP was increased very ­dramatically (27-fold, compared to sham-operated controls) after spinal cord transection.

Proteomic Analysis of CNS Injury and Recovery

517

A similar indication of attempts of the mammalian spinal cord to regenerate was found in the study by Ding et al. (2006). Changes in protein abundance were examined 5 days after complete transection of the spinal cord. However, two major drawbacks of this investigation are that the authors analyzed only increases, but not decreases, in protein abundance. Furthermore, they included in their samples spinal cord tissue both from the site of injury and from uninjured neighboring regions. Thus, the specific localization of any changes in protein abundance remains unclear. Compared with sham-operated controls, 20 protein spots in the soluble fraction and 24 protein spots in the insoluble fraction showed a £1.5-fold increase in intensity. One interesting group among the proteins with increased abundance is comprised by iron-saturated transferrin, galectin-3, and apolipoprotein A-I, which are factors potentially mediating neuronal survival and regeneration. In the peripheral nervous system, induction of transferrin receptor expression by axotomy has been implicated in neuronal repair (Graeber, Raivich, & Kreutzberg, 1989; Raivich, Graeber, Gehrmann, & Kreutzberg, 1991). The oxidized form of galectin-1, a member of the b-galactosidase-binding lectin family, plays a role in the initial stage of outgrowth of peripheral nerves after axotomy (Horie et al., 1999; Inagaki, Sohma, Horie, Nozawa, & Kadoya, 2000). Apolipoprotein E is involved in lipid delivery for growth and regeneration of axons after injury (for review, see Vance, Campenot, & Vance, 2000). Based on this potential of the mammalian spinal cord to initiate repair, Tsai et al. (2008) have examined, through proteomic analysis, one of the factors – acidic fibroblast growth factor – that might be important in overcoming the limits that restrict successful spinal cord regeneration in mammals. Using behavioral paradigms, the authors demonstrated improved locomotor recovery after SCI in rats treated with this growth factor. The local administration of acidic fibroblast growth factor led, in addition to changes in the abundance of other proteins, to a downregulation of the expression of GFAP, S100b, and keratan sulfate proteoglycans. This contrasts with the up-regulation of expression of these proteins found after SCI without such treatment. GFAP, S100b, and keratan sulfate proteoglycans are thought to be major contributing factors causing the pathological conditions of secondary injury, including tissue inflammation and scar formation. Acidic fibroblast growth factor, therefore, appears to improve functional recovery by reducing the extent of secondary injury.

3 Proteome Analysis of Injury-Related Plasticity of the Mammalian Brain 3.1 Vestibular Compensation: The Restoration of Vestibular Function After Injury Whereas the mammalian brain lacks the ability to fully recover from brain injuries, there is some potential for neural plasticity after injury to partially restore

518

G.K.H. Zupanc and M.M. Zupanc

c­ ompromised functions. However, while the structural changes and the behavioral recovery mediating this type of plasticity have been intensively studied, the opportunities offered by a large-scale identification of the proteins involved in the underlying molecular mechanisms are still almost entirely unexplored. One exception is the type of brain plasticity associated with vestibular compensation. Through this process, the functioning of the vestibular system is, at least, partially restored after head injuries or vestibulopathy (for reviews, see Dieringer, 2003; Lacour, 2006). In humans, characteristic symptoms of vestibular dysfunction include oculomotor and postural deficits and the sensation of vertigo and incorrect motion, even in the absence of any motion. In animal models, vestibular compensation can be induced by deafferentiation of the vestibular receptors of one inner ear, most commonly done through unilateral labyrinthectomy. This causes severe ocular and motor postural symptoms, which can be divided into two categories. The first category called static includes circular walking, head and upper body yaw-and-roll tilt, a tendency to fall toward the side of the lesioned ear, and spontaneous nystagmus. The second category called dynamic consists of deficits in the gain of the vestibulo-ocular and vestibulo-spinal reflexes that result in oscillopsia. Most of the static deficits subside rapidly within a few hours or days. Recovery of the dynamic vestibular functions occurs more gradually over several months and is incomplete. This process of restoration of vestibular function – the vestibular compensation – is mediated by a number of cellular changes in the central vestibular system, particularly in the medial vestibular nucleus, as well as in the cerebellum. Among the changes observed in this nucleus are alterations in synaptic input, intrinsic excitability of cells, and activity-dependent reorganization of the connectivity in the vestibular network of the brainstem.

3.2 Differential Proteome Analysis of Proteins Involved in Vestibular Compensation Differential proteome analysis of the changes in protein abundance in the ipsi-lesioned and contra-lesioned medial vestibular nucleus has revealed a number of protein candidates potentially involved in vestibular compensation (Paterson et al., 2006). One week after rats underwent unilateral labyrinthectomy, changes in protein abundance in the ipsi-lesional and contra-lesional medial vestibular nucleus were compared to tissue collected from corresponding regions of intact or sham-operated animals. A number of proteins were affected by unilateral labyrintectomy only. Proteins that showed an increase in protein abundance include neuropilin-2 and collapsin response mediator protein-2. Neuropilin-2 is a high-affinity receptor for factors belonging to the class-3 semaphorin family, which participate in the guidance of growing axons to their targets in the developing nervous system (for review, see Nakamura, Kalb, & Strittmatter, 2000). Collapsin response mediator protein-2 is a  cytosolic protein involved in microtubule assembly. It is required for neuronal  process elongation and growth cone motility (see Sects. 2, 2.1, and 2.1.2).

Proteomic Analysis of CNS Injury and Recovery

519

These findings suggest that both neuropilin-2 and collapsin response mediator protein-2 are involved in axonal re-wiring of the neural network in the medial vestibular nucleus. This remodeling of synaptic connectivity might include a reorganization of the synaptic connectivity between the vestibular nuclei in the two hemispheres and an expansion of input from other areas to the vestibular neurons, for example proprioceptive afferents from neck muscles.

4 Biomarkers of CNS Injury 4.1 Criteria for Useful Biomarkers The availability of biomarkers of CNS injury is of critical importance for the assessment of the severity of the injury, and thus for the choice of treatment of the patient. Pineda, Wang, & Hayes (2004) have summarized the criteria that should guide the search for useful biomarkers. They should be based on readily accessible biological material, predict the magnitude of injury, exhibit high sensitivity and specificity, and appear in a time-locked sequence after the insult. Ideally, such ­biomarkers should be unique to the CNS. Up to now, no biomarker has been found that meets all these criteria. Most of the research to identify better biomarkers is based on sampling from the cerebrospinal fluid (CSF). In the following, the principal approaches used in this search will be discussed.

4.2 Production, Composition, and Function of the Cerebrospinal Fluid The CSF is produced by the choroid plexus and, to a lesser extent, the ependymal lining of the ventricles, at a rate of 600–700 ml per day in humans. It totals approximately 140 ml in an adult person. Most of this volume occupies the subarachnoid space, less the ventricular lumina, of the brain. The main function of the CSF is to cushion and protect the brain (for reviews, see Segal, 1993, 2000). Due to the blood–brain barrier, macromolecules are not able to enter the CSF from the peripheral circulatory system. However, proteins secreted by brain cells can be found in the CSF. Thus far, using 2D PAGE, almost 200 proteins have been identified in the embryonic human CSF, and more than 500 protein spots could be distinguished in the adult human CSF (Sickmann et al., 2000, 2002; Yuan, Russell, Wood, & Desiderio, 2002; Zappaterra et  al., 2007). A major challenge in the ­analysis has been the presence of impurities caused, for example, by lysed blood; the high salt concentration; the low abundance of most of the proteins; and the dominance of a few other proteins, such as albumin, immunoglobulins, and

520

G.K.H. Zupanc and M.M. Zupanc

t­ransferrin, which represent more than 70% of the total protein content. The latter difficulty has been addressed by including in the sample preparation concentration and desalting steps, as well as prefractionation by chromatographic methods (Davidsson, Paulson, Hesse, Blennow, & Nilsson, 2001; Sickmann et al., 2002).

4.3 Toward a Molecular Identification of Biomarkers of CNS Injury in Humans To analyze proteins in the CSF, samples are most commonly obtained by lumbar puncture. Changes in composition and relative amount of CSF proteins provide valuable diagnostic and prognostic information for a number of neurological ­conditions (for review, see Romeo et al., 2005). One of the most widely used markers is S100b. Its concentration correlates well with the extent of brain or spinal cord damage caused by neurodegenerative disorders, traumatic or focal insults, and stroke (Loy et al., 2005; Michetti & Gazzolo, 2002; Michetti, Massaro, & Murazio, 1979; Mitchetti, Massaro, Russo, & Rigon, 1980). Elevated levels of S100b may also indicate dysfunction of the blood–brain barrier (Kanner et al., 2003). Information on other potential markers is very limited, although a reliable assessment of the severity of brain damage, and the prognosis for recovery, are likely to require a battery of biomarkers, rather than just a single marker. Identification of such biomarkers is an area with enormous scope for the application of proteomics approaches, particularly with techniques becoming increasingly sensitive so that minimal amounts of protein can be detected in the CSF. However, to analyze differential abundance of CSF proteins after TBI in humans, a major problem has been to obtain samples collected under comparable conditions from patients with a defined injury and from a matching control group. To avoid this difficulty, a number of model systems have been tested. In one attempt, CSF was collected from postmortem brains of individuals without head injury, assuming that, in the stage of dying, and shortly after death, cellular processes in the normal brain are similar to those occurring during massive brain injury (Burgess et  al., 2006). Postmortem CSF was collected by ventricular puncture at autopsy, on average 6 h after death. As controls, CSF samples were obtained by lumbar puncture from patients who were under clinical treatment for disorders other than brain injuries or neurological diseases. However, the validity of this approach is questionable, since there is evidence that a rostrocaudal concentration gradient of CSF protein exists (Sommer, Gaul, Heckmann, Neundörfer, & Erbguth, 2002; Weisner & Bernhardt, 1978), thus making it difficult to compare samples taken from different regions of the CNS. Moreover, the extremely massive biochemical changes associated with death of the entire brain do not necessarily adequately reflect the changes occurring after more focal brain injuries, when major parts of the brain are left intact, or are less affected. It is, therefore, not surprising that 172 out of 299 proteins identified in this study have previously not been found in the CSF, and that more than 75% of the former are intracellular proteins presumably released from the dying cells.

Proteomic Analysis of CNS Injury and Recovery

521

It is unclear how many of these proteins could, indeed, serve as biochemical markers in brain injury patients, with the aim of assessing the severity of the injury and making a prognosis about the further course of recovery. Similar criticism applies to a second study performed on humans (Conti et al., 2004). CSF samples were collected from trauma patients by an external ventricular catheter inserted as part of the normal therapeutic intervention to limit intracranial brain pressure, whereas control CSF was obtained from the subarchnoid space through lumbar puncture. Due to the rostrocaudal concentration gradient of CSF proteins, this approach raises questions of whether or not the results of a proteome analysis of the two types of samples are comparable. Keeping these limitations in mind, the main finding of this study was that proteolytic degradation products of fibrinogen-b occur only in the CSF of TBI patients. Perhaps this reflects a posttraumatic increase in fibrinolysis following brain injury. Due to the severe limitations in obtaining comparable samples from human patients with brain injury and from control subjects, an alternative approach to identify potential markers for the assessment of the severity of brain damage was proposed by Siman et al. (2004). This approach is based on the assumption that the same proteins that are released by neurons after brain injury (and which can be detected after some time in the CSF) are also released by cultured neurons into the medium after treatment with a pro-necrotic agent. By combining proteome analysis with Western blotting, a number of proteins released from necrotic neurons into the culture medium could be identified. Evidence in favor of the applicability of this approach was obtained through immunodetection of some of these proteins in the CSF of rats employing experimental models of TBI or transient global forebrain ischemia. These proteins include the b- and V-isoforms of protein 14-3-3 (which is involved in cell cycle progression and regulation of apoptosis; for reviews, see Hermeking & Benzinger, 2006; Porter et al., 2006), as well as calpain- and caspase-cleavage products of the microtubule protein tau and the actin-binding protein a-spectrin. The occurrence of protein 14-3-3, as well as the cleavage products of protein tau and a-spectrin, in the CSF is likely to reflect the cell death in tissue damaged through the traumatic insult or the transient ischemia. The differential expression of these proteins appears to be specific, as 72 other proteins detected by 2D PAGE did not show any alteration after TBI.

5 Comparative Proteomic Analysis of RegenerationCompetent vs. Regeneration-Deficient Systems 5.1 Comparative Proteomics as a Strategy to Identify Novel Signals Associated with Regeneration An exciting, yet largely unexplored strategy to identify natural signals that promote or inhibit regeneration is to employ a comparative analysis of regeneration-­ competent vs. regeneration-deficient systems. This can be achieved by comparing,

522

G.K.H. Zupanc and M.M. Zupanc

after experimentally applied lesions, the proteome of two different animal species – one with an intrinsic ability to regenerate CNS tissue after injury vs. the other that lacks such a capability. A similar approach can be applied using developmental stages differing in regenerative potential. Anurans, for example, are able to regenerate amputated hindlimbs during early tadpole stages, but gradually loose this capacity during subsequent development (for review, see Suzuki et al., 2006). In the following, we will discuss the results of a proteomic analysis of injuryassociated proteins in the adult teleost fish brain – a regeneration-competent system – and compare these findings with the changes in abundance of specific proteins observed in the adult mammalian brain – a regeneration-deficient system – after TBI.

5.2 Regenerative Potential in the Teleost Fish Brain The brain of adult teleost fish is one of the best-examined regeneration-­competent systems among vertebrates (for reviews, see Hitchcock & Raymond, 1992; Hitchcock, Ochocinska, Sieh, & Otteson, 2004; Otteson & Hitchcock, 2003; Zupanc, 2006a, 2006b, 2008a, 2008b; Zupanc & Zupanc, 2006). This capacity includes regeneration of both axons (“axonal regeneration”) and whole neurons (“neuronal regeneration”). The latter is closely linked to the enormous neurogenic potential in the intact brain. In the two teleostean species in which a quantitative analysis has been performed thus far, the brown ghost knifefish (Apteronotus leptorhynchus) and the zebrafish (Danio rerio), the rate of constitutive cell proliferation, relative to the total number of brain cells, has been estimated to be at least one, if not two, orders of magnitude higher than in the intact mammalian brain (Hinsch & Zupanc, 2007; Zupanc, Hinsch, & Gage, 2005; Zupanc & Horschke, 1995; Zupanc & Zupanc, 2006). These new cells originate from pluripotent adult stem cells harbored in dozens of specific ­proliferation zones within the brain (Hinsch & Zupanc, 2006). Approximately half of the new cells persist for the rest of the fish’s life and develop into a ­variety of cell types, including neurons and glial cells (Hinsch & Zupanc, 2007; Ott, Zupanc, & Horschke, 1997; Zupanc, Horschke, Ott, & Rascher, 1996; Zupanc et al., 2005). After application of mechanical lesions to the cerebellum (Fig. 1), cell loss is restricted to an initial wave of apoptotic cell death, thus lacking a secondary wave of cell death common in the mammalian brain (Zupanc, Kompass, Horschke, Ott, & Schwarz, 1998). Further, there is no evidence for necrosis, a common type of cell death in mammals after TBI. The cells lost to injury are replaced by new cells within a few weeks. These new cells are recruited by inducing an increase in mitotic activity of stem cells in areas exhibiting constitutive neurogenesis, and by activating stem cells that are quiescent in the intact brain (Zupanc & Ott, 1999).

Proteomic Analysis of CNS Injury and Recovery

523

Fig. 1  The cerebellar lesion paradigm. (a) Side view of the brain of A. leptorhynchus. Cerebellar lesions are applied by puncturing the skull of the fish with a surgical blade in a defined location of the head, thus generating a stab wound in one hemisphere in the rostral part of the corpus ­cerebelli (CCb). (b) Transverse section through the corpus cerebelli taken at the level indicated by the arrow in (a). The lesion path (arrowheads) travels through the dorsal molecular layer [CCbmol(d)] and the granular layer (CCb-gra) roughly halfway between the midline of the brain and the lateral edge of the granular layer. BS brain stem; CCb-mol(v) ventral molecular layer of corpus cerebelli; EGp eminentia granularis pars posterior; SC spinal cord; Tel telencephalon; TeO optic tectum; TL torus longitudinalis; VCb-gra granule cell layer of valvula cerebelli; VCb-mol ­molecular layer of valvula cerebelli (from Zupanc & Zupanc, 2006)

5.3 Proteome Analysis of Proteins Involved in Repair of the Teleostean Brain To identify proteins potentially involved in the various steps of the regenerative process, application of a well-defined cerebellar lesion paradigm was combined with differential proteome analysis (Fig. 2) (Zupanc et al., 2006). The abundance of proteins in tissue from the site of the lesion was compared with the abundance of the same proteins in tissue from an equivalent region of the intact cerebellum 3 days after the lesion. This time point is characterized by a transition from degenerative processes, such as apoptotic cell death, to regenerative processes, such as recruitment of new cells destined to replace the cells lost to injury (Zupanc & Ott, 1999; Zupanc et al., 1998).

524

G.K.H. Zupanc and M.M. Zupanc

Fig.  2  Analysis of changes in protein abundance after traumatic brain injury via differential d­ isplay. Tissue samples are collected from the site of injury in lesioned brains and from an equivalent region in intact (control) brains. Protein extracts are separated by two-dimensional (2D) gel electrophoresis. After staining, the gels are scanned, and the images analyzed for differences in protein spot pattern using specific computer software. Protein spots of interest are excised from the gels and subjected to in-gel digestion with trypsin. The peptide mass fingerprint for every

Proteomic Analysis of CNS Injury and Recovery

525

2D PAGE of lysates from these two types of tissue samples revealed nearly 800 protein spots (Fig.  3a). Spot intensity was statistically significantly increased at least twofold in 30 of these spots, and decreased to at least half the intensity of intact tissue in 23 spots (Fig. 3b). The proteins associated with 24 of this total of 53 spots could be identified by peptide mass fingerprinting and MS/MS fragmentation. The identified proteins can be divided into three groups. The first group includes cytoskeletal proteins essential for the formation of new cells, such as b-actin. The transient up-regulation of b-actin expression might relate to the regeneration of injured axons in the area of the lesion and the development of axons of the newly generated cells, as suggested by immunohistochemical staining of brain sections after application of mechanical lesions to the corpus cerebelli (Fig. 4). Such a proposed function is in line with the results of regeneration studies in mammals, which have demonstrated an up-regulation of actin within a few days after axotomy, reflecting the emergence of axonal sprouts and the concomitant need to lengthen the axon shaft and to support membrane expansion at the growth cone (Lund, Machado, & McQuarrie, 2002; Lund & McQuarrie, 1996; Tetzlaff, Alexander, Miller, & Bisby, 1991). Also included in this first group are proteins that mediate the correct assembly of these structural proteins, such as: chaperonin containing tailless-complex polypeptide 1, subunit e, which might mediate chaperoning of cytoskeletal proteins (Gao, Thomas, Chow, Lee, & Cowan, 1992; Yaffe et  al., 1992); tropomodulins-3 and -4, which are known to perform a capping of slowgrowing ends of actin filaments (for review, see Fischer & Fowler, 2003); and bullous pemphigoid antigen 1, which has been proposed to link actin with intermediate filament proteins (Leung, Zheng, Prater, & Liem, 2001; Yang et al., 1996). The second group is comprised of a variety of proteins: keratin-10, which in keratinocytes has been shown to negatively regulate cell proliferation (Paramio et  al., 1999; Santos, Paramio, Bravo, Ramirez, & Jorcano, 2002); myosin heavy chain, which has been implicated in cellular motility in the developing nervous system (for review, see Brown & Bridgman, 2004); 78,000-Da glucose-regulated protein, glutamine synthetase, and cytosolic aspartate aminotransferase, which appear to be involved in various aspects of neuroprotection (Gorovits, Avidan, Avisar, Shaked, & Vardimon, 1997; Kiang & Tsokos, 1998; Lai, Murthy, Cooper, Hertz, & Hertz, 1989); two isoforms of enolase, whose increase in abundance might not only reflect a higher metabolic demand of cells during brain repair but also their putative function as neurotrophic factors promoting survival of neurons (Hattori, Takei, Mizuno, Kato, & Kohsaka, 1995; Li, Lane, Johnson, Chader, & Tombran-

Fig. 2 (continued)  digested protein is analyzed through matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) and matched against protein databases to identify protein candidates. This identification step is validated by acquisition of tandem mass spectrometry (MS/MS) spectra of selected peptides and analysis of their sequences

526

G.K.H. Zupanc and M.M. Zupanc

Fig. 3  (a) Images of 2D gels of proteins from the intact corpus cerebelli (control) and lesioned corpus cerebelli (lesion) at a post-lesion survival time of 3 days in A. leptorhynchus. (b) Scatter plot of the fold changes of 772 polypeptide spots in the corpus cerebelli of A. leptorhynchus 3 days after a lesion compared to unlesioned controls. Decrease in protein abundance is indicated by negative changes, increase by positive changes. Proteins are sorted in ascending order of the spot intensity change. The analysis is based on a comparison of the averaged spot intensities of four 2D gels from lesioned brains and three 2D gels from intact brains. Dotted lines indicate the +2 fold and −2 fold threshold (after Zupanc et al., 2006)

Fig. 4 (continued)  protein in the brain of this species was independently confirmed, and the pattern of differential expression examined in more detail, by immunohistochemical techniques. Immunohistochemical sections of the intact (a) and injured (b–d) corpus cerebelli are shown on the left, the corresponding DAPI counterstains on the right (a¢–d¢). In the injured cerebellum, the lesion path, indicated by arrowheads, runs through the dorsal molecular layer (d) into the granular layer (gra). Whereas in the intact corpus cerebelli immunolabeling is virtually absent (a), immunoreactive fibers have developed 1 day after lesioning in the area around the lesion path within the dorsal molecular layer (b). Three days after lesioning, the size of the fiber plexus and the intensity of immunostaining of fibers has increased (c). Fifteen days after lesioning of the cerebellum, both the number and the intensity of immunolabeling has declined, compared to the pattern observed at the 3-day survival time (d) (after Zupanc et al., 2006)

Proteomic Analysis of CNS Injury and Recovery

527

Fig.  4  Validation and expansion by immunohistochemistry of results obtained through proteome analysis. Differential proteome analysis had indicated an increase in abundance of b-actin 3 days after application of a stab-wound lesion to the cerebellum of A. leptorhynchus. The existence of a b-actin-like

528

G.K.H. Zupanc and M.M. Zupanc

Tink, 1995; Lin & Matesic, 1994; Slemmer, Weber, & De Zeeuw, 2004; Takei et al., 1991); and F-ATP synthase, which catalyzes the formation of ATP from ADP and phosphate, thus playing a role in energy metabolism (for review, see Yoshida, Muneyuki, & Hisabori, 2001). The third group consists of a single protein, bone marrow zink finger 2. This protein was first identified from a screen of zinc finger proteins in the hematopoietic system (Han et al., 1999). Although nothing is known about its role in the nervous system, its molecular structure and some experimental ­evidence based on examination of Wilms’ tumor, a pediatric kidney cancer, suggest that bone marrow zinc finger 2 acts as a transcriptional regulator (Lee, Lwu, Kim, & Pelletier, 2002). During regeneration of the teleost cerebellum, it may exert a similar function – regulating the transcription of genes of other proteins. Proteins that differ in their expression pattern after TBI between regenerationcompetent systems and regeneration-deficient systems are particularly interesting from a comparative point of view. For example, the increase in abundance of glutamine synthetase 3 days after cerebellar lesions in fish (Zupanc et  al., 2006) contrasts with the down-regulation of this enzyme under traumatic conditions in the mammalian CNS (Grosche, Hartig, & Reichenbach, 1995; Härtig et  al., 1995; Lewis, Erickson, Guerin, Anderson, & Fisher, 1989; Lewis, Guerin, Anderson, Matsumoto, & Fisher, 1994; Oliver et  al., 1990; Smith et  al., 1991). Glutamine synthetase is a glia-specific enzyme that converts synaptically released glutamate into the nontoxic amino acid glutamine. Under normal conditions, this mechanism prevents the accumulation of neurotoxic amounts of glutamate in neural tissue and thus protects neurons from cell death. However, under traumatic conditions, the amount of glutamine synthetase is insufficient to catalyze the excessive amounts of glutamate released by damaged cells. The additional down-regulation of glutamine synthetase after brain trauma in ­mammals aggravates the situation in mammals, thus failing to limit the spread of damage caused by continuous over-excitation of post-synaptic glutamate receptors. On the other hand, the up-regulation of glutamine synthetase in the teleostean brain is likely to enable fish to reduce the neurodegenerative effect induced by glutamate neurotoxicity. Although largely unexplored, such differences between regeneration-competent species and regeneration-incompetent species, revealed by comparative proteome analysis, are likely to prove extremely powerful to identify target mechanisms for therapeutic intervention after TBI in humans.

6 Potential Problems and Limitations of Proteome Analysis of Brain Injury-Associated Proteins Proteome analysis by 2D PAGE enables investigators to resolve proteins associated with CNS injury on a global scale. However, there are a number of potential problems and limitations of this approach that have to be taken into account when designing a study or interpreting results.

Proteomic Analysis of CNS Injury and Recovery

529

First, a given method of sample preparation allows the investigator to resolve only certain types of proteins. Different methods are required to visualize cytoplasmic vs. membrane-bound proteins, or proteins in the medium molecular-weight range vs. high molecular proteins, or more acidic/basic proteins vs. proteins in the standard pI range of 4–7. Second, for certain low-abundance proteins, the sensitivity of stains used in 2D PAGE may be too low to be visualized. For example, the minimum amount of protein required for detection with Coomassie Brilliant Blue G-250 is 30–100 ng, but this threshold can be lowered to 8–10  ng if the dye is allowed to form colloidal microprecipitates (Neuhoff, Arold, Taube, & Ehrhardt, 1988). For comparison, the detection threshold of proteins is 1–10 ng when using silver nitrate (Lilley, Razzaq, & Dupree, 2002; Tonge et  al., 2001) and approximately 500  pg when employing CyDye DIGE fluors (Marouga, David, & Hawkins, 2005). These limitations may be (partially) overcome by pre-fractionation of samples (Barnea, Sorkin, Ziv, Beer, & Admon, 2005; Lopez et al., 2000; Serna-Sanz, Rairdan, & Peck, 2007; Tang et al., 2005; Yuan & Desiderio, 2005). Third, to obtain sufficient amount of protein, it is often necessary to pool samples from different individuals. This averaging obliterates possible inter-individual differences, which, for example, may be related to sex or age. Care has to be taken in the experimental design to avoid possible bias in sampling, or to specifically address such hypothetical effects of sex, age, or other factors on the regenerative potential. Fourth, sometimes careful consideration is required to define proper controls for differential proteome analysis. For example, in the case of application of a lesion to a certain brain region in one hemisphere, the investigator has to decide whether the control tissue is taken from the corresponding region in the contralateral hemisphere, or from the equivalent region in intact or sham-operated animals. After unilateral lesion to the teleostean cerebellum, for example, most changes in cellular processes were found to be restricted to the ipsilateral hemisphere, with one exception. The number of microglia/macrophages was elevated not only ipsilaterally, but also – although to a lesser extent – contralaterally (Zupanc et al., 2003). In such a case, differential proteome analysis will evidently yield different results depending on whether control tissue is taken from the part of the cerebellum contralateral to the lesion, or from the cerebellum of an intact animal. Fifth, after application of local brain lesions, tissue samples taken from the site of injury will, in most cases, be heterogenous in terms of their composition. Different parts of the sample may have been affected differently by the traumatic insult, and different cell types are likely spread throughout the piece of tissue. This will have a negative effect on the signal-to-noise ratio when performing differential proteome analysis. In some instances, it may be possible to improve the purity of the sample by use of laser-assisted microdissection methods (Sirivatanauksorn, Drury, Crnogorac-Jurcevic, Sirivatanauksorn, & Lemoine, 1999). Sixth, with the large number of protein spots examined simultaneously by 2D PAGE, careful statistical analysis is required to avoid erroneous conclusions. For example, if one compares the ratio of spot intensity between two repetitions of the same experiment analyzing a gel with 1,000 protein spots, and the error in the

530

G.K.H. Zupanc and M.M. Zupanc

measurement is normally distributed, 50 of the ratios are expected to be more than two standard deviations from the mean value of 1. Thus, if the standard deviation is 0.5, a good number of protein spots will exhibit more than twofold changes in intensity – just as a consequence of statistical variability! Seventh, as it is possible that some changes in protein spot intensity are caused by statistical variability, it is important to verify alterations in protein abundance by independent approaches, such as Western blotting, immunohistochemistry, and/or in situ hybridization. Such validation experiments may be expanded to include analysis of the spatiotemporal pattern of gene expression or protein abundance by examination of tissue at different time points post injury (Fig. 4). Eighth, changes in protein spot intensity in brain samples taken post-injury may not only indicate altered levels of translation, but also posttranslational modifications or proteolytic degradation, the latter two resulting in altered protein mass and/ or charge. Proteolysis is encountered frequently in cases in which cells undergo apoptotic or necrotic cell death following the injury. Ninth, changes in protein spot intensity after injury provide only correlative evidence that the protein associated with the spot is involved in degeneration or regeneration. The exact role of the respective protein has to be examined employing different approaches, such as the modification of the level of expression of the gene encoding this protein, followed by in vivo and/or in vitro examination of the effect of this experimental manipulation.

7 Perspectives Proteomics allows for a large-scale identification of proteins involved in processes occurring after CNS injury. As detailed above, among the proteins found with altered abundance after injury in the mammalian brain and spinal cord are several that are known to inhibit cell death, promote survival, and mediate regeneration of damaged neurons. These changes in protein abundance appear to reflect the attempt of the mammalian CNS to initiate repair programs. However, with the exception of a few CNS regions that exhibit a certain degree of plasticity, degenerative processes prevail in adult mammals, leading to the death of whole neurons and the loss of function. This contrasts with the enormous regenerative potential of other vertebrates, particularly teleost fish and certain developmental stages of anurans. Proteome analysis based on such model systems has revealed a number of proteins that potentially mediate the capability for CNS repair. These proteins appear to define a permissive environment that promotes regeneration in the CNS of such species. The same or similar proteins may also be able to initiate repair processes in the mammalian CNS, since experimental evidence has shown that modification of the local microenvironment can stimulate cell proliferation and subsequent development of endogenous, under normal conditions quiescent, neural precursors in mammals (for review, see Zupanc & Zupanc, 2006). Thus, comparative proteome analysis based on CNS systems differing in their intrinsic regenerative potentials could open new vistas to identify potential targets for therapeutic intervention to increase the repair capacity of the human brain.

Proteomic Analysis of CNS Injury and Recovery

531

References Arvidsson, A., Collin, T., Kirik, D., Kokaia, Z., & Lindvall, O. (2002). Neuronal replacement from endogenous precursors in the adult brain after stroke. Nature Medicine, 8, 963–970. Barnea, E., Sorkin, R., Ziv, T., Beer, I., & Admon, A. (2005). Evaluation of prefractionation ­methods as a preparatory step for multidimensional based chromatography of serum proteins. Proteomics, 5, 3367–3375. Brown, M. E., & Bridgman, P. C. (2004). Myosin function in nervous and sensory systems. Journal of Neurobiology, 58, 118–130. Burgess, J. A., Lescuyer, P., Hainard, A., Burkhard, P. R., Turck, N., Michel, P., et  al. (2006). Identification of brain cell death associated proteins in human post-mortem cerebrospinal fluid. Journal of Proteome Research, 5, 1674–1681. Busch, S. A., & Silver, J. (2007). The role of extracellular matrix in CNS regeneration. Current Opinion in Neurobiology, 17, 120–127. Conti, A., Sanchez-Ruiz, Y., Bachi, A., Beretta, L., Grandi, E., Beltramo, M., et  al. (2004). Proteome study of human cerebrospinal fluid following traumatic brain injury indicates fibrin(ogen) degradation products as trauma-associated markers. Journal of Neurotrauma, 21, 854–863. Darling, D. L., Yingling, J., & Wynshaw-Boris, A. (2005). Role of 14-3-3 proteins in eukaryotic signaling and development. Current Topics in Developmental Biology, 68, 281–315. Davidsson, P., Paulson, L., Hesse, C., Blennow, K., & Nilsson, C. L. (2001). Proteome studies of human cerebrospinal fluid and brain tissue using a preparative two-dimensional electrophoresis approach prior to mass spectrometry. Proteomics, 1, 444–452. Dieringer, N. (2003). Activity-related postlesional vestibular reorganization. Annals of the New York Academy of Sciences, 1004, 50–60. Ding, Q., Wu, Z., Guo, Y., Zhao, C., Jia, Y., Kong, F., et al. (2006). Proteome analysis of up-­regulated proteins in the rat spinal cord induced by transection injury. Proteomics, 6, 505–518. Doucet, A., & Overall, C. M. (2008). Protease proteomics: Revealing protease in vivo functions using systems biology approaches. Molecular Aspects of Medicine, 29, 339–358. Fischer, R. S., & Fowler, V. M. (2003). Tropomodulins: Life at the slow end. Trends in Cell Biology, 13, 593–601. Gao, Y., Thomas, J. O., Chow, R. L., Lee, G.-H., & Cowan, N. J. (1992). A cytoplasmic chaperonin that catalyzes b-actin folding. Cell, 69, 1043–1050. Gorovits, R., Avidan, N., Avisar, N., Shaked, I., & Vardimon, L. (1997). Glutamine synthetase protects against neuronal degeneration in injured retinal tissue. Proceedings of the National Academy of Sciences of the United States of America, 94, 7024–7029. Graeber, M. B., Raivich, G., & Kreutzberg, G. W. (1989). Increase of transferrin receptors and iron uptake in regenerating motor neurons. Journal of Neuroscience Research, 23, 342–345. Grosche, J., Hartig, W., & Reichenbach, A. (1995). Expression of glial fibrillary acidic protein (GFAP), glutamine synthetase (GS), and Bcl-2 protooncogene protein by Müller (glial) cells in retinal light damage of rats. Neuroscience Letters, 185, 119–122. Han, Z. G., Zhang, Q. H., Ye, M., Kan, L. X., Gu, B. W., He, K. L., et al. (1999). Molecular ­cloning of six novel Krüppel-like zinc finger genes from hematopoietic cells and identification of a novel transregulatory domain KRNB. Journal of Biological Chemistry, 274, 35741–35748. Härtig, W., Grosche, J., Distler, C., Grimm, D., el-Hifnawi, E., & Reichenbach, A. (1995). Alterations of Müller (glial) cells in dystrophic retinae of RCS rats. Journal of Neurocytology, 24, 507–517. Hattori, T., Takei, N., Mizuno, Y., Kato, K., & Kohsaka, S. (1995). Neurotrophic and neuroprotective effects of neuron-specific enolase on cultured neurons from embryonic rat brain. Neuroscience Research, 21, 191–198. Hedgecock, E. M., Culotti, J. G., Thomson, J. N., & Perkins, L. A. (1985). Axonal guidance mutants of Caenorhabditis elegans identified by filling sensory neurons with fluorescein dyes. Developmental Biology, 111, 158–170. Hermeking, H., & Benzinger, A. (2006). 14-3-3 proteins in cell cycle regulation. Seminars in Cancer Biology, 16, 183–192.

532

G.K.H. Zupanc and M.M. Zupanc

Hinsch, K., & Zupanc, G. K. H. (2006). Isolation, cultivation, and differentiation of neural stem cells from adult fish brain. Journal of Neuroscience Methods, 158, 75–88. Hinsch, K., & Zupanc, G. K. H. (2007). Generation and long-term persistence of new neurons in the adult zebrafish brain: A quantitative analysis. Neuroscience, 146, 679–696. Hitchcock, P., Ochocinska, M., Sieh, A., & Otteson, D. (2004). Persistent and injury-induced neurogenesis in the vertebrate retina. Progress in Retinal and Eye Research, 23, 183–194. Hitchcock, P. F., & Raymond, P. A. (1992). Retinal regeneration. Trends in Neurosciences, 15, 103–108. Horie, H., Inagaki, Y., Sohma, Y., Nozawa, R., Okawa, K., Hasegawa, M., et al. (1999). Galectin-1 regulates initial axonal growth in peripheral nerves after axotomy. Journal of Neuroscience, 19, 9964–9974. Inagaki, Y., Sohma, Y., Horie, H., Nozawa, R., & Kadoya, T. (2000). Oxidized galectin-1 promotes axonal regeneration in peripheral nerves but does not possess lectin properties. European Journal of Biochemistry, 267, 2955–2964. Jenkins, L. W., Peters, G. W., Dixon, C. E., Zhang, X., Clark, R. S. B., Skinner, J. C., et al. (2002). Conventional and functional proteomics using large format two-dimensional gel electrophoresis 24 hours after controlled cortical impact in postnatal day 17 rats. Journal of Neurotrauma, 19, 715–740. Johansson, B. B. (2007). Regeneration and plasticity in the brain and spinal cord. Journal of Cerebral Blood Flow and Metabolism, 27, 1417–1430. Kang, S. K., So, H. H., Moon, Y. S., & Kim, C. H. (2006). Proteomic analysis of injured spinal cord tissue proteins using 2-DE and MALDI-TOF MS. Proteomics, 6, 2797–2812. Kanner, A. A., Marchi, N., Fazio, V., Mayberg, M. R., Koltz, M. T., Siomin, V., et  al. (2003). Serum S100beta: A noninvasive marker of blood-brain barrier function and brain lesions. Cancer, 97, 2806–2813. Kiang, J. G., & Tsokos, G. C. (1998). Heat shock protein 70 kDa: Molecular biology, biochemistry, and physiology. Pharmacology & Therapeutics, 80, 183–201. Kobeissy, F. H., Ottens, A. K., Zhang, Z., Liu, M. C., Denslow, N. D., Dave, J. R., et al. (2006). Novel differential neuroproteomics analysis of traumatic brain injury in rats. Molecular & Cellular Proteomics, 5, 1887–1898. Kochanek, A. R., Kline, A. E., Gao, W.-M., Chadha, M., Lai, Y., Clark, R. S. B., et al. (2006). Gel-based hippocampal proteomic analysis 2 weeks following traumatic brain injury to ­immature rats using controlled cortical impact. Developmental Neuroscience, 28, 410–419. Lacour, M. (2006). Restoration of vestibular function: Basic aspects and practical advances for rehabilitation. Current Medical Research and Opinion, 22, 1651–1659. Lai, J. C. K., Murthy, C. R. K., Cooper, A. J. L., Hertz, E., & Hertz, L. (1989). Differential effects of ammonia and b-methylene-DL-aspartate on metabolism of glutamate and related amino acids by astrocytes and neurons in primary culture. Neurochemical Research, 14, 377–389. Lee, T. H., Lwu, S., Kim, J., & Pelletier, J. (2002). Inhibition of Wilms tumor 1 transactivation by bone marrow zinc finger 2, a novel transcriptional repressor. Journal of Biological Chemistry, 277, 44826–44837. Leung, C. L., Zheng, M., Prater, S. M., & Liem, R. K. (2001). The BPAG1 locus: Alternative splicing produces multiple isoforms with distinct cytoskeletal linker domains, including predominant isoforms in neurons and muscles. Journal of Cell Biology, 154, 691–697. Lewis, G. P., Erickson, P. A., Guerin, C. J., Anderson, D. H., & Fisher, S. K. (1989). Changes in the expression of specific Müller cell proteins during long-term retinal detachment. Experimental Eye Research, 49, 93–111. Lewis, G. P., Guerin, C. J., Anderson, D. H., Matsumoto, B., & Fisher, S. K. (1994). Rapid changes in the expression of glial cell proteins caused by experimental retinal detachment. American Journal of Ophthalmology, 118, 368–376. Li, A., Lane, W. S., Johnson, L. V., Chader, G. J., & Tombran-Tink, J. (1995). Neuron-specific enolase: A neuronal survival factor in the retinal extracellular matrix? Journal of Neuroscience, 15, 385–393.

Proteomic Analysis of CNS Injury and Recovery

533

Lifshitz, J., Sullivan, P. G., Hovda, D. A., Wieloch, T., & McIntosh, T. K. (2004). Mitochondrial damage and dysfunction in traumatic brain injury. Mitochondrion, 4, 705–713. Lilley, K. S., Razzaq, A., & Dupree, P. (2002). Two-dimensional gel electrophoresis: Recent advances in sample preparation, detection and quantitation. Current Opinion in Chemical Biology, 6, 46–50. Lin, R. C. S., & Matesic, D. F. (1994). Immunohistochemical demonstration of neuron-specific enolase and microtubule-associated protein 2 in reactive astrocytes after injury in the adult forebrain. Neuroscience, 60, 11–16. Lopez, M. F., Kristal, B. S., Chernokalskaya, E., Lazarev, A., Shestopalov, A. I., Bogdanova, A., et al. (2000). High-throughput profiling of the mitochondrial proteome using affinity fractionation and automation. Electrophoresis, 21, 3427–3440. López-Otín, C., & Overall, C. M. (2002). Protease degradomics: A new challenge for proteomics. Nature Reviews Molecular Cell Biology, 3, 509–519. Loy, D. N., Sroufe, A. E., Pelt, J. L., Burke, D. A., Cao, Q.-L., Talbott, J. F., et al. (2005). Serum biomarkers for experimental acute spinal cord injury: Rapid elevation of neuron-specific ­enolase and S-100b. Neurosurgery, 56, 391–397. Lund, L. M., Machado, V. M., & McQuarrie, I. G. (2002). Increased b-actin and tubulin polymerization in regrowing axons: Relationship to the conditioning lesion effect. Experimental Neurology, 178, 306–312. Lund, L. M., & McQuarrie, I. G. (1996). Axonal regrowth upregulates b-actin and Jun D mRNA expression. Journal of Neurobiology, 31, 476–486. Marouga, R., David, S., & Hawkins, E. (2005). The development of the DIGE system: 2D fluorescence difference gel analysis technology. Analytical and Bioanalytical Chemistry, 382, 669–678. Michetti, F., & Gazzolo, D. (2002). S100B protein in biological fluids: A tool for perinatal ­medicine. Clinical Chemistry, 48, 2097–2104. Michetti, F., Massaro, A., & Murazio, M. (1979). The nervous system-specific S-100 antigen in cerebrospinal fluid of multiple sclerosis patients. Neuroscience Letters, 11, 171–175. Michetti, F., Massaro, A., Russo, G., & Rigon, G. (1980). The S-100 antigen in cerebrospinal fluid as a possible index of cell injury in the nervous system. Journal of the Neurological Sciences, 44, 259–263. Monteoliva, L., & Albar, J. P. (2004). Differential proteomics: An overview of gel and non-gel based approaches. Briefings in Functional Genomics & Proteomics, 3, 220–239. Nakamura, F., Kalb, R. G., & Strittmatter, S. M. (2000). Molecular basis of semaphorin-mediated axon guidance. Journal of Neurobiology, 44, 219–229. Neuhoff, V., Arold, N., Taube, D., & Ehrhardt, W. (1988). Improved staining of proteins in polyacrylamide gels including isoelectric focusing gels with clear background at nanogram sensitivity using Coomassie Brilliant Blue G-250 and R-250. Electrophoresis, 9, 255–262. Nicholls, D. G., & Budd, S. L. (2000). Mitochondria and neuronal survival. Physiological Reviews, 80, 315–360. Oliver, C. N., Starke-Reed, P. E., Stadtman, E. R., Liu, G. J., Carney, J. M., & Floyd, R. A. (1990). Oxidative damage to brain proteins, loss of glutamine synthetase activity, and production of free radicals during ischemia/reperfusion-induced injury to gerbil brain. Proceedings of the National Academy of Sciences of the United States of America, 87, 5144–5147. Opii, W. O., Nukala, V. N., Sultana, R., Pandya, J. D., Day, K. M., Merchant, M. L., et al. (2007). Proteomic identification of oxidized mitochondrial proteins following experimental traumatic brain injury. Journal of Neurotrauma, 24, 772–789. Ott, R., Zupanc, G. K. H., & Horschke, I. (1997). Long-term survival of postembryonically born cells in the cerebellum of gymnotiform fish, Apteronotus leptorhynchus. Neuroscience Letters, 221, 185–188. Otteson, D. C., & Hitchcock, P. F. (2003). Stem cells in the teleost retina: Persistent neurogenesis and injury-induced regeneration. Vision Research, 43, 927–936.

534

G.K.H. Zupanc and M.M. Zupanc

Paramio, J. M., Casanova, M. L., Segrelles, C., Mittnacht, S., Lane, E. B., & Jorcano, J. L. (1999). Modulation of cell proliferation by cytokeratins K10 and K16. Molecular and Cellular Biology, 19, 3086–3094. Paterson, J. M., Short, D., Flatman, P. W., Seckl, J. R., Aitken, A., & Dutia, M. B. (2006). Changes in protein expression in the rat medial vestibular nuclei during vestibular compensation. Journal of Physiology, 575, 777–788. Payne, B. R., & Lomber, S. G. (2002). Plasticity of the visual cortex after injury: What’s different about the young brain? The Neuroscientist, 8, 174–185. Pekny, M., & Nilsson, M. (2005). Astrocyte activation and reactive gliosis. Glia, 50, 427–434. Pineda, J. A., Wang, K. K. W., & Hayes, R. L. (2004). Biomarkers of proteolytic damage ­following traumatic brain injury. Brain Pathology, 14, 202–209. Porter, G. W., Khuri, F. R., & Fu, H. (2006). Dynamic 14-3-3/client protein interactions integrate survival and apoptotic pathways. Seminars in Cancer Biology, 16, 193–202. Raivich, G., Graeber, M. B., Gehrmann, J., & Kreutzberg, G. W. (1991). Transferrin receptor expression and iron uptake in the injured and regenerating rat sciatic nerve. European Journal of Neuroscience, 3, 919–927. Reaume, A. G., Elliott, J. L., Hoffman, E. K., Kowall, N. W., Ferrante, R. J., Siwek, D. R., et al. (1996). Motor neurons in Cu/Zn superoxide dismutase-deficient mice develop normally but exhibit enhanced cell death after axonal injury. Nature Genetics, 13, 43–47. Rego, A. C., & Oliveira, C. R. (2003). Mitochondrial dysfunction and reactive oxygen species in excitotoxicity and apoptosis: Implications for the pathogenesis of neurodegenerative diseases. Neurochemical Research, 28, 1563–1574. Romeo, M. J., Espina, V., Lowenthal, M., Espina, B. H., Petricoin, E. F., III, & Liotta, L. A. (2005). CSF proteome: A protein repository for potential biomarker identification. Expert Review of Proteomics, 2, 57–70. Rowland, L. P., & Sciarra, D. (1989). Trauma. In L. P. Rowland (Ed.), Merritt’s textbook of neurology (8th ed., pp. 369–393). Philadelphia, London: Lea & Febiger. Santos, M., Paramio, J. M., Bravo, A., Ramirez, A., & Jorcano, J. L. (2002). The expression of keratin K10 in the basal layer of the epidermis inhibits cell proliferation and prevents skin tumorigenesis. Journal of Biological Chemistry, 277, 19122–19130. Segal, M. B. (1993). Extracellular and cerebrospinal fluids. Journal of Inherited Metabolic Disease, 16, 617–638. Segal, M. B. (2000). The choroid plexuses and the barriers between the blood and the cerebrospinal fluid. Cellular and Molecular Neurobiology, 20, 183–196. Serna-Sanz, A., Rairdan, G., & Peck, S. C. (2007). Preparative denaturing isoelectric focusing for enhancing sensitivity of proteomic studies. Methods of Molecular Biology, 354, 99–104. Sickmann, A., Dormeyer, W., Wortelkamp, S., Woitalla, D., Kuhn, W., & Meyer, H. E. (2000). Identification of proteins from human cerebrospinal fluid, separated by two-dimensional ­polyacrylamide gel electrophoresis. Electrophoresis, 21, 2721–2728. Sickmann, A., Dormeyer, W., Wortelkamp, S., Woitalla, D., Kuhn, W., & Meyer, H. E. (2002). Towards a high resolution separation of human cerebrospinal fluid. Journal of Chromatography B, 771, 167–196. Siddiqui, S. S., & Culotti, J. G. (1991). Examination of neurons in wild type and mutants of Caenorhabditis elegans using antibodies to horseradish peroxidase. Journal of Neurogenetics, 7, 193–211. Siman, R., McIntosh, T. K., Soltesz, K. M., Chen, Z., Neumar, R. W., & Roberts, V. L. (2004). Proteins released from degenerating neurons are surrogate markers for acute brain damage. Neurobiology of Disease, 16, 311–320. Sirivatanauksorn, Y., Drury, R., Crnogorac-Jurcevic, T., Sirivatanauksorn, V., & Lemoine, N. R. (1999). Laser-assisted microdissection: Applications in molecular pathology. Journal of Pathology, 189, 150–154. Slemmer, J. E., Weber, J. T., & De Zeeuw, C. I. (2004). Cell death, glial protein alterations and elevated S-100b release in cerebellar cell cultures following mechanically induced trauma. Neurobiology of Disease, 15, 563–572.

Proteomic Analysis of CNS Injury and Recovery

535

Smith, C. D., Carney, J. M., Starke-Reed, P. E., Oliver, C. N., Stadtman, E. R., Floyd, R. A., et al. (1991). Excess brain protein oxidation and enzyme dysfunction in normal aging and in Alzheimer disease. Proceedings of the National Academy of Sciences of the United States of America, 88, 10540–10543. Sofroniew, M. V. (2005). Reactive astrocytes in neural repair and protection. The Neuroscientist, 11, 400–407. Sommer, J. B., Gaul, C., Heckmann, J., Neundörfer, B., & Erbguth, F. J. (2002). Does lumbar cerebrospinal fluid reflect ventricular cerebrospinal fluid? A prospective study in patients with external ventricular drainage. European Neurology, 47, 224–232. Sullivan, P. G., Rabchevsky, A. G., Waldmeier, P. C., & Springer, J. E. (2005). Mitochondrial permeability transition in CNS trauma: Cause or effect of neuronal cell death? Journal of Neuroscience Research, 79, 231–239. Suzuki, M., Yakushiji, N., Nakada, Y., Satoh, A., Ide, H., & Tamura, K. (2006). Limb regeneration in Xenopus laevis froglet. ScientificWorld Journal, 6(suppl 1), 26–37. Takei, N., Kondo, J., Nagaike, K., Ohsawa, K., Kato, K., & Kohsaka, S. (1991). Neuronal survival factor from bovine brain is identical to neuron-specific enolase. Journal of Neurochemistry, 57, 1178–1184. Tang, H.-Y., Ali-Khan, N., Echan, L. A., Levenkova, N., Rux, J. J., & Speicher, D. W. (2005). A novel four-dimensional strategy combining protein and peptide separation methods enables detection of low-abundance proteins in human plasma and serum proteomes. Proteomics, 5, 3329–3342. Tetzlaff, W., Alexander, S. W., Miller, F. D., & Bisby, M. A. (1991). Response of facial and ­rubrospinal neurons to axotomy: Changes in mRNA expression for cytoskeletal proteins and GAP-43. Journal of Neuroscience, 11, 2528–2544. Tonge, R., Shaw, J., Middleton, B., Rowlinson, R., Rayner, S., Young, J., et al. (2001). Validation and development of fluorescence two-dimensional differential gel electrophoresis proteomics technology. Proteomics, 1, 377–396. Tsai, M.-C., Shen, L.-F., Kuo, H.-S., Cheng, H., & Chak, K.-F. (2008). Involvement of acidic fibroblast growth factor in spinal cord injury repair processes revealed by a proteomics approach. Molecular & Cellular Proteomics, 7, 1668–1687. Tsuchiya, D., Hong, S., Matsumori, Y., Kayama, T., Swanson, R. A., Dillman, W. H., et al. (2003). Overexpression of rat heat shock protein 70 reduces neuronal injury after transient focal ­ischemia, transient global ischemia, or kainic acid-induced seizures. Neurosurgery, 53, 1179–1188. Vance, J. E., Campenot, R. B., & Vance, D. E. (2000). The synthesis and transport of lipids for axonal growth and nerve regeneration. Biochimica et Biophysica Acta, 1486, 84–96. Weisner, B., & Bernhardt, W. (1978). Protein fractions of lumbar, cisternal, and ventricular ­cerebrospinal fluid: Separate areas of reference. Journal of the Neurological Sciences, 37, 205–214. Yaffe, M. B., Farr, G. W., Miklos, D., Horwich, A. L., Sternlicht, M. L., & Sternlicht, H. (1992). TCP1 complex is a molecular chaperone in tubulin biogenesis. Nature, 358, 245–248. Yang, Y., Dowling, J., Yu, Q.-C., Kouklis, P., Cleveland, D. W., & Fuchs, E. (1996). An essential cytoskeletal linker protein connecting actin microfilaments to intermediate filaments. Cell, 86, 655–665. Yin, M., Wheeler, M. D., Connor, H. D., Zhong, Z., Bunzendahl, H., Dikalova, A., et al. (2001). Cu/Zn-superoxide dismutase gene attenuates ischemia-reperfusion injury in the rat kidney. Journal of the American Society of Nephrology, 12, 2691–2700. Yoshida, M., Muneyuki, E., & Hisabori, T. (2001). ATP synthase: A marvellous rotary engine of the cell. Nature Reviews Molecular Cell Biology, 2, 669–677. Yuan, X., & Desiderio, D. M. (2005). Proteomics analysis of prefractionated human lumbar ­cerebrospinal fluid. Proteomics, 5, 541–550. Yuan, X., Russell, T., Wood, G., & Desiderio, D. M. (2002). Analysis of the human lumbar ­cerebrospinal fluid proteome. Electrophoresis, 23, 1185–1196. Zappaterra, M. D., Lisgo, S. N., Lindsay, S., Gygi, S. P., Walsh, C. A., & Ballif, B. A. (2007). A comparative proteomic analysis of human and rat embryonic cerebrospinal fluid. Journal of Proteome Research, 6, 3537–3548.

536

G.K.H. Zupanc and M.M. Zupanc

Zupanc, G. K. H. (2006a). Neurogenesis and neuronal regeneration in the adult fish brain. Journal of Comparative Physiology A, 192, 649–670. Zupanc, G. K. H. (2006b). Adult neurogenesis and neuronal regeneration in the teleost fish brain: Implications for the evolution of a primitive vertebrate trait. In T. H. Bullock & L. R. Rubenstein (Eds.), The evolution of nervous systems in non-mammalian vertebrates (pp. 485–520). Oxford: Academic. Zupanc, G. K. H. (2008a). Adult neurogenesis in teleost fish. In F. H. Gage, G. Kempermann, & H. Song (Eds.), Adult neurogenesis (pp. 571–592). New York: Cold Spring Harbor Laboratory Press. Zupanc, G. K. H. (2008b). Adult neurogenesis and neuronal regeneration in the brain of teleost fish. Journal of Physiology (Paris), 102, 357–373. Zupanc, G. K. H., Clint, S. C., Takimoto, N., Hughes, A. T. L., Wellbrock, U. M., & Meissner, D. (2003). Spatio-temporal distribution of microglia/macrophages during regeneration in the cerebellum of adult teleost fish. Apteronotus leptorhynchus: A quantitative analysis. Brain Behavior and Evolution, 62, 31–42. Zupanc, G. K. H., Hinsch, K., & Gage, F. H. (2005). Proliferation, migration, neuronal differentiation, and long-term survival of new cells in the adult zebrafish brain. Journal of Comparative Neurology, 488, 290–319. Zupanc, G. K. H., & Horschke, I. (1995). Proliferation zones in the brain of adult gymnotiform fish: A quantitative mapping study. Journal of Comparative Neurology, 353, 213–233. Zupanc, G. K. H., Horschke, I., Ott, R., & Rascher, G. B. (1996). Postembryonic development of the cerebellum in gymnotiform fish. Journal of Comparative Neurology, 370, 443–464. Zupanc, G. K. H., Kompass, K. S., Horschke, I., Ott, R., & Schwarz, H. (1998). Apoptosis after injuries in the cerebellum of adult teleost fish. Experimental Neurology, 152, 221–230. Zupanc, G. K. H., & Ott, R. (1999). Cell proliferation after lesions in the cerebellum of adult teleost fish: Time course, origin, and type of new cells produced. Experimental Neurology, 160, 78–87. Zupanc, G. K. H., & Zupanc, M. M. (2006). New neurons for the injured brain: Mechanisms of neuronal regeneration in adult teleost fish. Regenerative Medicine, 1, 207–216. Zupanc, M. M., Wellbrock, U. M., & Zupanc, G. K. H. (2006). Proteome analysis identifies novel protein candidates involved in regeneration of the cerebellum of teleost fish. Proteomics, 6, 577–696.

MALDI Imaging of Formalin-Fixed Paraffin-Embedded Tissues: Application to Model Animals of Parkinson Disease for Biomarker Hunting Isabelle Fournier, Julien Franck, Céline Meriaux, and Michel Salzet

Abstract  After 10 years of important technical developments, MALDI Imaging Mass Spectrometry appears to be a sufficiently mature technology to be introduced in laboratories as a practical approach to exploring tissue properties at the molecular level, particularly in the comparison of normal vs. pathological states to study neurological diseases such as Parkinson’s. In this report, we will present the technology and the development to enable the use of formalin fixed and paraffin embedded tissue (FFPE) tissue in order to examine banked hospital clinical samples. Keywords  MALDI imaging • Pathologies • Developments • Proteomics

1 What is MALDI Imaging Mass Spectrometry Procedure? Mass spectrometry analyzes proteins based on the mass-to-charge (m/z) ratio of ionized peptides/proteins. A typical mass spectrometer consists of an ionization source, a mass analyzer, and a detector for counting the number of analytes at each m/z ratio. In 1994, Spengler was the first to describe the concept of ion imaging and confocal microscopy with a new Scanning UV-laser microprobe (Spengler, 1994), at the American Society of Mass Spectrometry (ASMS) meeting. Newly emerging mass spectrometry technologies have now demonstrated that direct tissue analysis is feasible using matrix-assisted laser desorption/ionization (MALDI) sources (Chaurand et al., 1999; Dreisewerd et al., 1997; Fournier et al., 2003; Jimenez et al., 1998; Li et  al., 1994). A major advantage of direct MALDI analysis is to avoid time-consuming extraction, purification or separation steps, which have the potential for producing artifacts. The studies referenced above and performed by different M. Salzet (*) Laboratoire de Neuroimmunologie et Neurochimie Evolutives, FRE CNRS 3249, MALDI Imaging Team, Université Lille Nord de France, Université de Lille1, Bâtiment SN3, 1er étage, 59655 Villeneuve d’Ascq Cedex, France e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_21, © Springer Science+Business Media, LLC 2011

537

538

I. Fournier et al.

groups, highlight that the acquisition of cellular expression profiles while maintaining the cellular and molecular integrity was feasible (Fournier et al., 2008) and that coupled with automation, the reconstruction of complex spectral data with imaging software, was possible (Stoeckli et al., 1999), producing multiplex imaging maps of selected biomolecules within tissue sections (Caprioli et al., 1997, Stoeckli et al., 2001, Stoeckli et  al., 1999). Thus, direct MALDI analysis obtained from tissue sections can be converted into imaging maps, a method now known as MALDIimaging. MALDI-imaging combines the power of mass spectrometry, namely exquisite sensitivity and unequivocal structural information, within an intact and unaltered morphological context. The procedure of MALDI-Imaging is described as follows (Fig. 1): tissue sections from the fresh organ or biopsy are laid out on conductive glasses (such as nickel or ITO), and treated with organic solvents (chloroform, acetone, xylene) to remove lipids (Lemaire, Wisztorski, et al., 2006). Tissue staining is performed as in classical histochemistry for cell classification (Chaurand et al., 2004). Several dyes have been tested and some are compatible with MALDI processes, e.g., coumarin dyes (Arafah et al., 2009) or cresyl violet (Chaurand et al., 2004). Sections are then covered with a solid ionic matrix (Lemaire, Tabet, et al., 2006) followed by gold sputtering in order to remove charge effects (Wisztorski et al., 2006) and are then introduced in the MALDI-TOF for analysis. Next, the MALDI laser is used to scan each point of the surface area and the mass spectra representative of the biomolecules present in each point are analyzed. Automated analysis of the complete tissue is obtained by performing mass spectra every 50–200 mm, providing representative information of selected ions (each ion is a specific bio-molecule). Each collected spectrum represents the average of several laser shots in order to obtain a statistical representation of the analyzed area. MALDI-MSI (mass spec imaging) in its most current form is a point-to-point analysis. Analysis is obtained within 2–6  h and images are reconstructed using imaging software (Jardin-Mathe et  al., 2008; Stoeckli et al., 1999). MSI offers the possibility to detect in a surface area of 100 mm on average at least several hundred to a thousand different ions (Stoeckli et al., 2001). Thanks to MALDI imaging software, it is possible to obtain for each identified ion, a molecular image of its representation in the tissue (Jardin-Mathe et  al., 2008; Stoeckli et  al., 1999). Ions can be lipids, peptides or proteins (Fournier et  al., 2008; McDonnell & Heeren, 2007; Woods & Jackson, 2006). The mass range offered by the technique is from 0 to 30,000 m/z (McDonnell & Heeren, 2007). However, some work in our laboratory has tended to push the limit to around 100,000 m/z with specific ionic matrices coupled to detergent and hydrophobic solvents (Bonnel et  al., 2008; Franck et al., 2010) and high mass detector (van Remoortere et al., 2010). Fractionation on tissue will open the door to get access to the deep proteome (Bonnel et al., 2008). This last result coincides with tissue chemical derivatization, also developed in the laboratory in order to get access to the primary peptide sequence directly in tissue without extraction (Franck, El Ayed, et al., 2009). De novo sequencing is a promising technique for determining both the localization of the biomolecules and

m/z1

m/z m/z2

m/z3

V. Data processing and image reconstruction

x3

x4

… … … … … … …

y1 I1;1 I1;2 I1;3 I1;4 y2 I2;1 I2;2 I2;3 I2;4 y3 I3;1 I3;2 I3;3 I3;4 y4 I4;1 I4;2 I4;3 I4;4 y5 I5;1 I5;2 I5;3 I5;4 y6 I6;1 I6;2 I6;3 I6;4 y7 I7;1 I7;2 I7;3 I7;4

…

…

…

…

…

…

…

…

x3

x2

1 I I I I x1 x2 x3 x4 …

x2

x1 …

x1

x4

IV. MALDI analysis

(Cryostat, Microtome)

I7;n

I6;n

I5;n

I4;n

I3;n

I2;n

I1;n

xn

…

…

7;n

6;n

5;n

4;n

3;n

2;n

I1;n

xn

…

7;n

6;n

5;n

4;n

3;n

2;n

1;n

xn

m/z1

m/z2

m/z3

III. Matrix deposition

II. Sectioning and transfer on sample plate

Fig. 1  Schematic representation of MALDI imaging procedures. Tissue sections from fresh organ or biopsy are transferred onto an ITO slide (indium tin oxide). Sections are then covered with a specific matrix, depending on the nature of the bio-molecule under study. For peptides/proteins, very intense signals are obtained with a-cyano-4-hydroxycinnamic acid (a-CHCA) as a matrix. Once the sections are covered with a-CHCA (or another matrix), the target is introduced in the MALDI-TOF instrument for analysis. Next, the MALDI laser scans through a set of pre-selected locations (10–50 mm apart) in the tissue and mass spectra representative of the peptides/proteins (or lipids) present at each location are collected. Automated data collection takes 2–6 h, depending on the number of points assayed, and followed by analysis and image reconstruction using Flex-imaging software

Intensity (a.u.)

I. Sacrifice and organ dissection

MALDI Imaging of Formalin-Fixed Paraffin-Embedded Tissues 539

540

I. Fournier et al.

their primary sequence as well as identifying the molecules surrounding them. Acquiring the sequence of the molecule and its tissue localization can provide useful information on its putative roles. A new concept of using tagged molecules for performing multiplex studies was introduced by Tag-Mass development (Lemaire, Stauber, et  al., 2007). Tag-Mass uses labeled probes for specific identification in mass spectrometry. The concept consists to an oligonucleotide sequence (or probe) that hybridizes mRNA in a tissue section similar to in situ hybridization techniques. In Tag-Mass, the oligonucleotide probe sequence is attached to a photo-cleavable group linked up to a “Tag” marker, which is an amino acid sequence with a defined mass. This “Tag” marker can be modified to generate different known masses. Following hybridization of the TagMass probe to the tissue section, MALDI analysis is performed, however, the pulse laser cleaves the photo-cleavable group to release the Tag. The signals obtained for specific biomolecules in Tag-Mass will be much higher and will yield unique signatures. As a proof-of-concept, a modified uracil base bearing the photo-cleavable linker was developed, allowing multiplex in situ hybridization using MALDI technology. A photo-cleavable linker can also be used in other applications including antibodies, lectins or aptamers for use in tagged-specific MALDI-imaging. Thus, we can anticipate targeting specific disease-marker-gene RNA transcripts, follow their expression within tissues and then confirming their translation by targeting their specific protein products or metabolites. Disease/health states will thus be precisely and rapidly monitored at a multi-molecular level.

2 Application of MALDI-MSI to Brain Analyses Because its anatomy has been extensively characterized, the rat brain was the first biological model used in MALDI MSI studies, and several molecular maps of different biomolecules have been reported (Fournier et al., 2003, 2008; Jespersen et al., 1999; Wisztorski et al., 2008). These maps accurately match the corresponding distributions obtained using immunocytochemistry, but are not restricted to only those molecules for which antibodies are available. In fact, more than a thousand molecules can be detected in a single study. Beside peptides and proteins, molecular distributions of different classes of lipids (Murphy et al., 2009) or drugs (Reyzer & Caprioli, 2007; Rubakhin et al., 2005) have been mapped using MALDI MSI. MALDI MSI has also been used in differential display studies of animals injected with lipopolysaccharides (LPS) in order to mimic bacterial challenges. The presence of vasopressin in the supraoptic nucleus before LPS injection, and its decrease afterwards, confirm the feasibility of accessing such information by MALDI MS profiling (Fournier et  al., 2003). However, the first MALDI tissue profiling studies have also been performed on invertebrate nervous systems. For example, neuronal somata are often quite large (up to a millimeter in diameter in the mollusc Aplysia), and neurons can be reliably identified and repeatedly studied

MALDI Imaging of Formalin-Fixed Paraffin-Embedded Tissues

541

in different experimental contexts, neural functions are frequently segregated in separate ganglia or regions of ganglia, and neural tissues can be readily and quickly prepared for study in a MALDI Imaging instrument (Furukawa et al., 2001; Garden et al., 1998; Kruse et al., 2001; Li et al. 1999; Li et al., 1998; Li et al., 2000; Painter et al., 1998; Rubakhin et al., 2000; Rukakhin et al., 1999; Schein et  al., 2001; Sweedler et al., 2000; Sweedler et al., 2002). MSI studies complement more traditional studies of peptides and proteins purified from selected regions of the invertebrate nervous system (Dani et al., 2008; DeKeyser et al., 2007; Hofer et al., 2005; Hummon et al., 2006), and can be carried out in developing stages to obtain 4D (spatial + temporal) maps of specific molecules of interest. This level of analysis can be extended to studies of degeneration, regeneration and repair (Kruse & Sweedler, 2003; Rubakhin et al., 2003). All these extensive studies have been performed on frozen tissue.

3 How to Apply MALDI-MSI to Archived Tissue in FFPE The use of archived material in paraffin blocks from hospital pathology departments represents a “gold mine” of existing information. However, the major technical hurdle is cross-linking due to formalin fixation, and embedding in paraffin (FFPE tissue). Formalin fixation provokes the formation of protein–nucleic acid and protein–protein cross-linking in the intracellular environment arising from the reactivity of formaldehyde with the side-chains of lysyl, argininyl, tyrosyl, aspartyl, histidyl, and seryl residues (Shi, et al., 1995). Two procedures have been developed taking into account the age of the tissue. In the case of tissue, stored for less than 6 months, an active matrix, namely 2,4-dinitrophenylhydrazine (DNPH) is used. DNPH neutralizes formalin excess and allows the analysis of embedded tissues such as those from paraffin sections (Lemaire, Desmons, et al., 2006). In the case of tissue stored for more than 6 months, micro-digestion with trypsin must be performed. In this context, two adjacent slices are subjected to trypsin digestion using an automatic spotter (Franck, Arafah, et  al., 2009). One slice is subjected to a molecular MSI image and the other one used for a classical proteomic “bottom-up” strategy (Lemaire, Desmons, et al., 2007). After tryspin digestion, C18 beads are deposited on top of the slices. The beads are then collected from the tissue slices and all the digested peptides attached to the C18 beads are then removed from the beads using an extraction buffer prior to nanoLC-IT-MS in MS/MS mode analyses for shotgun protein characterization (Lemaire, Desmons, et al., 2007). The ions of the protein found to change between normal or pathologic samples are then investigated on the adjacent slices in order to create an image of their localization (Stauber et al., 2008). All ions deriving from the same protein after digestion can then be localized and submitted to MS/MS analysis after N-terminal derivatization (Franck, El Ayed, et al., 2009) (Fig. 2). Besides, these pioneer developments other

542

I. Fournier et al.

groups have improved the technology by adding additional steps to the procedure, such as the addition of a heat induced antigen retrieval (HIAR) step in the presence of EDTA (Ronci et al., 2008), or the addition of detergent in the digestion protocol to increase the yield of tryptic peptides (Djidja, Francese, et al., 2009). Instrumental development such as the addition of ion mobility separation in order to improve the selectivity and specificity of the method, have also been recently reported (Djidja, Claude, et  al., 2009). These developments have been applied to high-throughput proteomic analysis of FFPE tissue microarrays (TMA) using on-tissue tryptic digestion followed by MSI (Groseclose et al., 2008). TMA sections containing 112-needle core biopsies from lung-tumor patients were analyzed using MS and the data were correlated to a serial hematoxylin and eosin (H&E)-stained section with various histological regions marked, including cancer, non-cancer, and normal ones. By correlating each mass spectrum to a defined histological region, statistical classification models were generated that can sufficiently distinguish biopsies from adenocarcinoma from those from squamous cell carcinoma. The ability to detect and characterize tumor marker proteins for a large cohort of FFPE samples in a high-throughput approach will be of significant benefit to clinicians for diagnostic and prognostic purposes (Groseclose et al., 2008).

4 Application of MALDI-MSI to Neurodegenerative Diseases: Parkinson’s Disease Neurodegeneration induces various changes in the brain, changes that may be investigated using neuroimaging techniques. For example, Parkinson’s disease (PD) is the second most common neurological disease after Alzheimer’s disease. PD is characterized by a selective degeneration of dopaminergic neurons in the substantia nigra pars compacta and by cytoplasmic inclusions (Lewy Bodies) where specific proteins are stored such as a-synuclein (Beal & Hantraye, 2001). The clinical symptoms are severe motor dysfunctions, including rigidity, postural imbalance, slowness of movements, and uncontrollable tremor. Mutations in genes encoding a-synuclein, parkin and ubiquitin carboxy-terminal hydrolase L1 (UCH L1) have been identified in sporadic familial forms of PD (Giasson & Lee, 2003). Several proteomic studies using 2D gel analyses in animal models and humans, have identified biomarkers implicated in this pathology (Basso et al., 2003, 2004; De Iuliis et al., 2005; Palacino et al., 2004; Strey et al., 2004) (Table 1). In Parkin knockdown models, pyruvate dehydrogenase; NADH ubiquinone oxyreductase 30 kDa, cytochrome c oxydase, peroxiredoxin 1, 2 and 6, lactoylglutathione lyase, vacuolar protein sorting 29, crystalline chain b and heterogeneous nuclear ribonucleoprotein 1 are all down-regulated (Palacino et al., 2004). Similarly, in 6-OHDA animal models (Ungerstedt, et al., 1974), as well as in humans, a-enolase, a-actin, Lasp-1, neurofilament triplet L and M are also down-regulated (Basso et al., 2003, 2004; De Iuliis et al., 2005; Palacino et al., 2004). In contrast, human peroxiredoxin 2, complexin I, fatty acid binding protein, L type calcium chanel d subunit,

200

y1

y2

400

y3

y4

600

b3

y5 y6

800 mz

y7

y8 y10

b8 y11

1000 1200 1400

y9

Online Nano-LC nanoESI/IT MS/MS

2 mm

hν

Protein ID

MALDI images of identified proteins

Enzymatic in situ micro-digestion

x104 8

0

2

4

6

5%

200

a1 y1 b2

400

y4

600 800 m/z

b5-17 a6-17 y8 z7

y10

a1

1000 1200

b8

?

m/z 728

2 mm 0%

On tissue MALDI-TOF/TOF MS/MS

m/z 643

80 % 80 %

MALDI Imaging

MALDI images acquisition on the micro-spots

matrix in situ Micro-spotting

Fig. 2  Strategies for identifying and imaging proteins from preserved tissue. Schematic diagram of protocols used for applying MS imaging and identifying proteins in FFPE (formalin-fixed and paraffin-embedded) tissues sections, which was successfully carried out independent of storage length. Trypsin microdigestion are performed before matrix micro-spotting. Half part of the brain is transferred to an eppendorf and subjected to enzymatic digestion. Digest is analysed by nanoLC-nanoESI/IT and MS/MS analyses. The other half is subjected to a MALDI imaging analyses. Ions detected from the nanoLC ESI/IT analyses are then imaged by MALDI. With permission from Journal of Proteome Research (Lemaire, Desmons, et al. 2007)

0.0

0.2

0.4

0.6

0.8

1.0

x105

Online Nano-LC nanoESI/IT MS

Pre-purification

Intens ty (a u )

Transfer

Buffer extraction from de whole section

Tissue Section global enzymatic digestion

Intensity (a u )

FFPE tissue section after paraffin removal

MALDI Imaging of Formalin-Fixed Paraffin-Embedded Tissues 543

Table 1  Specific 6-OHDA biomarkers found after trypsic digestion by comparing 6-OHDA FFPE tissues to control FFPE tissues of rat brains after nanoLCnanoESI-IT MS followed by MALDI direct analyses of digested tissues (in sequence box, matched peptides are indicated by a bar and are underlined in red) Digestion peptide Name & Sequence mass (Da) coverage Peptides match [M+H]+ m/z Sequences ESTLHLVLR Ubiquitin 43.4% 1067.62 TLSDYNIQK M = 8559.62 1081.56 ESTLHLVLRLR 1336.67 IQDKEGIPPDQQR 1523.84 MSILK a-Enolase 16.8% 591.08 EIFDSR M = 47098.24 766.46 IAKAAGEK 786.47 SFRNPLAK 932.53 TGAPCRSER 976.49 IGAEVYHNLK 1143.64 IDQLMIEMDGTENK 1636.85 EIFDSRGNPTVEVDLYTAK 2153.16 PEP 19 37.1% 1049.53 TSGDNDGQKK M = 6803.29 1336.67 MSERQSAGATNGK ITSDR CRMP2 15.0% 591.08 IVAPPGGR M = 62238.59 766.46 MSYQGKK 841.52 SRLAELR 844.48 SAAEVIAQARK 1143.64 VMSKSAAEVIAQAR 1460.79 NLHQSGFSLSGAQIDDNIPR 2169.17 GLYDGPVCEVSVTPKTVTPASSAK 2406.33 Peroxidoxin 11.1% 788.47 SVDEALR M = 21770.06 1636.85 GVLRQITVNDLPVGR ESVHGGLINK 40.7% 1052.62 Profilin CYEMASHLR M = 14977.52 1109.70 MAGWNAYIDSLMADGTCQ 3959.72 DAAIVGYKDSPSVWAAVPGK

544 I. Fournier et al.

HSP27 M = 22807.57

20.5%

26.0% Ferritin heavy chain M = 20982.27 Apoliporotein E 25.3% M = 35731.33

591.080 932.532 988.549 1358.716 1481.856

726.23 888.44 968.53 1162.14 1183.80 1238.70 1505.86 1523.84

1481.8 4105.41

SIKELGDHVTNLR QNYHQDSEAAINRQINLELYAS YVYLSMSCYFDR LSTHLR GVSAIRER LGPLVEQGR MEEQTQQIR LAKEVQAAQAR TANLGAGAAQPLRGWF EPLVEDMQR GWFEPLVEDMQR QRTANLGAGAAQPLR QTADR QTADRWR RVPFSLLR AQIGGPESEQSGAK EGVVEITGKHEER

MALDI Imaging of Formalin-Fixed Paraffin-Embedded Tissues 545

546

I. Fournier et al.

mitochondrial complex III and ATP synthase D chain are up-regulated (Basso et  al., 2004). These results have been confirmed by Strey et  al. (2004) in SOD1 gene studies, where a-enolase as well as HSP25, HSP27, phosphatidyl insitol transfer protein, apoliprotein E and ferritin heavy chain are up-regulated. These contradictory studies for peroxidoxin 2 and profilin show the difficulty in getting real biomarkers from these classical techniques. The first tissue profiling studies on the 6-OHDA Parkinson model have been performed by Per Andrén’s group (Pierson et al., 2004). Several differences were found in the dopamine-depleted side of the rat brain when compared to the corresponding intact side, in calmodulin, cytochrome c, and cytochrome c oxidase, for example, implicating denervation of dopamine neurons per se in the regulation of ubiquitin pathways, at least in a classical animal model of PD (Pierson et al., 2005). This study also emphasizes the utility of molecular profiling with MSI, because it has the capacity to distinguish between metabolic fragments, conjugated proteins, and posttranslational modifications (Pierson et al., 2005). Most antibodies used in immunocytochemical studies for assessing the presence of ubiquitin in PD-associated Lewy bodies are directed towards a protein-bound form of ubiquitin. Free monomeric ubiquitin is not immunogenic in most mammalian species used to produce antibodies (Langstrom et al., 2007; Pierson et al., 2005). These antibodies are also capable of cross-reacting with free monomeric ubiquitin, and therefore it is unclear which form of ubiquitin is detected using these immunocytochemical techniques. MSI, by comparison, easily discriminates between these two forms (Langstrom et al., 2007). Recently, we examined MALDI tissue profiling combining the use of automatic spotting of MALDI matrix with in situ tissue enzymatic digestion of FFPE brain of 6-OHDA unilaterally treated animals (Lemaire, Desmons, et al., 2006). 6-OHDA is known to inhibit the mitochondrial transport chain and this, along with the resulting production of reactive oxygen species, contributes to neuronal death (Grunblatt, Mandel, & Youdim, 2000). This agent appears to induce neuronal death by activating transcription as well as DNA repair enzymes (Herceg & Wang, 2001). It also induced a large number of genes involved in endoplasmic reticulum stress and unfolded protein response (UPR) such as ER chaperones and elements of the ubiquitin-proteasome system (Daniel et al., 2000). Identification procedures performed on the compounds obtained from tissue slices after in situ tissue digestion followed by nanoLC/MS–MS analysis confirmed the transcriptomic data (Table 1). These analyses confirmed that ubiquitin, transelongation factor 1, hexokinase and Neurofilament M are down-regulated, as previously shown in both human and animal model tissues, whereas peroxidoredoxin 6, F1 ATPase and a-enolase symbol are up-regulated (Wisztorski, Lemaire, Stauber, Ait Menguellet, et al., 2007; Wisztorski, Lemaire, Stauber, Menguelet, et al., 2007) (Figs. 3 and 4) which is consistent with previous studies performed with classical proteomic or DNA microarrays approaches (Table 2). Neurofilament M protein has been described as down-regulated (Basso et  al., 2004; Liu et al., 2004) and its down-regulation is correlated with decreased mRNA during the disease and is dependent upon disease severity (Liu et  al., 2004).

MALDI Imaging of Formalin-Fixed Paraffin-Embedded Tissues

547

Fig. 3  Molecular image reconstructed with MITICS on ubiquitin trypsic specific peptides (1,063, 1,527 and image based on the sum of all tryspin digested peptide). Insert pictures correspond to scan of the tissue slide before MALDI analyses, expression of ubiquitin mRNA in mouse brain (data from Allen Brain Atlas http://www.brain-map.org) and map of the corresponding tissue section (Bregma Index). With permission from Journal of Proteomic (Jardin-Mathe et al., 2008)

Fig.  4  Comparison of the images based on apolipoprotein E trypsin (ApoE) specific peptides (m/z 726 and 1,523) using either the peak area or the maximum peak intensity with a threshold of 20 and a margin of error of 2 Da. Chromatic scale is also presented (Jardin-Mathe et al., 2008)

548

I. Fournier et al.

Table 2  Comparison of Parkinson biomarkers identified using different proteomics approaches on different models Proteomic Model approach Protein Modification References Parkin −/− 2D gel Pyruvate deshydrogenase Down-regulated Pierson et al. (2005) NADH ubiquinone Up-regulated oxyreductase 24 kDa NADH ubiquinone Down-regulated oxyreductase 30 kDa Cytochrome c oxydase Down-regulated Peroxiredoxin 1 Down-regulated Peroxiredoxin 2 Down-regulated Peroxiredoxin 6 Down-regulated Lactoylglutathione lyase Down-regulated Profilin Down-regulated Vacuolor protein sorting 29 Down-regulated a-Crystalin chain b Down-regulated Heterogeneous nuclear Down-regulated ribonucleoprotein 1 Lasp-1 Down-regulated a-Enolase Up-regulated Pierson et al. (2004) b Actin Down-regulated 6-OHDA MDAa Calmodulin Down-regulated Rubakhin et al. (2003) Cytochrome C Down-regulated Cytochrome C oxydase Down-regulated Ubiquitin Up-regulated Human Blood Serum creatine kinase Up-regulated 2D gel Neurofilament triplet L Down-regulated Reyzer & Caprioli Neurofilament triplet M Down-regulated (2007) Peroxiredoxin 2 Up-regulated

MTP G93 A (gene SOD1)

MDA 2D Gel

Mitochondrial complex III ATP synthase D chain Profilin L type calcium chanel d subunit Fatty acid binding protein PEP 19 a-Enolase HSP 25 HSP 27 Phosphatidinylinositol transfer protein a Apolipoprotein E Ferritin heavy chain

Up-regulated Up-regulated Up-regulated Up-regulated Up-regulated Up-regulated Down-regulated Up-regulated Up-regulated Up-regulated

Ronci et al. (2008)

Up-regulated Up-regulated

MDA MALDI direct analysis MS

a

Our  results using direct analysis of FFPE tissues support these findings (Fig.  6). Due to 6-OHDA treatment provoking oxidative stress, several molecules were found to be up-regulated under these conditions, such as Peroxydoxin 6, known as

MALDI Imaging of Formalin-Fixed Paraffin-Embedded Tissues

549

an anti-oxidative protein. Similar results were found by Strey et  al. (2004) using classical proteomic studies. In contrast, peroxydoxin 2 is described as being downregulated in the parkin knockout model but is upregulated in human Parkinson’s disease studies. These data reaffirm the need to consider multiplex biomarkers for pathology. a-Enolase was also increased in our study as previously observed elsewhere in PD and AD (De Iuliis et al., 2005; Strey et al., 2004). This molecule is known to be the target of specific oxidation or nitrozation (De Iuliis et al., 2005). We also observed up-regulation of F1 ATPase as previously demonstrated by Seo et al. (2004). The ubiquitin complex was observed to be upregulated in our study on FFPE tissue as shown by Pierson et al. (2004). It is known that high levels of ubiquitin and ubiquinated proteins are present in Lewy bodies indicating that protein degradation is impaired in PD (Giasson & Lee, 2003). Proteins conjugated with a chain of ubiquitin moieties are targeted to the ubiquitin–proteasome system complex, where they undergo proteolytic degradation. Genetic studies of PD have identified mutations in the genes coding for proteins involved in the ubiquitin– proteasome degradation pathway (Leroy et al., 1998). In addition, we identified three novel putative biomarkers, Trans elongation factor 1 (eEF1) and the Collapsin response mediator proteins, CRMP-1 and -2, using protein libraries (Wisztorski, Lemaire, Stauber, Ait Menguellet, et  al., 2007; Wisztorski, Lemaire, Stauber, Menguelet, et al., 2007). We focused our attention on the CRMP family. Several digestion fragments of the CRMP1 and CRMP2 have been found by nanoESI, MALDI on tissue and databank interrogation using the MS/MS experiments confirm the identification of such proteins. All the data combined identified this up-regulated protein as the CRMP2 splice variant B of the CRMP2 protein (Rattus norvegicus, P47492). Two splices variants have been recently found (Yuasa-Kawada et al., 2003). The CRMP2A is the long N-terminal isoform (75 kDa) and induces oriented microtubule patterns in cultured fibroblasts, a pattern also observed in axons. Conversely, CRMP2B, the shortest variant (64 kDa), induces disorientation of microtubule patterns in cultured fibroblasts and reduces axon length when over-expressed in retinal explants (Yuasa-Kawada et al., 2003). In adult brain, it is known that expression of CRMPs is dramatically down-regulated (Ricard et  al., 2000). In our study, an increase of CRMP2B in 6-OHDA-treated animals has been observed, which is in agreement with previous molecular data (Barzilai et al., 2000). In order to identify the CRMP2 protein and its localization in 6-OHDA-treated animals, MALDI molecular images were obtained after reconstruction of the different ions corresponding to different digestion fragments of the protein (Fig.  5). All these ions globally present an equivalent localization in the rat brain section, although some of them present more contrasted images. The last image is a composite image of all digested fragments of CRMP2 protein (Fig.  5). The most striking feature is the localization of the protein to very specific regions of the brain with a highly contrasted signal. This is especially true for the corpus callosum where ions are always found to be very intense. Figure 6 shows the map of the corresponding section and compares the MALDI images from the 6-OHDA treated animals to the expression of the CRMP2 mRNA in normal adult mouse. As found in the literature for the rat,

550

I. Fournier et al.

Fig.  5  Detection of CRMP-2 protein with antibodies or MALDI imaging (MALDI molecular images reconstructed from the data collected after acquisition on a 6-OHDA treated rat brain section originally conserved after formalin fixation and paraffin embedding over 9 years after in situ automatic trypsin digestion). (a) Image of ion at m/z 643, (b) m/z 728, (c) 1,083 and (d) composite image of all ions corresponding to digestion fragments of CRMP2 protein and detected in the MALDI experiment. With permission from Journal of Proteome Research (Stauber et al., 2008)

this variant of CRMP2 is normally not present in the corpus callosum. It is located predominantly in dendrites of specific neuronal populations, such as cortical pyramidal neurons, hippocampal CA1 pyramidal cells, or Purkinje cerebellar cells (Bretin et  al., 2005). Thus localization in the corpus callosum is in line with an involvement in neurodegenerative diseases. In fact, the corpus callosum is known to be a brain area implicated in dementia in several neurodegenerative diseases (Charrier et al., 2003). In PD, mRNAs encoding CRMP (intracellular protein mediating Semaphorin3A) and the mitochondrial stress protein HSP60 are known to be up-regulated. It has been hypothesized that these proteins are positive mediators of Dopamine (DA)induced neuronal apoptosis in PD. In PD, nigral neuronal death could be due to excessive oxidative stress generated by auto- and enzymatic oxidation of DA, the formation of neuromelanin and the presence of high concentrations of iron. The hippocampus of patients with AD also express phosphorylated MAP1B, collapsin-response mediator protein 2 (CRMP-2), Plexins A1 and A2, and a processed form of Sema3A (Good et al., 2004). In Alzheimer’s disease, CRMP-2 is known to be implicated in neurite degeneration, acting on the assembly and polymerization of microtubules (Gu et al., 2000). Accumulation of Sema3A overlaps with the appearance of phosphorylated MAP1B and tau in many neurons, suggesting that Sema3A signaling at some level may be coupled to these previously identified cytoskeletal markers of neurodegeneration (Good et  al., 2004). Taking into account all the data, we speculate that CRMP factors are good

MALDI Imaging of Formalin-Fixed Paraffin-Embedded Tissues

551

Relative Intensity (%)

100

Σm/z (m/z 643 + 728 + 813 + 1083+ 1460)

m/z 728

80 %

5%

10 %

80

6-OHDA treated 1770.8

60 40 20

0

80 %

2 mm

Controls

1750

1770

M/z

1790

1810

2 mm

Fig. 6  (a) composite MALDI molecular image reconstructed using all detected digestion fragments of CRMP2 protein on a 6-OHDA treated FFPE tissue rat brain section after paraffin removal and in situ trypsin digestion. (b) Optical image of the tissue section after automatic spotting of trypsin and solid ionic matrix HCCA/ANI, (c) Superposition of six mass spectra obtained by MALDI direct analysis in the same region of FFPE tissue sections after in situ trypsin digestion for three 6-OHDA treated animals vs. three control animals. Arrows indicate ion at m/z 1770.3 identified as a digestion fragment of neurofilament triplet M protein which is a potential biomarker (d) MALDI molecular images reconstructed from the data collected after acquisition on a 6-OHDA-treated rat brain section originally conserved after formalin fixation and paraffin embedding over 9 years after in situ automatic trypsin digestion. Image of ion at m/z 728. With permission from Journal of Proteome Research (Stauber et al., 2008)

biomarkers for neurodegenerative diseases like PD or AD, and that MALDI MSi is a well-suited technology to investigate proteome changes during the course of neurologic patholologies. Acknowledgments  Supported by grants from Centre National de la Recherche Scientifique (CNRS), Ministère de L’Education Nationale, de L’Enseignement Supérieur et de la Recherche, Agence Nationale de la Recherche (ANR to I.F.).

References Arafah, K., Wisztorski, M., Croix, D., Fournier, I., & Salzet, M. (2009). Dye-Assisted Laser Desorption Ionisation (DALDI): Introduction to a new field of investigation in lipid mass spectrometry imaging. Proceedings of the 18th International Mass Spectrometry Conference, Bremen, Germany. Barzilai, A., Zilkha-Falb, R., Daily, D., Stern, N., Offen, D., Ziv, I., et al. (2000). The molecular mechanism of dopamine-induced apoptosis: Identification and characterization of genes that mediate dopamine toxicity. Journal of Neural Transmission. Supplementum, 60, 59–76. Basso, M., Giraudo, S., Corpillo, D., Bergamasco, B., Lopiano, L., & Fasano, M. (2004). Proteome analysis of human substantia nigra in Parkinson’s disease. Proteomics, 4(12), 3943–3952.

552

I. Fournier et al.

Basso, M., Giraudo, S., Lopiano, L., Bergamasco, B., Bosticco, E., Cinquepalmi, A., et al. (2003). Proteome analysis of mesencephalic tissues: Evidence for Parkinson’s disease. Neurological Sciences, 24(3), 155–156. Beal, M. F., & Hantraye, P. (2001). Novel therapies in the search for a cure for Huntington’s disease. Proceedings of the National Academy of Sciences of the United States of America, 98(1), 3–4. Bonnel, D., Franck, J., Wisztorski, M., El ayed, M., Salzet, M., & Fournier, I. (2008). On tissue fractionation: Unmasking difficult proteins for MALDI profiling. 56th Annual Conference on Mass Spectrometry and Allied Topics, Denver, Colorado. Bretin, S., Reibel, S., Charrier, E., Maus-Moatti, M., Auvergnon, N., Thevenoux, A., et al. (2005). Differential expression of CRMP1, CRMP2A, CRMP2B, and CRMP5 in axons or dendrites of distinct neurons in the mouse brain. The Journal of Comparative Neurology, 486(1), 1–17. Caprioli, R. M., Farmer, T. B., & Gile, J. (1997). Molecular imaging of biological samples: Localization of peptides and proteins using MALDI-TOF MS. Analytical Chemistry, 69(23), 4751–4760. Charrier, E., Reibel, S., Rogemond, V., Aguera, M., Thomasset, N., & Honnorat, J. (2003). Collapsin response mediator proteins (CRMPs): Involvement in nervous system development and adult neurodegenerative disorders. Molecular Neurobiology, 28(1), 51–64. Chaurand, P., Schwartz, S. A., Billheimer, D., Xu, B. J., Crecelius, A., & Caprioli, R. M. (2004). Integrating histology and imaging mass spectrometry. Analytical Chemistry, 76(4), 1145–1155. Chaurand, P., Stoeckli, M., & Caprioli, R. M. (1999). Direct profiling of proteins in biological tissue sections by MALDI mass spectrometry. Analytical Chemistry, 71(23), 5263–5270. Dani, F. R., Francese, S., Mastrobuoni, G., Felicioli, A., Caputo, B., Simard, F., et  al. (2008). Exploring proteins in Anopheles gambiae male and female antennae through MALDI mass spectrometry profiling. PLoS ONE, 3(7), 2822. Daniel, R., He, Z., Carmichael, K. P., Halper, J., & Bateman, A. (2000). Cellular localization of gene expression for progranulin. The Journal of Histochemistry and Cytochemistry, 48(7), 999–1009. De Iuliis, A., Grigoletto, J., Recchia, A., Giusti, P., & Arslan, P. (2005). A proteomic approach in the study of an animal model of Parkinson’s disease. Clinica Chimica Acta, 357(2), 202–209. DeKeyser, S. S., Kutz-Naber, K. K., Schmidt, J. J., Barrett-Wilt, G. A., & Li, L. (2007). Imaging mass spectrometry of neuropeptides in decapod crustacean neuronal tissues. Journal of Proteome Research, 6(5), 1782–1791. Djidja, M. C., Claude, E., Snel, M., Scriven, P., Francese, S., Carolan, V. A., et  al. (2009). MALDI-ion mobility separation-mass spectrometry imaging of glucose-regulated protein 78 kDa (Grp78) in human formalin fixed paraffin embedded pancreatic adenocarcinoma tissue sections. Journal of Proteome Research, 8, 4876–4884. Djidja, M. C., Francese, S., Loadman, P. M., Sutton, C. W., Scriven, P., Claude, E., et al. (2009). Detergent addition to tryptic digests and ion mobility separation prior to MS/MS improves peptide yield and protein identification for in situ proteomic investigation of frozen and formalin-fixed paraffin-embedded adenocarcinoma tissue sections. Proteomics, 9(10), 2750–2763. Dreisewerd, K., Kingston, R., Geraerts, W. P., & Li, K. W. (1997). Direct mass spectrometric peptide profiling and sequencing of nervous tissues to identify peptides involved in male copulatory behavior in Lymnaea stagnalis. International Journal of Mass Spectrometry, 169, 291–299. Fournier, I., Day, R., & Salzet, M. (2003). Direct analysis of neuropeptides by in situ MALDITOF mass spectrometry in the rat brain. Neuro Endocrinology Letters, 24(1–2), 9–14. Fournier, I., Wisztorski, M., & Salzet, M. (2008). Tissue imaging using MALDI-MS: A new frontier of histopathology proteomics. Expert Review of Proteomics, 5(3), 413–424. Franck, J., Arafah, K., Barnes, A., Wisztorski, M., Salzet, M., & Fournier, I. (2009). Improving tissue preparation for matrix-assisted laser desorption ionization mass spectrometry imaging. Part 1: Using microspotting. Analytical Chemistry, 81(19), 8193–8202. Franck, J., El Ayed, M., Wisztorski, M., Salzet, M., & Fournier, I. (2009). On-tissue N-terminal peptide derivatizations for enhancing protein identification in MALDI mass spectrometric imaging strategies. Analytical Chemistry, 81(20), 8305–8317.

MALDI Imaging of Formalin-Fixed Paraffin-Embedded Tissues

553

Franck, J., Longuespée, R., Wisztorski, M., van Remoortere, A., van Zeijl, R., Deelder, A., et al. (2010). MALDI mass spectrometry imaging of proteins exceeding 30,000 daltons. Medical Science Monitor, 16(9), BR293–BR299. Furukawa, Y., Nakamaru, K., Wakayama, H., Fujisawa, Y., Minakata, H., Ohta, S., et al. (2001). The enterins: A novel family of neuropeptides isolated from the enteric nervous system and CNS of Aplysia. The Journal of Neuroscience, 21(20), 8247–8261. Garden, R. W., Shippy, S. A., Li, L., Moroz, T. P., & Sweedler, J. V. (1998). Proteolytic processing of the Aplysia egg-laying hormone prohormone. Proceedings of the National Academy of Sciences of the United States of America, 95(7), 3972–3977. Giasson, B. I., & Lee, V. M. (2003). Are ubiquitination pathways central to Parkinson’s disease? Cell, 114(1), 1–8. Good, P. F., Alapat, D., Hsu, A., Chu, C., Perl, D., Wen, X., et al. (2004). A role for semaphorin 3A signaling in the degeneration of hippocampal neurons during Alzheimer’s disease. Journal of Neurochemistry, 91(3), 716–736. Groseclose, M. R., Massion, P. P., Chaurand, P., & Caprioli, R. M. (2008). High-throughput proteomic analysis of formalin-fixed paraffin-embedded tissue microarrays using MALDI imaging mass spectrometry. Proteomics, 8(18), 3715–3724. Grunblatt, E., Mandel, S., & Youdim, M. B. (2000). MPTP and 6-hydroxydopamine-induced neurodegeneration as models for Parkinson’s disease: Neuroprotective strategies. Journal of Neurology, 247(suppl 2), II95–II102. Gu, Y., Hamajima, N., & Ihara, Y. (2000). Neurofibrillary tangle-associated collapsin response mediator protein-2 (CRMP-2) is highly phosphorylated on Thr-509, Ser-518, and Ser-522. Biochemistry, 39(15), 4267–4275. Herceg, Z., & Wang, Z. Q. (2001). Functions of poly(ADP-ribose) polymerase (PARP) in DNA repair, genomic integrity and cell death. Mutation Research, 477(1–2), 97–110. Hofer, S., Dircksen, H., Tollback, P., & Homberg, U. (2005). Novel insect orcokinins: Characterization and neuronal distribution in the brains of selected dicondylian insects. The Journal of Comparative Neurology, 490(1), 57–71. Hummon, A. B., Amare, A., & Sweedler, J. V. (2006). Discovering new invertebrate neuropeptides using mass spectrometry. Mass Spectrometry Reviews, 25(1), 77–98. Jardin-Mathe, O., Bonnel, D., Franck, J., Wisztorski, M., Macagno, E., Fournier, I., et al. (2008). MITICS (MALDI Imaging Team Imaging Computing System): A new open source mass spectrometry imaging software. Journal of Proteomics, 71(3), 332–345. Jespersen, S., Chaurand, P., van Strien, F. J., Spengler, B., & van der Greef, J. (1999). Direct sequencing of neuropeptides in biological tissue by MALDI-PSD mass spectrometry. Analytical Chemistry, 71(3), 660–666. Jimenez, C. R., Li, K. W., Dreisewerd, K., Spijker, S., Kingston, R., Bateman, R. H., et al. (1998). Direct mass spectrometric peptide profiling and sequencing of single neurons reveals differential peptide patterns in a small neuronal network. Biochemistry, 37(7), 2070–2076. Kruse, R. A., Rubakhin, S. S., Romanova, E. V., Bohn, P. W., & Sweedler, J. V. (2001). Direct assay of Aplysia tissues and cells with laser desorption/ionization mass spectrometry on porous silicon. Journal of Mass Spectrometry, 36(12), 1317–1322. Kruse, R., & Sweedler, J. V. (2003). Spatial profiling invertebrate ganglia using MALDI MS. Journal of the American Society for Mass Spectrometry, 14(7), 752–759. Langstrom, B., Andren, P. E., Lindhe, O., Svedberg, M., & Hall, H. (2007). In vitro imaging techniques in neurodegenerative diseases. Molecular Imaging and Biology, 9(4), 161–175. Lemaire, R., Desmons, A., Ducoroy, P., Tabet, J. C., Salzet, M., & Fournier, I. (2006) Direct analysis and MALDI imaging on formalin fixed paraffin embedded tissue (FFPE): Application to Parkinson disease. Proceedings of the 54th American Society for Mass Spectrometry Conference, Seattle, Washington. Lemaire, R., Desmons, A., Tabet, J. C., Day, R., Salzet, M., & Fournier, I. (2007). Direct analysis and MALDI imaging of formalin-fixed, paraffin-embedded tissue sections. Journal of Proteome Research, 6(4), 1295–1305.

554

I. Fournier et al.

Lemaire, R., Stauber, J., Wisztorski, M., Van Camp, C., Desmons, A., Deschamps, M., et al. (2007). Tag-mass: Specific molecular imaging of transcriptome and proteome by mass spectrometry based on photocleavable tag. Journal of Proteome Research, 6(6), 2057–2067. Lemaire, R., Tabet, J. C., Ducoroy, P., Hendra, J. B., Salzet, M., & Fournier, I. (2006). Solid ionic matrixes for direct tissue analysis and MALDI imaging. Analytical Chemistry, 78(3), 809–819. Lemaire, R., Wisztorski, M., Desmons, A., Tabet, J. C., Day, R., Salzet, M., et  al. (2006). MALDI-MS direct tissue analysis of proteins: Improving signal sensitivity using organic treatments. Analytical Chemistry, 78(20), 7145–7153. Leroy, E., Boyer, R., Auburger, G., Leube, B., Ulm, G., Mezey, E., et al. (1998). The ubiquitin pathway in Parkinson’s disease. Nature, 395(6701), 451–452. Li, L., Garden, R. W., Romanova, E. V., & Sweedler, J. V. (1999). In situ sequencing of peptides from biological tissues and single cells using MALDI-PSD/CID analysis. Analytical Chemistry, 71(24), 5451–5458. Li, K. W., Hoek, R. M., Smith, F., Jimenez, C. R., van der Schors, R. C., van Veelen, P. A., et al. (1994). Direct peptide profiling by mass spectrometry of single identified neurons reveals complex neuropeptide-processing pattern. The Journal of Biological Chemistry, 269(48), 30288–30292. Li, L., Moroz, T. P., Garden, R. W., Floyd, P. D., Weiss, K. R., & Sweedler, J. V. (1998). Mass spectrometric survey of interganglionically transported peptides in Aplysia. Peptides, 19(8), 1425–1433. Li, L., Romanova, E. V., Rubakhin, S. S., Alexeeva, V., Weiss, K. R., Vilim, F. S., et al. (2000). Peptide profiling of cells with multiple gene products: Combining immunochemistry and MALDI mass spectrometry with on-plate microextraction. Analytical Chemistry, 72(16), 3867–3874. Liu, Q., Xie, F., Siedlak, S. L., Nunomura, A., Honda, K., Moreira, P. I., et  al. (2004). Neurofilament proteins in neurodegenerative diseases. Cellular and Molecular Life Sciences, 61(24), 3057–3075. McDonnell, L. A., & Heeren, R. M. (2007). Imaging mass spectrometry. Mass Spectrometry Reviews, 26(4), 606–643. Murphy, R. C., Hankin, J. A., & Barkley, R. M. (2009). Imaging of lipid species by MALDI mass spectrometry. Journal of Lipid Research, 50(suppl), S317–S322. Painter, S. D., Clough, B., Garden, R. W., Sweedler, J. V., & Nagle, G. T. (1998). Characterization of Aplysia attractin, the first water-borne peptide pheromone in invertebrates. The Biological Bulletin, 194(2), 120–131. Palacino, J. J., Sagi, D., Goldberg, M. S., Krauss, S., Motz, C., Wacker, M., et  al. (2004). Mitochondrial dysfunction and oxidative damage in parkin-deficient mice. The Journal of Biological Chemistry, 279(18), 18614–18622. Pierson, J., Norris, J. L., Aerni, H. R., Svenningsson, P., Caprioli, R. M., & Andren, P. E. (2004). Molecular profiling of experimental Parkinson’s disease: Direct analysis of peptides and proteins on brain tissue sections by MALDI mass spectrometry. Journal of Proteome Research, 3(2), 289–295. Pierson, J., Svenningsson, P., Caprioli, R. M., & Andren, P. E. (2005). Increased levels of ubiquitin in the 6-OHDA-lesioned striatum of rats. Journal of Proteome Research, 4(2), 223–226. Reyzer, M. L., & Caprioli, R. M. (2007). MALDI-MS-based imaging of small molecules and proteins in tissues. Current Opinion in Chemical Biology, 11(1), 29–35. Ricard, D., Stankoff, B., Bagnard, D., Aguera, M., Rogemond, V., Antoine, J. C., et al. (2000). Differential expression of collapsin response mediator proteins (CRMP/ULIP) in subsets of oligodendrocytes in the postnatal rodent brain. Molecular and Cellular Neurosciences, 16(4), 324–337. Ronci, M., Bonanno, E., Colantoni, A., Pieroni, L., Di Ilio, C., Spagnoli, L. G., et  al. (2008). Protein unlocking procedures of formalin-fixed paraffin-embedded tissues: Application to MALDI-TOF imaging MS investigations. Proteomics, 8(18), 3702–3714.

MALDI Imaging of Formalin-Fixed Paraffin-Embedded Tissues

555

Rubakhin, S. S., Garden, R. W., Fuller, R. R., & Sweedler, J. V. (2000). Measuring the peptides in individual organelles with mass spectrometry. Nature Biotechnology, 18(2), 172–175. Rubakhin, S. S., Greenough, W. T., & Sweedler, J. V. (2003). Spatial profiling with MALDI MS: Distribution of neuropeptides within single neurons. Analytical Chemistry, 75(20), 5374–5380. Rubakhin, S. S., Jurchen, J. C., Monroe, E. B., & Sweedler, J. V. (2005). Imaging mass spectrometry: Fundamentals and applications to drug discovery. Drug Discovery Today, 10(12), 823–837. Rubakhin, S. S., Li, L., Moroz, T. P., & Sweedler, J. V. (1999). Characterization of the Aplysia californica cerebral ganglion F cluster. Journal of Neurophysiology, 81(3), 1251–1260. Schein, C. H., Nagle, G. T., Page, J. S., Sweedler, J. V., Xu, Y., Painter, S. D., et al. (2001). Aplysia attractin: Biophysical characterization and modeling of a water-borne pheromone. Biophysical Journal, 81(1), 463–472. Seo, B. B., Nakamaru-Ogiso, E., Cruz, P., Flotte, T. R., Yagi, T., & Matsuno-Yagi, A. (2004). Functional expression of the single subunit NADH dehydrogenase in mitochondria in  vivo: A potential therapy for complex I deficiencies. Human Gene Therapy, 15(9), 887–895. Shi, S. R., Imam, S. A., Young, L., Cote, R. J., & Taylor, C. R. (1995). Antigen retrieval immunohistochemistry under the influence of pH using monoclonal antibodies. The Journal of Histochemistry and Cytochemistry, 43(2), 193–201. Spengler, B. (1994). Ion imaging and confocal microscopy with a new scanning UV-laser microprobe. In 42nd Annual Conference on Mass Spectrometry and Allied Topics, Chicago, IL. Stauber, J., Lemaire, R., Franck, J., Bonnel, D., Croix, D., Day, R., et al. (2008). MALDI imaging of formalin-fixed paraffin-embedded tissues: Application to model animals of Parkinson disease for biomarker hunting. Journal of Proteome Research, 7(3), 969–978. Stoeckli, M., Chaurand, P., Hallahan, D. E., & Caprioli, R. M. (2001). Imaging mass spectrometry: A new technology for the analysis of protein expression in mammalian tissues. Natural Medicines, 7(4), 493–496. Stoeckli, M., Farmer, T. B., & Caprioli, R. M. (1999). Automated mass spectrometry imaging with a matrix-assisted laser desorption ionization time-of-flight instrument. Journal of the American Society for Mass Spectrometry, 10(1), 67–71. Strey, C. W., Spellman, D., Stieber, A., Gonatas, J. O., Wang, X., Lambris, J. D., et al. (2004). Dysregulation of stathmin, a microtubule-destabilizing protein, and up-regulation of Hsp25, Hsp27, and the antioxidant peroxiredoxin 6 in a mouse model of familial amyotrophic lateral sclerosis. The American Journal of Pathology, 165(5), 1701–1718. Sweedler, J. V., Li, L., Floyd, P., & Gilly, W. (2000). Mass spectrometric survey of peptides in cephalopods with an emphasis on the FMRFamide-related peptides. The Journal of Experimental Biology, 203(Pt 23), 3565–3573. Sweedler, J. V., Li, L., Rubakhin, S. S., Alexeeva, V., Dembrow, N. C., Dowling, O., et al. (2002). Identification and characterization of the feeding circuit-activating peptides, a novel neuropeptide family of aplysia. The Journal of Neuroscience, 22(17), 7797–7808. Ungerstedt, U., Ljungberg, T., & Steg, G. (1974). Behavioral, physiological, and neurochemical changes after 6-hydroxydopamine-induced degeneration of the nigro-striatal dopamine neurons. Advances in Neurology, 5, 421–426. van Remoortere, A., van Zeijl, R. J., van den Oever, N., Franck, J., Longuespee, R., Wisztorski, M., et al. (2010). MALDI imaging and profiling MS of higher mass proteins from tissue. Journal of American Society for Mass Spectrometry. Wisztorski, M., Brunet, L., Dreisewerd, K., Hillenkamp, F., Berkenkamp, S., Salzet, M., et al. (2006). Effect of metals coating for UV MALDI-a-TOF mass spectrometry imaging (MALDI MSI) and direct tissue analysis in UV/IR MALDI-o-TOF mass spectrometry. Proceedings of the 54th American Society for Mass Spectrometry Conference, Seattle, Washington. Wisztorski, M., Croix, D., Macagno, E., Fournier, I., & Salzet, M. (2008). Molecular MALDI imaging: An emerging technology for neuroscience studies. Developmental Neurobiology, 68(6), 845–858.

556

I. Fournier et al.

Wisztorski, M., Lemaire, R., Stauber, J., Ait Menguellet, S., Jardin-Mathe, O., Day, R., et  al. (2007). MALDI imaging: A new technology to discover and validate new biomarkers. Medical Science (Paris), 23(Spec no. 1), 31–36. Wisztorski, M., Lemaire, R., Stauber, J., Menguelet, S. A., Croix, D., Mathe, O. J., et al. (2007). New developments in MALDI imaging for pathology proteomic studies. Current Pharmaceutical Design, 13(32), 3317–3324. Woods, A. S., & Jackson, S. N. (2006). Brain tissue lipidomics: Direct probing using matrixassisted laser desorption/ionization mass spectrometry. The AAPS Journal, 8(2), E391–E395. Yuasa-Kawada, J., Suzuki, R., Kano, F., Ohkawara, T., Murata, M., & Noda, M. (2003). Axonal morphogenesis controlled by antagonistic roles of two CRMP subtypes in microtubule organization. The European Journal of Neuroscience, 17(11), 2329–2343.

Comparative Proteomic Analysis as a Method to Investigate Inflammatory and Neuropathic Pain Ellen Niederberger

Abstract  Patients suffering from pathological pain due to neuropathy or chronic inflammation are dependent on treatment with efficient and highly specific drugs. Although a number of analgesics are available many patients cannot be adequately treated because of lacking efficacy or severe side effects. The pain hypersensitivity in inflammatory and neuropathic pain is associated with changes of protein expression in the CNS which are not completely clarified at the moment. To develop novel treatment strategies it is necessary to elucidate the mechanisms of both pain states, and to find proteins that are specifically regulated in either neuropathic or inflammatory pain and that may become drug targets. Proteomics by 2D-gel electrophoresis combined with MALDI-TOF mass spectrometry might help identifying regulated proteins in the nervous system in inflammatory or neuropathic pain models and therefore facilitate the development of novel analgesics. In these chapter a number of proteomic approaches in the field of inflammatory pain and nerve injury are reviewed which might provide starting points for further research in this field. Keywords  Proteomics • Inflammation • Neuropathy • Spinal cord Abbreviations CCI IEF IPG MALDI-TOF MS

chronic constriction injury isoelectric focusing immobilized pH gradient matrix-assisted laser desorption ionization time-of-flight mass spectrometry

E. Niederberger (*) Pharmazentrum frankfurt/ZAFES, Institut für Klinische Pharmakologie, Klinikum der Johann Wolfgang Goethe-Universität Frankfurt, Theodor Stern Kai 7, 60590 Frankfurt am Main, Germany e-mail: [email protected] J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5_22, © Springer Science+Business Media, LLC 2011

557

558

MW NC SNI SNL

E. Niederberger

molecular weight nerve crush spared nerve injury spinal nerve ligation

1 Introduction Chronic pain can be classified according to its triggering mechanisms and its drug treatment strategies as either nociceptive or neuropathic pain, associated with peripheral tissue damage and inflammation or nerve injury, respectively. This pathological pain is characterized by persistent pain hypersensitivity which is believed to be mediated by sensitization of nociceptors and spinal dorsal horn neurons leading to hyperalgesia and allodynia. Changes of protein expression and/or phosphorylation are known to contribute to the development of this hyperexcitability of the nociceptive system (Dubner & Ruda, 1992; Ji & Woolf, 2001; Woolf & Costigan, 1999; Yaksh, Hua, Kalcheva, Nozaki-Taguchi, & Marsala, 1999). Despite the large number of available analgesics such as opioids or nonsteroidal anti-inflammatory drugs (NSAIDs), treatment of chronic pain is still often hampered by the occurrence of adverse drug reactions and/or poor activity of available drugs (Kingery, 1997; Koltzenburg, 1998). Both the targets of NSAIDs (cyclooxygenases) and the targets of opioids (opioid receptors), are among the proteins that are known to be regulated in the spinal cord in nociceptive models – ­cyclooxygenase-2 is increased in inflammatory models and opioid receptors are diminished in neuropathic pain models (Niederberger et  al., 2001; Rashid, Inoue, Toda, & Ueda, 2004; Tegeder et  al., 2004) – suggesting that regulated proteins play an important role in pain control. Proteomics by 2D gel electrophoresis combined with MALDI-TOF mass spectrometry can be applied to investigate differences in the protein expression pattern and to identify regulated proteins in the nervous system in inflammatory or neuropathic pain models. This proteomic approaches may therefore reveal potential novel targets for analgesics.

2 Nociceptive Pain Nociceptive pain arises from mechanical, chemical or thermal irritation of peripheral sensory nerves, particularly following inflammation. Signals of primary afferent fibers are transmitted to secondary sensory neurons in the dorsal horn of the lumbar spinal cord (Yaksh et al., 1999). Persistent afferent input caused, e.g., by peripheral inflammation or incomplete nerve lesions may lead to an increased excitability of nociceptive neurons in the spinal cord. This central sensitization relies on the flexibility of the nociceptive system. It allows for an adaptive facilitation of synaptic transmission and thereby development of hyperalgesia and allodynia. Various intracellular signaling pathways that are linked to various kinases

Comparative Proteomic Analysis as a Method to Investigate Inflammatory and Neuropathic Pain

559

such as protein kinase A, C, and G are involved in this process. The fast and transient phosphorylation-dependent mechanisms are followed by changes of gene transcription due to transcription factor activation or inhibition. The signal cascade results in immediate early gene (IEG) induction, which is considered as the starting event of a widespread change in protein synthesis. The alteration of protein expression is thought to cause the transition from short-term adaptive processes to longterm hyperexcitability of nociceptive neurons that probably contributes to the development of chronic pain (Beiche, Scheuerer, Brune, Geisslinger, & GoppeltStruebe, 1996; Dubner & Ruda, 1992; Ji et al., 1994, 1995; Mannion et al., 1999; McCarson & Krause, 1994; Woolf & Costigan, 1999).

2.1 Animal Models of Nociceptive Pain A large number of nociceptive behavioral models in animals have been described (Dubner, 1994; Dubner & Ren, 1999; Le Bars, Gozariu, & Cadden, 2001; Tjølsen & Hole, 1997). These models differ in the nociceptive stimuli (e.g., ­thermal, mechanical, chemical), the form and the duration of nociceptive response, and the analysis of the nociceptive response. In this chapter, only the zymosaninduced paw inflammation model is described in more detail since proteomic analysis has only been investigated after application of this model.

2.1.1 Zymosan-Induced Paw Inflammation Unilateral hind paw inflammation and hyperalgesia is induced by subcutaneous injection of 1.25 mg zymosan suspended in 100 ml phosphate buffered saline into the midplantar region of the right hind paw (Meller & Gebhart, 1997). This treatment results in an immediate inflammation leading to paw edema and hyperalgesic reactions to thermal or mechanical stimulation.

3 Neuropathic Pain Neuropathic pain is defined as maladaptive pain which results from damage to the nervous system. A wide variety of insults to peripheral or central nerves including metabolic disorders, traumatic injury, inflammation and neurotoxicity can initiate neuropathic pain which is characterized by spontaneous pain, hyperalgesia and allodynia. These disturbances often persist long after the initial injury is resolved (Woolf & Mannion, 1999). Common causes of neuropathy are diabetes, herpes zoster infection, chronic trauma (such as repetitive motion disorders) or acute

560

E. Niederberger

trauma (including surgery), and various neurotoxins. Neuropathic pain is frequent in cancer as a direct result of compression of peripheral nerves by a tumor or as a side effect of many chemotherapy drugs. The underlying mechanisms are still not completely understood, and as a consequence, treatment is often disappointing (Chen et al., 2004; Hansson & Dickenson, 2005; Sindrup & Jensen, 1999). Pharmacological treatment includes tricyclic antidepressants such as amitriptyline, anticonvulsants such as gabapentin and pregabalin, and serotonin norepinephrine reuptake inhibitors (SSNRI) such as duloxetine. However, the pain relief mechanisms of these drugs are not completely understood. Neuropathic pain reflects both peripheral and central sensitization mechanisms; that is, transcriptional and post-transcriptional modifications are occurring in sensory nerves (Campbell & Meyer, 2006; Woolf & Mannion, 1999). Proteomic investigation of these changes might help to find new proteins which can be used as drug targets and may improve the treatment conditions for patients suffering from ­neuropathic pain.

3.1 Animal Models of Neuropathic Pain Most neuropathic pain models are based on injury of the spinal cord or the peripheral­ nerves (Wang & Wang, 2003). According to the models applied in the proteomic studies summarized in this chapter, only a short description of these respective models is given. 3.1.1 Central Models Spinal cord injury can be performed by weight drop (Anderson, 1982; Greenberg, McKeever, & Balentine, 1978), spinal cord compression (Tarlov, 1972), crushing (Rivlin & Tator, 1978), photochemically induced injury, excitatory neurotoxin methods, and spinal hemisection (Bennett, Chastain, & Hulsebosch, 2000; Christensen, Everhart, Pickelman, & Hulsebosch, 1996; Christensen & Hulsebosch, 1997). These models lead to spontaneous and evoked pain as well as allodynia and hyperalgesia. Weight Drop Model In this model, spinal cord injury is produced by dropping a weight on the exposed spinal cord surface at the thoracic-lumbar level which is simulating the human clinical condition of traumatic injury of the spinal cord. This treatment leads to paraplegia and complete segmental necrosis (Anderson, 1982; Greenberg et al., 1978).

Comparative Proteomic Analysis as a Method to Investigate Inflammatory and Neuropathic Pain

561

Spinal Cord Hemisection In the case of spinal cord hemisection, the spinal cord is only partially dissected at the Th13 level thus leading to an immediate flaccid paralysis of the ipsilateral hindpaw which is recovered 15 days after injury. Concurrently, noxious stimulation of this paw induces signs of hyperalgesia and allodynia (Bennett et  al., 2000; Christensen & Hulsebosch, 1997; Christensen et al., 1996). 3.1.2 Peripheral Models Common models of peripheral neuropathic pain include spinal nerve ligation (SNL), chronic constriction injury (CCI), nerve crush (NC) and partial nerve ligation (PNL) (Fig. 1).

Dorsal Root Ganglion

L4 L5

Sciatic Nerve

Spinal nerve ligation (SNL)

L6

Partial nerve injury (PNL) Chronic constriction injury (CCI)

Fig. 1  Three different nerve injury models are shown. In the spinal nerve ligation (SNL) model, one or more spinal nerves going to the foot are ligated and cut (Kim & Chung, 1992). In the partial nerve ligation (PNL) model, a portion of the sciatic nerve is tightly ligated (Seltzer, Dubner, & Shir, 1990). The CCI model involves placement of four loose suture ligatures on the sciatic nerve. An immune response to the sutures leads to nerve swelling and nerve constriction (Bennett & Xie, 1988; George, Marziniak, Schafers, Toyka, & Sommer, 2000). In each model, only a portion of the afferents going to the foot are lesioned (adapted from (Campbell & Meyer, 2006)). Reproduced from Expert Review of Proteomics 5(6), 799–818 (2008) with permission of Expert Reviews Ltd

562

E. Niederberger

Spinal Nerve Ligation This experimental model of mononeuropathy exists in several different forms. Kim and Chung reported the first model with L5 and L6 unilateral SNL (Choi, Yoon, Na, Kim, & Chung, 1994; Kim & Chung, 1992). In this model, hyperalgesia and allodynia develop quickly and last for at least 4 months without signs of autotomy (self-mutilation of the ipsilateral paw). A variant of this model is the ligation of only the L5 spinal nerve (Kim & Chung, 1992). L5-ligated rats also exhibit longlasting allodynia and hyperalgesia, but the method is easier to perform in comparison with combined L4/L5-ligation. Partial Nerve Ligation To achieve partial nerve injury, about half the sciatic nerve high in the thigh is unilaterally ligated. Within a few hours after the operation, and for several months thereafter, the rats develop spontaneous pain characterized by guarding behavior of the ipsilateral hind paw and licking. No autotomy has been observed (Seltzer et al., 1990). Chronic Constriction Injury of the Sciatic Nerve This model has been reported as painful peripheral mononeuropathy where the sciatic nerve is constricted on the left or right side, with loose ligatures at the midthigh level (Bennett & Xie, 1988; George et al., 2000). CCI rats show behavioral signs of spontaneous pain, as well as hyperalgesia, due to noxious thermal and mechanical stimuli. Furthermore, they develop cold and tactile allodynia. All pain signs last over a period of at least 2 months.

4 Proteomics The proteome quantitatively acquires the composition of proteins in an organism derived from an individual’s full genetic information. Most physiological body functions are dependent on the integrity of these proteins. Therefore, a number of pharmacologically active drugs are focused on proteins because of their pathophysiological relevance. Proteomic-based approaches are designed to investigate the proteome of a cell, a tissue, or an organism. Proteomics is complementary to genomics which examine DNA and RNA. In contrast to genomics, proteomics deliver information about protein isoforms, post-translational protein modifications such as glycosylations and phosphorylations, protein–protein interactions, protein stability and degradation. While the genome of an organism is rather static, the proteome is dynamic, differs strongly from cell to cell, and is constantly modulated through biochemical interactions with the genome and the environment. In different parts of a body, the

Comparative Proteomic Analysis as a Method to Investigate Inflammatory and Neuropathic Pain

563

protein expression is very variable and depends on a number of parameters such as age, different environmental conditions or diseases. Regarding the essential role of proteins in the life of an organism, proteomics might be a useful instrument for the discovery of biomarkers, which indicate a particular disease. Proteomics can be utilized to generate protein maps of certain tissues (profiling proteomics), to study pathophysiology by investigation of aberrant proteins (functional proteomics) and to correlate nucleic acid levels with proteins (reviewed in Choudhary & Grant, 2004; Gorg, Weiss, & Dunn, 2004).

4.1 2D PAGE Two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) combined with protein identification by mass spectrometry is currently the main utilized technique in proteomics. 2D-PAGE consists of isoelectric focusing (IEF) in the first dimension and sodium dodecyl sulfate gel electrophoresis in the second dimension, and it allows the separation and visualization of complex protein mixtures according to their isoelectric point (pI), molecular weight, solubility and relative abundance (O’Farrell, 1975). Depending on the pH gradient and the gel size, 2D-PAGE allows the separation of thousands of protein spots in a gel which can be visualized by staining with a variety of chemical dyes or fluorescent markers. To quantify protein levels, the intensity of the protein stain is measured and quantitatively analyzed. Regulated protein spots are subsequently separated and identification is generally performed by mass spectrometry, specifically matrix-assisted laser desorption-ionization time-of-flight (­MALDI-TOF) mass spectrometry (Matsumoto & Komori, 1999). In the first dimension of 2D-PAGE which is an IEF, proteins are separated on the basis of their net charge (pI). Therefore, they are loaded onto gel strips with an immobilized pH gradient (IPG) and then separated by high-voltage IEF (Gorg et al., 2000). Proteins migrate along this gradient until their overall charge is zero (neutral). After IEF, the IPG strips are placed onto the top of a sodium dodecyl sulfate (SDS)-polyacrylamide gel and a standard electrophoresis is performed. Proteins on the gels are separated according to their molecular weight, subsequently stained with Coomassie Brilliant Blue, silver or fluorescent dyes, and then analyzed for differences in protein expression between two different samples. The comparison of protein spots between different 2D gels includes detection of proteins, matching of proteins in the respective gels, as well as quantification of matching proteins. The differences can be compared manually; however, regarding the large number spot on the gels, this can be difficult and might decrease the accuracy and thereby the quality of the results. A number of image analysis software have been developed which include differential integration gel electrophoresis. By subtraction processes, differences between protein spots (e.g., intensity, size, absence) of healthy and diseased tissues can be visualized. Depending on the gel size and pH gradient, 2D PAGE typically allows for the separation of hundreds to thousands (typically

564

E. Niederberger

1,000–3,000) protein spots onto one gel and the detection and quantification of <1 ng protein per spot (Gorg et al., 2004). Thus, it enables the separation of complex mixtures of proteins according to pI, Mr, solubility, and relative abundance.

4.2 Identification of 2D-Separated Proteins Spots of interest are cut out of the gels, destained and dried in a speed vac. The dried gel pieces are rehydrated in an appropriate digestion-buffer containing trypsin and digested overnight at 37°C. The fragmented peptides are extracted from the gel with 50% acetonitrile/0.5% trifluoro-acetic-acid (TFA). The proteins are identified by matrix-assisted laser desorption ionization timeof-flight mass spectrometry (MALDI-TOF MS) on the basis of peptide mass matching (Burre et  al., 2006; Rosenfeld, Capdevielle, Guillemot, & Ferrara, 1992).

5 Protein Expression Following Peripheral Inflammation Although inflammatory pain is a field of intensive research, to date there has been only one published study from our group, which investigated the differential regulation of protein expression in the CNS after peripheral inflammatory stimulation. We applied the model of zymosan-induced paw inflammation as described above. Injection of the inflammatory stimulus zymosan resulted in a massive edema and hyperalgesia in the inflamed paw. Then, 96 h after zymosan injection animals were killed, the lumbar spinal cord was rapidly dissected and protein extracts were prepared. 2D-PAGE allowed for a separation of more than 500 protein spots on each gel with an apparent range of molecular masses from 10 to 100 kDa and pI values from 3 to 10. Quantitative analysis of the spots revealed that 9 protein spots were significantly altered 96 h after the zymosan injection. These spots were identified using MALDI-TOF mass spectrometry. Seven proteins were down- and 2 proteins were up-regulated in the zymosan-treated group as compared to control animals (Table 1) (Kunz et al., 2005). Among the down-regulated proteins during zymosan-induced paw inflammation was neurofilament light (NF-L). In another study, we confirmed that NF-Ldegradation in the spinal cord is a consequence of peripheral inflammatory stimulation and that the NF-L regulation is mediated, at least in part, by the action of the protease calpain (Kunz et al., 2004). Two neighboring spots in the 2D gels with the same molecular weight but different isoelectric points were identified as aldose reductase (AR) indicating an phosphorylation event. AR has previously been suggested to be involved in the pathogenesis of secondary complications associated with diabetes, e.g., retinopathy, nephropathy and neuropathy (Yabe-Nishimura, 1998). Furthermore, it has been

Comparative Proteomic Analysis as a Method to Investigate Inflammatory and Neuropathic Pain

565

Table 1  Regulated proteins in rat spinal cord 96 h after induction of a chronic paw inflammation (Kunz et al., 2005) Protein identification Expression General function pI MW Regulated protein number folda Aldose reductase GI 6978491 Aldo-keto reductase 6.3 35.8 0.13 Aldose reductase GI 6978491 Aldo-keto reductase 6.1 35.8 3.44 ATP synthase D GI 9506411 Proton-transport in the 6.2 18.8 0.36 respiratory chain a-B-Crystallin GI 117388 Small heat shock protein 6.8 20.1 0.21 Cytochromoxidase GI 117100 Electron-transport 6.1 16.1 0.33 Va in the respiratory chain Diazepam binding GI 13937379 Endogenous ligand 8.8 10.0 3.01 inhibitor for GABA-receptor Neurofilament light GI 13929098 Intermediate filament 4.6 61.3 0.12 (NF-L) b-Synuclein GI 18373326 Presynapticphos­ 4.5 14.5 0.15 phoneuroprotein 5.8 17.3 0.48 Stathmin GI 8393696 Microtubule regulatory phosphoprotein a Indication of the relative protein expression level in the zymosan-treated group compared to untreated control animals

implicated with inflammatory responses in various cell culture models and in mice (Ramana, Bhatnagar, & Srivastava, 2004, Ramana et  al. 2006a, 2006b). Phosphorylation of AR by PKC has already been described (Varma et al., 2003) but the physiological function of this phosphorylation remained unclear.

6 Protein Expression Following Neuropathic Pain A number of studies have investigated the regulation of the protein expression pattern in the nervous system in several different models of neuropathy. Kang et al. studied the differential regulation of proteins after traumatic injury of the spinal cord (SCI) (Kang, So, Moon, & Kim, 2006). Among 947 protein spots on the 2D gel, they found over 39 up-regulated and 29 down-regulated proteins 24  h after spinal cord injury. The down-regulated proteins in injured spinal cord tissue included septin 2, chaperonin-containing TCP1, pyruvate dehydrogenase beta, Phgdh protein, programmed cell death 6 interacting protein, dynamin1, Fscn 1 protein, reticulocalbin 2, and neuronal differentiation-related gene. In the case of up-regulated proteins in injured spinal cord tissue, proteins such as those similar to ATPase H1transportingV1 subunitA, RhoGDI-1, peroxiredoxin 2, pyruvate kinase 3, apolipoprotein A-I precursor, annexin 5, GFAP delta, tubulin alpha 1, and neurofilament 3 were identified.

566

E. Niederberger

In the same study, additional immunohistochemical analysis confirmed the regulation of a number of neural lineage proteins as well as apoptotic signaling proteins, following damage to the spinal cord. These results indicated that secondary events after SCI not only cause apoptotic cell death but also help restore chronic function by increasing the local levels of growth factors that stimulate the migration, proliferation, gliogenesis, and neurogenesis of endogenous neural progenitor cells in spinal cord. In a model of L4/L5 nerve ligation, five proteins with different expression levels after nerve injury were identified (Lee et al., 2003). The proteins guanine nucleotide-binding protein G(0) alpha subunit1, proteasome component C8, L-lactate dehydrogenase H chain were up-regulated, while creatine kinase B was down-regulated. Since creatine has been found to reduce glutamate levels and to exhibit neuroprotective properties (Klivenyi et  al., 1999; Sullivan, Geiger, Mattson, & Scheff, 2000; Xu et al., 1996), the authors concluded that the downregulation of creatine kinase B might be particularly important for the development and maintenance of neuropathic pain and therefore a valuable therapeutic target for neuropathic pain. After applying the same neuropathy model, another group investigated the differential protein expression in the brainstem (Alzate et al., 2004). They found regulation of 21 proteins 7 days after nerve ligation. For 14 proteins, the levels were higher in rats with neuropathic pain in comparison to sham-operated rats, while 7 showed a reduced expression. Interestingly, none of the proteins found to be regulated in the spinal cord were also regulated in the brainstem and vice versa, indicating that the nociceptive transmission involves different proteins in the different tissues. A third study investigated protein regulation in the L4 and L5 dorsal root ganglia after L5 SNL (Komori et al., 2007). The results revealed regulations of over 60 proteins which are involved in a number of neuronal cell functions. The protein expression profile of the rat sciatic nerve has been investigated in a model of experimental NC. The authors found at least 121 regulated proteins with respect to the time points investigated. From these proteins, 82 could be identified (Jimenez et al., 2005). In a model of partial nerve injury, Katano et al. reported the exclusive occurrence of 12 proteins in the spinal nerves peripheral to the DRG and 3 in the central region of this nerve. The peripherally expressed proteins included collagen a1, a-tubulin and collapsin response mediator 2, and the centrally expressed proteins tubulin b3 and b15 (Katano et al., 2006). Our group investigated the protein expression in the lumbar spinal cord of rats after applying the CCI model. We detected an average of 500 protein spots on the 2D gels. 14 days after induction of the CCI of the sciatic nerve, five protein spots were significantly regulated and subsequently identified by MALDI-TOF mass spectrometry. Protein disulfide-isomerase A3 precursor was decreased while protein disulfide isomerase (PDI) was increased. Creatine kinase MM, ubiquinolcytochrome c reductase iron-sulfur subunit precursor and a-B-crystallin were also decreased (Kunz et al., 2005).

Comparative Proteomic Analysis as a Method to Investigate Inflammatory and Neuropathic Pain

567

7 Conclusion Patients suffering from pathological pain due to neuropathy or chronic inflammation are dependent on treatment with efficient and highly specific drugs. Although a number of analgesics are available, many patients cannot be adequately treated because of lack of efficacy or severe side effects. Therefore, it is necessary to elucidate the mechanisms of both pain states, and to find proteins that are specifically regulated in either neuropathic or inflammatory pain and that may become drug targets (Scholz & Woolf, 2002). Despite the importance of achieving new data about signaling pathways in pain, there are only a few studies which deal with the regulation of protein expression in the central nervous system after nociceptive stimulation. However, studies have been designed to investigate the differences in protein expression in the nervous system following neuropathic pain, and one study investigated protein regulation in the spinal cord after peripheral inflammation. The proteins identified as potential regulators of pain can be roughly subdivided in fundamental categories such as heat shock proteins/chaperones and antioxidants, neuronal function proteins, proteins related to cellular homeostasis and metabolism, proteins related to the immune system, signaling proteins, proteins related to cell cycle/apoptosis and neurodegeneration, and proteins related to protein synthesis and processing (Table 2). Some of the proteins have already been related to pain, but a number of regulated proteins have not been previously implicated in this context and may therefore provide interesting new fields of pain research. Except for the small heat shock protein (Hsp) a-crystallin, the changes in protein expression were specific for the neuropathic or inflammatory model, respectively. However, a-crystallin protein expression was decreased in the ­spinal cord in both models. Alpha-crystallin is a molecular chaperone that ­stabilizes partially or totally unfolded proteins. Failure of correct protein folding occurs particularly under endoplasmatic reticulum (ER) stress conditions (MacRae, 2000). In a model of exerciseinduced muscle pain, alpha-crystallin levels in the muscle were increased (Feasson et  al., 2002). These data are consistent with the observations made in both pain models described here, and might indicate that a-crystallin is unspecifically upregulated after a number of painful conditions. Neurofilament light (NF-L) protein was also regulated in both models; however, not in the same direction. In the paw inflammation model, it was down-regulated while up-regulation was observed in experimental neuropathy. NF-L is an important protein for axonal architecture and transport. We have previously shown that NF-L-degradation after peripheral inflammatory stimulation is mediated at least in part by the action of the protease calpain (Kunz et al., 2004). In addition, it has been shown that mutations in the NF-L gene are involved in autosomal dominant neuropathies (Fabrizi et al., 2006). The observed differences in protein regulation between the inflammatory and neuropathic models exemplify the differential physiology of pain responses in the

568

E. Niederberger

spinal cord during inflammatory stimulation and nerve injury. This might contribute to the different therapeutic efficacy of analgesics in the treatment of inflammatory and neuropathic pain. Interestingly, in the models of neuropathic pain, among over 150 regulated proteins found, there were only about 30 overlapping protein regulations. This might indicate that the different models induce differential protein regulation and that these regulations are again variably pronounced in different neuronal tissues. Another explanation might be that the different approaches to investigate pain are very heterogeneous. The different animal strains and nociceptive models used, the differential samples preparations, various time points for tissue dissection, differing conditions for IEF (pH-range, separation protocol, etc.), and SDS-PAGE might contribute to the wide variety of regulated proteins. The proteins that were regulated in more than one neuropathy model are indicated in Table 2. The role of several of these proteins in pain models is discussed below. PDI acts either as a redox catalyst or as a molecular chaperone that prevents protein aggregation and degradation (Gruber, Cemazar, Heras, Martin, & Craik, 2006). In the nervous system, up-regulation of PDI has been shown as a result of hypoxia and brain ischemia, thus protecting cells from apoptosis (Tanaka, Uehara, & Nomura, 2000). Concerning neuropathic pain, this might also constitute a protective mechanism against apoptotic cell death induction. Hsp 27 is a small Hsp. The Hsps are stress proteins that mediate protein stabilization in various tissues and protect cells from environmental stress. A number of Hsps are up-regulated in the nervous system in response to stress or injury (Sharma, Gordh, Wiklund, Mohanty, & Sjoquist, 2006; Yenari, Giffard, Sapolsky, & Steinberg, 1999). Novel evidence suggests that overexpression of Hsp27 in neurons protects against neurotoxic stimuli and may act as an inhibitor of neurodegeneration (Costigan et al., 1998; Kretz, Schmeer, Tausch, & Isenmann, 2006; Latchman, 2005). Surprisingly, two proteomic studies revealed a down-regulation of Hsp27 after NC and spinal cord injury, respectively (Jimenez et  al., 2005; Kang et  al., 2006). This might suggest that the nerves are irreversibly damaged. Voltage-dependent anion channels are major constituents of the outer mitochondrial membrane where they control membrane permeability and the subsequent release of apoptosis-promoting factors (Granville & Gottlieb, 2003). Therefore, in the context of nerve injury they might be involved in the degradation of neurons. Changes in albumin expression in tissues of the central nervous system indicate a dysfunction of the blood–brain or blood–spinal cord barriers, respectively. It has been shown that the integrity of these barriers is disturbed after nerve injury, resulting in an increased immunoreactivity of albumin in spinal cord tissue (Gordh, Chu, & Sharma, 2006). Collapsin response mediator protein 2 (CRPM2) is already well known as a regulator of neuronal polarity, axonal growth and regeneration after nerve injury (Arimura, Menager, Fukata, & Kaibuchi, 2004; Seltzer et  al., 1990). One of the groups which found a regulation in CRPM2 expression showed that this protein is regulated in a truncated form only in the peripheral part of the primary afferent neuron after partial nerve injury (Katano et al., 2006).

11177910 8394331 13592133 62655196 13242318 61556900

Heat shock 70-kDa protein 2 Superoxid dismutase 2

SNL SNL

Neuronal function and structural proteins SNL Actin b or SNL Actin related protein 2 SNL Advillin/pervin SNL Capping protein gelsolin-like

14010865

SCI/NC/SNL

5.3 7.7 n.i. 6.1

5.4 9.0

6.1

42.1 40.0 n.i. 39.1

69.8 24.9

22.8

Increased Increased Decreased Increased

(continued)

(Komori et al., 2007) (Komori et al., 2007) (Komori et al., 2007) (Komori et al., 2007)

−1.64/decreased/ (Kang et al., 2006)/(Jimenez increased et al., 2005; Komori et al., 2007) Increased (Komori et al., 2007) Increased (Komori et al., 2007)

Increased Increased −2.65

SNL NC SCI

20.9 69.5 26.4

(Kunz et al., 2005)/(Jimenez et al., 2005) (Alzate et al., 2004) (Jimenez et al., 2005) (Kang et al., 2006)

0.75/decreased

7.6 5.4 5.7

(Kang et al., 2006)

−2.78

P02529 92355 6978873

(Kang et al., 2006)

−4

Gamma crystallin C DnaK-type molecular chaperone hst 70 Guanidinoacetate methyltransferase Adenylyl cyclase-associated protein (CAP) homo Heat shock 27 kDa protein 1

References

Change in expression

Table 2  Characterization of proteins differentially expressed in the rat spinal cord following nerve lesion Protein Identification Animal number pI MW model Regulated protein Heat shock proteins/chaperones/antioxidants SCI Chaperonin containing TCP1, 40018616M 6.2 60.6 subunit 3 34872057 6.6 58.02 SCI Similar to CCT (chaperonin containing TCP-1) zeta subunit CCI/NC a-B-Crystallin 117388 6.8 20.1

Comparative Proteomic Analysis as a Method to Investigate Inflammatory and Neuropathic Pain 569

6981416 131803 16924010 38328248 9507125 92930 21746161

Collapsin response mediator protein 2

Collapsin response mediator protein 3 Collapsin response mediator protein 4 Dynamin 1 F-actin capping protein Z beta Fascin 1 Glial fibrillary acidic protein delta GTP-binding protein Rab 3A Lamin A Myelin protein zero Neurofilament, light polypeptide Neurofilament triplet M protein Neurofilament 3, medium

Neuronal differentiation-related gene Periaxin

Peripherin Ras-related protein Rab-1B Septin 2 Tubulin, alpha 1

a-synuclein Tubulin beta chain 15 Tubulin b1

NC/PNL/SNL

SNL SNL SCI NC SNL SCI SNL SNL SNL SCI SCI SCI/SNL

SCI NC/SNL

SCI SNL SCI SCI/PNL

NC SCI NC

21326455 9506999

34861163 14518293 18093102 4826659 30023548 5030428 7689363 1072002 8393778 13929098 P12839 8393823

135126

Regulated protein

Protein Identification number

Animal model

Table 2  (continued)

4.7 4.8 4.8

5.4 5.6 6.1 4.9

6.2 6.4

6.5 6.0 6.3 5.7 6.6 5.8 4.8 6.3 9.5 4.6 4.8 4.8

5.9

pI

14.5 49.9 50.7

53.6 22.1 41.6 50.2

55.2 16.1

62.7 62.4 95.9 31.1 52.2 48.8 24.9 74.6 27.9 61.3 95.8 95.7

62.8

MW

Decreased 2.67 Increased

Only SCI Increased −2.19 Only SCI

−3.55 Decreased

Decreased Decreased −2.98 Decreased Increased 26.71 Increased Increased Increased 3.13 4.83 5/decreased

Decreased

Change in expression

(Jimenez et al., 2005; Katano et al., 2006; Komori et al., 2007) (Komori et al., 2007) (Komori et al., 2007) (Kang et al., 2006) (Jimenez et al., 2005) (Komori et al., 2007) (Kang et al., 2006) (Alzate et al., 2004) (Komori et al., 2007) (Komori et al., 2007) (Kang et al., 2006) (Kang et al., 2006) (Kang et al., 2006; Komori et al., 2007) (Kang et al., 2006) (Jimenez et al., 2005; Komori et al., 2007) (Kang et al., 2006) (Alzate et al., 2004) (Kang et al., 2006) (Kang et al., 2006; Katano et al., 2006) (Jimenez et al., 2005) (Kang et al., 2006) (Jimenez et al., 2005)

References

570 E. Niederberger

19705431

Albumin

Adenine phosphoribosyltransferase

Aldehyde dehydrogenase Aldehyde reductase 1 Aldolase A Aldose C

Mitochondrial H1-ATP synthase alpha subunit40538742

Similar to ATPase, H1 transporting, V1 subunit A, isoform 1 ATP synthase beta subunit

SCI/NC/SNL

NC/SNL

NC NC NC NC/SNL

SCI/NC

SCI

NC/SNL

40538860

Proteins related to cellular homeostasis and metabolism SCI/NC Aconitase 2, mitochondrial

92350

34869154

25990263 6978491 6978487 1334163

114075

62078997 860908

Trp-Asp-repeat protein 1 Vimentin

SNL NC/SNL

10998840 61557028 20178269 29336093 6981672

Tubulin beta 3 Transgelin 2 Tropomyosin 2b Tropomyosin 3 Tropomyosin 4, alpha

PNL SNL SNL SNL NC/SNL

4.9

5.4

9.2

5.7 6.3 8.3 6.8

6.3

6.1

7.9

6.2 4.8

4.8 8.4 4.7 4.8 4.7

50.9

68.3

59.8

53.9 36.3 39.8 39.7

19.7

68.7

85.4

67.0 44.7

50.4 22.6 33.0 29.3 28.7

n.i.

Only SCI/ decreased 2.66

Decreased Decreased Decreased Decreased

Increased

15.24/increased

3/increased

Increased Increased

n.i. Increased Increased Increased Increased

(continued)

(Jimenez et al., 2005; Komori et al., 2007)

(Kang et al., 2006)/(Jimenez et al., 2005) (Kang et al., 2006)/(Jimenez et al., 2005)/(Komori et al., 2007) (Alzate et al., 2004; Jimenez et al., 2005) (Jimenez et al., 2005) (Jimenez et al., 2005) (Jimenez et al., 2005) (Jimenez et al., 2005; Komori et al., 2007) (Jimenez et al., 2005; Kang et al., 2006) (Kang et al., 2006)

(Katano et al., 2006) (Komori et al., 2007) (Komori et al., 2007) (Komori et al., 2007) (Jimenez et al., 2005; Komori et al., 2007) (Komori et al., 2007) (Jimenez et al., 2005; Komori et al., 2007)

Comparative Proteomic Analysis as a Method to Investigate Inflammatory and Neuropathic Pain 571

6912328 27704430 50926833M 54035288 P55051 71824

Carbonic anhydrase 3 Collagen alpha 1

Collagen alpha 2 Creatine kinase B Creatine kinase, mitochondrial Creatin kinase MM (muscle form)

Similar to dihydropyrimidinase related protein-2 Dimethylargininase 1 Dimethylargininase 2 Eno1 protein

b-Enolase, muscle-type Fatty acid-binding protein, brain (FABP-B) Fibrinogen alpha chain

Fibrinogen beta chain Fibrinogen gamma chain Gamma enolase GAPDH Galectin 3

Glucose-6-phosphate dehydrogenase Glutathione peroxidase 1 Glutathione synthetase

SNL PNL/SNL

SNL SNL SNL CCI/SNL

SCI

NC NC SCI/SNL

SNL SNL NC/SNL

SNL SNL NC NC NC/SNL

SNL SNL SNL

8393381 2654236 25742757

56971493 61098186 182118 27661099 1346429

3 34874349

62665835 417208 60678254 6978661

31377484 27688933

Regulated protein

Protein Identification number

Animal model

Table 2  (continued)

6.0 7.7 5.5

7.9 5.3 5 8.4 8.9

7.1 5.4 6.6

5.5 5.7 6.2

6.3

5.7 5.4 5.2 6.6

6.9 5.7

pI

59.9 22.5 52.7

55.0 53 47.6 36.2 25.6

47.4 15.0 60.6

31.1 30.1 47.1

81.4

99.6 42.9 93.5 43.2

29.8 139.1

MW

Increased Increased Increased

Increased Increased Decreased Decreased Increased

Increased Increased Increased

Decreased Decreased 2.1/increased

3.13

Increased 0.1 Decreased 0.63

Increased n.i./increased

Change in expression

(Jimenez et al., 2005) (Jimenez et al., 2005) (Kang et al., 2006; Komori et al., 2007) (Komori et al., 2007) (Alzate et al., 2004) (Jimenez et al., 2005; Komori et al., 2007) (Komori et al., 2007) (Komori et al., 2007) (Jimenez et al., 2005) (Jimenez et al., 2005) (Jimenez et al., 2005; Komori et al., 2007) (Komori et al., 2007) (Komori et al., 2007) (Komori et al., 2007)

(Komori et al., 2007) (Katano et al., 2006; Komori et al., 2007) (Komori et al., 2007) (Lee et al., 2003) (Komori et al., 2007) (Komori et al., 2007; Kunz et al., 2005) (Kang et al., 2006)

References

572 E. Niederberger

Guanine deaminase Hemopexin

Hydroxymethylglutaryl-CoA synthase 1 L-lactate dehydrogenase Lactate dehydrogenase B

Similar to leucine aminopeptidase Alpha-1 macroglobulin Malat dehydrogenase Peroxiredoxin 2 Peroxiredoxin 6 NADH dehydrogenase 1 alpha subcomplex 10-like protein 3-Phosphoglycerate dehydrogenase Phosphoglycerate mutase type B subunit Phosphoglycerate kinase

Protein disulfide isomerase

Purine-nucleoside phosphorylase Pyruvate dehydrogenase beta Pyruvate kinase M1 Pyruvate kinase M2 Pyruvate kinase 3 Sialic acid synthetase Nervous system cytosolic sulfotransferase

SNL NC/SNL

SNL SNL NC/SNL

SCI NC NC SCI SNL SCI

SNL SCI NC/SNL

CCI/NC

SNL SCI SNL SNL SCI NC SNL

34869683 50925725M 56929 62665891 16757994M 27714479 14522868

1352384

55562727 8248819 16757986

34878080 202857 15100179 34849738 16758348 30171809M

8393538 Gi 120975 6981146

7533042 16758014

6.5 6.2 6.6 7.6 6.6 6.3 5.6

5.9

6.3 7.1 7.5

8.0 6.5 6.2 5.3 5.6 7.1

5.6 5.3 5.7

5.5 7.6

32.6 38.98 58.4 58.6 57.8 40 19.1

57.0

57.5 28.8 45.1

64.5 167.1 36.8 21.8 24.9 40.54

58.2 36.8 37

51.6 51.3

Increased −2.47 Increased Increased 2.86 Decreased Increased

Decreased 1.97 Decreased/ increased 3.08

1.81 Increased Decreased 2.581 Increased −1.55

Decreased 1.9 Decreased

Increased Increased

(continued)

(Komori et al., 2007) (Kang et al., 2006) (Jimenez et al., 2005; Komori et al., 2007) (Jimenez et al., 2005; Kunz et al., 2005) (Komori et al., 2007) (Kang et al., 2006) (Komori et al., 2007) (Komori et al., 2007) (Kang et al., 2006) (Jimenez et al., 2005) (Alzate et al., 2004)

(Komori et al., 2007) (Jimenez et al., 2005; Komori et al., 2007) (Komori et al., 2007) (Lee et al., 2003) (Jimenez et al., 2005; Komori et al., 2007) (Kang et al., 2006) (Jimenez et al., 2005) (Jimenez et al., 2005) (Kang et al., 2006) (Komori et al., 2007) (Kang et al., 2006)

Comparative Proteomic Analysis as a Method to Investigate Inflammatory and Neuropathic Pain 573

Transferrin 6175089 Triose phosphate isomerase 12621074 Ubiquinol-cytochrome c reductase iron-sulfur GI 136708 subunit, mitochondrial precursor (Rieske iron-sulfur protein) (RISP) Ubiquitin C-terminal hydrolase 92934 UDP glucose dehydrogenase 13786146

SNL NC CCI

Immune system SNL Complement component 3 SNL Alpha-2-HS-glycoprotein SNL Immunoglobulin heavy chain SNL Immunoglobulin g heavy chain V region NC Interferon inducible protein 10 receptor SNL Interleukin-1 SNL Alpha-1-macroglobulin SNL Major histocompatibility complex (MHC) class 1 SNL T-cell receptor (TCR)

NC/SNL

Vitamin D binding protein

Thioredoxin Peroxidase 2

NC

NC NC/SNL

Thioredoxin Peroxidase 1

NC

6.1 6.3 5.2 5.7 5.1 5.1 6.5 5.2 5.7

23194212

5.6

5.1 7.5

6.9 6.5 8.9

8.3

5.2

pI

8393024 6978477 1685249 5051137 20984919 Q9NZH8 202857 1906678M

476569

16923958

2499469

Regulated protein

Protein Identification number

Animal model

Table 2  (continued)

10.9

188.2 39.0 13.3 13.7 89.7 18.7 168.7 19.1

55.3

24.7 54.9

79.1 26.9 28.0

22.4

21.8

MW

Decreased

Increased Increased Increased Increased Decreased Increased Increased Increased

Increased

Decreased Increased

Increased as well as decreased Increased as well as decreased Increased Decreased 0.48

Change in expression

(Alzate et al., 2004)

(Komori et al., 2007) (Komori et al., 2007) (Alzate et al., 2004) (Alzate et al., 2004) (Jimenez et al., 2005) (Alzate et al., 2004) (Komori et al., 2007) (Alzate et al., 2004)

(Jimenez et al., 2005) (Jimenez et al., 2005; Komori et al., 2007) (Jimenez et al., 2005; Komori et al., 2007)

(Komori et al., 2007) (Jimenez et al., 2005) (Kunz et al., 2005)

(Jimenez et al., 2005)

(Jimenez et al., 2005)

References

574 E. Niederberger

Apolipoprotein AI

SNL

6978515

5.5

5.5

Apolipoprotein A-I precursor

SCI/NC

113997

7.0 7.5 6.0 5.5 4.9

Proteins related to cell cycle/apoptosis and nerve degeneration SNL Annexin A1 6978501 SCI Annexin A2 9845234M SNL Annexin A3 51980303 NC Annexin A9/31 4502103 NC Annexin V 1421099

5.3 6.0 5.5 5.0 5.9 5.1 8.6

Gi 120975

5.9 5.1 5.3 6.8

4.1

3747079 8393910 56789330 71051053 34875656 13786200M

SNL NC SNL SNL SCI SCI/NC

SNL

13929166 50657380 220698 116416M

SNL SNL SNL SNL

Chloride intracellular channel 1 Chloride intracellular channel 4 Contrapsin-like protease inhibitor 21 Delayed rectifier potassium channel subunit IsK Guanine nucleotide-binding protein G(o), alpha subunit1 Nitric oxide synthase (NOS) 3 Phosphatidylethanolamine binding protein Rho GDP dissociation inhibitor beta Serin proteinase inhibitor B-1a Similar to RhoGDI-1 Voltage dependent anion channel 1

P02593

Signaling proteins SNL/NC Calmodulin (CaM)

30.1

30.1

39.2 38.7 36.6 37.6 35.8

13.2 20.9 22.9 42.9 23.5 30.8

40.6

28.9 27.4 46.7 14.6

16.7

Increased

1.71/increased

Increased 2.52 Increased Increased 3.11/decreased

Decreased Decreased Increased Increased 5.75 3.457/increased

2.2

Decreased/ decreased Increased Increased Increased Increased

(continued)

(Komori et al., 2007) (Kang et al., 2006) (Komori et al., 2007) (Jimenez et al., 2005) (Jimenez et al., 2005; Kang et al., 2006) (Jimenez et al., 2005; Kang et al., 2006) (Komori et al., 2007)

(Alzate et al., 2004) (Jimenez et al., 2005) (Komori et al., 2007) (Komori et al., 2007) (Kang et al., 2006) (Jimenez et al., 2005; Kang et al., 2006)

(Lee et al., 2003)

(Alzate et al., 2004)/(Jimenez et al., 2005) (Komori et al., 2007) (Komori et al., 2007) (Komori et al., 2007) (Alzate et al., 2004)

Comparative Proteomic Analysis as a Method to Investigate Inflammatory and Neuropathic Pain 575

5.7 4.8

Major vault protein 40S ribosomal protein SA

Calumenin Cathepsin B FK506 binding protein 4 Fscn1 protein

SNL NC

Miscellaneous NC NC SCI SCI 6680840 1127276 22324680 30023548

4.5 5.1 5.7 6.6

56.5 39.6 49.2 60.9

6.3 7.1 5.9 6.7

37.1 27.7 45.9 51.4

100.1 32.7

24.7 28.6 46.4

4.5 6.2 5.3

7514005 125970

100.2

21.6 35.8

8.6

4.9 5.2

44.4

MW

47.3 70.5

6978523 114041

5.1

pI

9.0 7.5

Apolipoprotein D Apolipoprotein E

NC NC/SNL

114008

Protein Identification number

Cyclic nucleotide phosphodiesterase 1 34873724M Similar to programmed cell death 6 interacting 34866400 protein SNL Transforming growth factor b induced 62663187 Protein synthesis and processing NC Elongation factor 1b 461991 NC Endoplasmatic reticulum protein 29 16758848 SCI Similar to translation initiation factor eIF-4A 55716055M II SCI Phgdh protein 55562727M SNL Poly (rC) binding protein 3 62665831 NC Heterogeneous nuclear ribonucleoprotein H1 10946928 NC/SNL Heterogeneous nuclear ribonucleoprotein L 20824058

Apolipoprotein A-IV

NC/SNL

SCI SCI

Regulated protein

Animal model

Table 2  (continued)

Increased Increased 1.63 −2

Increased Increased

−3.06 Decreased Decreased Increased

Increased Increased −2.27

Increased

Only SCI −2.25

Increased Increased

Increased

Change in expression

(Jimenez et al., 2005) (Jimenez et al., 2005) (Kang et al., 2006) (Kang et al., 2006)

(Kang et al., 2006) (Komori et al., 2007) (Jimenez et al., 2005) (Jimenez et al., 2005; Komori et al., 2007) (Komori et al., 2007) (Jimenez et al., 2005)

(Jimenez et al., 2005) (Jimenez et al., 2005) (Kang et al., 2006)

(Komori et al., 2007)

(Jimenez et al., 2005; Komori et al., 2007) (Jimenez et al., 2005) (Jimenez et al., 2005; Komori et al., 2007) (Kang et al., 2006) (Kang et al., 2006)

References

576 E. Niederberger

Haptoglobin Similar to Beta-soluble NSF attachment protein Proteasome subunit Reticulocalbin 2 Reticulocalbindin TCR a-chain precursor V region TCR-b Variable region Translationally controlled tumor protein Tumor rejection antigen gp96 Gi 3914438 8394171 6677691 88674 P04435 27686473 34862435

6981042 34859344 5.3 4.3 4.3 7.7 6.9 5 4.7

6.3 5.8 28.6 31.2 37.2 14.5 14.9 15.6 92.8

38.5 38.97 2.6 −2.21 Increased Increased Decreased Increased 4.21

Increased −1.6 (Lee et al., 2003) (Kang et al., 2006) (Jimenez et al., 2005) (Alzate et al., 2004) (Alzate et al., 2004) (Jimenez et al., 2005) (Kang et al., 2006)

(Jimenez et al., 2005) (Kang et al., 2006)

The changes in protein expression are indicated as described in the respective publication and show the fold increase or decrease (indicated by “−” or values <1) of the protein level n.i. not indicated; the gray fields show proteins that are regulated in more than one model

SNL SCI NC SNL SNL NC SCI

NC SCI

Comparative Proteomic Analysis as a Method to Investigate Inflammatory and Neuropathic Pain 577

578

E. Niederberger

Up-regulation of apolipoproteins in response to nerve injury have already been described, and it has been suggested that they serve as vehicles to transport lipids between cells during regeneration and degeneration of neurons (Boyles, Notterpek, & Anderson, 1990; Ignatius et al., 1986). Taken together, 2D PAGE can deliver potentially useful new information concerning “pain-associated” proteins and may therefore provide a reasonable technique for the identification of regulated proteins in the spinal cord following inflammation or nerve injury. However, it should be noted that there are a number of limitations with proteomics in neuroscience. Firstly, sample preparation of brain or spinal cord delivers a complex and heterogeneous mix of cells, which cannot be distinguished on 2D gels. Single, defined cells can only be investigated from neuronal cell cultures. Laser micro-dissection of single cells might yield a new method to analyze small groups of cells from neuronal tissues. Secondly, regulatory proteins expressed at low levels cannot be detected; protein analysis is substrate limited since no amplification methods are available as yet. Similarly, high and low molecular weight proteins as well as hydrophobic membrane proteins are difficult to separate by 2D-gel electrophoresis. However, the great advantage of 2D PAGE is the evaluation of protein patterns under physiological and pathophysiological conditions, respectively. This functional proteomic analysis delivers a substantial volume of data which might serve as a basis for the development of further research approaches. The studies summarized here, together with previous studies that identified single molecular participants in inflammatory or neuropathic pain, may improve the understanding of the molecular mechanisms which are involved in pain processes. This might facilitate the overview of pathophysiological signaling pathways in pain and thus the development of new pain therapeutics.

References Alzate, O., Hussain, S. R., Goettl, V. M., Tewari, A. K., Madiai, F., Stephens, R. L., Jr., et  al. (2004). Proteomic identification of brainstem cytosolic proteins in a neuropathic pain model. Brain Research. Molecular Brain Research, 128, 193–200. Anderson, T. E. (1982). A controlled pneumatic technique for experimental spinal cord contusion. Journal of Neuroscience Methods, 6, 327–333. Arimura, N., Menager, C., Fukata, Y., & Kaibuchi, K. (2004). Role of CRMP-2 in neuronal polarity. Journal of Neurobiology, 58, 34–47. Beiche, F., Scheuerer, S., Brune, K., Geisslinger, G., & Goppelt-Struebe, M. (1996). Up-regulation of cyclooxygenase-2 mRNA in the rat spinal cord following peripheral inflammation. FEBS Letters, 390, 165–169. Bennett, A. D., Chastain, K. M., & Hulsebosch, C. E. (2000). Alleviation of mechanical and thermal allodynia by CGRP(8-37) in a rodent model of chronic central pain. Pain, 86, 163–175. Bennett, G. J., & Xie, Y. K. (1988). A peripheral mononeuropathy in rat that produces disorders of pain sensation like those seen in man. Pain, 33, 87–107. Boyles, J. K., Notterpek, L. M., & Anderson, L. J. (1990). Accumulation of apolipoproteins in the regenerating and remyelinating mammalian peripheral nerve. Identification of apolipoprotein D, apolipoprotein A-IV, apolipoprotein E, and apolipoprotein A-I. The Journal of Biological Chemistry, 265, 17805–17815.

Comparative Proteomic Analysis as a Method to Investigate Inflammatory and Neuropathic Pain

579

Burre, J., Beckhaus, T., Corvey, C., Karas, M., Zimmermann, H., & Volknandt, W. (2006). Synaptic vesicle proteins under conditions of rest and activation: analysis by 2-D difference gel electrophoresis. Electrophoresis, 27, 3488–3496. Campbell, J. N., & Meyer, R. A. (2006). Mechanisms of neuropathic pain. Neuron, 52, 77–92. Chen, H., Lamer, T. J., Rho, R. H., Marshall, K. A., Sitzman, B. T., Ghazi, S. M., et al. (2004). Contemporary management of neuropathic pain for the primary care physician. Mayo Clinic Proceedings, 79, 1533–1545. Choi, Y., Yoon, Y. W., Na, H. S., Kim, S. H., & Chung, J. M. (1994). Behavioral signs of ongoing pain and cold allodynia in a rat model of neuropathic pain. Pain, 59, 369–376. Choudhary, J., & Grant, S. G. (2004). Proteomics in postgenomic neuroscience: the end of the beginning. Nature Neuroscience, 7, 440–445. Christensen, M. D., Everhart, A. W., Pickelman, J. T., & Hulsebosch, C. E. (1996). Mechanical and thermal allodynia in chronic central pain following spinal cord injury. Pain, 68, 97–107. Christensen, M. D., & Hulsebosch, C. E. (1997). Chronic central pain after spinal cord injury. Journal of Neurotrauma, 14, 517–537. Costigan, M., Mannion, R. J., Kendall, G., Lewis, S. E., Campagna, J. A., Coggeshall, R. E., et al. (1998). Heat shock protein 27: developmental regulation and expression after peripheral nerve injury. The Journal of Neuroscience, 18, 5891–5900. Dubner, R. (1994). Methods of assessing pain in animals. In P. D. Wall & R. Melzack (Eds.), Textbook of pain (pp. 293–302). Edinburgh: Churchill Livingston. Dubner, R., & Ren, K. E. (1999). Assessing transient and persistent pain in animals. In P. D. Wall & R. Melzack (Eds.), Textbook of pain (pp. 359–369). Edinburgh: Churchill Livingston. Dubner, R., & Ruda, M. A. (1992). Activity-dependent neuronal plasticity following tissue injury and inflammation. Trends in Neurosciences, 15, 96–103. Fabrizi, G. M., Cavallaro, T., Angiari, C., Cabrini, I., Taioli, F., Malerba, G., et al. (2006). CharcotMarie-Tooth disease type 2E, a disorder of the cytoskeleton. Brain, 130(Pt 2), 394–403. Feasson, L., Stockholm, D., Freyssenet, D., Richard, I., Duguez, S., Beckmann, J. S., et al. (2002). Molecular adaptations of neuromuscular disease-associated proteins in response to eccentric exercise in human skeletal muscle. Journal de Physiologie, 543, 297–306. George, A., Marziniak, M., Schafers, M., Toyka, K. V., & Sommer, C. (2000). Thalidomide treatment in chronic constrictive neuropathy decreases endoneurial tumor necrosis factor-alpha, increases interleukin-10 and has long-term effects on spinal cord dorsal horn met-enkephalin. Pain, 88, 267–275. Gordh, T., Chu, H., & Sharma, H. S. (2006). Spinal nerve lesion alters blood-spinal cord barrier function and activates astrocytes in the rat. Pain, 124, 211–221. Gorg, A., Obermaier, C., Boguth, G., Harder, A., Scheibe, B., Wildgruber, R., et al. (2000). The current state of two-dimensional electrophoresis with immobilized pH gradients. Electrophoresis, 21, 1037–1053. Gorg, A., Weiss, W., & Dunn, M. J. (2004). Current two-dimensional electrophoresis technology for proteomics. Proteomics, 4, 3665–3685. Granville, D. J., & Gottlieb, R. A. (2003). The mitochondrial voltage-dependent anion channel (VDAC) as a therapeutic target for initiating cell death. Current Medicinal Chemistry, 10, 1527–1533. Greenberg, J., McKeever, P. E., & Balentine, J. D. (1978). Lysosomal activity in experimental spinal cord trauma: an ultrastructural cytochemical evaluation. Surgical Neurology, 9, 361–364. Gruber, C. W., Cemazar, M., Heras, B., Martin, J. L., & Craik, D. J. (2006). Protein disulfide isomerase: the structure of oxidative folding. Trends in Biochemical Sciences, 31, 455–464. Hansson, P. T., & Dickenson, A. H. (2005). Pharmacological treatment of peripheral neuropathic pain conditions based on shared commonalities despite multiple etiologies. Pain, 113, 251–254. Ignatius, M. J., Gebicke-Harter, P. J., Skene, J. H., Schilling, J. W., Weisgraber, K. H., Mahley, R. W., et al. (1986). Expression of apolipoprotein E during nerve degeneration and regeneration. Proceedings of the National Academy of Sciences of the United States of America, 83, 1125–1129.

580

E. Niederberger

Ji, R. R., & Woolf, C. J. (2001). Neuronal plasticity and signal transduction in nociceptive neurons: implications for the initiation and maintenance of pathological pain. Neurobiology of Disease, 8, 1–10. Ji, R. R., Zhang, X., Wiesenfeld-Hallin, Z., & Hokfelt, T. (1994). Expression of neuropeptide Y and neuropeptide Y (Y1) receptor mRNA in rat spinal cord and dorsal root ganglia following peripheral tissue inflammation. The Journal of Neuroscience, 14, 6423–6434. Ji, R. R., Zhang, X., Zhang, Q., Dagerlind, A., Nilsson, S., Wiesenfeld-Hallin, Z., et al. (1995). Central and peripheral expression of galanin in response to inflammation. Neuroscience, 68, 563–576. Jimenez, C. R., Stam, F. J., Li, K. W., Gouwenberg, Y., Hornshaw, M. P., De Winter, F., et  al. (2005). Proteomics of the injured rat sciatic nerve reveals protein expression dynamics during regeneration. Molecular & Cellular Proteomics, 4, 120–132. Kang, S. K., So, H. H., Moon, Y. S., & Kim, C. H. (2006). Proteomic analysis of injured spinal cord tissue proteins using 2-DE and MALDI-TOF MS. Proteomics, 6, 2797–2812. Katano, T., Mabuchi, T., Okuda-Ashitaka, E., Inagaki, N., Kinumi, T., & Ito, S. (2006). Proteomic identification of a novel isoform of collapsin response mediator protein-2 in spinal nerves peripheral to dorsal root ganglia. Proteomics, 6, 6085–6094. Kim, S. H., & Chung, J. M. (1992). An experimental model for peripheral neuropathy produced by segmental spinal nerve ligation in the rat. Pain, 50, 355–363. Kingery, W. S. (1997). A critical review of controlled clinical trials for peripheral neuropathic pain and complex regional pain syndromes. Pain, 73, 123–139. Klivenyi, P., Ferrante, R. J., Matthews, R. T., Bogdanov, M. B., Klein, A. M., Andreassen, O. A., et al. (1999). Neuroprotective effects of creatine in a transgenic animal model of amyotrophic lateral sclerosis. Natural Medicines, 5, 347–350. Koltzenburg, M. (1998). Painful neuropathies. Current Opinion in Neurology, 11, 515–521. Komori, N., Takemori, N., Kim, H. K., Singh, A., Hwang, S. H., Foreman, R. D., et al. (2007). Proteomics study of neuropathic and nonneuropathic dorsal root ganglia: altered protein regulation following segmental spinal nerve ligation injury. Physiological Genomics, 29, 215–230. Kretz, A., Schmeer, C., Tausch, S., & Isenmann, S. (2006). Simvastatin promotes heat shock protein 27 expression and Akt activation in the rat retina and protects axotomized retinal ­ganglion cells in vivo. Neurobiology of Disease, 21, 421–430. Kunz, S., Niederberger, E., Ehnert, C., Coste, O., Pfenninger, A., Kruip, J., et  al. (2004). The calpain inhibitor MDL 28170 prevents inflammation-induced neurofilament light chain breakdown in the spinal cord and reduces thermal hyperalgesia. Pain, 110, 409–418. Kunz, S., Tegeder, I., Coste, O., Marian, C., Pfenninger, A., Corvey, C., et al. (2005). Comparative proteomic analysis of the rat spinal cord in inflammatory and neuropathic pain models. Neuroscience Letters, 381, 289–293. Latchman, D. S. (2005). HSP27 and cell survival in neurones. International Journal of Hyperthermia, 21, 393–402. Le Bars, D., Gozariu, M., & Cadden, S. W. (2001). Animal models of nociception. Pharmacological Reviews, 53, 597–652. Lee, S. C., Yoon, T. G., Yoo, Y. I., Bang, Y. J., Kim, H. Y., Jeoung, D. I., et al. (2003). Analysis of spinal cord proteome in the rats with mechanical allodynia after the spinal nerve injury. Biotechnological Letters, 25, 2071–2078. MacRae, T. H. (2000). Structure and function of small heat shock/alpha-crystallin proteins: established concepts and emerging ideas. Cellular and Molecular Life Sciences, 57, 899–913. Mannion, R. J., Costigan, M., Decosterd, I., Amaya, F., Ma, Q. P., Holstege, J. C., et al. (1999). Neurotrophins: peripherally and centrally acting modulators of tactile stimulus-induced inflammatory pain hypersensitivity. Proceedings of the National Academy of Sciences of the United States of America, 96, 9385–9390. Matsumoto, H., & Komori, N. (1999). Protein identification on two-dimensional gels archived nearly two decades ago by in-gel digestion and matrix-assisted laser desorption ionization time-of-flight mass spectrometry. Analytical Biochemistry, 270, 176–179.

Comparative Proteomic Analysis as a Method to Investigate Inflammatory and Neuropathic Pain

581

McCarson, K. E., & Krause, J. E. (1994). NK-1 and NK-3 type tachykinin receptor mRNA expression in the rat spinal cord dorsal horn is increased during adjuvant or formalin-induced ­nociception. The Journal of Neuroscience, 14, 712–720. Meller, S. T., & Gebhart, G. F. (1997). Intraplantar zymosan as a reliable, quantifiable model of thermal and mechanical hyperalgesia in the rat. European Journal of Pain, 1, 43–52. Niederberger, E., Tegeder, I., Vetter, G., Schmidtko, A., Schmidt, H., Euchenhofer, C., et  al. (2001). Celecoxib loses its anti-inflammatory efficacy at high doses through activation of NF-kappaB. The FASEB Journal, 15, 1622–1624. O’Farrell, P. H. (1975). High resolution two-dimensional electrophoresis of proteins. The Journal of Biological Chemistry, 250, 4007–4021. Ramana, K. V., Bhatnagar, A., & Srivastava, S. K. (2004). Inhibition of aldose reductase attenuates TNF-alpha-induced expression of adhesion molecules in endothelial cells. The FASEB Journal, 18, 1209–1218. Ramana, K. V., Fadl, A. A., Tammali, R., Reddy, A. B., Chopra, A. K., & Srivastava, S. K. (2006). Aldose reductase mediates the lipopolysaccharide-induced release of inflammatory mediators in RAW264.7 murine macrophages. The Journal of Biological Chemistry, 281, 33019–33029. Ramana, K. V., Willis, M. S., White, M. D., Horton, J. W., DiMaio, J. M., Srivastava, D., et al. (2006). Endotoxin-induced cardiomyopathy and systemic inflammation in mice is prevented by aldose reductase inhibition. Circulation, 114, 1838–1846. Rashid, M. H., Inoue, M., Toda, K., & Ueda, H. (2004). Loss of peripheral morphine analgesia contributes to the reduced effectiveness of systemic morphine in neuropathic pain. The Journal of Pharmacology and Experimental Therapeutics, 309, 380–387. Rivlin, A. S., & Tator, C. H. (1978). Effect of duration of acute spinal cord compression in a new acute cord injury model in the rat. Surgical Neurology, 10, 38–43. Rosenfeld, J., Capdevielle, J., Guillemot, J. C., & Ferrara, P. (1992). In-gel digestion of proteins for internal sequence analysis after one- or two-dimensional gel electrophoresis. Analytical Biochemistry, 203, 173–179. Scholz, J., & Woolf, C. J. (2002). Can we conquer pain? Nature Neuroscience, 5(Suppl), 1062–1067. Seltzer, Z., Dubner, R., & Shir, Y. (1990). A novel behavioral model of neuropathic pain disorders produced in rats by partial sciatic nerve injury. Pain, 43, 205–218. Sharma, H. S., Gordh, T., Wiklund, L., Mohanty, S., & Sjoquist, P. O. (2006). Spinal cord injury induced heat shock protein expression is reduced by an antioxidant compound H-290/51. An experimental study using light and electron microscopy in the rat. Journal of Neural Transmission, 113, 521–536. Sindrup, S. H., & Jensen, T. S. (1999). Efficacy of pharmacological treatments of neuropathic pain: an update and effect related to mechanism of drug action. Pain, 83, 389–400. Sullivan, P. G., Geiger, J. D., Mattson, M. P., & Scheff, S. W. (2000). Dietary supplement creatine protects against traumatic brain injury. Annals of Neurology, 48, 723–729. Tanaka, S., Uehara, T., & Nomura, Y. (2000). Up-regulation of protein-disulfide isomerase in response to hypoxia/brain ischemia and its protective effect against apoptotic cell death. The Journal of Biological Chemistry, 275, 10388–10393. Tarlov, I. M. (1972). Acute spinal cord compression paralysis. Journal of Neurosurgery, 36, 10–20. Tegeder, I., Niederberger, E., Schmidt, R., Kunz, S., Guhring, H., Ritzeler, O., et  al. (2004). Specific Inhibition of IkappaB kinase reduces hyperalgesia in inflammatory and neuropathic pain models in rats. The Journal of Neuroscience, 24, 1637–1645. Tjølsen, A., & Hole, K. (1997). Animal models of analgesia. In A. H. Dickenson & J. M. Besson (Eds.), Pharmacology of pain (pp. 1–20). Berlin: Springer. Varma, T., Liu, S. Q., West, M., Thongboonkerd, V., Ruvolo, P. P., May, W. S., et  al. (2003). Protein kinase C-dependent phosphorylation and mitochondrial translocation of aldose reductase. FEBS Letters, 534, 175–179. Wang, L. X., & Wang, Z. J. (2003). Animal and cellular models of chronic pain. Advanced Drug Delivery Reviews, 55, 949–965.

582

E. Niederberger

Woolf, C. J., & Costigan, M. (1999). Transcriptional and posttranslational plasticity and the generation of inflammatory pain. Proceedings of the National Academy of Sciences of the United States of America, 96, 7723–7730. Woolf, C. J., & Mannion, R. J. (1999). Neuropathic pain: aetiology, symptoms, mechanisms, and management. Lancet, 353, 1959–1964. Xu, C. J., Klunk, W. E., Kanfer, J. N., Xiong, Q., Miller, G., & Pettegrew, J. W. (1996). Phosphocreatine-dependent glutamate uptake by synaptic vesicles. A comparison with atpdependent glutamate uptake. The Journal of Biological Chemistry, 271, 13435–13440. Yabe-Nishimura, C. (1998). Aldose reductase in glucose toxicity: a potential target for the prevention of diabetic complications. Pharmacological Reviews, 50, 21–33. Yaksh, T. L., Hua, X. Y., Kalcheva, I., Nozaki-Taguchi, N., & Marsala, M. (1999). The spinal biology in humans and animals of pain states generated by persistent small afferent input. Proceedings of the National Academy of Sciences of the United States of America, 96, 7680–7686. Yenari, M. A., Giffard, R. G., Sapolsky, R. M., & Steinberg, G. K. (1999). The neuroprotective potential of heat shock protein 70 (HSP70). Molecular Medicine Today, 5, 525–531.

Index

A Acetyl CoA acetyltransferase, 384, 390, 394 Acidic fibroblast growth factor, 517 Adult disease and fetal origins cardiovascular disease, 88 environmental demand, 88 hypothalamic-pituitary-adrenal (HPA) stress responses, 88, 90 intrauterine deprivation, 88 long-term programming, 91, 94 mother-offspring interactions, 89 overall fetal body growth, 88 perinatal adaptation, 88 phenotypic plasticity, 88 stress resilience, 88 Adult stem cells. See Stem cells, adult AIDA-1, 237, 238 Alcoholism, proteomics of prefrontal cortex in, 387–388 Alpha-internexin, 387, 408 Alternative splicing of FOXP2 alternative transcripts and protein isoforms, 529 functional properties of alternative isoforms, 529 Alzheimer’s disease amyloid beta protein, 457 amyloid precursor protein (APP), 49, 323, 490–492, 502–505 apolipoprotein E, 498, 499 cerebrospinal fluid, 457 glutamine synthetase, 324 mouse model of, 489–506 proteome of, 490–492, 505–506, 550 Alzheimer’s disease related proteome ACAT2 as part of, 499, 502–503 apolipoprotein E precursor as part of, 499 CaMK-II alpha subunit as part of, 499

complement C1q subcomponent as part of, 499 dihydropyrimidinase related protein-2 as part of, 499 dynamin 1-like as part of, 500 glial fibrillary acidic protein as part of, 499 moesin as part of, 499 N-etylmaleimide sensitive fusion protein as part of, 499, 503 olfactory marker protein as part of, 499, 501 peroxiredoxin 6 as part of, 498, 499, 501–502 pyruvate kinase 3 / M2 as part of, 500 serotransferrin precursor as part of, 500 serum albumin precursor as part of, 500 synaptotagmin I as part of, 499, 503 Aminergic neurotransmitters, 401 AMPA receptors activity-dependent trafficking of, 321 in the formation of associative learning and memory, 322 insertion and removal of, 321 responsible for altering synaptic strength, 321 specifically regulated by dynamin-dependent endocytosis, 321 Amplified developmental instability hypothesis, developmental homeostasis, 28 Amygdala an essential neuroanatomical circuit, 306 fear learning and behavior, 306 formation of fear memory, 306, 308 locus of plasticity, 306 in regulating the acquisition, storage, and expression of fear memory, 306 role in fear recognition, 8 Amyloid precursor protein (APP) Alzheimer’s disease-related gene, 323 kinesin light chain 1 (KLC1), 323 U2AF65, 323

J.D. Clelland (ed.), Genomics, Proteomics, and the Nervous System, Advances in Neurobiology 2, DOI 10.1007/978-1-4419-7197-5, © Springer Science+Business Media, LLC 2011

583

584 Animal models, rat model of depression, 430 Annexin A1, 390, 394, 575 Annexin A5, 389, 414 Anterior cingulate cortex abnormal neurotransmitters of, 395 functional changes in schizophrenia, 382 neuropathology of in schizophrenia, 382, 383, 395, 396 specific protein changes in schizophrenia, 389–396 Antidepressants clorgyline, 404, 405 CRF receptor 1 antagonist, 422 desipramine, 402, 404, 405 DMP696, 405, 422 escitalopram (ESC), 402, 405, 422 fluoxetine, 402, 405, 406, 421–422, 426, 432–433 GR205171, 405, 422 L-000760735, 405, 421 monoamine oxidase inhibitors (iMAO), 400, 401 NK1 receptor antagonist, 421–422 nortriptyline (NOR), 400, 402, 431 paroxetine, 402, 406, 426, 432 selective serotonin re-uptake inhibitors (SSRI), 400–402, 421, 422, 426, 431, 433 tricyclic antidepressants (TCA), 400–402, 420, 560 venlafaxine, 402, 405, 421 AP-2 clathrin-associated, 315, 321, 324 in regulating the ligand-induced endocytosis of AMPA receptors, 321 Apolipoprotein, 120, 308, 458, 463, 498, 499, 502, 517, 547, 548, 565, 575, 576, 578 Apoptosis, 32, 38, 49, 61, 66, 115, 123, 146, 147, 168, 212, 218, 308, 407, 409–412, 414, 415, 417–420, 470, 473, 477, 478, 480, 513, 514, 516, 521, 550, 567, 568, 575 Apoptotic cell death. See Apoptosis Associative memories activation of NMDA receptors, 305 amygdala, 305, 322, 324 fear conditioning conditioned stimulus (CS), 305 contextual, 305–309, 323 CS-US association, 305–306, 308 cued, 305, 308, 309

Index interface of memory and emotion, 305 unconditioned stimulus (US), 305 ATP synthase, 388, 389, 409, 414, 415, 420, 427, 516, 528, 546, 548, 565, 571 Attention, 4, 6, 10, 13, 128, 148, 209, 382, 549 Attention deficit hyperactivity disorder, possible role for X-linked genes in, 13 Autism, male vulnerability to, 8, 11, 14, 470 Axonal transport, 426 B b-actin, 126, 269, 513, 525, 527, 548 Basal forebrain cholinergic neurons (BFCN) aging-dependent neurodegeneration of, 25 Alzheimer disease, 26 Behavior defensive, 305 molecular and cellular mechanisms of, 304 Behavioural flexibility, 4, 11 Biomarkers, CNS injury, 519–521 Bone marrow zink finger 2, 528 Brain aging, behavioral and cognitive changes during, 305 Brodmann area 9, proteome of in alcoholism, 387–388 Bullous pemphigoid antigen, 525 C Calcium homeostasis, 515 CaMKII-actinin-actin complex, anchoring AMPA receptors at synapses, 321 Ca2+ signaling cascade intracellular stores, 165–166 NMDA receptor, 163, 165–168 voltage dependent Ca2+ channel, 165 Cell cycle, 52, 208, 411, 412, 433, 513, 521, 567, 575 Cell differentiation, 53–55, 407, 408, 411, 415–419, 426 Cell proliferation, 10, 32, 52, 66, 145, 410, 411, 414, 418, 420, 421, 470, 522, 525, 530 Cell signalling, 412, 433 Cell survival, 146, 147, 164, 433 Cellular assembly, 432–434, 436 Cellular function and maintenance, 433, 436 Cellular metabolism, 54, 381, 396, 412, 432, 473, 477 Cellular morphogenesis, 431 Cerebellar lesion paradigm, 523 Cerebrospinal fluid composition, 456, 460, 519

Index pathological prion protein (PrPSc), 456–459 Chaperones, 321, 392, 411, 421, 427, 451, 472, 475, 514, 546, 567–569 Chaperonin containing tailless-complex polypeptide 1, 525 Choline acetyltransferase, 25, 148, 392, 395 Chromatin structure acetylation, 92, 97–98 histones and nucleosomes, 91–92 remodeling, 92, 93, 95 transcriptional activation, 92, 95 Citrate synthase, 385, 388 CNTNAP2 is a direct target of FOXP2 CNTNAP2/CASPR2 expression, 271 CNTNAP2 in neurological disorders, 271, 446, 490 CNTNAP2 mutant mouse, 268, 269, 273 shotgun-ChIP, PCR and EMSA validation, 270 Collapsin response mediator protein-2, 514, 515, 518, 519, 549, 550, 566, 568, 570 Controlled cortical impact, 513–514 Copy number variation the clinical consequences of submicroscopic, 184, 187–188 in the general population, 186–187, 334 Creatinine kinase ubiquitous mitochondrial 1, 389 Cu/Zn superoxide dismutase, 46–48, 513 Cylindromatosis, 237 Cytoarchitectural proteins, 390, 392, 394 Cytogenetics conventional, 178–179 chromosome aberrations numerical, 179 structural, 178, 179 chromosome banding techniques, 179 karyotyping, 179 light microscopy, 179 Cytoskeleton reorganisation, 426, 431 Cytosolic aspartate aminotransferase, 525 Cytosolic chaperonin CCT in amygdala, 305, 315, 319, 321 a molecular chaperone required for the folding of a-actin and b-tubulin, 321 D 78,000-Da glucose-regulated protein, 410, 525 78,000-Da glucose-regulated protein precursor, 513

585 Degeneration, 26, 37, 64, 450, 451, 512, 530, 541, 542, 550, 575, 578 Degradomics, 515 Developmental verbal dyspraxia and the KE family developmental verbal dyspraxia (DVD), 255, 256 KE family phenotype, 255 KE pedigree, 256 Differential analysis, 428–429 Differential in-gel electrophoresis (DIGE), 405, 421, 428, 459, 489–506, 529 Dihydropyrimidinase related protein 2 (DRP2), 384, 407, 422, 498, 499, 503, 515, 572 Dimethylarginine dimethylaminohydrolase 1 (DDAH-1), 384, 387, 388, 410 Disorders of speech and language autism spectrum disorder (ASD), 254 specific language impairment (SLI), 254, 255 DJ protein 1, 384 DNA demethylation maternal behavior, 94, 95 mechanisms, 93, 94 methyl-CpG-binding protein isoform (MBD2b, demethylase), 93 DNA methylation cytosine-phosphodiester-guanine (CpG) dinucletides, 93 de novo methylation (DNMT-3a, DNMT-3b), 93 direct mechanism (transcription factor binding), 93 indirect mechanism (transcriptional repressors), 93 maintenance methyltransferase (DNMT-1), 93 DNA microarrays cDNA arrays, 307 genechips, 307 hybridization, 307 multiple probe pairs, 307 oligonucleotide arrays, 307 spotted arrays, 307 Dopamine (DA), 53, 99, 101, 127, 148, 149, 199, 201, 204–207, 217, 348, 350, 382, 394, 395, 401, 450, 546, 550 Dorsal striatum, 385, 386, 389, 396 Dosage-sensitive gene hypothesis, critical region function, 27

586 Down syndrome candidate genes of, 46–59 critical region, 26, 44, 52 genetic origin of, 28 intelligence quotient decrease in, 23 learning deficits in, 37, 38, 52 memory deficits in, 26, 37, 38, 44, 45, 59, 66 mental retardation in, 21–67 molecular and cellular mechanisms of, 59–61 molecular pathways in, 59–64 phenotypes of, 23, 26–28, 30, 32, 33, 36–38, 42–45, 53, 59, 62, 63 transgenic mouse models of, 43–46 trisomic mouse models of, 37–44, 52, 63 2D polyacrylamide gel electrophoresis. See also 2D Gel electrophoresis 2D-gel, 430 2D maps, 430 Dynamin 1 brain, 387 E EH-Domain containing protein 3, 387 2D Electrophoresis. See also 2D Gel electrophoresis 2D difference gel electrophoresis (DIGE) technique, 428, 459, 495 differential in-gel electrophoresis (DIGE), 428, 459 fluorescent dyes, 459, 492, 505, 563 silver stain, 428, 459 Energy metabolism, energy production, 421, 427, 431, 516, 528 Enolase, 143, 404, 407, 410, 458, 514, 516, 525, 544, 546, 548, 549, 572 Epigenetics chromatin structure, 91–92, 95 DNA methylation, 92–99, 102 epigenome and epigenotype, 91 non-coding RNAs (microRNA), 91 regulation in mammals, 91 Estrogen aging, 169 Alzheimer’s disease, 168 cognition, effect on, 162 neuroprotection, 168–169 signaling cascades, 162, 163, 165–169 Exercise, 402, 406, 427, 432 Extreme male brain theory, 11 F F-ATP synthase, 528 FOXP2 and common forms of language disorder

Index association studies, 272 autism, 272 specific language impairment, 272 FOXP2 function, regulatory capacity, 263 FOX transcription factors accepted nomenclature, 257 in developmental disorders, 258 forkhead-box DNA binding domain, 257 FOXP2 conservation between species, 258 FOXP dimerisation, 263 FOXP2 expression pattern, 258 FOXP subfamily, 258 FOXP unusual characteristics, 258 FOX subfamilies, 257 Fragile-X-related gene (FXR1) ribosome-associated, RNA-binding protein, 322 in translational regulation of selective messenger RNA transcripts, 322 Fructose bisphosphate aldolase C, 385, 414 Functional consequences of FOXP2 mutations, 262–263 effects of etiological point mutations, 263 FOXP2 mutation screening, 272 FOXP2.Q17L substitution, 263 FOXP2.R553H missense mutation, 263–264 FOXP2.R328X nonsense, 263, 264, 268 Functional properties of FOXP2 FOXP2 transcripts, 259–260 genomic locus, 257, 259 G 2D Gel electrophoresis (2DE), 383, 384, 427–428, 459, 462, 493, 524, 558, 578 Gel matching, 428–429 Gene expression age-related changes in, 307 cellular functions, 316, 318 cluster analysis of, 316, 317, 475, 476 down-regulated, 61, 121, 314, 316, 322, 323, 476, 477 dynamical changes of, 311 enriched environments, 307, 308, 324 overall patterns of, 56, 314 regulated by behavioral learning paradigm, 304–305 transcriptional responses, 28, 63, 324 up-regulated, 121, 308, 314, 316, 322, 324, 476, 477, 496 Gene ontology (GO), 267, 310, 407–420, 433 Gene overdosage gene overexpression, 30, 33, 44, 46, 48, 50–52, 55, 56, 58–63, 66

Index primary gene effects, 28–29, 60 secondary gene effects, 28–29, 60, 61 Gene/protein networks, 433 Genome-wide screening for DNA copy-number variation diagnostic yield of, 185 1 Mb resolution microarray, 185 oligonucleotide microarray, 181–183, 310 tiling path resolution BAC arrays, 185 Genomic disorders low-copy repeats (LCRs), 185 non-allelic homologous recombination (NAHR), 185 17p11.2 microduplication syndrome, 186 Potocki–Lupski syndrome, 186 17q21.31 microdeletion syndrome, 186 segmental duplications, 185 Genomic imprinting, X-linked, 6, 9, 11–15 Genomic process estrogen receptor (ER) alpha, 164, 165 estrogen receptor beta, 164, 165 estrogen response element (ERE), 164 GFAP. See Glial fibrillary acidic protein Glial activation markers F2-isoprostane, 458 prostaglandins, 458 Glial fibrillary acidic protein (GFAP), 123, 144, 165, 236, 391, 407, 427, 435, 436, 498, 499, 513, 569 Glial specific proteins, 390, 394 Glucocorticoid receptor (GR) exon 17 promoter chromatin immunoprecipitation (ChIP), 95 cross-fostering and DNA methylation, 95 maternal care, 96–98 methylation, 95, 96, 98, 100 transcription factor NGFI-A binding site, 95–98, 103 Glucose metabolism, 88, 382, 394 Glutamate receptor 1 (GluR1) in the amygdala, 317, 318, 320, 321 immunohistochemistry, 317 one of the amino-3-hydroxy-5-methyl-4isoxalone propionic acid (AMPA) receptor subunits, 237, 318, 321, 471 quantitative immunohistological analysis of, 317 Glutamine synthetase, 165, 314, 324, 416, 525, 528, 572 Glycerol-3 phosphate dehydrogenase cytosolic (GPDH-C), 394 Glycogen phosphorylase, 392, 394 G-protein regulators, 231, 237, 238, 244

587 Grey matter, alterations in schizophrenia of, 383–387, 389, 390, 394–396 Guanine nucleotide binding protein polypeptide beta 1, 387 H Heat shock 70kDa protein 1, 385 High-density oligonucleotide microarrays advantages of, 182–183 loss of heterozygosity, 183 oligonucleotides, 181–183 one-color hybridization, 183 single nucleotide polymorphisms, 182, 183 two-color hybridization, 183 High-resolution chromosome banding, 179, 180 High throughput analysis of FOXP2 target genes ChIP-chip, 265, 268 Foxp2.S321X mutant mouse, 268 promoter microarrays, 265 in vitro and in vivo FOXP2 target identification, 268 High throughput screening technologies in mental retardation identification of genes in known mental retardation syndromes CHARGE, 183 Peters-Plus syndrome, 184 Pitt-Hopkins syndrome, 184 X-linked mental retardation, 184 Hippocampus in contextual fear conditioning, 306–308 in the encoding and retrieval of specific types of information, 306–307 spatial and episodic memories, 307 Histone acetyltransferases (HATs), 87, 92, 95, 97, 101 codes, 92 deacetylases (HDACs), 92, 95, 96 modifications, 92, 93 Homer, 231, 237, 238, 240, 242 Human epigenomics aging and epigenetic drift, 102 epigenetic analysis during life [Peritheral blood mononuclear cells (PBMCS)], 102 maternal behavior/mood and glucocorticoid promoter (GR) epigenotype, 103 stress responses and mental health, 102, 103

588 Human speech and language neural circuits subserving language, 254 neurological components necessary for speech and language, 254 Hydroxyacylglutathione hydrolase, 390, 394 I Identification, 7–11, 14, 21, 24, 29, 33, 43, 44, 46, 59, 64, 66, 67, 114, 178–180, 183–186, 228–232, 244, 245, 253, 265, 268–270, 307, 361, 365, 383, 384, 386, 390–393, 403, 404, 421, 423–425, 428–430, 458, 459, 461, 469, 475, 478, 489–491, 493–498, 504–506, 511, 512, 515, 518, 520–521, 525, 530, 540, 546, 549, 563–565, 569–578 Identifying a gene underlying verbal dyspraxia CS translocation, 256 FOXP2, 256, 257 FOXP2.R553H mutation, 257 FOX superfamily, 257 SPCH1 region, 256 Immobilised pH gradient (IPG), 423, 425, 428, 563 Ingenuity pathway analysis (IPA), 433, 434, 474, 475, 477, 478 Inositol monophosphate, 384, 392 Isocitrate dehydrogenase, 390, 394 Isoelectric focusing (IEF), 231, 423–425, 427, 493, 563, 568 K Keratin-10, 525 L Lactate dehydrogenase (LDH), 141, 146, 394, 514 Learning and memory disorders Alzheimer’s disease, 323 amyloid precursor protein (APP), 323 neurofibromatosis type 1 gene (NF1), 323 Limbic system amygdala, 162 hippocampus, 162 l-methionine anxiety related behaviors, 96 cytosine (re)methylation, 96 hypothalamic-pituitary-adrenal (HPA) stress response, 96

Index M Macromolecule biosynthesis, 421 Major depression. See Major depressive disorder Major depressive disorder (MDD), 284, 285, 400 Mass-spectrometry MALDI-TOF-MS protein identification, 496 nanoLC–MS–MS protein identification, 496 Maternal behavior altricial development, 98 anxiety-mediated behaviors, 91, 100 glucocorticoid receptor (GR) expression, 89, 97 hippocampus transcriptome, 90 licking and grooming (LG) and archedback nursing (ABN) posture, 89, 97 neuroendocrine and Hypothalamicpituitary-adrenal (HPA) stress responses, 88 nurturing environment, 89 Maternal programming of HPA stress responses NGFI-A and DNA demethylase (MBD2b), 95, 96, 98 NGFI-A bimodal role, 98 Maternal separation (MS), 431 Memory deficits, 26, 37–39, 54, 59, 66, 308, 323, 492 Mental retardation, 8, 12, 21–67, 177–189, 209, 210, 255, 322, 470 Metabolic proteins, 394 MiRNAs fundamental biological roles of, 66 regulation of protein-coding genes by, 32 therapeutic targets, 33, 59, 66 Molecular fluorescence in situ hybridization microdeletion syndromes, 180 multiplex-FISH, 180 spectral karyotyping, 180 subtelomeric chromosome anomalies, 180 polymerase chain reaction-based methods multiplex amplifiable probe hybridization, 180 multiplex ligation-dependent probe amplification, 180 Molecular karyotyping array-based comparative genomic hybridization (array CGH), 181

Index clone-based genomic microarrays large-insert genomic clones, 181 resolution of, 181 comparative genomic hybridization, 181 Myosin heavy chain, 525 N NADH oxidoreductase, 390, 394 Necrosis, 512, 522, 560 Nervous system development, 56, 407, 408, 411, 414, 417–419, 421, 431, 433, 434, 436 N-ethylmaleimide sensitive fusion protein, 241, 387, 499, 503, 504 Neurodegenerative diseases, Alzheimer’s disease, 542 Neurofilament, 231, 232, 235, 392, 408, 412, 421, 422, 435–437, 516, 542, 546, 548, 551, 564, 565, 567, 570 Neurogenesis, 26, 67, 115, 116, 149, 419, 421, 426, 431, 432, 437, 478, 522, 566 Neuroleptics, effects on animal brain proteome, 388–389 Neuroligin, 8, 237, 238 Neuronal plasticity, 49, 307, 422, 517 Neuron-related markers H-FABP, 460 neuron-specific enolase (NSE), 458 14-3-3 protein, 458 S-100b, 458, 460 tau protein, 458 Neuropathic pain animal models of central models of, 560–561 chronic constriction injury of the sciatic nerve (CCI-model), 561, 562 partial nerve injury (PNL), 562 peripheral models of, 561–562 spinal cord hemisection, 561 spinal nerve ligation (SNL), 561, 562 weight drop, 560 protein expression following, 565–566 Neuropilin-2, 518, 519 Neuroprotection, 146–147, 149, 166, 426, 525 Neurotransmission, 57, 64, 148, 149, 165, 266, 360, 395, 396, 404, 432, 471, 477 NFATc activation of target genes by dephosphorylated proteins of, 62, 63 DYRK1A and DSCR1 overexpression dyregulates pathways of, 62, 63

589 similar phenotypes in trisomic mouse models and mice deficient of, 63 NMDA receptor, 49, 65, 66, 143, 148, 149, 163, 165–168, 237, 240, 305, 321, 352, 395, 471 Nociceptive pain animal models of, 559 zymosan induced paw inflammation, 559 Norepinephrine (NE), 401, 404, 421, 560 O Oxidative stress, 48, 169, 394–396, 501, 503, 506, 548, 550 Oxidative stress markers lactate dehydrogenase, 146, 575 lipid peroxidation composition, 460 malondialdheide, 460 ubiquitin, 549 P Pantophysin related to synaptophysin and usually co-distributed with VAMP, 52 a vesicle protein, 323 Pathway analysis, 433, 434, 474, 476, 477 Pavlovian conditioning, fear conditioning, 305 Pearsons correlation, 384 Peptidyl-prolyl cis-trans isomerise A, 387 Peripheral inflammation, protein expression following, 564 Pharmacotherapy of Down syndrome mental retardation GABAA antagonists, 314 NMDA receptor antagonist, 65 Phosphatidylethanolamine-binding protein, 436 Phosphatidylinositol-4-phosphate-5-kinase (PIP5K) cellular signaling processes, 322 synthesis of phosphatidylinositol 4,5-bisphosphate (PIP2), 322 Phosphatidylinositol transfer protein-a, 513 Phosphoglycerate mutase, 384, 387–389 Physical treatments, 403 Postsynaptic density electron micrograph, 57 post mortem modification, 242

590 Postsynaptic density fraction affinity purification, 234–236, 245 contaminants, 229 from hippocampal slices, 242–245 mass spectrometric analyses, 230–231 proteins identified, 231–232 TritonX-100-derived, 236 Post-translational modifications, 29, 92, 268, 387, 404, 427, 505 Pre-and post mortem factors, 384 Prefrontal cortex, proteome of in alcoholism, 387–388 Profiling gene expression, 308–324 Prohibitin, 387, 417, 426, 517 Protein folding, 58, 408, 409, 411, 412, 415, 416, 420, 426, 431, 472, 475, 567 isoelectric point (pI), 404 isoforms, 231, 259, 260, 498, 562 molecular mass, 240, 404, 429, 459, 564 networks, 403, 433 14-3-3 Protein, 408, 426, 432, 437, 458, 461, 513, 521 Protein misfolding cyclic amplification (PMCA), 458 Protein phosphatases (PP), Ca2+-independent activity of CaMKII, 322 Proteins with unknown function, 394 Proteolysis, 240, 393, 418, 490, 514, 515, 530 Proteolytic proteins, 394 Proteome balancers protein function, 36 of grey matter of ACC in schizophrenia, 383, 385–388 homeostasis, 36 protein alterations as polymorphic variations, 35 of white matter of ACC in schizophrenia, 385, 386 Proteome analysis. See Proteomics Proteomics advantages, 361 Alzheimer’s disease related proteome, 491 comparative, 521–522 differential, 421, 463, 512 2D PAGE, 405, 406, 459, 461, 463, 513, 516, 519, 521, 525, 528, 529, 563–564, 578 identification of 2D-separated proteins, 564 limitations, 7, 335, 347–348, 359, 395, 403, 481, 521, 528–529, 578 mass spectrometry protein identification, 421, 566

Index proteome, 36, 59, 139, 149, 232, 384–389, 394–396, 399–437, 490, 491, 517–518, 528–530, 562 two-dimensional difference gel electrophoresis (DIGE) technique, 421, 481–505 PSD-93, 231, 237, 238, 244 PSD-95, 149, 231, 232, 234, 235, 237–241, 244 Psychotherapies, 402 R Rapid signaling cascades kinase activity, 166 phosphatase activity, 166, 167 Rat brain cholinergic basal forebrain, 427 cortical neurons, 426 hippocampus, 148, 404, 513 neural cells, 426 parietal cortex, 404 prefrontal/frontal cortex, 431, 432 synaptosomes, 431 Reactive gliosis, 514 Real-time RT-PCR, TaqMan, 311 Regeneration axonal, 516, 522, 525, 568 neuronal, 517, 522 Regions of the brain amygdala accessory basal nucleus, 306, 320 basal lateral nucleus, 306, 320 basolateral amygdaloid complex, 306 central nucleus, 306, 320 lateral nucleus, 306, 320 cortex, 306 hippocumpus, 306 hypothalamus, 209 neuroanatomical circuit, 306 thalamus, 306 Regulatory networks associated with human speech and language chromatin immunoprecipitation (ChIP), 265, 266, 268 targets of FOXP2, 265–268, 270–272 R553H affects multiple aspects of FOXP2 function DNA binding R553H affects multiple aspects of, 263 localisation, 264 Risperidone, 283–389, 396 R328X yields an unstable non-functional product, 264–265

Index S SAPAPs, 231, 237, 240, 242 S100b, 38, 47, 50, 517, 520 Scanning for intensely fluorescent targets (SIFT), 457 Schizophrenia aetiology and symptoms of, 128 cholinergic system, 395 glutamate activity, 395 metabolic changes, 390, 392, 394 role of dopamine, 395 Secondary injury, 517 Serine/threonine protein phosphatase, 390, 395 Serotonin (5HT), 25, 37, 48, 49, 97, 122, 127, 148, 167, 348, 356, 395, 401, 560 Serotonin biosynthesis, 426 Shanks, 231, 237, 238, 240, 242, 244, 480 Signaling proteins, 143, 566, 567, 575 Signal transduction, 29, 51, 57, 58, 61, 120, 121, 123, 145, 161–169, 228, 307, 322, 407, 410, 411, 413–416, 418, 421, 431–433, 472, 475, 504 Sleep deprivation, 285, 402, 406, 426, 427 Social cognition, 4–7, 14 aII Spectrin, 514 b-III Spectrin in amygdale, 315, 321 a golgi-and vesicle-associated protein, 321 NMDA receptor, interacts with, 321 Spinal cord injury, 146, 512, 516–517, 560, 565, 566, 568–573, 575–577 Sporadic Creutzfeldt–Jakob Disease clinical phenotypes, 456 diagnosis, 456 methionine/valine codon 129 polymorphisms, 456 molecular analysis, 459 Spot, 34, 184, 231, 241, 307, 384, 387, 390–394, 421, 423–425, 427–429, 431, 432, 451, 461, 493–498, 505, 513, 514, 516, 517, 519, 524–526, 529–530, 541, 543, 546, 551, 563–566 Sprague–Dawley rats, 389, 405–406 Staining, 35, 165, 243, 244, 259, 310, 311, 320, 421, 423–425, 427–429, 459, 461, 493, 494, 496, 525, 538, 542, 563 Stem cells, adult, 522 Stress, 34, 48, 64, 87–103, 128, 168, 169, 200, 305, 390, 392, 394–396, 401, 405, 407, 411, 412, 414–416,

591 419–420, 422, 426, 431–433, 460, 478, 499, 501–503, 506, 513, 546, 548, 550, 567, 568 Succinate dehydrogenase, 384, 392, 394 Superior temporal gyrus, 4–5, 14, 25, 271, 472, 475 Superoxide dismutase [Cu-Zn] (SOD), 46–48, 390, 394, 419, 513 Synapse electron micrograph, 57, 228, 239 immuno-EM for PSD-95 and Synapsin 1, 238–239 Synaptic plasticity, 37, 39, 40, 48, 49, 51–53, 63, 66, 162–168, 207, 266, 273, 307, 322, 323, 427, 481 Synaptic plasticity, long-term potentiation (LTP), 39, 40, 48, 49, 51, 53, 63, 163, 166 Synaptic proteins, 243, 244, 315, 394, 395 Synaptosomal associated protein 25, 34, 123, 390, 395, 418 Synaptotagmin, 122, 308, 312, 314, 319, 323, 499, 503, 515 Synaptotagmin, in synaptic vesicle trafficking and neurotransmitter release, 323 Syne-1 (CPG2), 237 T Targeted microarray applications, 184 TARP, 237 Teleost fish, 522–523, 530 Trafficking proteins, 394, 503 Transcription factor NGFI-A glucocorticoid receptor GR exon 17 promoter binding site, 95–98 interaction with cAMP response element binding protein-binding protein CBP, 95, 97 transcriptional activation, 95 Transcription factors, 12, 28, 38, 46, 49, 51, 62, 63, 92, 93, 95, 97, 99, 101, 103, 121, 125, 144, 163, 164, 167, 208, 257–259, 261, 263–266, 269, 270, 272, 281, 317, 318, 361, 559 Transcriptome compensatory mechanisms, 30 dosage-dependent genes, 61 eQTLs, 31 euploid genes, 30 gene decreased expression, 29, 30 inter-individual gene expression variations, 31 transcriptional variations, 29–33

592 Transferrin, 125, 391, 463, 504, 514, 517, 520, 574 Transgenic mouse model of Alzheimer’s disease b-amyloid precursor protein (APP) transgenic mouse line as, 491 Thy1-APP751SL transgenic mice as, 491 Transmission of maternal behavior environmental influences, 100–102 epigenetic mechanisms, 99–100 ER-a bimodal role, 98 estrogen sensitivity and maternal responsivity, 100 maternal brain circuit [medial preoptic area (MPOA) ER-a expression], 99 Traumatic brain injury, 512–516 Trichostatin A (TSA) anxiety related behaviors, 96 cytosine demethylation, 96 hippocampal transcriptome, 96 histone deacetylase inhibitor (HDACi), 95 hypothalamic-pituitary-adrenal (HPA) stress responses, 96 transcription factor binding, 93 Trisomy 21 chromosome imbalance, 26, 43 partial trisomy 21, 26 Trophic Factors BDNF, 168 IGF-1, 143 Tropomodulin, 525 b-Tubulin in amygdala, 321 cytoskeletal structure, undergo learning induced changes of, 321, 427 Turner syndrome, 3–15, 179, 210

Index U Ubiquitin carboxy-terminal hydrolase L1, 387, 414, 514, 542 Unbiased identification of FOXP2 binding CNTNAP2, 268–270 shotgun-ChIP, 268–270 V Vesicle-associated membrane protein (VAMP), in synaptic vesicle trafficking and neurotransmitter release, 323 Vesicular trafficking, 431 Vestibular compensation, 517–519 Vestibulopathy, 518 Vimentin, 312–314, 319, 324, 420, 426, 513, 514, 571 W Western blotting (WB), 123, 127, 261, 426, 471, 473, 521, 530 White matter, alterations in schizophrenia of, 130 X X chromosome, 3–9, 11, 12, 178, 184, 210, 361 X-inactivation, escape from, 5, 7–10, 12–14 39,XO mouse, 7, 15 Y Y chromosome, 12, 13, 125