Parkinson's Disease (Solvay Pharmaceuticals Reserach Forum, 1)

PARKINSON'S DISEASE Solvay Pharmaceuticals Conferences Series Editors Werner Cautreels, Claus Steinborn and Lechoslaw...

Author: E. Ronken

52 downloads 798 Views 13MB Size Report

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

Report copyright / DMCA form

DOWNLOAD PDF

PARKINSON'S DISEASE

Solvay Pharmaceuticals Conferences Series Editors Werner Cautreels, Claus Steinborn and Lechoslaw Turski

Volume 1 25-27 October 2000, Como, Italy

ISSN: 1566-7685

Parkinson's Disease Edited by E. Ronken Solvay Pharmaceuticals, Weesp, The Netherlands and

G.J.M. van Scharrenburg Solvay Pharmaceuticals, Weesp, The Netherlands

/OS Press

Ohmsha

Amsterdam • Berlin • Oxford • Tokyo • Washington, DC

© 2002. The authors mentioned in the Table of Contents All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted. in any form or by any means, without the prior written permission from the publisher. ISBN I 58603 207 0(IOS Press) ISBN 4 274 90486 5 C3047 (Ohmsha) Library of Congress Control Number: 2001098020

Publisher IOS Press Nieuwe Hemweg 6B 1013BG Amsterdam The Netherlands fax:+31 206203419 e-mail: [email protected]

Distributor in the UK and Ireland IOS Press/Lavis Marketing 73 Lime Walk Headington Oxford OX3 7AD England fax:+44 1865750079

Distributor in the USA and Canada IOS Press, Inc. 5795-G Burke Centre Parkway Burke, VA 22015 USA fax:+l 7033233668 e-mail: [email protected]

Distributor in Germany. Austria and Switzerland IOS Press/LSL.de Gerichtsweg 28 D-04103 Leipzig Germany fax:+49 341 9954255

Distributor in Japan Ohmsha. Ltd. 3-1 Kanda Nishiki-cho Chiyoda-ku. Tokyo 101 Japan fax:+81 332332426

LEGAL NOTICE The publisher is not responsible for the use which might be made of the following information. PRINTED IN THE NETHERLANDS

Preface "The Solvay Pharmaceuticals Conferences: where industry meets academia in a search for novel therapies"

The Solvay Pharmaceuticals Conferences allow leading representatives of academic and industrial research an exchange of views about how to make significant progress towards the discovery and development of novel medicines. Emerging insights into the pathophysiology of human diseases are addressed, together with any new therapeutic opportunities these might give. The design of novel therapies, from the bench through to the clinic is debated. Their aim is to bring together interdisciplinary groups of scientists working on all aspects of target discovery, molecular design, drug discovery technologies and pharmacology, as well as physicians working at the bedside and clinicians designing trials with potential therapies. Past decades were marked by significant advances in the understanding of human diseases. The cloning of several families of proteins has produced new strategies for the modulation of disordered processes. The engineering of transgenic animals with modified expression of target proteins has created new insights into human pathology. Enabling technologies in target discovery, medicinal chemistry, toxicology and novel forms of organisation of drug discovery or clinical trials were implemented to support discovery efforts. The resulting advances lead to clinical trials addressing therapeutic value of new medicines (proof of principle) and subsequently introduction of breakthrough therapies. Progress in medical sciences not only deserves the highest public attention, it also triggers expectations with regard to quality of new therapies, speed of discovery, efficiency of marketing, and overall business ethics. These are challenges for contemporary pharmaceutical industry and applied academic sciences. The mission of the Solvay Pharmaceuticals Conferences is to facilitate meetings between industry and academia where avenues of ongoing work can be discussed in an open and interactive manner. The first Solvay Pharmaceuticals Conference on "Parkinson's Disease" was held in Como from October 25 to 27, 2000, a city in the picturesque area of Lake Como, Italy. The scientific sessions covered most important aspects of the work on movement disorders. Presentations reviewed the role of receptors, ion channels and proteins in the etiology of Parkinson's disease, and reported on clinical experience with novel drug candidates. This book is a collection of the presentations of interest to scientists from a variety of disciplines including neuroscience, physiology, medicinal chemistry, pharmacology, toxicology, genetics, molecular biology and medicine. W. Cautreels C. Steinborn L. Turski

List of contributors Aebischer, P. Division of Surgical Research and Gene Therapy Center, Lausanne University Medical School, Lausanne, Switzerland Beal, M.F. Department of Neurology and Neuroscience, Weill Medical College of Cornell University and the New York Hospital-Cornell Medical Center, New York, U.S.A. Bronstein, J.M. UCLA School of Medicine, Department of Areurology, 710 Westwood Plaza, Los Angeles. CA 90095, U.S.A. Brooks, D.J.

MRC Cyclotron Building, Imperial College School of Medicine, Hammersmith Hospital, Du Cane Road, London W12 ONN, United Kingdom. P. Caviedes Program of Molecular & Clinical Pharmacology, JCBM Faculty of Medicine, University of Chile, Santiago, Chile Drukarch. B. Section of Experimental Neurology, Department of Neurology, Research Institute Neurosciences, Vrije Universiteit. v.d. Boechorststraat 7, 1081 BT Amsterdam, The Netherlands Earl, C. Klinikum der Philipps Universitat, Klinik fur Neurologie. Rudolf-Bultmann Strasse 8, D-35033 Marburg, Germany Jenner, P. GKT School of Biomedical Sciences, King's College London, Guy's Campus. London Bridge. London SE1 1UL, United Kingdom Klockgether, T. Department of Neurology. Friedrich-Wilhelm University. Bonn, Germany McCreary, A. Solvay Pharmaceuticals, Weesp, The Netherlands Mount, H.T.J.

Centre for Research in Neurodegenerative Diseases and Division of Neurology, Department of Medicine, University of Toronto, 6 Queens Park Crescent West, Toronto, Ontario M5S 3H2. Canada Oertel, W. Klinikum der Philipps Universitat, Klinik fur Neurologie, Rudolf-Bultmann Strasse 8. D-35033 Marburg, Germany

Richardson, P.J. Department of Pharmacology, University of Cambridge, Tennis Court Road, Cambridge, CB2 JQJ, United Kingdom Ronken, E. Solvay Pharmaceuticals, Weesp, The Netherlands Sautter, J. Klinikum der Philipps Universitat, Klinik fiir Neurologie, Rudolf-Bultmann Strasse 8, D-35033 Marburg, Germany Scharrenburg, G.J.M. van Solvay Pharmaceuticals, Weesp, The Netherlands Schlegel, J. Neuropathology, Institute of Pathology, Munich Technical University, Miinchen, Germany Standaert, D.G. Department of Neurology, Massachusetts General Hospital and Harvard Medical School, Boston MA 02114, U.S.A. Turski, L. Solvay Pharmaceuticals, Weesp, The Netherlands

Contents Preface, W. Cautreels, C. Steinborn and L. Turski List of Contributors Conference Preface, G.J.M. van Scharrenburg and E. Ronken Key-note lecture - Movement Disorders: An Overview, D.J. Brooks Part I: Parkinson's Disease - Symptoms, Models and Treatment Neurochemical Changes in Parkinson's Disease, D.G. Standaert Psychosis and Depression in Parkinson's Disease, D. Brandstadter and W.H. Oertel Experimental Models of Parkinson's Disease, P. Jenner SLV308: A Novel Antiparkinsonian Agent with Antidepressant and Anxiolytic Properties, A.C. McCreary, E. Ronken, J. van der Heyden, A. Herremans, T. Tuinstra, S. Long and G.J.M. van Scharrenburg Adenosine A2A Receptor Antagonists and Parkinson's Disease, P.J. Richardson Part II: Understanding and Influencing Neuro-Degeneration Bioenergetics in Neurodegeneration, M.F. Beal Neuronal Cell Death and Apoptosis in Neurodegenerative Diseases, T. Klockgether andU.Wullner Animal Models of Neurodegeneration in Parkinson's Disease, C. Earl and J. Sautter Parkinsonian Features of the Ataxia-Telangiectasia and of the Atm-Deficient Mouse, H.TJ. Mount, Y. Wu, P. Fluit, Q. Bi, T.O. Crawford and A.S. Mandir Neuroprotection in Parkinson's Disease through Selective Gene Induction, B. Drukarch, J. Flier and F.L. van Muiswinkel Part III: Towards Neuroprotective Efficacy The RET-Dependent Neuroprotective Effects of GDNF are Mediated by Activation of Proteinkinase B (PKB)/AKT, J. Schlegel, F. Neffand K. Eggert An Immortalized Neuronal Cell Line Derived from the Substantia Nigra of an Adult Rat: Application to Cell Transplant Therapy, C. Arriagada, J. Salazar, T. Shimahara, R. Caviedes and P. Caviedes Lentiviral Vector Delivery of GDNF in Primate Models of Parkinson's Disease Prevents Neurodegeneration, J.H. Kordower, M.E. Emborg, J. Bloch, S.Y. Ma, Y. Chu, L. Leventhal, J. McBride, E.-Y. Chen, S. Palfi, B.Z. Roitberg, J.E. Holden, W.D. Brown, R. Pyzalski, M.D. Taylor, P. Carvey, Z.D. Ling. D. Trono, P. Hantraye, N. Deglon and P. Aebischer Design of Clinical Studies for Neuroprotective Efficacy, J.M. Bronstein Regulation of Glial Cell Line-Derived Neurotrophic Factor (GDNF) for Dopamine Conservation in Models for Parkinson's Disease, E. Ronken, D. McCrossan and J. Venema

51 59 71 73 81 88 94 105 113 \ \5

120

\ 33 143

148

Concluding Remarks, E. Ronken and G.J.M. van Scharrenburg

157

Author Index

159

Conference Preface and

Key-note Lecture

This page intentionally left blank

Conference Preface Parkinson's Disease (PD) tends to be regarded as a disease linked to ageing. Most cases are diagnosed between the ages 50 and 60 years and less than 10% is diagnosed below the age of 40. Parkinson's Disease is now recognised as the most common cause of long term disability in the elderly. The WHO estimated in 1996, that there were about 3.5 million cases world-wide. One percent of the population over the age of 60 suffer from the disease; this number rises to two percent for those over 80. It is clear that the number of PD cases will grow with the longevity of the population. By 2025 one out of four people in the western world will be aged 60 or more. Most likely 250 people per 100.000 of the population will develop the disease by that time. Parkinson's Disease, known as "shaking palsy" is characterised in most patients by the primary motor symptoms such as progressively developing tremor, rigidity, slowness of movement and postural instability. The major secondary symptom associated with the disease is depression, which may be present in more than 50% of the population and is sometimes more disabling and affecting the quality of life as the primary symptoms. Parkinson's Disease is considered to be a degenerative disease. The motor symptoms of PD were discovered to arise as a result of the degeneration of nerve cells in the substantia nigra located in the mid brain, which use dopamine for signalling and project to the caudate nucleus 8 putamen (corpus striatum). As a consequence of the degeneration, dopamine levels are reduced by 70% at the moment the first symptoms appear. Why these dopamine neurons in the substantia nigra are so sensitive and lose their dopamine phenotype or even die is unclear. At present there is support for the hypothesis that these dopamine synthesising nerve cells of patients are more sensitive to oxidative insult due to the relative high amount of free radicals generated during dopamine synthesis. Another reason could be mitochondrial defects, which compromise the energy supply of the sensitive dopamine neurons to such an extent that it eventually leads to e.g. apoptotic cell death. All current therapy is based on reducing the symptoms of PD and improving the quality of life. At present no marketed products can claim to slow or halt the progression of the disease. The ultimate challenge is to develop pharmacotherapy treating the fundamental cause of the disease rather than the symptoms. However, there is still a clear need for better symptomatic treatment capable to keep the quality of life of PD patients on a higher level for a longer period of time. Solvay Pharmaceuticals has dedicated since a few years a substantial part of its R&D efforts to neurology with focus on Parkinson's Disease. In order to take notice of recent developments relevant for PD, leading clinical and preclinical scientists active in the field of movement disorders and/or neurodegeneration were invited to present their latest results and views during the 1st Solvay Pharmaceuticals Conference.

The wish to preserve this information valuable for everyone actively involved in the R&D for new, innovative therapies for Parkinson's Disease has led to these Proceedings containing all contributions. I hope this book will indeed stimulate both scientists and clinicians in their efforts to improve therapy for Parkinson's Disease and related neurological disorders. Guus van Scharrenburg and Eric Ronken

Parkinson's Disease E. Ronken and G.J.M. van Scharrenburg (Eds.) IOS Press, 2002

Movement disorders: An overview David J. Brooks* Division of Neuroscience, Imperial College School of Medicine, Hammersmith Hospital, London, UK Keywords: Parkinson's disease, Huntington's disease, genetics, treatment, positron emission tomography, deep brain stimulation, transplantation.

Introduction Movement disorders can broadly be broken down into syndromes associated with poverty of movement, that is parkinsonian disorders, and those associated with involuntary movements - see Tables 1,2. This division is somewhat arbitrary as Parkinson's disease can be associated with involuntary movements such as tremor, levodopa-induced dyskinesias (a variant of chorea), dystonia, and myoclonus. Conversely, while the primary feature of Huntington's disease is chorea, it is also associated with dystonia, and parkinsonism. Genetic causes of dystonia also result in tremor and myoclonic jerking. Nevertheless, this division provides a useful clinical approach for categorising patients. Table 1: Movement disorders

Akinetic-rigid disorders

Involuntary movements

Parkinson's disease Parkinson-plus syndromes (progressive supranuclear palsy, multiple system atrophy) tremor, dystonia, chorea, tics, myoclonus, startle syndromes

Table 2: Involuntary Movements

Dystonia Chorea Myoclonus Tics Tremor

Sustained or repetitive muscle spasms caused by co-contraction of agonists and antagonists leading to characteristic posturing. Irregular, unpredictable, jerky movements that flit from one body part to another Rapid, shock-like, muscle jerks which are often repetitive. Repetitive, stereotyped, movements that seem purposeful and can be imitated. Often preceded by compulsion and followed by a feeling of relief. A rhythmic, sinusoidal movement

Address for correspondence: David J. Brooks MD DSc FRCP, Hartnett Professor of Neurology, MRC Cyclotron Building, Imperial College School of Medicine, Hammersmith Hospital, Du Cane Rd, London W12 ONN, UK

6

DJ. Brooks /Movement Disorders: An Overview

Parkinson's disease Some of the questions that are still unresolved in PD are detailed and considered below: (a) Is PD a single entity ? A difficulty with addressing this question is that there is currently no agreed definition of Parkinson's disease (PD). A clinician would define PD as a patient having a combination of bradykinesia, rigidity that was "lead pipe" in character, and a 3-5 Hz resting tremor. Supportive features include an asymmetrical onset, a good response to dopaminergic medication, and subsequent postural instability. By contrast, a pathologist generally defines PD as neuronal Lewy body inclusion disease targeting the substantia nigra compacta and other brainstem nuclei. These clinical and pathological definitions are only partially reconcilable. First, there are a variety of other subcortical and cortical degenerations that are associated with parkinsonism and that can clinically mimic PD, as can neuroleptic exposure or brain damage associated with certain toxins - see table 3. Second, Lewy body disease presents more frequently as dementia due to cortical involvement than as PD which is the second most common manifestation. Tremor, dystonia, autonomic failure and supranuclear gaze palsies have also been described in association with Lewy body pathology [1].

Table 3: Causes of parkinsonism

Parkinson's disease Subcortical degenerations

Cortical degenerations

Metabolic disorders

Basal ganglia lesions Encephalitis Toxins Drugs

(Lewy body disease) striatonigral degeneration multiple system atrophy progressive supranuclear palsy corticobasal degeneration Huntington's disease Hallervorden-Spatz disease Pallidal atrophy Alzheimer's disease Creuzfeldt- Jakob disease Hydrocephalus dopa-responsive dystonia Wilson's disease Chronic liver failure Multi-infarct disease Tumours Encephalitis lethargica, Other viral infections MPTP, CO, hypoxia, Mn II . CS2, solvents, pesticides dopamine receptor blockers dopamine depleting agents

D.J. Brooks / Movement Disorders: An Overview

1

Two recent clinicopathological series have examined the brains of patients clinically diagnosed in life as having PD by experienced physicians [2,3]. They reported similar findings: Only around three quarters of cases had brainstem Lewy body disease and, of these, half had mixed pathology with coincidental Alzheimer and small vessel disease. The rest proved to have alternative conditions such as multiple system atrophy, progressive supranuclear palsy, Alzheimer's disease, or vascular disease. Given this, it is probably more appropriate to decribe Parkinson's disease as a syndrome rather than a single entity. (b) Is PD genetic or environmental in origin ? This is difficult to answer given the uncertainty in the definition of Parkinson's disease. Several gene mutations associated with familial parkinsonism have now been identified see Table 4.

Table 4: Familial parkinsonism

Inheritance PARK1 PARK2 PARK3

Locus AD AR AD AD AD

Gene 4q 6q 2p 4p 4q

Lewy bodies a-synuclein parkin ? ? UCHL-1

Yes No Yes 7 ?

a-synuclein gene mutations have been reported in seven kindreds originating from Italy, Greece, and Germany and are associated with onset of PD in the 4th-6th decades and Lewy body pathology. Dementia and a poor levodopa response are also features so the condition is not entirely clinically typical of sporadic PD [4]. Parkin gene abnormalities have been reported in 70% of PD cases with onset before the fourth decade, most of whom were thought to have a sporadic disorder [5,6]. These parkinsonian patients are levodopa responsive but generally lack brainstem Lewy bodies and so, pathologically, have a different disorder from PD. If one defines PD as brainstem Lewy body disease than another problem arises: For every case of diagnosed PD there are probably another 10-15 cases of subclinical nigral Lewy body disease [7]. In the absence of a marker for PD it is difficult to determine the cause of the disorder if in practice one can only detect a minority of affected subjects. I8Fdopa PET studies on at-risk asymptomatic adult relatives of PD patients with familial disease have reported a 25% prevalence of subclinical dopaminergic dysfunction [8]. Assuming 50% of these relatives to be gene carriers this represents a 50% pick up rate with PET. Twin surveys have concluded that clinical concordance for PD is equally low in mono- and dizygotic twin pairs and that inheritance is unlikely to play a role in the "sporadic" condition. The most recent and largest survey [9] screened 19,842 white male twins for PD by questionaire from the US Veterans Register. 71 MZ and 90 DZ PD pairs were identified and pairwise concordance was similar (0.155 MZ, 0.111 DZ; relative risk, 1.39; 95% confidence interval, 0.63-3.1). In 16 pairs diagnosed by the age of 50, MZ concordance was significantly higher than DZ concordance: MZ 1.0 (4 pairs), DZ 0.167 (relative risk, 6.0; 95% confidence interval, 1.69-21.26). The authors concluded that genetic factors are important when sporadic disease begins before the age of 50 but that no genetic component is evident when the disease begins after that age.

8

D.J. Brooks / Movement Disorders: An Overview

A problem with the above conclusion is that if only one in ten affected cases can be detected clinically many concordant late-onset twin pairs may have been missed in twin surveys. A recent l8F-dopa PET study detected subclinical dopaminergic dysfunction in 50% of asymptomatic adult MZ co-twins compared with 18% of DZ co-twins of patients with apparently sporadic disease and an average age of 58 [10]. This suggests that a genetic predisposition to PD may well be present in all sporadic cases though additional environmental factors may act to trigger expression. (c) What are the mechanisms of cell death in PD ? It is still being debated whether Lewy bodies per se are cytotoxic or protective to neurones undergoing degeneration. There is strong evidence for increased oxidative stress in PD substantia nigra as glutathione levels are reduced by 30-40%, Mn" SOD activity is raised, FeIII is increased but ferritin levels are normal, lipid peroxidation is raised, and mitochondrial complex 1 activity is 40% down [11,12]. Additionally, reduced striatal dopamine results in overactivity of subthalamic-pallidal glutamate projections which potentially leads to excessive activation of nitric oxide synthase and microglia with resultant cytokine release. Microglial activation in the nigra and pallidum has now been imaged in vivo in PD with "C-PK11195 PET [13]. Increased Ca11 transport into cells and activation of the caspase and stress kinase cascades then results potentially inducing cell apoptosis. A number of neuroprotective strategies have consequently evolved to try and modify PD progression - see Table 5. (d) Can we objectively measure PD progression ? Subjective rating scales such as the Unified Parkinson's Disease Rating Scale (UPDRS) have been validated for assessing the clinical severity of PD and its rate of progression. Difficulties with the UPDRS are that it is subjective, non-linear, and tends to overemphasise bradykinesia. Additionally, as PD patients are usually started on symptomatic agents within a year of diagnosis these can effectively mask disease progression.

Table 5: Neuroprotective strategies

Disease modifying agents Glutamate release inhibitors and receptor blockers Free radical scavengers Anti-oxidants Complex 1 stimulators NOS inhibitors Ca11 channel blockers Anti-inflammatory agents: propentofylline, pentoxifylline Anti-apoptotic agents Desmethyl selegeline (DSM) and CyA upregulate synthesis of GSH, SOD1, bcl-2 and bclXL-

Benzodiazepines, DSM, and CyA stabilise mitochondrial DYM and maintain closure of the mitochondrial permeability transition pore. Nerve growth factors (GDNF,BDNF,CNTF) and immunophillins block release of apoptosis inducing factors and activate the MAPK pathway. Caspase, ICE, protein kinase, GAPDH, and c-jun inhibitors.

D.J. Brooks / Movement Disorders: An Overview

9

Washout can be employed to try and dissociate treatment effects from disease progression but patients have difficulty tolerating this and the correct length of washout is still unclear. Timed motor tests are more objective than the UPDRS but in practice are often less sensitive to disease progression [14]. A final problem with clinical assessments of neuroprotective efficacy is defining a suitable primary endpoint. It is unclear whether time to a define absolute or percentage change in UPDRS, a life event, or requirement for a change in treatment is the most appropriate endpoint. In order to overcome some of these difficulties the use of a biological marker of disease progression such as imaging dopamine terminal function has been advocated. 18Fdopa PET is a marker of terminal dopa decarboxylase activity while PET and SPECT tropane tracers such as 123I-ß-CIT, I23I-FP-CIT, and 18F-CFT, measure dopamine transporter (DAT) binding. l8F-dopa and 18F-CFT PET and 123I-ß-CIT SPECT have all now been validated for following loss of dopaminergic function in PD and have shown that in early disease there is around a 10% per annum loss of putamen terminal acticity [15-17]. The potential is, therefore, there to use these modalities to determine whether putative neuroprotective agents can slow the course of PD. (e) What determines predisposition to dementia and depression in PD ? It is well recorded that around 20% of patients with PD develop frank dementia and this appears to be related to direct cortical Lewy body involvement and/or coincidental Alzheimer pathology [18]. Additionally, there is a loss of mesial-frontal dopamine and nucleus basalis - cortical cholinergic projections in PD. The question arises as to whether one can identify in advance the at-risk PD population for dementia. Clearly, effective neuroprotection would ideally halt dementing as well as locomotor aspects of PD. With the advent of high-resolution 3D cameras, 18F-dopa PET is now capable of imaging cingulate and prefrontal along with striatal dopaminergic function. A recent series has demonstrated that, if groups of demented and non-demented PD patients with equivalent disability are compared, the former show a significant loss of cingulate and prefrontal 18F-dopa uptake [19]. 18FDG PET is a marker of brain glucose utilisation (rCMRGlc). In Alzheimer's disease there is a characteristic pattern of loss of resting rCMRGlc involving parieto-temporal association areas. Interestingly, about a third of nondemented PD patients also show such a pattern of loss suggesting that this population may be at particular risk of developing dementia in years to come [20]. Around 30% of PD patients are also prone to develop depression requiring treatment though this differs from endogenous depression in that guilt and suicidal ideation are uncommon aspects. Recent surveys have suggested that the presence or absence of depression is a principle determinant of quality of life in PD [21]. It is known that PD is associated with Lewy body degeneration of serotonin cell bodies in the median raphe but what is unclear is whether serotonin cell loss correlates with depressive features of the disorder. "C-WAY100635 PET is a marker of serotonin HT1a sites which are found presynaptically on serotonin cell bodies in the median raphe and post-synaptically on cortical pyramidal neurones. Patients with PD, whether depressed or non-depressed show a 20% reduction in median raphe uptake of "C-WAY100635 [22]. In contrast, there is only significantly reduced cortical binding of "C-WAY100635 in PD patients with a current or prior history of depression. This finding suggests that depression in PD requires both a loss of serotonin projections along with altered cortical responsiveness to this monoamine. (f) How should we treat de novo PD ? Levodopa remains the most effective symptomatic treatment for PD but has a number of obvious disadvantages. It is a prodrug with low extraction influenced by levels of other

10

D.J. Brooks / Movement Disorders: An Overview

aminoacids and a short half life of 90-120'. It has to be administered with a peripheral dopa decarboxylase blocker to increase bioavailability and reduce nausea and hypotension. Chronic exposure to pulsatile plasma levels is now thought to prime patients to fluctuating motor responses and dyskinesias and, additionally, use of high doses of levodopa may well increase levels of oxidative stress in the PD brain [11]. In an attempt to smooth the plasma profile of levodopa de novo PD patients have been treated with controlled release preparations but this has not proved an effective strategy in practice for delaying motor complications [23]. As a consequence there has been a move to introduce dopamine agonists as de novo treatment. These agents have a longer half-life than levodopa, do not require transport by the amino acid carrier, act directly primarily on D2type receptors, have anti-oxidant and free radical scavenging properties, and reduce subthalamic glutamate output [11]. It can, therefore, be argued that agonists should both act to delay onset of motor complications and possibly have a neuroprotective action. Trials in early PD have now established that initial use of either the ergot-based agonists bromocriptine and cabergoline, or the non-ergot agonists ropinirole and pramipexole, as monotherapy can delay onset of fluctuating motor responses and dyskinesias [24-28]. Their efficacy, however, has not proved as strong as levodopa though patient groups were adequately treated in all these studies. To date, there is no hard evidence that agonist use is neuroprotective though functional imaging studies are running to test this hypothesis. (g) What are the mechanisms underlying L-dopa associated complications ? As mentioned earlier, it is now thought that chronic exposure to pulsatile levels of levodopa primes PD patients to develop fluctating motor responses and dyskinesias. The mechanism of this priming remains uncertain but studies using animal lesion models of PD have thrown light and some possible mechanisms. Rats exposed to pulsatile levodopa develop hyperphosphorylation of their NMDA receptors affecting glutamate transmission in the striatum and pallidum [29]. This is avoided by the use of continuous dopaminergic stimulation and the NMDA antagonist amantadine has also proven to be an effective antidyskinetic agent [30]. Another pharmacological abnormality associated with onset of dyskinesias in primed non-human primates is excessively raised levels of basal ganglia enkephalin and dynorphin [31]. This can indirectly be shown to occur in PD as PET studies on dyskinetic cases have shown reduced putamen "C-diprenorphine binding due to increased opioid receptor occupancy by endogenous opioids [32]. Interestingly, the opioid antagonist naloxone has been shown to have weak anti-dyskinetic activity [33]. One intriguing possible future means of tackling dyskinesias may be the use of dopamine partial agonists. These agents act to normalise dopamine tone in the basal ganglia and effectively reverse parkinsonism where dopamine deficiency is present [34]. The hope is that they will also act as anti-dyskinetic agents by becoming antagonists when excessive dopaminergic stimulation is present. (h) Can we develop effective treatments for autonomic failure, bulbar dysfunction, postural instability, and dementia in PD ? Current medical therapy for PD is effective for reversing limb bradykinesia and rigidity and. to some degree, tremor, but is largely ineffective in treating the axial symptoms of dysarthria, dysphagia. and gait instability. Acetylcholinesterase inhibitors have shown some short term efficacy in relieving confusion and dementia in PD. Bladder instability responds in part to anticholinergic agents though these may worsen dementia and constipation, the lack of effective medication for these aspects of PD is a major problem as they sare often the prime determinants of quality of life in this condition.

D.J. Brooks /Movement Disorders: An Overview

11

What is the role of stereotactic surgery in PD ? There is now clear evidence that bilateral pallidal and subthalamic (STN) high frequency electrical stimulation are both effective in relieving PD symptomatology [35]. The mechanisms by which deep brain stimulation (DBS) works remain uncertain but probably involves depolarising conduction block of the output from the target nuclei. Pallidal stimulation appears to be most effective for relieving dyskinesias while STN stimulation effectively relieves parkinsonism and allows reduction of levodopa dosage so reducing dyskinesias. In contrast to medication, DBS also appears to help postural instability . An intriguing observation is that STN stimulation appears to protect against tosic nigral damage in the rat, possibly by reducing glutamate output [36]. If DBS is truly neuroprotective one would like to employ it early rather than late, however, the small but significant morbidity associated with this procedure and cost considerations are likely to restrict its use to end-stage cases. Stereotactic approaches also allow infusion of nerve growth factors and implantation of dopaminergic cell lines into target areas. There is now a considerable body of evidence that implantation of human fetal mesencephalic cells into putamen can improve parkinsonism, the degree of locomotor improvement correlating with increases in putamen 18 F-dopa uptake [37]. Transplantation appears to work most effectively in younger patients and improves rigidity, bradykinesia, and gait. It also allows reduction of levodopa requirements by an average of 50%. A possible complication of transplantation that is now emerging is the development of "off dyskinesias which in some patients have proved disabling. It is likely that, in the future, human stem cells or cells to release dopamine or growth factors rather than human fetal mesencephalic cells will be used for implantation procedures to increase supply. Xenografts are also being explored as a further possibility.

Huntington's disease This is a dominantly transmitted disorder with full penetrance which arises when the CAG repeat expansion in the IT15 locus on chromosome 4 exceeds 38 repeats [38]. The IT15 gene produces the protein huntingtin, its mRNA being expressed in all cells [39], and the CAG repeat expension results in an excess of terminal glutamine aminoacids. The pathology of Huntington's disease (HD) is marked striatal atrophy with formation of cytoplasmic and nuclear inclusions [40,41]. Lesser volume loss is seen in pallidum, subthalamus, thalamus, brainstem, and cortex. Low levels of striatal GABA, glutamic acid decarboxylase, dopamine Dl and D2 sites are found while GABA, enkephalin, and dynorphin levels are reduced in the pallidum. The classical clinical presentation of HD is distal limb chorea spreading proximally with late rigidity and akinesia. Depression, dementia, and psychosis are also accompanying features. HD provides an ideal model system for testing putative non-toxic neuroprotective agents as there are currently no effective treatments for this condtion though recently partial dopamine agonists have demonstrated anti-choreic action [42]. Questions that remain unanswered in HD and other CAG repeat disorders such as the spinocerebellar ataxias (SCA syndromes and DRPLA) include: (1) How are the relevant expanded proteins cytotoxic? (2) Are the cytoplasmic and nuclear aggregates that form cytotoxic or protective? There is now some evidence from both human pathology and transgenic mouse models of CAG repeat disorders that nuclear aggregate formation is not required for cell loss and that it is likely to be the truncated fragments of the polyglutamine expanded proteins that lead to cytotoxicity [43]. Having said that, the mechanism of this cytotoxicity still remains uncertain though may involve activation of the caspase cascade and induction of apoptosis. Recently, attempts to alleviate HD by implantation of fetal striatal eminence

12

DJ. Brooks / Movement Disorders: An Overview

cells into striatum have been attempted [44,45]. Anecdotal reports have suggested some early patients respond symptomatically to this procedure in the short term but the long term role of transplantation in HD remains unclear. Conclusions Movement disorders is an exciting area for development of pharmacotherapy. In PD effective therapies are urgently needed that are non-dyskinesogenic and avoid the side effects of confusion and psychosis as are treatments that tackle the associzated dementia, depression, and autonomic problems. Effective treatments for tremor, dystonia, and Huntington's disease are also lacking. PD and HD also provide ideal model subcortical degenerations for trials of putative neuroprotective agents as there are now validated clincal rating scales for following their progression while functional imaging approaches provide sensitive surrogate biological markers of nigral and striatal dysfunction. References [1] [2] [3] [4j [5]

[6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] 118] [ 19]

D.J. Brooks, Parkinson's disease - a single clinical entity ? Quart.J.Med. 88 (1995):81-91. A.J. Hughes et al.. The accuracy of the clinical diagnosis of Parkinson's disease: a climcopathological study of 100 cases. J.Neurol.Neurosurg.Psychiatr. 55 (1992): 181-184. A.H. Rajput e.al., Accuracy of clinical diagnosis in Parkinsonism - a prospective study. Can.J.Neurol.Sci. 18 (1991) 275-278. M.H. Polymeropoulos et al.. Mutation in the alpha-synuclein gene identified in families with Parkinson's disease. Science 276 (1997) 2045-2047. N. Hattori et al., Molecular genetic analysis of a novel Parkin gene in Japanese families with autosomal recessive juvenile parkinsonism: Evidence for variable homozygous deletions in the Parkin gene in affected individuals. Ann.Neurol. 44 (1998) 935-941. C.B. Lucking et al.. Association between early-onset Parkinson's disease and mutations in the parkin gene. New England Journal of Medicine 342 (2000)1560-1567. Golbe L.I. The genetics of Parkinson's disease: A reconsideration. Neurology 40 Supplement 3 (1990) 7-16. P. Piccini et al., Dopaminergic function in familial Parkinson's disease: A clinical and l8F-dopa PET study. Ann.Neurol. 41 (1997) 222-229. C.M. Tanner et al., Parkinson disease in twins - an etiologic study. JAMA -Journal of the American Medical Association 281 (1999) 341-346. P. Piccini et al.. The role of inheritance in sporadic Parkinson's disease: Evidence from a longitudinal study of dopaminergic function in twins. Ann.Neurol. 45 (1999) 577-582. C.W. Olanow et al., Dopamine agonists and neuroprotection in Parkinson's disease. Ann.Neurol. 44 Suppl (1998)S167-S174. P. Jenner, and C.W. Olanow, Understanding cell death in Parkinson's disease. Annals of Neurology 44 (1998)S72-S84. R. Banati et al.. Imaging microglial activation in idiopathic Parkinson's disease. Movement Disorders 15 Supp 3 (2000) 216 (abstr). P.K. Morrish et al.. An [l8F]dopa PET and clinical study of the rate of progression in Parkinson's disease. Brain 119 (1996) 585-591. P.K. Morrish et al.. Measuring the rate of progression and estimating the preclinical period of Parkinson's disease with [l8F]dopa PET. J.Neurol.Neurosurg.Psychiat. 64(1998) 314-319. E.M. Nurmi et al.. The rate of progression in Parkinson's disease: A [l8F]Dopa PET study. Neurology 52 Supp 2 (1999) A91 (abstr). K.L. Marek et al.. Assessment of Parkinson's disease progression with b-CIT and SPECT imaging. Movement Disorders 13 Supp 2 (1998) 238-240. I. McKeith et al.. Neuroleptic sensitivity in patients with senile dementia of Lewy body type. BMJ. 305 (1992)673-678. A.S. Nagano et al.. The MRI-aided spatial normalization of F-dopa Ki images enable us to distinguish the DLB from PD on images. Journal of Nuclear Medicine 40 (1999) 1 179 (abstr).

D.J. Brooks

[20]

[21] [22]

[23] [24] [25]

[26] [27]

[28] [29] [30] [31] [32] [33] [34] [35]

[36] [37] [38] [39] [40] [41] [42] [43] [44] [45]

Disorders: An Overview

13

M.T.M. Hu, Taylor-Robinson SD, Chaudhuri KR et al. Cortical dysfunction in non-demented Parkinson's disease patients: A combined 31Phosphorus MRS and 18FDG PET study. Brain 123 (2000) 340-352. A. Schrag et al., What contributes to quality of life in patients With Parkinson's disease? Journal of Neurology Neurosurgery and Psychiatry 69 (2000) 308-312. M. Doder et al., Brain serotonin HT1A receptors in Parkinson's disease with and without depression measured by positron emission tomography and 11C-WAY100635. Movement Disorders 15 Supp 3 (2000)213(abstr). W.C. Koller et al., Immediate-release and controlled-release carbidopa/levodopa in PD - a 5-year randomized multicenter study. Neurology 53 (1999) 1012-1019. D.J. Brooks, Dopamine agonists - their role in the treatment of Parkinson's disease. J. Neurol. Neurosurg. Psychiat. 68 (2000) 685-689. J.L. Montastruc et al., A randomised controlled study comparing bromocriptine to which levodopa was later added, with levodopa alone in previously untreated patients with Parkinson's disease: a five year follow-up. J.Neurol.Neurosurg.Psychiat. 57 (1994)1034-1038. O. Rascol et al., Dyskinesia in Parkinson's Disease: A 5-Year Study of ropinirole versus levodopa. New Eng.J Med. 342 (2000) 1484-1491. U.K. Rinne et al., Early treatment of Parkinson's disease with cabergoline delays the onset of motor complications. Results of a double-blind levodopa controlled trial. The PKDS009 study group. Drugs 55(1998)23-30. R. Holloway et al., Pramipexole vs levodopa as initial treatment for Parkinson disease - a randomized controlled trial. Journal of the American Medical Association 284 (2000) 1931-1938. T.N. Chase, and J.D. Oh. Striatal dopamine- and glutamate-mediated dysregulation in experimental parkinsonism. Trends in Neurosciences 23 (2000) S86-S91. L.V. Metman et al.. Amantadine for levodopa-induced dyskinesias - a 1-year follow- up study. Archives of Neurology 56 (1999) 1383-1386. J. Jolkkonen et al., L-dopa reverses altered gene expression of substance P but not enkephalin in the caudate-putamen of common marmosets treated with MPTP. Mol. Brain Res. 32 (1995) 297-307. P. Piccini et al., Opioid receptor binding in Parkinson's patients with and without levodopa-induced dyskinesias. Ann.Neurol. 42 (1997) 720-726. B. Henry and J.M. Brotchie, Potential of opioid antagonists in the treatment of levodopa-induced dyskinesias in Parkinson's disease. Drugs and Aging. 9 (1996) 149-158. A. Ekesbo et al., Motor effects of (-)-OSU6162 in primates with unilateral 6-hydroxydopamine lesions. European Journal of Pharmacology 389 (2000) 193-199. M. Hallett and I. Litvan, Evaluation of surgery for Parkinson's disease - a report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Neurology 53 (1999) 1910-1921. B. Piallat et al., Neuroprotective effect of chronic inactivation of the subthalamic nucleus in a rat model of Parkinson's disease. Journal of Neural Transmission-Supplement (1999) 71-77. O. Lindvall, Cerebral implantation in movement disorders: State of the art. Movement Disorders 14 (1999)201-205. The Huntington's disease collaborative research group. A novel gene containing a trinucleotide repeat that is expanded and unstable on Huntington's disease chromosomes. Cell 72 (1993) 971-983. E. Sapp et al., Huntingtin localization in brains of normal and Huntington's disease patients. Annals of Neurology 42 (1997) 604-612. C.M. Kosinski et al., Intranuclear inclusions in subtypes of striatal neurons in Huntington's disease transgenic mice. Neuroreport 10 (1999) 3891-3896. J.B. Martin and J.F. Gusella, Huntington's disease: pathogenesis and management. New.Engl.J.Med. 315(1986) 1267-1276. J. Tedroff et al., Long-lasting improvement following (-)-OSU6162 in a patient With Huntington's disease. Neurology 53 (1999) 1605-1606. D.P. Huynh et al., Nuclear localization or inclusion body formation of ataxin-2 are not necessary for SCA2 pathogenesis in mouse or human. Nature Genetics 26 (2000) 44-50. A.C. Bachoud-Levi et al., Safety and tolerability assessment of intrastriatal neural allografts in five patients with Huntington's disease. Experimental Neurology 161 (2000) 194-202. V. Gaura et al., Follow-up with positron emission tomography of Huntington's disease patients grafted with fetal neuronal cells. Neurology 54 (2000) A153.

This page intentionally left blank

PartI Parkinson's Disease Symptoms, Models and Treatment Chair: L. Turski

This page intentionally left blank

Parkinson's Disease E. Ronken and G.J.M. van Scharrenburg (Eds.) IOS Press, 2002

17

Neurochemical Changes in Parkinson's Disease David G. Standaert, MD., PhD Department of Neurology, Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114, USA Abstract. The primary neurochemical abnormality in the brain in Parkinson's disease, the deficiency of dopamine, is well established and is the basis for most current therapies for the disease. The loss of dopamine, however, induces a complex set of changes in the neurochemistry and function of the basal ganglia network, affecting glutamatergic, GABAergic, cholinergic and other types of synapses. Some of these changes can be accounted for by the existing models of basal ganglia function. These models have proven quite useful in understanding the effects of drugs and lesions on the activity of the basal ganglia, but clearly are not adequate to explain some of the most troubling symptoms of Parkinson's disease, particularly the long-term consequences of dopaminergic therapy: dyskinesias and wearing off effects. Recent work has demonstrated that the basis of these adverse effects of dopaminergic drugs may be modifications in glutamatergic neurotransmission, particularly the N-methyl-D-aspartate class of glutamate receptors. These receptors may represent an important new target for therapeutics in Parkinson's disease. Keywords: neuroanatomy, basal ganglia, glutamate, NMDA.

1. Parkinson's Disease Parkinson's disease is a common and disabling neurologic disorder. It affects more than 3% of persons over the age of 65. Currently, about 500,000 people in the United States have Parkinson's disease, and this number is expected to increase as the population ages. The symptoms of Parkinson's disease, which include tremor, rigidity, slowness of movement and loss of postural balance leading to falls, result from loss of dopaminergic neurons in the substantia nigra pars compacta which innervate the striatum (caudate and putamen). The cause of the cell death in the substantia nigra is unknown. It is a disease of aging, in that few cases occur under the age of 40 while the frequency increases rapidly over the age of 60. A few families have been identified with autosomal dominant or recessive forms of parkinsonism, but in the vast majority of cases there does not appear to be a clear pattern of inheritance or a single gene defect. There is also evidence for the participation of environmental agents in the pathogenesis of Parkinson's disease, but no single agent accounts for the majority of the risk. At present, the most effective therapy for the disease is replacement of dopamine, using either the biosynthetic precursor of dopamine Ldihydroxyphenylalanine (L-DOPA, levodopa) or direct agonists of dopamine receptors. Although helpful, these treatments are not entirely satisfactory. Virtually all patients treated with levodopa for extended periods develop motor complications of dopaminergic therapy: "wearing off," the abrupt loss of effectiveness of the medication at the end of the dosing interval, and "dyskinesias," abnormal involuntary choreiform movements. These symptoms frequently leave patients disabled for a large portion of their waking hours [50,51]. There is an urgent need for better treatments for this common and disabling disease.

18

D.G. Standaert /Neurochemical Changes in Parkinson's Disease

2. The Neural Basis of Parkinson's Disease The last two decades have seen remarkable advances in the understanding of the neural basis of Parkinson's disease. A key element has been elucidation of the basic anatomical connectivity of the motor circuits of the brain. To this has been added a great deal of information on the neurochemistry of the structures, and the regulation of genes expressed. More recently, it has become possible to study the activity of the many neurons involved in parallel, and examine the interrelationships between groups of neurons. This work has lead to models which are the framework for much of the current experimental work on the basal ganglia. In addition, these studies have revealed new opportunities for treating human disorders of movement. They have lead to the revitalization of surgical therapies for Parkinson's disease. In addition, they have revealed that transmitter systems other than dopamine, in particular the NMDA glutamate receptors, may be important targets of therapeutic efforts. 3. Models of Basal Ganglia Function Models of neural circuit function are necessary for hypothesis driven investigations of the basal ganglia. The existing models of basal ganglia circuits have been criticized because they are imperfect. This is certainly true, and ongoing investigations have revealed many discrepancies between behaviors predicted by the models and experimental observation. It is important to recognize that these disparities not as failures of the model, but rather opportunities for improvement. Current conceptions of the functional organization of the basal ganglia and their role in the genesis of the symptoms of movement disorders have their roots in the models proposed by Albin, Penney and Young [2] and Alexander, Crutcher and Delong [3,4]. They are built around two central concepts. First, that there are several parallel re-entrant circuits, conveying information from different regions of the cortex, through the basal ganglia and then via the thalamus back to the cortex. Second, that within each of these circuits, there are parallel pathways for information flow, and the balance of activity in these pathways is crucial for regulating movement. The motor circuit is most directly related to the pathogenesis of parkinsonism. The principal components of the motor circuit are: a) b) c)

the primary input structure, the striatum (caudate and putamen). the output structures, the GPi and SNpr, These can be considered a single structure divided during development by the internal capsule. The intrinsic nuclei, the GPe, the STN, and the SNpc.

These interconnnected nuclei modulate the flow of motor information from many regions of the cerebral cortex through the ventral thalamus and back to cortical primary motor and motor association areas (figure 1).

D.G. Standaert/Neurochemical Changes in Parkinson's Disease

19

Figure 1. Basal Ganglia motor circuit, as proposed in 1989. The cerebral cortex provides excitatory glutamatergic input to the striatum. Within the striatum are projections bearing dopamine Dl and/or D2 receptors, as well as several types of intemeurons: cholinergic (ACh); somatostatin/nNOS (SS); and GABAergic. The projection neurons provide GABAergic output to the GPe and the GPi/SNpr.

4. The Striatum and Afferent Input to the Basal Ganglia The striatum is the principal input structure of the basal ganglia [29]. It receives excitatory glutamatergic afferent input from the neocortex as well as a smaller excitatory input from the intralaminar nuclei of the thalamus. It also receives a dense dopaminergic input from the substantia nigra pars compacta (SNpc). Although the striatum has a relatively homogeneous appearance in routine histological preparations, it in fact has a complex internal organization. Two principal classes of cells are present in the striatum. The most numerous are the projection neurons, which make up about 90% of the striatal neurons. These medium sized (15-20 µm in diameter) cells are GABAergic, have dendrites which are densely studded with spines, and send axons to structures outside the striatum. The remaining 10% of striatal neurons are composed of several types of intemeurons, including large aspiny cholinergic cells, medium aspiny cells containing somatostatin and neuronal nitric oxide synthase (nNOS), and several types of medium aspiny cells containing GABA and calcium binding proteins [20]. The bulk of the excitatory input to the cortex is targeted to the projection neurons. These inputs terminate on the heads of the dendritic spines. The dopaminergic inputs terminate principally on the shafts of the spines, where they may exert a strong modulatory action on the cortical inputs [47]. The projection neurons of the striatum have two principal targets: the globus pallidus externa (GPe) and the Gpi/SNpr. Both projections are GABAergic and inhibitory. Although the neurons which project to these two targets are morphologically indistinguishable, they have distinct neurochemical characteristics: the neurons projecting to the GPi/SNpr have a preponderance of dopamine Dl receptors and express the neuropeptide Substance P, while the neurons projecting to the GPe have a preponderance of dopamine D2 receptors, and express the neuropeptide enkephalin [15].

D.G. Standaert /Neurochemical Changes in Parkinson's Disease

5. Convergence and Divergence Afferent input to the striatum is divergent, in that a single region of motor cortex usually innervates multiple contiguous but non-overlapping regions within the striatum [13]. Divergence is also apparent at the cellular level. Within the volume encompassed by the dendrites of one medium spiny striatal neuron, there are approximately 2840 neighboring medium spiny cells, and 380,000 corticospinal axons. A single corticospinal axon makes only 10-40 synapses with a single medium spiny neuron, which receives 5000 inputs from other axons [21]. Thus the influence of a single corticostriatal neuron on any individual striatal neuron is very small, and significant alterations in striatal output will be accomplished only by highly coordinated activation of the inputs. Striatal efferent flow is highly convergent. Small areas of the GPe and GPi/SNpr receive input from multiple regions of the striatum; at least in some cases these have been shown to correspond to the multiple targets of single cortical sites [13,16]. In the target nuclei, the striatofugal fibers form dense networks of "wooly fibers" around a relatively limited number of target neurons. Convergence is also evident from a numerical perspective, in that there are estimated to be about 31 million neurons in the squirrel monkey striatum, but only 400,000 target neurons in target nuclei [16]. 6. Direct and Indirect Pathways, and the Role of Dopamine The essential features of the motor circuit model proposed by Albin, Penney and Young [2] are illustrated in Figure 1. The most important conceptual feature is the identification of two distinct pathways between the striatum and the Gpi/SNpr. The direct pathway consists of GABAergic striatal projection neurons forming inhibitory synapses directly with target neurons in the GPi and SNpr. The striatal neurons giving rise to the indirect pathway express predominantly Substance P. The indirect pathway was proposed to consist of a three step link: a) the GABAergic projection from striatal neurons to the GPe, arising from neurons expressing predominantly dopamine D2 receptors; b) an inhibitory GABAergic projection from the GPe to the STN, and c) an excitatory, glutamatergic projection from the STN to the Gpi/SNpr. This model proposes that the net output from the basal ganglia (that is, the extent of inhibitory output from GPI/SNpr to the ventral thalamus) is regulated by the relative balance of activity in the direct and indirect pathways. The direct pathway is linked to dopamine Dl receptors. Since these activate the synthesis of cAMP, they can be construed as "excitatory," and their activation tends to increase the inhibitory input to the Gpi/SNpr. The striatal neurons giving rise to the indirect pathway are regulated by "inhibitory" D2. Stimulation of these D2 receptors is predicted to lead, through a double inhibitory pathway, to a reduction in the excitatory drive from the STN to the GPi/SNpr.

D.G. Standaert /Neurochemical Changes in Parkinson's Disease

21

7. Models and the Pathogenesis of Movement Disorders The existing model can be used to account for some of the features of clinically important movement disorders [2] [11]:

Figure 2. The motor circuit in Parkinson's disease. Structures activated are in black, while structures with reduced activity are indicated by diagonal shading. Loss of the dopaminergic input to the striatum results in increased activation of the direct pathway, and relative inhibition of the indirect pathway. Both effects tend to increase the inhibitory output of the GPi/SNpr.

a)

In Parkinson's disease (Figure 2), the principal defect is loss of dopaminergic neurons in the substantia nigra, leading to a loss of dopaminergic input to the striatum. This should have a differential effect on the direct and indirect pathways. The direct pathway, under the control of Dl receptors, will be less active, leading to reduced inhibition of the GPi and SNpr. At the same time, activity in the indirect pathway will be augmented, leading to increased excitatory input to the Gpi/SNpr. Together, these effects will produce strong inhibition of the ventral thalamus, and reduced excitation of motor cortex. A key feature of this scheme is that it emphasizes the role of the STN in maintaining the abnormal motor state, an observation that has fueled much subsequent laboratory and clinical research.

b)

Lesions of the STN are predicted to result in diminished excitation of the Gpi/SNpr and consequently markedly increased motor activity, in agreement with common clinical observation.

c)

In Huntington's disease (Figure 3), there is degeneration of striatal projection neurons. In typical adult onset disease, there is more severe impairment of the enkephalinergic neurons of the indirect pathway, and relative preservation of the neurons of the direct pathway. This leads to unopposed inhibition of GPi/SNpr, and increased motor output.

22

D.G. Standaert /Neurochemical Changes in Parkinson's Disease

8. Problems with the Existing Models

Figure 3. The motor circuit in Huntington's disease. Selective dysfunction of the projection neurons giving rise to the indirect pathway results in reduced inhibitory output from the GPi/SNpr.

In the ten years since they were first proposed, several flaws in the existing models have been identified [25,33,37]). From an anatomical perspective, the connections proposed are too simple. Recent studies have revealed that projections from the striatum and other structures are highly collateralized. In addition, the GPe is postulated as a simple relay, providing afferent input to the STN. Recent evidence suggests that this is a great oversimplification [38,39,48]. Finally, the models do not consider in much detail the role of brainstem structures such as the PPN, which are intimately linked with the forebrain structures of the basal ganglia [46]. From a physiological perspective, the "box and line" approach considers only the quantity of neural activity linking structures, rather than the pattern or coherence of such activity. Striatal neurons have intrinsic membrane oscillations ("up" and "down" states), dependent on potassium currents and correlated among different striatal neurons. These are not in themselves sufficient to excite cells to fire, but determine whether or not afferent input can excite the cells to threshold [41]. Using cross-correlation techniques. Bergman and colleagues have demonstrated that one of the principal abnormalities of the basal ganglia output structures in primate models of PD is not an increase in rate of firing, but rather an increase in the correlation of firing in different neurons in the nuclei [5]. The models have also fallen short in attempting to explain some clinical features of basal ganglia disease. In particular, the adverse effects of dopaminergic therapy, such as wearing off and dyskinesias, are not well accounted for by the current models.

D.G. Standaert/Neurochemical Changes in Parkinson's Disease

23

Figure 4. An updated model of basal ganglia circuitry, incorporating recent anatomical and physiological data.

9. Motor Complications of Dopaminergic Therapy Treatment with levodopa or other dopamine agonists is highly effective in early Parkinson's disease, but most patients eventually develop motor complications as a result of these treatments. These complications include wearing off, the abrupt loss of efficacy at the end of each dosing interval, and dyskinesias, involuntary and sometimes violent writhing movements. Wearing off and dyskinesias produce substantial disability, and frequently prevent effective therapy of the disease. Although wearing off and dyskinesias often appear related to the timing of medication doses, they are not simply a consequence of the pharmacokinetics of levodopa. Motor complications are virtually never observed early in the treatment of PD; they appear only after prolonged treatment, usually several years. Furthermore, individuals who do not have PD but receive levodopa for other indications do not develop motor complications. Experiments using controlled administration of dopaminergic drugs support these clinical observations [6,32]. From this work it is clear that motor complications are not simply a passive manifestation of pharmacokinetics, but rather are the result of actively induced changes in brain function. 10. NMDA Receptors and the Treatment of Parkinson's Disease Recently, glutamate receptors have emerged as an important therapeutic target in Parkinson's disease. Glutamate is the principal excitatory neurotransmitter in the mammalian brain. Of the several different classes of receptors which mediate the actions of glutamate, the N-methyl-D-aspartate (NMDA) type are of particular interest because they are involved in the long-term processes which underlie neural adaptation and memory. NMDA receptors are assembled from 12 distinct proteins from two gene families. In vitro, receptors composed of different combinations of subunits have distinct properties. The functions of the receptors are further regulated by differential trafficking and phosphorylation of the subunit proteins.

24

D.G. Standaert INeurochemical Changes in Parkinson's Disease

NMDA glutamate receptors play a particularly important role in the regulation of movement by the striatum. NMDA binding sites are very abundant in this brain region [1]. Striatal NMDA receptors are involved in the regulation of GABA, acetylcholine, neuropeptide and glutamate release, and NMDA activation causes striatal neurons to dephosphorylate the dopamine receptor associated protein DARRP-32 [8,10,14,17,30,45,49,56,57]. Direct injection of small amounts of NMDA agonists into the striatum causes behavioral activation with contralateral rotation, while bilateral injection causes parkinsonism [23,53]. NMDA antagonists also potentiate the effect of dopamine on striatal function in several animal model systems [19,22,31,36,52]. A variety of recently evidence indicates that: 1) NMDA antagonists are highly effective in attenuating motor complications of dopaminergic therapy in animal models of parkinsonism; and 2) the basis for this effect may be that long-term dopaminergic treatment modifies the properties of striatal NMDA receptors [12,28,34,35]. Recent work has revealed a potentially important molecular mechanism for NMDA receptor regulation: Dl receptor agonists produce a rapid and striking enhancement of NMDA-induced depolarization of striatal cells, while D2 agonists attenuate NMDA responses [9]. In vitro studies suggest that these alterations result from Dl-receptor driven phosphorylation of NMDA receptor subunits [7]. 10. Alterations in NMDA Receptor Subunits in Models of Parkinsonism: Abundance, Composition and Phosphorylation In a recently published study, we used biochemical methods to study the subunit abundance, composition, and the serine and tyrosine phosphorylation of NMDA receptor subunits present in the normal rat striatum, and determine how these properties are altered in the rat unilateral 6-OHDA model of parkinsonism and by chronic treatment with the nonselective dopamimetic agent L-DOPA. Since NMDA receptor subunit proteins were known to be present in both the cytoplasm as well as in association with the cell membrane of neurons [40] and trafficking of receptor proteins from the cytoplasm to the cell surface is a potentially important means of functional regulation [58], we analyzed separately the receptors present in total striatal homogenate and those associated with neuronal membranes. In rats with unilateral 6-OHDA lesions of the nigrostriatal pathway, we have found that there is a reduction in the abundance of NR1 and NR2B subunits in the membrane fractions of lesioned striatum, while the abundance of NR2A is not altered. Coimmunoprecipitation of receptor assemblies under non-denaturing conditions revealed that these alterations arose from a selective decrease in the number of membrane NMDA receptors composed of NR1/NR2B subunits. Interestingly, these changes were seen only in NMDA receptors present on striatal membranes and not receptors found in total striatal homogenate. This finding implies a redistribution of receptor subunits from the membrane to the cytoplasmic compartment. The phosphorylation state of the NMDA subunits was also modified (Figure 5): there was a decrease in the serine phosphorylation of NR1 at residues 890 and 896, while the tyrosine phosphorylation of NR2B but not NR2A was altered. Finally, chronic treatment with L-DOPA resulted in normalization of the abundance and composition of striatal NMDA receptors in the membrane fraction, but produces marked increases in the phosphorylation of NR1 at serine890, serine896 and serine897. and tyrosine phosphorylation of NR2A and NR2B subunits.

D.G. Standaert /Neurochemical Changes in Parkinson's Disease

25

Figure 5. Changes in NMDA receptor subunit phosphorylation in the rat 6-OHDA lesion model. Similar increases in tyrosine phosphorylation of NR2B subunit have been seen in other animal models of neural plasticity, including long term potentiation in the rat dentate gyrus [44] and taste learning in the insular cortex [43]. The phosphorylation of NR1 protein at all the three serine residues suggests that this dopaminergic treatment may involve the activation of both protein kinase A and protein kinase C pathways. A potential mechanism for this effect is suggested by the report of Snyder et al, [59] showing that agonists of the dopamine Dl receptor increased the phosphorylation of NR1 subunit by regulating a phosphoprotein (DARPP-32) that selectively inhibits the protein phosphatase-1. The alterations in NMDA subunit abundance, composition, and phosphorylation observed in the lesioned striata may have important effects on the functional properties of striatal receptor channels. In particular, the reduction in the proportion of NMDA receptors composed of NRland NR2B, relative to receptors containing NR1 and NR2A, would be expected to result in a population of receptors with high sensitivity for competitive NMDA antagonists, and reduced affinity for glutamate [24,26]. The relative enrichment of NR1/NR2A receptors would also be predicted to lead to a corresponding increase in receptors with fast deactivation kinetics, and reduced affinity of the receptors for the noncompetitive polyamine antagonists such as ifenprodil and haloperidol [27,55]. The precise role of phosphorylation in modulating the properties of NMDA channels is still unclear. Electrophysiological studies of spinal dorsal horn neurons have demonstrated that protein tyrosine kinases and protein tyrosine phosphatase inhibitors potentiate NMDA receptor currents [60,61]. Also, phosphorylation of NMDA receptors at serine and threonine residues has been suggested to regulate the subcellular redistribution and targeting of intracellular NMDA receptors to the synaptic membrane [18,42,54].

26

D.G. Standaert/Neurochemical Changes in Parkinson's Disease

12. Summary The existing models of basal ganglia function are imperfect, but they are nevertheless a valuable framework for hypothesis driven investigations of parkinsonism and motor circuits, and have lead to therapeutic advances. Problems with the models with respect to anatomical, physiological and clinical issues have been identified. One outcome of this work has been the recognitition of the important role of glutamate receptors, particulary those of the NMDA class, in regulating motor control. NMDA receptors may also make an important contribution to the pathogenesis of the adverse motor effects of dopaminergic therapy, an important source of disability. References [1] [2] [3] [4]

[5] [6] [7]

[8] [9]

[10] [11] [12] [13] [14]

[15] [16] [17] [18] [19] [20]

R.L. Albin el al. Excitatory amino acid binding sites in the basal ganglia of the rat: a quantitative autoradiographic study. Neuroscience, 46 (1992) 35-48. R.L. Albin el al.,. The functional anatomy of basal ganglia disorders. Trends in Neurosciences, 12 (1989)366-375. G.E. Alexander and M.D. Crutcher, Functional architecture of basal ganglia circuits: neural substrates of parallel processing, Trends in Neurosciences, 12 (1990)366-371. G.E. Alexander el al., Basal ganglia-thalamocortical circuits-parallel substrates for motor oculomotor 'prefrontal' and 'limbic' functions. In H.B.M. Uylings, C.G. Vaneden, J.P.C. Debruin. M.A. Corner and F.M.G.P. (Eds.), Prefrontal Cortex. Its Structure, Function, and Pathology, Vol 8S. Elsevier Science. Amsterdam (1990) pp.119-146. H. Bergman el al., Physiological aspects of information processing in the basal ganglia of normal and parkinsonian primates, Trends in Neurosciences, 21 (1998) 32-38. P.S. Blanchet et al. Modulation of levodopa-induced motor response complications by NMDA antagonists in Parkinson's disease, Neurosci Biobehav Rev, 21 (1997) 447-53. T. Blank et al. The phosphoprotein DARPP-32 mediates cAMP-dependent potentiation of striatal Nmethyl-D-aspartate responses. Proceedings of the National Academy of Sciences U.S.A.. 94 (1997) 14859-14864. O. Bustos et al. Regulation of excitatory amino acid release by N-methyl-D-aspartate receptors in rat striatum: HI vivo microdialysis studies, Brain Research 585 (1992) 105-15. C. Cepeda et al, Neuromodulatory actions of dopamine in the neostriatum are dependent upon the excitatory amino acid receptor subtypes activated. Proceedings of the National Academy of Sciences U.S.A., 90 (1993) 9576-9580. G. Damsma et al, Dopaminergic regulation of striatal acetylcholine release: importance of Dl and Nrnethyl-D-aspartate receptors, Journal of Pharmacolog\ and Experimental Therapeutics. 259 (1991) 1064-72. M. DeLong, Primate models of movement disorders of basal ganglia origin. Trends in Neurosciences. 13(1990)281-285. T.M. Engber et al., NMDA receptor blockade reverses motor response alterations induced by levodopa, Neuroreport, 5 (1994) 2586-2588. A.W. Flaherty and A.M. Grasybiel, Input-output organization of the sensorimotor striatum in the squirrel monkey, Journal of Neuroscience, 14 (1994) 599-610. T. Galli et al., Modulation of GABA release by a-amino-3-hydroxy-5-melhylisoxazole-4-propionate and N-methyl-D-aspartate receptors in matrix-enriched areas of the rat striatum. Neuroscience. 50 (1992)769-780. C.R. Gerfen, The neostriatal mosaic: multiple levels of compartmental organization. Trends in Neurosciences, 15(1992) 133-139. A.M. Graybiel et al., The basal ganglia and adaptive motor control, Science, 265(1994)1826-31. S. Halpain et al., Activation of NMDA receptors induces dephosphorylation of DARPP-32 in rat striatal slices. Nature 343 (1990) 369-372. C Hisatsune et al., Phosphorylation-dependenl regulation of N-methyl-D-aspartate receptors by calrnodulin. J Biol Chem. 272 (1997) 20805-20810. S. Kaur. And M.S. Starr, Differential effects of intrastriatal and intranigral injections of glutamate antagonists on motor behaviour in the reserpine-treated rat, Neuroscience, 76 (1997) 345-354. Y. Kawaguchi et al., Striatal interneurons: chemical, physiological, and morphological characteri/ation. Trends in Neurosciences. 18 (1995 ) 527-535.

D.G. Standaert/Neurochemical

Changes in Parkinson's Disease

27

A.E. Kincaid et al., Connectivity and convergence of single corticostriatal axons, J Neurosci, 18 (1998) 4722-4731. T. Klockgether and L. Turski, NMDA antagonists potentiate antiparkinsonian actions of L-dopa in monoamine-depleted rats, Annals of Neurology, 28 (1990) 539-546. T. Klockgether and L. Turski,, Toward an understanding of the role of glutamate in experimental Parkinsonism: agonist-sensitive sites in the basal ganglia. Annals of Neurology, 34 (1993) 585-593. D.J. Laurie and P.H. Seeburg, Ligand affinities at recombinant N-methyl-D-aspartate receptors depend on subunit composition, European Journal of Pharmacology, 268 (1994) 335-345. R. Levy et al., Re-evaluation of the functional anatomy of the basal ganglia in normal and Parkinsonian states, Neuroscience, 76 (1997) 335-343. D.R. Lynch et al., N-methyl-D-aspartate receptors - different subunit requirements for binding of glutamate antagonists, glycine antagonists, and channel-blocking agents, Molecular Pharmacology, 45 (1994)540-545. D.R. Lynch and M.J. Gallagher, Inhibition of N-methyl-D-aspartate receptors by haloperidol: developmental and pharmacological characterization in native and recombinant receptors, Journal of Pharmacology and Experimental Therapeutics, 279 (1996) 154-161. C. Marin et al., MK-801 prevents levodopa-induced motor response alterations in parkinsonian rats, Brain Res, 736 (1996) 202-205. J.W. Mink, The basal ganglia: focused selection and inhibition of competing motor programs, Prog Neurobiol, 50 (1996) 381-425. M. Morari et al., N-methyl-D-aspartic acid differentially regulates extracellular dopamine, GAB A, and glutamate levels in the dorsolateral neostriatum of the halothane-anesthetized rat: an in vivo microdialysis study, Journal of Neurochemistry, 60 (1993) 1884-1893. M. Morelli et al., Opposite effects of NMDA receptor blockade on dopaminergic Dl- and D2-mediated behavior in the 6-hydroxydopamine model of turning: relationship with c-fos expression, Journal of Pharmacology and Experimental Therapeutics, 260 (1992) 402-408. M.M. Mouradian et al., Modification of central dopaminergic mechanisms by continuous levodopa therapy for advanced Parkinson's disease, Annals of Neurology, 27 (1990) 18-23. J.A. Obeso et al., Basal ganglia pathophysiology. A critical review, Adv Neural, 74 (1997) 3-18. S.M. Papa et al, Reversal of levodopa-induced motor fluctuations in experimental parkinsonism by NMDA receptor blockade, Brain Research, 701 (1995) 13-18. S.M. Papa and T.N. Chase, Levodopa-induced dyskinesias improved by a glutamate antagonist in parkinsonian monkeys, Annals of Neurology, 39 (1996) 574-578. S.M. Papa et al., Opposite effects of NMDA and AMPA receptor blockade on catalepsy induced by dopamine receptor antagonists, European Journal of Pharmacology, 232 (1993) 247-253. A. Parent and F. Cicchetti, The current model of basal ganglia organization under scrutiny, Mov Disord, 13(1998) 199-202. A. Parent and L.N. Hazrati, Functional anatomy of the basal ganglia . 1. The cortico-basal gangliathalamo-cortical loop, Brain Research Reviews, 20 (1995) 91-127. A. Parent and L.N. Hazrati, Functional anatomy of the basal ganglia .2. The place of the subthalamic nucleus and external pallidum in basal ganglia ciruitry, Brain Research Reviews, 20 (1995) 128-154 R.S. Petralia et ai, Light and electron microscope distributions of the NMDA receptor subunit NMDAR1 in the rat nervous system using a selective anti-peptide antibody, Journal of Neuroscience, 14(1994)667-696. D. Plenz and S.T. Kitai, Up and down states in striatal medium spiny neurons simultaneously recorded with spontaneous activity in fast-spiking interneurons studied in corex-striatum-substantia nigra organotypic cultures, J Neurosci, 18 (1998) 266-283. I.M. Raman et al., Beta- adrenergic regulation of synaptic NMDA receptors by camp-dependent protein kinase, Neuron, 16 (1996) 415-421. K. Rosenblum et al., NMDA receptor and the tyrosine phosphorylation of its 2B subunit in taste learning in the rat insular cortex. Journal of Neuroscience, 17 (1997) 5129-5135. J.A. Rostas et al., Enhanced tyrosine phosphorylation of the 2B subunit of the N-methyl-fl- aspartate receptor in long-term potentiation, Proceedings of the National Academy of Sciences U.S.A., 93 (1996) 10452-10456. B.B. Ruzicka and K. Jhamandas, Met-enkephalin release from slices of the rat striatum and globus pallidus: stimulation by excitatory amino acids, Journal of Pharmacology and Experimental Therapeutics, 257(1991) 1025-1033. D.B. Rye et al., Pedunculopontine tegmental nucleus of the rat: cytoarchitecture, cytochemistry, and some extrapyramidal connections of the mesopontine tegmentum, J Comp Neurol, 259 (1987) 483-528. Y. Smith et al., Synaptic relationships between dopaminergic afferents and cortical or thalamic input in the sensorimotor territory of the striatum in monkey, J Comp Neurol, 344 (1994)1-19.

28

[48]

D.G. Standaert/Neurochemical

Changes in Parkinson's Disease

Y. Smith et al., Microcircuitry of the direct and indirect pathways of the basal ganglia, Neuroscience. 86(1998)353-387. [49] D.L. Somers and R.M. Beckstead,. N-methyl-D-aspartate receptor antagonism alters substance P and met5-enkephalin biosynthesis in neurons of the rat striatum. Journal of Pharmacology and Experimental Therapeutics. 262 (1992) 823-833. [50] D.O. Standaert and M.B. Stern, Update on the management of Parkinson's disease, Medical Clinics of North America, 77 (1993) 169-183. [51] D.G. Standaert and A.B. Young, Treatment of central nervous system degenerative diseases.. Goodman and Oilman's Pharmacological Basis of Therapeutics, 1995, pp. in press. [52] M.S. Starr and B.S. Starr, Facilitation of dopamine-Dl receptor-dependent but not dopamine DI/D2 receptor-dependent locomotion by glutamate antagonists in the reserpine-treated mouse. European Journal of Pharmacology, 250(1993) 239-246. [53] P.K. Thanos et al., N-methyl-D-aspartate unilaterally injected into the dorsal striatum of rats produces contralateral circling: antagonism by 2-amino-7-phosphonoheptanoic acid and cis-flupenthixol. Brain Research, 589(1992)55-61. [54] W.G. Tingley et al.. Characterization of protein kinase A and protein kinase C phosphorylation of the N-methyl-D-aspartate receptor NR1 subunit using phosphorylation site-specific antibodies. Journal of Biological Chemistry, 272 (1997) 5157-5166. [55] K. Williams et al.. Sensitivity of the N-methyl-D-aspartate receptor to polyamines is controlled by NR2 subunits, Molecular Pharmacology, 45 (1994) 803-809. [56] A.M.J. Young and H.F. Bradford, N-methyl-D-aspartate releases excitatory amino acids in rat corpus striatum in vivo. Journal of Neurochemistry. 56(1991) 1677-1783. [57] A.M.J. Young and H.F. Bradford, N-methyl-D-aspartate releases g-aminobutyric acid from rat striatum in vivo: a microdialysis study using a novel preloading method. Journal of Neurochemistry. 60 (1993) 487-492. [58] S.H. Shi et al.. Rapid spine delivery and redistribution of AMPA receptors after synaptic NMDA receptor activation, Science 284 (1999) 1755-1757. [59] G.L. Snyder et al.. A dopamine/DI receptor/proteinkinase A/dopamine- and cAMP-regulated phosphoprotein (Mr32K/Da)/protein phosphotase-1 pathway regulates dephosphorylation of the NMDA receptor. J. of Neuroscience 18 (1998) 10297-10303. [60] Y.T. Wang and M.W. Salter, Regulation of NMDA receptors by tyrosinekinases and phosphatases Nature 369 (1994) 233-235. [61 ] X.M. Yu et al., NMDA channel regulation by channel-associated protein tyrosine kinase. Src. Science 275(1997)674-678.

Parkinson's Disease E. Ronken and G.J.M. van Scharrenburg (Eds.) IOS Press, 2002

29

Psychosis and Depression in Parkinson's Disease D. Brandstadter and W.H. Oertel Department of Neurology, Philipps-University, Rudolf-Bultmannstrasse 8, D-35033 Marburg, Germany Abstract. Depression and drug-induced psychosis are frequently associated with Parkinson's disease (PD) and have a great impact on quality of life in patients with PD. Their recognition and treatment plays an important role in the management of PD. Drug-induced psychosis occurs in approximately 20-30% of the patients with PD. Vivid dreams and nightmares are thought to be early signs. Visual hallucinations represent the most prominent feature of psychosis in PD. Clozapine is the best studied and effective drug for psychosis in PD without worsening motor symptoms. Several newly developed atypical neuroleptics for the treatment of psychosis in PD such as quetiapine or olanzapine may represent an alternative to clozapine. The prevalence of depression in PD is approximately 40%. Tricylic antidepressants (TCAs) represent a traditionally available group of medication for the treatment of depression in PD and many studies, although of limited quality, have suggested the efficacy of TCAs. Recently introduced selective serotonin reuptake inhibitors (SSRIs) have equal antidepressant efficacy to TCAs, however, they may have a better side-effect profile. Increased motor disability during treatment with SSRIs has rarely been reported. This article reviews the characteristics and treatment options for drug-induced psychosis and depression in Parkinson's disease. Keywords: psychosis, visual hallucinations, depression, neuroleptics, antidepressants

Due to improved management of motor symptoms, psychiatric disturbances associated with Parkinson's disease (PD) have received increased focus. Drug-induced psychosis and depression represent two important nonmotor features of PD and may probably have a greater impact on quality of life in PD than central motor symptoms. The recognition and treatment of these features should therefore gain increased importance in the current management of PD. 1.

Psychosis

1.1 Characteristics Psychosis is a frequent side-effect of levodopa and dopaminomimetic treatment in Parkinson's disease (PD). Psychosis is defined as gross impairment in reality testing and the creation of a new reality [1]. The occurrence of sleep disruption, vivid dreams or nightmares are believed to be early signs of psychosis in PD. The most common features of psychosis in PD are visual hallucinations, in contrast to schizophrenia in which auditory hallucinations predominate. Hallucinations in PD are commonly differentiated into several types: Minor hallucinations and illusions, formed visual hallucinations, auditory hallucinations and tactile hallucinations. Minor hallucinations are characterised by the

30

D. Brandstadter and W.H. Oertel/Psychosis and Depression in Parkinson's Disease

presence or "sideaway passage" of persons or animals [2] . Illusions are characterised by a transformation of an object in another object or animal. Formed visual hallucinations consist of animals, persons or objects, which can typically be well described. Auditory and especially tactile hallucinations are less common in PD. The different types of hallucinations may occur together or alone. Psychotic patients in PD often present with delusions, which are characterised by unrealistic complex beliefs. Most common thoughts are spousal infidelity, persecution and jealousy [3, 4]. Psychosis in PD typically presents with a clear sensorium, whereas psychosis with a clouded sensorium characterised by disorientation and impaired concentration is defined as delirious state. 1.2 Prevalence and risk factors Due to variable definitions of psychosis in PD, prevalence rates in published studies range from 3-85% [5]. In the only population based study of psychotic symptoms in PD of Aarsland et al. [6], 15,8% of the patients had hallucinations or delusions one week prior examination. In a recent study of Fenelon and colleagues [3], including 216 parkinsonian patients, 39,8% of the patients showed hallucinations during the previous 3 months. However, the generally accepted prevalence of psychotic symptoms in PD is approximately 20-30% [3-5]. The main risk factors reported for the development of psychosis in PD are advanced age, long duration of the disease, vascular comorbidity, cognitive impairment and prior history of psychotic disorder or drug-induced psychosis and daytime somnolence [5. 7]. 1.3 Etiology The etiology of drug-induced psychosis in PD remains poorly understood. Abnormalities in dopamimetic and serotonergic systems have been discussed to explain the development of psychosis. In a post mortem study, Birkmeyer and Riederer [8] found elevated levels of 5hydroxytryptamine and noradrenaline in the caudal substantia nigra, raphe and red nucleus and globus pallidus of delusional parkinsonian patients. Goetz et al. [9] proposed that overstimulation of mesocorticolimbic dopamine receptors in analogy to motor dyskinesia may cause dysfunction of limbic areas. Using single-cell recordings from the ventral tegmental area (VTA), Svensson and colleagues [10] showed that burst firing in cells mainly located in the paranigral nucleus, a subdivision of the VTA projecting in limbic areas, was increased after the administration of phencyclidine. a schizophrenomimetic N-methyl-D-aspartate receptor antagonist. However. burst firing was reduced in the parabrachial pigmented nucleus of the VTA, projecting mainly in prefrontal areas. This reduction within the prefrontal dopaminergic projections could be responsible for a false interpretation of sensory inputs of the association cortices. The authors suggested that 5-HT2 receptor antagonists, such as clozapine, produce activation of the prefrontal dopaminergic neurons, which may be responsible for therapeutic effects. Sleep disturbances are frequent in parkinsonian patients with and without hallucinations. Moskovitz and colleagues [11] first described the association between hallucinations and sleep disturbances in PD. They postulated a "kindling" mechanism secondary to chronic levodopa treatment, which progresses from vivid dreams to hallucinations to delusion and finally to a confused state. Polysomnografic sleep measure in PD showed a lower efficiency in patients with hallucinations and reduced total REM sleep time, indicating strong association between lack of REM sleep time and hallucinations [12].

D. Brandstddter and W.H. Oertel / Psychosis and Depression in Parkinson's Disease

31

1.4 Treatment options The treatment of psychosis in PD consists of different steps. First, several general factors triggering psychosis in PD, e.g. infection, should be eliminated. In a further step antiparkinsonian drugs other than levodopa, e.g. anticholinergics, amantadine and dopamine agonists should be taken off. Finally, levodopa should be reduced. If motor symptoms get too worse after reduction of antiparkinsonian medications or psychotic symptoms can not be controlled, a therapy with an atypical neuroleptic should be initiated. So far the term "atypical neuroleptic" is not exactly defined [13]. Atypical neuroleptics are characterised by a failure of inducing catalepsy and not increasing serum prolactin levels in animal models [14]. Furthermore, a greater affinity of atypical neuroleptics to serotonin receptors resulting in a higher 5-HT 2A/ D2 ratio may be important for the low extrapyramidal side-effects. Clozapine was the first available neuroleptic with a failure of extrapyramidal side effects. However, because of the potential to induce agranulocytosis, clozapine was withdrawn for several years from Europe and the United States. The incidence of agranulocytosis is about 1% and is not dose-related. If detected early and clozapine is discontinued, agranulocytosis is reversible. Patients treated with clozapine must have a weekly blood cell count for the first 18 weeks and then monthly. In 1985, Scholz and Dichgans [15] reported the first psychotic patients in PD treated successfully with clozapine. Since then, there have been a large number of studies, case reports and recently two multicenter, double-blind placebo-controlled trials with clozapine, indicating the effective treatment of psychosis in PD without worsening parkinsonism [16-23]. However, the doses of clozapine given in PD were generally low compared to treatment of schizophrenia, ranging from 6.25 to 150 mg per day. Rapid dose elevation has been associated with side-effects such as severe orthostatic hypotension, tachycardia and sedation. Finally, clozapine can also be used as a last resort treatment of tremor in patients with PD [ 24, 25]. Risperidone is another neuroleptic which is thought to be atypical. There have been a few open label studies with risperidone in PD [26-28]. The authors used doses from 0.5-1.9 mg/day. The studies showed that risperidone may be an effective treatment for psychosis in PD, however, risperidone worsened parkinsonian symptoms in several patients. In animal models risperidone induces elevation of serum prolactin levels, indicating a potential of the drug for extrapyramidal side-effects [29]. Olanzapine is an atypical neuroleptic drug with greater affinity blocking 5-HT2A receptors compared to D2 receptors. However, in animal models olanzapine induces in higher, clinically not used doses, catalepsy and increased serum prolactin secretion [30, 31]. A few studies showed extrapyramidal side-effects in non parkinsonian patients [32]. Several open label studies have been published on psychotic parkinsonian patients. Some studies noted a beneficial effect of olanzapine in the treatment of psychotic symptoms in PD without extrapyramidal side-effects [33, 34], however, most of the studies indicated a worsening of motor symptoms [35-38]. A recent double-blind comparison trial of olanzapine and clozapine in hallucinating patients with PD documented significant declines in motor function in the olanzapine group compared to baseline function and to the patients receiving clozapine [39]. The authors of this study recommended to favour clozapine instead of olanzapine for the management of hallucinations in PD. Quetiapine is the newest atypical neuroleptic with closest pharmacologic resemblance to clozapine. Fortunately, quetiapine does not induce agranulocytosis like clozapine [40]. Quetiapine has greater affinity for 5-HT2A receptors compared to dopamine D2 receptors. In studies on patients with schizophrenia extrapyramidal side-effects and serum prolactin elevation have been indistinguishable from placebo [41]. Evatt and colleagues [42] reported

32

D. Brandstadter and W.H. Oertel / Psychosis and Depression in Parkinson's Disease

the first open-label study using 50 mg quetiapine per day in parkinsonian patients with psychosis. They showed a significant improvement of psychosis as measured by the Brief Psychiatric Rating Scale (BPRS). In a 12 months follow-up study with quetiapine the same group showed a mild decrement of motor function in the PD patients [43]. Further open label studies with quetiapine in PD indicated an improvement of psychosis without decline of motor function [44-46]. In a recent study Fernandez [47] tried to switch eight parkinsonian patients with clozapine and 3 with olanzapine to quetiapine. Only five patients (three on clozapine and two on olanzapine) tolerated the transition easily. The author concluded that a transition from another neuroleptic especially clozapine to quetiapine may be difficult. However, the author replaced clozapine by quetiapine adding 12,5-25mg quetiapine per day and reducing clozapine by the same amount. This replacement caused crossover failures resulting in confusion and hallucinations of the patients. A possible reason may be that quetiapine and clozapine were switched in equivalent doses. Based on our own clinical experience with these drugs, the equivalent dose of quetiapine is approximately three times the dose of clozapine. Odansetron, an expensive, selective serotonin receptor (5-HT3) antagonist has also been studied for its antipsychotic potential in PD [48, 49]. However, initially promising data could not been reproduced by our group [50]. In conclusion, based on the mentioned data above and our clinical experience risperidone, olanzapine and odansetron can so far not be recommended for the treatment of psychosis in PD. Quetiapine seems to be effective and well tolerated without the need of blood cell monitoring, however, placebo-controlled multicenter studies have to confirm the encouraging experience published so far in the literature. A subgroup of psychotic PD patients will require clozapine. According to the current literature, ciozapine is still the most effective antipsychotic drug without extrapyramidal side-effects for the treatment of psychosis in PD. 2.

Depression

2. /

Characteristics

Depression is the most common psychiatric symptom in patients with PD [51]. The main symptoms of depression as a major affective disorder in PD are: Loss of self-esteem, hopelessness, worthlessness, pessimism about the future and sadness (without guilt or self approach). Depressed parkinsonian patients often have suicidal ideation, however the suicide rate is low in these patients [52]. Furthermore anxiety symptoms and panic attacks are frequently seen in PD patients with depression and can often precede motor symptoms [53, 54]. In a study of Menza et al. [55] 67% of the patients with diagnosed depression had a comorbidity with an anxiety disorder. The authors suggested that anxiety and depression are related to neurochemical changes in PD and do not represent a psychologic reaction or a side-effect of levodopa. An important problem is the differential diagnosis between depression and dementia [56]. Depression and dementia have complex relation and depression may clinically present as dementia ("pseudo-dementia"). Population-based investigations proposed a frequent coexistence of depression and dementia in PD [57-59], suggesting a common underlying mechanism. Celesia and Wannamaker [60] found in a longitudinal study of parkinsonian patients evidence of depression in 37% of the patients. However, the prevalence of depression was not related to the stage of the illness. Depression was most common in stages I (38%), III (42%) and IV (50 %). but less frequent in stages II (18 %) and V (22%). A nearly similar pattern of results was published by Starkstein and colleagues [61]. The data from these two

D. Brandstadter and W.H. Oertel/Psychosis and Depression in Parkinson's Disease

33

studies failed to show a linear increase of depression with the severity of PD. In fact, there is growing evidence that different subgroups of parkinsonian patients may be more vulnerable to depression than others. Studies of Santamaria [62] and Starkstein [63] found a significant association between depression and duration of the illness in patients with early onset of PD, but not in those with later onset. Finally, these results support the assumption of a multifactorial model of depression in PD. 2.2 Prevalence and risk factors The prevalence rate for depression in different studies ranged from 4-70% due to different definitions of depression and assessment techniques [52]. Gotham and colleagues [64] estimated the mean prevalence of depression to be 46% and there is general agreement that the frequency of depression in PD is about 40% [54]. A relationship between the kind of motor deficits and the depressive syndromes has been frequently discussed [65-67]. Bradykinesia, rigidity and postural changes are suggested to be more prominent in PD patients with depression than in non-depressed patients [68, 69]. Other risk factors for depression in PD have been proposed, including early onset of parkinsonian symptoms, a positive family history and female gender [60, 62, 70]. 2.3 Etiology The etiology of depression in PD still remains unknown, however endogenous, reactive or both mechanisms are discussed [71, 72]. Degeneration of dopaminergic, noradrenergic, serotonergic and cholinergic nuclei have been described [51, 52, 54]. Especially changes in serotonergic transmitter pathways have been proposed to be associated with depression in PD. Several studies found decreased levels of the serotonin metabolite 5hydroxyindoleacetic acid (5-HIAA) in the cerebrospinal fluid (CSF) of depressed parkinsonian patients compared to parkinsonian patients without depression and normal controls [73, 74]. In a study of Mayeux et al. [74], including PD patients with depression and dementia a significant reduction in CSF 5-HIAA concentration in patients with dementia and or depression compared to those without dementia or depression was found. PD patients with depression and dementia had the lowest levels of 5-HIAA in the CSF, suggesting that dementia and depression may not only coexist but may share an common biological mechanism. Others, like Roos et al. [75 ] failed to replicate these results. Torrack and Morris [76] found severe degeneration of ventral tegmental dopamine neurons and proposed that a selective disruption of ventral tegmental afferents of the cortex may be responsible for the behavioural symptoms in PD. Data from functional imaging studies, using positron emission tomography (PET) showed relative hypometabolism involving the caudate and orbital-inferior area of the frontal lobe in PD patients with depression compared to patients without depression and normal controls [77, 78]. In addition the magnitude of hypometabolism correlated with the severity of depression. As dopaminergic projections from the ventral tegmental area show regional specifity for the orbitofrontal cortex, Mayberg and colleagues [79] proposed that a degeneration of the meso-cortico-limbic dopaminergic system in parkinsonian patients may secondarily cause a metabolic dysfunction in the orbitofrontal region of the cortex. 2.4 Treatment options Many studies showed an improvement of depressive syndroms in PD after controlling motor symptoms with antiparkinsonian drugs. Improvement of depression with levodopa and dopamine agonists has been described [80, 81]. These studies have often been

34

D. Brandstddter and W.H. Oertel/Psychosis and Depression in Parkinson's Disease

criticised, because of the possibility that improvement may be secondary to optimisation of motor symptoms. Despite this controversial debate, the published studies provide evidence of a beneficial role of dopamine replacement in the therapy of depression in PD [54]. Tricylic antidepressants (TCAs) represent a traditionally available group of medication for the treatment of depression in PD. There are several clinical studies with TCAs indicating an improvement of depression in PD. In a double-blind, placebocontrolled study (n=39) Laitinen et al. [82] showed the efficacy of 100 mg desipramine /day for the treatment of 39 depressed parkinsonian patients. In another double controlled trial with nortriptyline, Anderson and collegues [83] evaluated a significant benefit of nortriptyline at a dose of 150 mg/day on 22 patients with depression in PD. However. especially in the elderly the TCAs must be used with caution. The anticholinergic sideeffects of the TCAs can cause urinary retention, paralytic ileus, acute glaucoma, decrease of cognitive function and severe postural hypotension [52]. Furthermore, in cognitive impaired patients delusion can be induced. The more recently introduced selective serotonin reuptake inhibitors (SSRIs) have equal antidepressant efficacy to TCAs, however, they may have a favourable side-effect profile, particularly in the elderly [51]. So far conflicting data have been published regarding whether SSRIs worsen motor function in PD. Besides described efficacy of SSRIs in the treatment of depressed PD patients [84. 85], some investigators have reported increased motor disability after the use of fluoxetine [86], fluvoxamine [87] and paroxetine [88]. A recently published study with paroxetine suggested that paroxetine at the dosage of 20 mg/day does not aggravate parkinsonian symptoms, despite observing increased tremor in one patient [89]. Two open label studies on sertraline for the treatment of depression in PD found no change in motor function [90, 91]. SSRIs either alone or in combination with selegeline can cause the "serotonine syndrome" [92]. Criteria for the diagnosis include the presence of at least three of the following symptoms: tremor, diarrhea, myoclonus, hyperreflexia, fever, Parkinson Study Group, severe adverse experiences resulting from the combined use of selegeline and SSRIs and TCAs in parkinsonian patients are thought to be rare [93]. However, we recommend that a combination of SSRIs or TCAs with selegeline should be avoided. Novel antidepressant drugs like reboxetine. a selective noradrenalin reuptake inhibitor (NARI), venlafaxine, a norepinephrine and serotonin reuptake inhibitor (SNRI) and mirtazapine, the first noradrenergic and specific serotonergic antidepressants (NaSSA), indicated promising data in clinical studies for the treatment of major depression [94-96]. The drugs are considered to be as effective as TCAs and SSRIs. As yet only case reports have been published describing the use of these drugs in PD [97]. A possible advantage of mirtazapine may be the early antidepressant effect, which is thought to occur about one week after drug administration [98]. Furthermore, the sedative, sleep-inducing side-effect of mirtazapine may be useful for the treatment of agitated PD patients during night. We believe further studies to investigate the effect of these novel drugs on depressive symptoms in PD are warranted, current available data in PD is insufficient to recommend these novel drugs in routine care of depressed PD patients. Electroconvulsive therapy (ECT) has been used as an effective treatment of depression in PD [99], especially in patients without a treatment response to antidepressants [100]. Interestingly, a marked improvement on parkinsonian symptoms could be temporarily observed, although the published data are rare. On the basis of the current literature and our clinical experience in the treatment of depression in PD. we recommend first to control and improve motor symptoms using antiparkinsoman drugs like levodopa or dopaminergic agents. In a further step. TCAs or SSRIs

shiverin

D. Brandstiidter and W.H. OerteUPsychosis and Depression in Parkinson's Disease

35

should be administrated preferring those with low anticholinergic side-effects in older and cognitive impaired patients. In agitated patients the treatment benefit might be greater using TCAs, in apathetic PD patients SSRIs should be favourised. ECT should be considered in patients with a failure to antidepressant treatment as the last attempt. Acknowledgement: We thank Dr. MT Huber for helpful comments. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

[11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26]

American Psychiatric Association, Diagnostic and statistical manual of mental disorders, revised third edition. Washington D.C., American Psychiatric Association, 1987. G. Fenelon et al. , Hallucinations in Parkinson's disease - prevalence, phenomenology and risk factors, Brain 123 (2000) 733-745. E. Melamed et al., Psychosis - impact on the patient and family, Neurology, 52 (suppl 3) S14-S16. E.C. Wolters, Dopaminomimetic psychosis in Parkinson's disease patients, Neurology 52 (suppl 3 (1999)810-813. M. Doraiswamy et al., Psychosis in Parkinson's disease: Diagnosis and treatment, Prog NeuroPsychopharmacol & Biol Psychiat 19 (1995) 835-846. D. Aarsland et al., Prevalence and clinical correlates of psychotic symptoms in Parkinson's disease. A community-based study, Arch Neurol 56 (1999) 595-601. N. Giladi et al., Risk factors for dementia, depression and psychosis in long-term Parkinson's disease, J Neurol Transm 107 (2000) 59-71. W. Birkmayer and P. Riederer, Responsibility of extra-striatal area for appearance of psychotic symptoms J Neural Transm 35 (1975) 175-182. C.G. Goetz et al., Pharmacology of hallucinations induced by long-term drug therapy. Am J Psychiatry 139(1982)494-497. T.H. Svensson et al., Mode of action of atypical neuroleptics in relation to the phencyclidine model of schizophrenia: A role of 5-HT2 receptor and alpha l-adrenoceptor antagonism, J Clin Psychopharmacol 15 (suppl 1)(1995)S11-S18. C. Moskovitz et al., Levodopa-induced psychosis: A kindling phenomenon. Am J Psychiatry 135 (1978)669-675. C.L. Cornelia et al., Polysomnografic sleep measures in Parkinson's disease patients with treatmentinduced hallucinations, Ann Neurol 34 (1993) 710-714. J.F. Friedman et al., Atypical antipsychotics in the treatment of Parkinson's disease, Mov Disord 15 (2000)201-211. J. Arndt and T. Skarsfeld, Do novel antipsychotis have similar pharmacological characteristics? A review of the evidence, Neuropsychopharmacology 18 (1998) 63-101. E. Scholz and J. Dichgans, Treatment of drug-induced exogenous psychosis in parkinsonism with clozapine and fluperlapine. Eur Arch Psychiatry Neurol Sci 235 (1985) 60-64. J.H. Friedman and M.C. Lannon, Clozapine in the treatment of psychosis in Parkinson's disease, Neurology 39 (1989) 1219-1221. R.F. Pfeiffer et al., Clozapine for psychosis in Parkinson's disease, Mov Disord 5 (1990) 239-242. S.A. Factor and D. Brown, Clozapine presents recurrence of psychosis in Parkinson's disease, Mov Disord 7 (1992) 125-131. P. Greene et al., Treatment of drug-induced psychosis in Parkinson's disease with clozapine, Adv Neurol 60 (1993) 703-706. P. Gonski, Clozapine in Parkinson's disease, Lancet 345 (1995) 516-517. J.M. Rabey et al., Low-dose clozapine in the treatment of levodopa induced mental disturbances in Parkinson's disease, Neurology 45 (1995) 432-434. The Parkinson Study Group, Low-dose clozapine for the treatment of drug-induced psychosis in Parkinson's disease, N Engl J Med 340 (1999) 757-763. The French Clozapine Parkinson Study Group, Clozapine in drug-induced psychosis in Parkinson's disease, Lancet 353 (1999) 2041-2042. J.H. Friedmann and M.C. Lannon, Clozapine-responsive tremor in Parkinson's disease. Mov Disord 5 (1990)225-229. E.N.H. Jansen, Clozapine in the treatment of tremor in Parkinson's disease. Acta Neurol Scand, 73 (1986)295-297. G. Meco et al., Risperidone for hallucinations in levodopa treated Parkinson's disease patients, Lancet 343(1994)1370-1371.

36

[27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58]

D. Brandstadter and W.H. Oertel/Psychosis and Depression in Parkinson's Disease

R.J. Workman el al.. The use of nsperidone for psychosis and agitation in demented patients with Parkinson's disease. J Neuropsychiatry Clin Neurosci 9 (1997) 594-597. N.A. Leopold, Risperidone in patients with Parkinson's disease, Mov Disord 13 (suppl 2) (1998) 78. S.R. David et at.. The effects of olanzapine, risperidone and haloperidol on plasma prolactin levels in patients with schizophrenia, Clin Ther 22 (2000) 1085-1096. C.M. Beasly et al., Efficacy of olanzapine: An overview of pivotal trials. J Clin Psychiatry 58 (suppl 10)(I997)7-12. N.A. Moore et al., The behavioral pharmacology of olanzapine. a novel "atypical" antipsychotic agent. J Pharmacol Exp Ther 262 (1992) 545-551. K. Solomons and O. Geiger, Olanzapine use in the elderly: A retrospective analysis. Can J Psych 45 (2000)151-155. E.C. Wolters et al., Olanzapine in the treatment of dopaminomimetic psychosis in patients with Parkinson's disease, Neurology 47 (1996) 1085-1087. Churchyard and R. Jansek, Olanzapine as treatment of the neuropsychiatric complications of Parkinson's disease Mov Disord 13 (suppl 2) (1998) 188. J.M. Graham et al., Oianzapine in the treatment of hallucinosis in idiopathic Parkinson's disease. J Neurol Neurosurg Psychiatry 65 (1998) 774-777. F.J. Jimenez-Jimenez et al., Olanzapine can worsen parkinsonism. Neurology 50 (1998) 1183-1184. J.H. Friedmann and S. Goldstein, Olanzapine in the treatment of dopaminomimetic psychosis in Parkinson's disease. Neurology 50 (1998) 1195-1196. E.S. Molho and S.A. Factor. Worsening of motor features of Parkinson's disease with olanzapine. Mov Disord 12 (suppl 2) (1998) 486. C.G. Goetz et al.. Olanzapine and clozapine-comparative effects on motor function in hallucinating Parkinson's disease patients. Neurology 55 (2000) 789-794. C.F. Sailer and A.L. Salama, Seroquel: Biochemical profile of a potential atypical antipsychotic. Psychopharmacology 112 (1993) 285-282. J.G. Small et al.. Quetiapine in patients with schizophrenia. A high and low-dose double-blind comparison with placebo, Arch Gen Psychiatry 54 (1997) 549-557 M.L. Evatt and D. Jewart. "Seroquel" (ICI 204.636) treatment of psychosis in parkinsonism. Mov Disord 11(1996)595. J.L. Juncos et al., Long term effects of quetiapine fumarate in parkinsonism complicated by psychosis. Neurology, 50 (suppl 4) (1998) S70-S71. M.A. Parsa and B. Bastani. Quetiapine (Seroquel) in the treatment of psychosis in patients with Parkinson's disease J Neuroppsychiat Clin Neurosc 10 (1998) 216-219. S.D. Targum and J.L. Abbott, Efficacy of quetiapine in Parkinson's patients with psychosis, J Clin Psychopharmacol 20 (2000) 54-60. R.B. Dewey et al.. Treatment of drug-induced psychosis with quetiapine and clozapine in Parkinson's disease. Neurology. 55 (2000) 1753-1754. H.H. Fernandez et al.. Quetiapine for the treatment of drug-induced psychosis in Parkinson's disease. Mov Disord 14 (1999) 484-487. Y. Zoldan et al., Ondansetron for hallucinosis in advanced Parkinson's disease. Lancet 341 (1993) 562-563. J Zoldan et al.. Psychosis in advanced Parkinson's disease. Treatment with odansetron. a 5 HT3 receptor antagonist, Neurology 45 (1995) 1305-1308. T.E. Eichhorn et al., Odansetron treatment of L-dopa-induced psychosis. Neurology 47 (1996) 16081609. T.A Zesiewicz et al.. Current Issues in Depression in Parkinson's disease. Am J Geriatr Psychiatry 7 (1999) 110-118. J.L. Cummings and D.L. Masterman, Depression in patients with Parkinson's disease. Int J Geriat Psychiatry 14(1999)711-718. M. Stein et al., Anxiety disorders in patients with Parkinson's disease. Am J Psychiatry 147 (1990) 217-220. W. Poewe and E. Luginger, Depression in Parkinson's disease, Neurology 52 (suppl 3) (1999) S2-S6. M.A. Menza et al., Parkinson's disease and anxiety. Comorbidity with depression. Biol Psychiatry 34 (1993), 465-470. M.M.A. Derix et al.. Depressive syndrome in Parkinson's disease: Diagnostic pitfalls. In: E.C. Wolters, P. Scheltens (eds.). Mental dysfunction in Parkinson's disease. IOS Press. Amsterdam 1993. pp. 325-333. C.A. Biggins et al., A controlled longitudinal study of dementia in Parkinson's disease. J Neurol Neurosurg Psychiatry 55 (1992) 556-571. M. Sano et al., Coexisting dementia and depression in Parkinson's disease. Arch Neurol 46 (1989) 1284-1286.

D. Brandstadter and W.H. Oertel /Psychosis and Depression in Parkinson's Disease

[59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89]

37

A.I. Troster et ai., The influence of depression on cognition in Parkinson's disease: A pattern of impairment distinguishable from Alzheimer's disease. Neurology 45 (1995) 672-676. G. G. Celesia and W.M. Wanamaker, Psychiatric disturbances in Parkinson's disease, Dis Nerv Syst 33 (1972)577-583. S.E. Starkstein et al., A prospective longitudinal study of depression, cognitive decline and physical impairment in patients with Parkinson's disease, J Neurol Neurosurg Psychiatry 55 (1992) 377-382. J. Santamaria et al., Parkinson's disease with depression: A possible subgroup of idiopathic parkinsonism, Neurology 36 (1986)1130-1133. S.E. Starkstein et al., Depression in patients with early versus later onset of Parkinson's disease, Neurology 39 (1989) 1141-1145. A.M. Gotham et al., Depression in Parkinson's disease: A quantitative and qualitative analysis, J Neurol Neurosurg Psychiatry 49 (1986) 381-389. S.E. Starkstein et al., Neuropsychological and psychiatric differences between Alzheimer's disease and Parkinson's disease with dementia. J Neurol Neurosurg Psychiatry 61 (1996) 381-387. R.G. Brown et al., Depression and disability in Parkinson's disease: A follow-up study of 132 cases, Psychol Med 18 (1988) 49-55. R.H.S. Mindham et al., Psychiatric symptoms during levodopa therapy for Parkinson's disease and their relationship to physical disability, Psychol Med 6 (1976) 23-33. G. Kuzis et al., Cognitive functions in major depression and in Parkinson's disease, Arch Neurol 54 (1997)982-986. S.E. Starkstein et al., Depression in classic versus akinetic-rigidic Parkinson's disease. Mov Dis 13 (1998)29-33. C.J. Todes and A.J. Lees, The premorbid personality of patients with Parkinson's disease, J Neurol Neurosurg Psychiatry 48 (1985) 97-100. W.H. Oertel et al., Depression in Parkinson's disease: An update. In: D. Calne (ed.), Parkinson's disease, Advances in Neurology, Lippincott Williams & Wilkins, Philadelphia, in press. R. Brown and M. Jahashahi, Depression in Parkinson's disease: A psychosocial viewpoint. In: W.J. Weiner, A.F. Lang (eds.), Advances in Neurology, Raven Press, New York 1995, pp. 61-84. V.S. Kostic et al, Depression and Parkinson's disease: A possible role of serotonergic mechanisms. J Neurol 234 (1987) 94-96. R. Mayeux et al., Altered serotonin metabolism in depressed patients with Parkinson's disease, Neurology 34 (1984) 642-646. B.E. Roos and R. Sjostrom, 5-Hydroxyindoleacetic acid levels in the cerebrospinal fluid after probenecid application in patients with manic-depressive psychosis, Pharmacol Clin 1 (1969) 153-155. R.M. Torrack and J.C. Morris, The association of ventral tegmental area histopathology with adult dementia, Arch Neurol 45 (1988) 211-218. H.A. Ring et al., Depression in Parkinson's disease. A positron emission study, Br J Psychiatry 165 (1994)333-339. W. Paulus and C. Trenkwalder, Imaging of nonmotor symptoms in Parkinson-syndroms, Clin Neurosci 5 (1998) 115-120. S. H. Mayberg et a/Selective hypometabolism in the frontal lobe in depressed patients with Parkinson's disease, Ann Neurol 28 (1990) 57-64. R. Jouvent et al., Antiparkinsonian and antidepressant effects of high doses of bromocriptine, J Affect Disord 5(1983) 141-145. Barbeau, L-dopa therapy in Parkinson's disease: A critical review of nine years' experience, J Can Med Assoc 101 (1969) 791-800. L. Laitinen, Desipramine in treatment of Parkinson's disease. Acta Neurol Scand 45 (1969) 109-113. J. Anderson et al., Antidepressive treatment in Parkinson's disease: A controlled trial of the effect of nortriptyline in patients with Parkinson's disease treated with L-dopa, Acta Neurol Scand 62 (1980) 210-219. J.A. Simons, Fluoxetine in Parkinson's disease, Mov Disord 11 (1996) 581-582. W. Wittgens et al., Treatment of depressive syndromes in Parkinson's disease with paroxetine. Mov Disord 12(1997) 128. E. Jansen-Steur, Increase of Parkinson disability after fluoxetine medication, Neurology 43 (1993) 211-213. V. Wils, Extrapyramidal symptoms in a patient treated with fluvoxamine, J Neurol Neurosurg Psychiatry 55(1992)330-331. F.J. Jimenez-Jimenez et al., Parkinsonism exerbated by paroxetine, Neurology 44 (1994) 2406. R. Cevalo et A3., Paroxetine in Parkinson's disease: Effects on motor and depressive symptoms, Neurology 55 (2000) 1216-1218.

38

[90] [91]

D. Brandstadter and W.H. Oertel / Psychosis and Depression in Parkinson's Disease

R.A. Hauser and T.A. Zesiewicz, Sertraline for the treatment of depression in Parkinson's disease. Mov Disord 12 (1997) 756-757. [92] R.J. Mara et A3., An open uncontrolled study of the use of sertraline in the treatment of depression in Parkinson's disease, J Serotonin Research 4 (1996) 243-249. [93] J.L. Ritter and B. Alexander, Retrospective study of selegeline-antidepressant drug interactions and review of the literature, Ann Clin Psychiatry 9 (1997) 7-13. [94] I.H. Richard et A3., Serotonin syndrome and the combined use of deprenyl and an antidepressant in Parkinson's disease. Parkinson Study Group, Neurology 48 (1997) 1070-1077. [95] S.A. Montgomery, Reboxetine: Additional benefits to the depressed patient. J Psychopharmacol 11 (suppl 4)(!997)S9-S15. [96] S.A. Montgomery et A3., Mirtazapine versus amitryptiline in the longterm, treatment of depression: A double blind placebo-controlled study, Int Clin Psychopharmacology 13 (1998) 63-73. [97] R.M. Hirschfeld, Efficacy of SSRIs and newer antidepressants in severe depression: Comparison with TCAs. J Clin Psychiatry 60 (1999) 326-335. [98] M.R. Lemke, Reboxetine treatment of depression in Parkinson's disease, J Clin Psychiatry 61 (2000) 872. [99] O. Benkert et A3., Mirtazapine compared with paroxetine in major depression. J Clin Psychiatry 61 (2000) 656-663. [101] R. Faber and M.R. Trimble, Electroconvulsive therapy in Parkinson's disease and other movement disorders. Mov Disord 6 (1991) 293-303. [ 1 0 1 ] P.A. Fal and A.K. Granerus. Maintenance ECT in Parkinson's disease. J Neural Transm (106) 1999 737-741.

Parkinson's Disease E. Ronken and G.J.M. van Scharrenburg (Eds.) IOS Press, 2002

39

Experimental Models of Parkinson's Disease Peter Jenner Neurodegenerative Diseases Research Centre, Guy's, King's & St Thomas' School of Biomedical Sciences, King's College, London SE11UL, United Kingdom Telephone: 020 7848 6011, Fax: 020 7848 6034, Email: [email protected] Abstract. The treatment of Parkinson's disease (PD) is complicated by long term problems arising as the result of dopaminergic replacement therapy. There is a need to produce a new generation of antiparkinsonian agents which overcome these problems so providing long term control of the illness without a loss of drug efficacy or the onset of dyskinesia or psychosis. Classical rodent models of PD, such as the AMPT treated or reserpinised mouse or 6-hydroxydopamine lesioned rat are effective as an initial means of selecting antiparkinsonian drugs. The intravenous administration of the herbicide, rotenone to rats may produce a model of selective nigral degeneration with Lewy body-like inclusions but the drug-responsiveness of these animals is unknown. More recently, a variety of rodent transgenic and mutant models of PD have been reported but none have so far been shown to be an effective model of the disorder. One exception is the overexpression ofa-synuclein in Drosophila that appears to partially mimic the human disease but where drug responsiveness is again unknown. At this time, the most effective model of PD is the MPTP-treated primate. These animals exhibit motor deficits resembling those of PD and show responsiveness to all commonly used antiparkinsonian agents. In addition, repeated treatment with L-DOPA rapidly induces dyskinesia and 'wearing off phenomenon. The MPTP-treated primate is useful in examining drugs acting selectively on dopamine receptor sub-types and non-dopaminergic approaches to treatment, such as adenosine antagonists, as well as new routes of drug administration, such as transdermal application. The study of the genesis and avoidance of dyskinesias in MPTP treated primates has shown long acting dopamine agonists to induce far less dyskinesia than L-DOPA and has led to the development of the continuous dopaminergic stimulation concept as a means of avoiding involuntary movements. Effective animal models of PD exist for the examination of novel symptomatic treatments. However, experimental models are needed in which the effect of drugs which may also be neuroprotective or neurorestorative can be studied, so bringing a new facet to the treatment of PD.

Introduction Parkinson's disease (PD) is an age-related disorder that affects up to 1 in 500 of the general population and 1 in 100 of individuals over the age of 60.[1] The prevalence of PD will rise as life expectancy increases and the number of affected individuals will rise significantly over the next three decades. PD is primarily due to the destruction of dopaminergic neurones in the zona compacta of substantia nigra and the subsequent loss of caudateputamen dopamine content leading to the onset of the cardinal symptoms of akinesia, rigidity, and tremor.[2] Dopamine replacement therapy remains the mainstay of the treatment of PD with L-DOPA being the most commonly employed agent, but with increasing use of dopamine agonist drugs, such as ropinirole and pramipexole. [3,4] Early symptomatic control of PD disease is not problematic; rather it is the long term control of

40

P. Jenner / Experimental Models of Parkinson's Disease

motor symptoms that remains difficult, coupled with the onset of side effects and a lack of effect of current therapies on non-motor components of the illness.[5] Chronic treatment of PD is associated with a loss of drug efficacy in the form of 'wearing off and 'on/off phenomena. In addition, involuntary movements appear in the form of dystonia and/or chorea and hallucinations and psychotic episodes may occur. Dyskinesias and psychosis can become treatment-limiting, and are thought to affect some 30-40% of all patients with PD.[6] Long-term complications are a particular feature of the use of L-DOPA and the current use of dopamine agonist drugs is based on their ability to control the early motor symptoms of PD without extensive priming of basal ganglia for the appearance of dyskinesia.[3,4,7] Patients with PD suffer from other problems, including bladder and bowel dysfunction, sweating, drooling, and, in particular, postural instability which do not respond to dopaminergic therapy and that are currently unbeatable.[5] In addition, all current therapies provide symptomatic control of PD but mask the continuing worsening of the pathology. At this time, no therapy is available which can influence disease progression. Current research into PD is aimed at improving the symptomatic treatment of the illness by the production of a new generation of agents. The objectives of this research are to produce drugs that are effective throughout the illness without loss of efficacy, and which do not prime the basal ganglia for the production of dyskinesia, or provoke established involuntary movements. In addition, such drugs should not provoke psychosis, they should avoid some of the acute peripheral side effects of dopaminergic therapies such as nausea, vomiting and hypotension, and they should have effects on the currently untreatable symptoms of PD, such as postural instability. In addition it is hoped that novel agents for the treatment of PD will also have neuroprotective effects, so preventing or slowing the progression of disease. To achieve these ends, it is necessary to have effective experimental models of PD in which new agents can be evaluated before attempting clinical studies. A number of models exist and recent advances in the understanding of the genetic basis of familial PD have started to provide some transgenic and mutant models of the disorder. In this review, the value of experimental models for the development of new approaches to the treatment of PD, will be assessed. Attention will be focussed on the MPTP-treated primate as currently the most effective model of the disorder and its usefulness in evaluating new dopaminergic and non-dopaminergic approaches to the treatment of PD.

Rodent models of Parkinson's disease A range of rodent models of PD has existed for some time, and novel approaches based on genetic manipulation are under development. Early rodent models were based on the use of chemical depletion of dopamine or on the use of toxins to destroy the nigrostriatal pathway. However, the increasing use of mutants and transgenics is providing a new generation of models but which still require development and evaluation.

Reserpine and a-methyl-p-tyrosine treated rodents

The earliest approach to developing a model of PD used drugs which disrupted dopamine storage or synthesis.[8] Administration of reserpine to rodents disrupts the storage of monoamines in presynaptic vesicles.[9] The subsequent depletion of dopamine, as well as noradrenaline and 5HT, produces an animal that is akinetic, hunched and which shows little exploratory activity. The effects of reserpine treatment can be reversed by the administration of -DOPA and dopamine agonist drugs.[10] Similarly, the administrati

P. Jenner / Experimental Models of Parkinson's Disease

41

a-methyl-para-tyrosine (AMPT) as an inhibitor of tyrosine hydroxylase, the rate-limiting enzyme for dopamine formation, reduces dopamine levels but without affecting the noradrenaline or 5HT content of the brain.[11] Again, AMPT treated animals show a reduction in spontaneous locomotion which can be reversed by the administration of LDOPA and dopamine agonists. [12] Reserpine and AMPT have also been used in combination to doubly impair the synthesis and storage of dopamine and to ensure marked dopamine depletion and motor dysfunction. As a primary screen for novel antiparkinsonian agents, particularly those affecting dopaminergic systems, these models can be extremely useful. However, the use of AMPT and reserpine has a number of disadvantages. Neither agent selectively produces dopamine loss in the nigrostriatal pathway and there is also a marked depletion in both the mesolimbic and mesocortical regions of brain, which are relatively spared in PD. Neither agent causes destruction of the nigrostriatal pathway and so does not mimic the disorder as it occurs in man. Further, both reserpine and AMPT produce a reversible depletion of dopamine in the brain, and the timing of studies employing these agents is important, as it is critical to examine the effects of potential antiparkinsonian drugs at the point of maximal dopamine synthesis or storage inhibition.[13] Both agents also impair monoaminergic function in the periphery, and the use of reserpine is associated with changes in cardiovascular function and temperature control and marked diarrhoea. Reserpine does, however, mimic PD by affecting noradrenaline and 5HT levels as well as dopamine whereas AMPT is selective for dopaminergic systems.

Toxin-based models of Parkinson's disease in rodents The use of toxins that selectively destroy the nigrostriatal dopaminergic pathway is a common means of developing models of PD. Such models may utilise systemic toxin administration but commonly employ the stereotaxic injection of toxins directly into the nigrostriatal pathway. The most commonly utilised and well characterised model of PD is the unilateral 6hydroxydopamine (6-OHDA) lesioned rat model. 6-OHDA is a toxin presumed to act through free radical mechanisms,[14] which is selectively taken up into and destroys catecholamine-containing neurones. This classical model was first developed during the 1960s and has been successfully employed for the routine screening of potential antiparkinsonian agents.[15-17] Unilateral injection of 6-OHDA either directly into the substantia nigra or into the medial forebrain bundle is the most frequently used procedure. This results in a subsequent degeneration of the nigrostriatal pathway in one hemisphere over a period of days, associated with the development of post-synaptic dopamine receptor supersensitivity in the denervated striatum. 6-OHDA is routinely used in combination with desipramine pre-treatment to prevent the toxin being taken up into and destroying noradrenergic fibres that also ascend in the medial forebrain bundle. 6-OHDA has also been employed by direct injection into the striatum where it appears to undergo retrograde transport to the substantia nigra causing a gradual die-back of dopaminergic neurones which has been proposed as a model of progression of PD.[18] The latter may be useful both for the development of symptomatic treatments of PD and in detecting neuroprotective agents for this disorder. Following 6-OHDA induced destruction of the nigrostriatal pathway, animals exhibit an asymmetric motor response to the administration of direct or indirectly acting dopaminergic agonists in the form of rotational behaviour.[19] Administration of directly acting dopamine agonists or of L-DOPA results in turning away from the side of the lesion (contraversive rotation)) whereas indirectly-acting dopaminomimetics act through the intact hemisphere and so cause rotational response towards the side of the lesion (ipsiversive rotation). So this model not only shows

42

P. Jenner / Experimental Models of Parkinson's Disease

responsiveness to dopaminergic drugs, but it can also indicate the mechanism through which they produce their effects. All currently used dopaminergic agents in the treatment of PD result in a rotational response in this model. It is an excellent and effective highthroughput screening system for the development of antiparkinsonian drugs. Unfortunately, 6-hydroxydopamine cannot be used to produce a bilateral lesion model, since such animals show adipsia and aphagia and rapidly waste. The major disadvantage of the 6hydroxydopamine-lesioned rat is that false positives occur with drugs, such as SKF 38393. inducing rotation subsequently being shown to be ineffective in either primate models of PD or in the human disease. The discovery of the selective nigral toxicity of 1-methyl-4-phenyl-1, 2. 3, 6tetrahydropyridine (MPTP) provided a major opportunity for the development of an effective model of PD by systemic toxin administration.[20,21] The toxicity of MPTP is mediated by its active metabolite l-methyl-4-phenyl-pyridinium species (MPP+) which is produced as a result of the metabolism of MPTP by MAO B in glial cells. MPP+ acts to inhibit complex I of the mitochondrial respiratory chain following its selective uptake into dopaminergic neurones. Unfortunately, most rodent species are insensitive to the actions of MPTP with the exception of some specific strains of mice, for example black C57 mice and Swiss Webster mice.[23-27] Even in these strains, large doses of MPTP are needed to produce a loss of dopamine content in the striatum, and not all studies show the administration of MPTP to be associated with nigral damage. Rather, MPTP can also exert a reversible reserpine-like action so depleting striatal dopamine content without any effect on nigral dopaminergic cell number. However, in those experiments where mice have been successfully lesioned with MPTP, a model is produced in which dopaminergic agents are able to reverse the akinesia induced as a result of toxin treatment. The systemic administration of MPTP results in a bilateral degeneration of substantia nigra, so more closely mimicking events occurring in PD. However, damage is selective to the nigrostriatal pathway and there is no loss of noradrenaline or 5HT content in the brain as occurs in PD. Rather than using the systemic administration of MPTP, MPP+ can be stereotaxically injected in to brain in a manner similar to that of 6-OHDA. Indeed, direct intranigral injection of MPP+ causes the loss of dopaminergic neurones in substantia nigra presumably through its ability to inhibit mitochondrial function.[28-31] However, some caution is required since the intranigral administration of MPP+ can lead to toxicity to nondopaminergic neurones through its mitochondrial respiratory chain actions. As with 6OHDA, the use of MPP+ is restricted to producing a unilateral lesion, since bilateral lesions induce adipsia and aphagia and animals have poor survival rates. Another toxin-based model of PD reported recently resembles that produced by MPP+ in that it utilises mitochondrial inhibition as the mechanism underlying cell death but allows for systemic toxin administration. Rotenone, like MPP+, is an inhibitor of mitochondrial complex I activity and it is the active constituent in extracts of derris which are used as a natural herbicide. Greenamyre and colleagues [32] showed that the intravenous infusion of rotenone in rats causes the selective degeneration of the nigrostriatal pathway accompanied by the formation of Lewy body-like inclusions. The ability of rotenone to produce bilateral damage in the rat may provide a further means of assessing potential drug activity, although the response of these animals to current antiparkinsonian agents has not been reported. A further drawback of this model is the generalised toxicity produced by rotenone through mitochondrial inhibition that leads to a high degree of mortality.

P. Jenner / Experimental Models of Parkinson's Disease

43

Mutant and transgenic rodent models of Parkinson's disease Increasingly, mutations and genetic modification of rodent species are being employed to provide useful models of neurodegenerative disorders, including PD. A number of mutant rodents, such as the Weaver mouse and the Wobbler mouse, have been proposed as showing behavioural, biochemical and pathological deficits reflecting those occurring in PD but subsequently discounted. More recently, a spontaneous mutation in a rat strain (AS/AGU) has been proposed as a model of PD based on evidence for progressive striatal dopamine loss accompanied by nigral dopaminergic cell degeneration.[33] The rats develop motor dysfunction that appears to be reversed by L-DOPA. However, the motor abnormalities exhibited by the mutants do not resemble that of PD rather they exhibit a complex motor syndrome suggestive of pathology in other brain areas. The genetic basis to this model remains unknown and at this time its relevance to PD is unclear. There has been recent interest in Nurr-1 knockout mice as a model of PD. Nurr-1 is a member of the nuclear receptor gene super family and is essential for the development of dopaminergic neurones.[34,35] In homozygous Nurr-1 knockout mice, there are no dopaminergic neurones in the substantia nigra or in the ventral tegmental area, and this is accompanied by striatal dopamine depletion without any change in brain levels of 5HT or noradrenaline. [36-38] Such animals die shortly after birth but heterozygotes appear to show progressive nigral degeneration and have an increased vulnerability to MPTP toxicity.[39] Theoretically this would seem to provide an highly relevant mouse model of PD in relation to the primary pathology of PD. However, at this time, the drug responsiveness of these animals has not been reported, and so its usefulness as a means of assessing drug action in PD remains unknown. Intense interest has centred on producing a transgenic model of PD related to the discovery of two mutant forms of a-synuclein arising from gene defects in familial Parkinson's disease.[40-42] Overexpression of mutant -synuclein in cell lines leads to increased apoptotic cell death in response to toxic stimuli.[43-45] However, to date, attempts to produce a a-synuclein transgenic mouse, which closely resembles PD in terms of pathology, the presence of inclusions, and biochemical changes accompanied by motor deficits, has been relatively unsuccessful. One problem associated with the production of an a-synuclein transgenic is that the wild type found in mice is similar to one of the mutant forms associated with familial PD. Production of a a-synuclein knockout mouse did not result in loss of nigral dopaminergic cell bodies, fibres or synapsesa-synucleinand there was a normal dopamine release and re-uptake in the striatum on electrical stimulation of dopaminergic neurones.[46] There were, however, a decrease in striatal dopamine levels and a decreased locomotor response to amphetamine but otherwise these animals did not appear to have an impaired motor system. Overexpression of human wild type a-synuclein in mice resulted in the progressive accumulation of a-synuclein and ubiquitin immunoreactive inclusions in the cortex hippocampus and substantia nigra.[47] While a loss of dopaminergic terminals occurred in the striatum, this was not accompanied by cell loss in the substantia nigra. These mice showed motor deficits on a rotarod but again the major pathology of PD was not apparent. Overexpression of a human mutant form of A53T a-synuclein resulted in widespread brain stem pathology that includeda-synucleinnigral cell loss.[48] However, pathology also occurred in motor neurones, but was accompanied by Lewy body-type inclusions and by a widespread gliosis. Again, these mice showed impaired performance on a rotarod but once more the pathology and the biochemical deficits occurring do not replicate events in PD. There has, however, been one successful approach to producing a a-synuclein transgenic with characteristics similar to those of PD through the expression of human wild

44

P. Jenner / Experimental Models of Parkinson's Disease

type and mutant a-synuclein in Drosophila.[49] There is an adult onset loss of dopaminergic neurones in the nervous system and the eye accompanied by the occurrence of filamentous intra-neuronal inclusions containing a-synuclein. The flies exhibit motor dysfunction, but at this time their drug response is not known. So while this may be an effective model of PD that may has utility in understanding cell death, its role as a model for drug discovery is not known.

The MPTP-treated primate model of Parkinson's disease The discovery of the selective nigral toxicity of MPTP created a highly appropriate model of PD when it became apparent that the toxin was most effective in primates.[50-52] Administration of MPTP to a range of primate species causes selective nigral cell loss accompanied by a decrease in caudate-putamen dopamine content and the onset of major motor symptoms of PD, such as akinesia, bradykinesia and rigidity. Tremor is not a common component of the MPTP syndrome and while postural tremor does occur, rest tremor characteristic of PD is only seen in specific primate species, such as the green monkey. MPTP only partially mimics events occurring in PD and the characteristics of the model compared to the disorder affecting man are shown in Table 1. Most studies involving MPTP utilise systemic administration of MPTP to produce bilateral motor deficits. However, some investigators prefer to use unilateral intracarotid injection of MPTP to produce a unilateral model, which lessens the adverse effects of the toxin.[53-56] It is the drug responsiveness of the MPTP-treated primate that makes it a useful tool in the drug discovery process for assessing novel compounds for the treatment of PD. Administration of L-DOPA plus carbidopa or dopamine agonist to MPTP-treated primates results in a reversal of the akinesia or bradykinesia measurable as an increase in locomotor activity.[57] This is accompanied by a decrease in motor disabilities that can be assessed using an observer rated scoring system. In addition, the repeated administration of L-DOPA to drug naive MPTP-treated primates rapidly results in the appearance of marked dyskinesias which closely resemble those occurring in the long term treatment of patients with PD.[58-60] The rapid appearance of dyskinesia is related to the severity of the nigral lesions in MPTP treated primates which lowers the threshold for dyskinesia induction by LDOPA.[61] The long-term administration of L-DOPA to MPTP-treated animals also results in a shortening of drug affect so mimicking the 'wearing-off effect seen in patient populations.[62]

Novel approaches to the treatment of Parkinson's disease utilising the MPTP-treated primate The MPTP-treated primate responds to the administration of all currently used dopaminergic drugs for the treatment of PD, including L-DOPA, bromocriptine, pergolide, ropinirole, pramipexole and cabergoline. So far, all actions of these compounds occurring in MPTP-treated primates have turned out to be highly predictive of drug action in man. For this reason, other novel dopaminergic approaches are now being evaluated in the MPTP-treated primate for their utility in the therapy of PD. The MPTP model has been utilised to assess the relative roles of drugs acting on specific dopamine receptor sub-types. Agents which are active on the D-2-like family of receptors are all antiparkinsonian in the MPTP-treated primates but provoke established dyskinesia in these animals resulting from prior L-DOPA exposure, reflecting the response which occurs in patients with PD.[63.64] Drugs acting on D-l-like receptors, such as

P. Jenner / Experimental Models of Parkinson's Disease

45

dihydrexidine and ABT 431, also exert antiparkinsonian activity in MPTP-treated primates.[65,66] While they too will provoke established dyskinesia, this is less intense than occurs with D-2 agonist drugs, although recent clinical data suggest this is not reflected in their actions in man.[67] Novel dopamine re-uptake blockers, such as brasofensine and BTS 74-398 are also currently being assessed in MPTP-treated primates.[68,69] This class of drug increases locomotor activity and decreases motor disability but does so without provoking established involuntary movements. The mechanism by which dopamine re-uptake blockers can separate actions of benefit in PD from those which result in a major side effect are not clear at this time but may relate to enhancement of mesocortical and mesolimbic dopamine function rather than to a direct effect on the nigro-striatal pathway. It is not known if reuptake blockers are effective in treating PD but they are currently undergoing clinical evaluation. Very recently, interest has developed in the potential use of partial dopamine agonists in the treatment of PD. Partial agonists will act as full agonists in the denervated striatum so promoting an antiparkinsonian response. But in the relatively intact mesolimbic and mesocortical systems, they would act as antagonists by competing for receptor occupation with dopamine. Such an effect might prevent or subdue psychotic episodes occurring in patients with PD. In MPTP-treated primates, partial dopamine agonist drugs, such as SLV 308, can be highly effective in reversing parkinsonian symptoms and again a number of these compounds are now entering clinical evaluation for their utility in man.[70a,b] Studies in the MPTP-treated primate are also used to assess routes of drug delivery for existing and novel dopaminergic agents. In particular, transdermal application of dopaminergic agents is being employed to produce more continuous dopaminergic stimulation and longer period of mobility during the waking day. Studies in MPTP treated primates have shown agents such as PHNO and N-0437 to be highly effective by application to the skin and by delivery from transdermal patches.[71,72] This route of administration transforms drugs with relatively short half-lifes following oral or systemic administration into agents that can act throughout a 24-hour period or longer. Transdermal administration would be exceptionally useful in treating PD providing tolerance did not develop as a result of constant receptor stimulation. Based on studies carried out in MPTPtreated primates, transdermal patches for PD are currently under clinical evaluation.[73] Non-dopaminergic approaches to the treatment of Parkinson's disease studied in MPTP-treated primates For the past 40 years, the treatment of PD has been based on dopamine replacement therapy. Recently, the potential of a range of non-dopaminergic neuronal targets within the striatum and other nuclei of basal ganglia through which motor activity has been explored. An entire range of novel approaches to the treatment of PD based on non-dopaminergic drugs is now under active investigation and clinical development. The spectrum of approaches being examined is shown in Table 2. The potential advantages of a non-dopaminergic approach are illustrated by the actions of A2A antagonists in MPTP treated primates. The A2A adenosine receptor has a highly selective localisation on output neurones from the striatum making up the indirect output pathway. Administration of an A2A antagonist, such as KW 6002, to MPTP-treated primates produces a modest increase in locomotor activity but a more substantial reversal of motor disability.[74,75] However, this is not accompanied by the appearance of dyskinesia in animals which have been primed for involuntary movements and which would exhibit dyskinesia when challenged with a dopamine agonist. The use of A2A antagonists therefore may be useful in patients with PD who have been treated previously with L-DOPA but now

46

P. Jenner / Experimental Models of Parkinson's Disease

have incapacitating involuntary movements. Other non-dopaminergic approaches of promise include the suppression of dyskinesia by glutamate antagonists[76,77] and a similar ability of adrenergic antagonists, such as idazoxan, to suppress established dyskinesia without affecting the antiparkinsonian activity of L-DOPA.[78,79] So nondopaminergic approaches may overcome some of the problems associated with the chronic treatment of PD using dopamine replacement therapy but it is likely that all these new drug approaches will bring their own range of side effects. Dyskinesia and the MPTP-treated primate Perhaps the greatest utility of the MPTP-treated primate has been in advancing understanding of the genesis of dyskinesia and the ways by which involuntary movements might be avoided. MPTP-treated primates develop dyskinesia in response to repeated administration of L-DOPA, over a period of a few weeks. The rapidity of onset and the intensity of dyskinesia is related to the dose of L-DOPA employed, and the frequency of its administration.[61] In addition, increasing brain exposure to L-DOPA (for example, by combining L-DOPA with a peripheral COMT inhibitor such as entacapone) serves to induce dyskinesia more rapidly and more intensely.[80] Dyskinesia can be induced by the repeated stimulation of either D-l or D-2 dopamine receptors contrary to an earlier view that the D-l receptor was solely responsible for the genesis of involuntary movements.[81] Once dyskinesia has been established in L-DOPA treated primates, all directly acting dopaminergic agents will produce the same pattern of involuntary movements as also occurs in patients with PD.[63,63] Indeed, once priming for dyskinesia has occurred in response to L-DOPA exposure, it is difficult to reverse and it appears to be persistent if not permanent. In contrast, the repeated administration of long acting dopamine agonist drugs to MPTP-treated primates produces far less dyskinesia than occurs after treatment with equieffective antiparkinsonian doses of L-DOPA.[58,82,83] The reason why dopamine agonists produce less involuntary movement may relate to their pharmacokinetic and pharmacodynamic properties since short-acting dopamine agonists can produce equivalent dyskinetic responses to L-DOPA. Indeed, while pulsatile administration of the short acting D-2 agonist U-91356A by repeated subcutaneous injection results in marked dyskinesia, continuous subcutaneous infusion of this agent only induced mild involuntary movements.[84] Based on the actions of long acting dopamine agonists in MPTP treated primates, the concept of continuous dopaminergic stimulation as a means of avoiding dyskinesia has arisen.[85] This has been demonstrated in numerous studies employing a range of dopamine agonist drugs in otherwise drug naive MPTP treated primates and confirmed by a range of 5 year clinical studies in patients with PD emphasising the predictive nature of the animal model.[3,4,7]

Conclusions The current models of PD in rodents and primates based on chemical depletion of dopamine and the use of toxins to destroy the nigrostriatal pathway have provided important testbeds for the development of novel approaches to the symptomatic treatment of PD. There is no doubt of the utility of the MPTP-treated primate in this respect, and this has been a major asset in terms of drug discovery and the development of new therapeutic approaches to the treatment of the illness in man. Non-dopaminergic approaches to PD will clearly form part of the future therapy of this disorder, and the MPTP-treated primate will be highly effective in assessing the

P. Jenner / Experimental Models of Parkinson's Disease

47

activities of non-dopaminergic drags prior to their clinical evaluation. However, the future treatment of PD will be directed in towards producing neuroprotective strategies for the illness that will stop or slow disease progression. So we now need models which will be useful in understanding the mechanisms underlying nigral cell degeneration. This will occur when effective transgenic mice models of PD are devised and when models of progressive disease in which potential neuroprotective strategies can be evaluated. References [1]

B.S. Schoenberg, Descriptive Epidemiology of Parkinson's Disease: Disease Distribution and Hypothesis Formulation, Adv Neurol 45 (1986) 277-283. [2] J.A. Obeso, C. W. et al. Levodopa Motor Complications in Parkinson's Disease, Trends Neurosci (2000) 23 (suppl 10) S2-S7 [3] Parkinson Study Group. Pramipexole vs Levodopa as Initial Treatment for Parkinson's Disease. JAMA 284(2000) 1931-1938 [4] O. Rascol et al, A Five Year Study of the Incidence of Dyskinesia in Patients with early Parkinson's Disease who were Treated with Ropinirole or Levodopa, New Engl J Med 342 (2000) 1484-1491 [5] J. Jankovic and D. Marsden, Therapeutic Strategies in Parkinson's Disease, Parkinson's Disease and Mov Disord 10 (1998) 191-220 [61 J.E. Ahlskog and M.D. Muenter, Frequency of Levodopa-Related Dyskinesias and Motor Fluctuations as Estimated From the Cumulative Literature, Mov Disord 16 (2001) 448-458 [7] W.H. Oertel, Pergolide vs L-DOPA (PELMOPET), Mov Disord 15 (2000) (suppl. 3) S4 [8] C.D. Marsden, et al., Relationship Between Animal Models and Clinical Parkinsonism, Adv Neurol 9 (1975) 165-75 [9] A. Carlsson, et al., 3, 4-Dihydroxyphenylalanine and 5-Hydroxytryptophan as Reserpine Antangonists, Nature (Lond) 180 (1957) 1200 Pages? [10] A.C. Dolphin, et al., The Relative Importance of Dopamine and Noradrenaline Receptor Stimulation for the Restoration of Motor Activity in Reserpine or a-Methyl-p-Tyrosine Pre-Treated Mice, Pharmacol Biochem Behav 4 (1975) 661-670 [11] H. Corrodi, et al., Studies on Central and Peripheral Noradrenaline Neurones using a New Dopamineß-Hydroxylase Inhibitor, Eur J Pharmacol 12 (1970) 145-155 [12] B. Costall and R.J. Naylor, Actions of Dopaminergic Agonists on Motor Function, Adv Neurol 9 ( 1975)285-297. [13] J- Haggendal and M. Lindqvist, Brain Monoamine Levels and Behaviour During Long-Term Administration of Reserpine, Int J Neuropharmacol 3 (1964) 59-64 [14] D.G. Graham, et al., Autoxidation Versus Covalent Binding of Quinones as the Mechanism of Toxicity of Dopamine, 6-Hydroxydopamine, and Related Compounds toward C1300 Neuroblastoma Cells in Vitro, Mol Pharm 14 (1978) 644-453 [15] U. Ungerstedt, Striatal Dopamine Releases after Amphetamine or Nerve Degeneration Revealed by Rotational Behaviours, Acta Physiol Scand 82 (1971) (suppl.) 367: 49-68 [16] U. Ungerstedt, Postsynaptic Supersensitivity after 6-Hydroxydopamine Induced Degeneration of the Nigro-Striatal Dopamine System, Acta Physiol Scand 82 (1971) (suppl.)367): 69-93 [17] P.F. Von Voigtlander and K.E. Moore, Turning Behaviour in Mice with Unilateral 6Hydroxydopamine Lesions in the Striatum: Effects of Apomorphine L-DOPA, Amantadine, Amphetamine and Other Psychomotor Stimulants, Neuropharmacology 12 (1973) 451-462 [18] S. Przedborski, et al., Dose-Dependent Lesions of the Dopaminergic Nigrostriatal Pathway Induced by Intrastriatal Injection of 6-Hydrooxydopamine, Neuroscience 67 (1995) 631-647 [19] C. Reavill, et al., Differentiation of Dopamine Agonists Using Drug-Induced Rotation in Rats with Unilateral or Bilateral 6-hydroxydopamine Destruction of Ascending Dopamine Pathways, Biochem Pharmacol 32 (1983)865-870 [20] G.C. Davis, et al., Chronic Parkinsonism Secondary to Intravenous Injection of Meperidine Analogues, Psychiatry Research 1 (1979) 249-254 [21] J.W. Langston, et al., Chronic Parkinsonism in Humans due to a Product of Meperidine-Analog Synthesis, Science 219 (1983) 979-980 [22] P. Jenner and C.D. Marsden, MPTP-Induced Parkinsonism: A Model of Parkinson's Disease and Its Relevance to the Disease Process, Parkinson's Disease and Mov Disord 4 (1993) 55-75

P. Jenner / Experimental Models of Parkinson's Disease

R.A. Mayer, el al., l-Methyl-4-Phenyl- I, 2, 3, 6-Tetrahydropyridine (MPTP) Administration to C57Black Mice Leads to Parallel Decrements in Neostriatal Dopamine Content and Tyrosine Hydroxylase Activity, Eur J Pharmacol 120 (1986) 375-377 R.E. Heikkila, Differential Neurotoxicity of l-Methyl-4-Phenyl-l. 2, 3. 6-Tetrahydropyridine (MPTP) in Swiss-Webster Mice from Different Sources, Eur J Pharmacol 117 (1985) 131-133 A. Giovanni, et at., Correlation Between the Neostriatal Content of the 1-Methyl-4-Phenylpyridinium Species and Dopaminergic Neurotoxicity Following 1-Methyl-4-Phenyl-1, 2, 3, 6-Tetrahydropyridine Administration to Several Strains of Mice, Pharmacol Expt Ther 257 (1991) 691-697 A.J. Bradbury, et al., The Effect of l-Methyl-4-Phenyl-l, 2, 3, 6 - Tetrahydropyndine (MPTP) on Striatal and Limbic Catecholamine Neurones in White and Black Mice, Neuropharmacology 25 (1986) 897-904 P.K. Sonsalla and R.E. Heikkila, The Influence of Dose and Dosing Interval on MPTP-Induced Dopaminergic Neurotoxicity in Mice, Eur Jour Pharmacol 129 (1986) 339-345 A. J. Bradbury, et al., l-Methyl-4-Phenylpyridine is Neurotoxic to the Nigrostriatal Dopamine Pathway, Nature 319 (1986) 56-57 J.W. Langston, et al., l-Methyl-4-PhenyIpyridinium ION (MPP+): Identification of a Metabolite of MPTP, a Toxin Selective to the Substantia Nigra, Neuroscience 48 (1984) 87-92 D. Cavalla, et al., Intracerebroventricular Administration of 1 -Methyl-4-Phenyl-1, 2 3, 6Tetrahydropyridine (MPTP) and its Metabolite l-Methyl-4-Phenylpyridnium ION (MPP+) Decrease Dopamine and Increase Acetylcholine in the Mouse Neostriatum, Neuropharmacology 24 (1985) 585586 R.E. Heikkila, et al., Dopaminergic Toxicity after the Stereotaxic Administration of the l-Methyl-4Phenylpyridinium ION (MPP+) to Rats. Neurosci Lett 59 (1985) 135-140 R. Betarbet et al, Chronic Systemic Pesticide Exposure Reproduces Features of Parkinson's Disease. Mat Neurosci 3 (2000) 1301-1306 A.P. Payne, et al., The AS/AGU Rat: a Spontaneous Model of Disruption and Degeneration in the Nigrostriatal Dopaminergic System, J Anat 196 (2000) 629-633 C. Backman, et al., A Selective Group of Dopaminergic Neurons Express Nurrl in the Adult Mouse Brain. Brain Res 851 (1999) 125-132 P. Sacchetti. et al, Nurrl Enhances Transcription of the Human Dopamine Transporter Gene through a Novel Mechanism. J Neurochem 76 (2001) 1565-1572 W-D. Le. et al., Selective Agenesis of Mesencephalic Dopaminergic Neurons in Nurrl-Deficient Mice. Exp Neurol 159 (1999) 451 -458 R.H. Zetterstrom. et al., Dopamine Neuron Agenesis in Nurrl-Deficient Mice, Science 276 (1997) 248-250 J. Witta. et al., Nigrostriatal Innervation is Preserved in Nurrl-Null Mice, although Dopaminergic Neuron Precursors are Arrested from Terminal Differentiation, Mol Brain Res 84 (2000) 67-78 W-D. Le. et al., Reduced Nurrl Expression Increases the Vulnerability of Mesencephalic Dopamine Neurons to MPTP-Induced Injury. J Neurochem 73 (1999) 2218-2221 M.H. Poiymeropoulos, et al., Mapping of a Gene for Parkinson's Disease to Chromosome 4q21-q23. Science 274 (1996) 229-232 M.H. Polymeropoulos. et at., Mutation in the a-Synuclein Gene Identified in Families with Parkinson's Disease, Science 276 (1997) .2045-2047 R Kriiger, et al., Ala30Pro Mutation in the Gene Encoding a-Synuclein in Parkinson's Disease. Nat Genet 18(1998) 106-108 S.J. Tabrizi, et al., Expression of mutant a-Synuclein causes increased susceptibility to dopamaine toxicity. Hum Mol Genetics 9 (2000) 2683-2689 N. Ostrerova-Golts, et al., The A53T a-Synuclein Mutation Increases Iron-Dependent Aggregation and Toxicity. J Neurosci 20 (2000) 6048-6054 W. Zhou, et al., Overexpression of Human a-Synuclein Causes Dopamine Neuron Death in Rat Primary Culture and Immortalised Mesencephalon-Derived Dells, Brain Res 866 (2000) 33-43 A. Abeliovich. et al., Mice Lacking a-Synuclein Display Functional Deficits in the Nigrostriatal Dopamine System. Neuron 25 (2000) 239-252 E. Masliah, et al., Dopaminergic Loss and Inclusion Body Formation in a-Synuclein Mice: Implications for Neurodegenerative Disorders, Science 287 (2000) 1265-1269 H. van der Putten. et al., Neuropathology in Mice Expressing Human a-Synuclein. J Neurosci 20 (2000)6021-6029 MB Feany and W.W Bender. A Drosophila Model of Parkinson's Disease. Nature 404 (2000) 394398

P. Jenner / Experimental Models of Parkinson's Disease

49

R.S. Burns, a/A Primate Model of Parkinsonism: Selective Destruction of Dopaminergic Neurons in the Pars Compacta of the Substantia Nigra by N-Methyl-4-Phenyl-1, 2, 3, 6-Tetrahydropyridine, Proc Natl Acad Sci USA 80 (1983) 4546-4550 P. Jenner, et al., l-Methyl-4-Phenyl-l, 2, 3, 6-Tetrahydropyridine-Iduced Parkinsonism in the Common Marmoset, Neurosci Lett 50 (1984) 85-90 J.W. Langston, et al., Selective Nigral Toxicity After Systemic Administration of l-MethyI-4-Phenyl1, 2, 5, 6-Tetrahydropyridine (MPTP) in the squirrel monkey, Brain Res 292 (1984) 390-394 P. Barone, et al., Dopaminergic Mechanisms in Hemiparkinsonian Monkeys, Neurology 37 (1987) 1592-1595 C.E. Clarke, et al., Drug-Induced Dyskinesia in Primates Rendered Hemiparkinsonian by Intracarotid Administration of 1-Methyl-4-Phenyl-1, 2, 3, 6-Tetrahydropyridine (MPTP). J Neurol Sci 90 (1989) 307-314 M.E. Emborg-Knott and E.F. Domino, MPTP-Induced Hemiparkinsonism in Nonhuman Primates 6-8 Years after a Single Unilateral Intracarotid Dose, Exp Neurol 152 (1998) 214-220 J.L. Eberling, et al., A Novel MPTP Primate Model of Parkinson's Disease: Neurochemical and Clinical Changes, Brain Res 805 (1998) 259-262 N. Arai, et al., Differential Effects of Three Dopamine Receptor Agonists in MPTP-Treated Monkeys, J Neural Transm 10 (1995) 55-62 P.J. Bedard, et al., Chronic Treatment with Levodopa, but not Bromocriptine Induces Dyskinesia in MPTP-Parkinsonian Monkeys. Correlation with [3H] Spiperone Binding, Brain Res 329 (1986) 294299 S. Boyce, et al., Induction of Chorea and Dystonia in Parkinsonian Primates, Mov Disord5 (1990) 3-7 R.K.B. Pearce, et al., Chronic L-DOPA Administration Induces Dyskinesias in the 1-Methyl-4-Phenyl1, 2, 3, 6-Tetrahydropyridine-Treated Common Marmoset (Callithrix Jacchus), Mov Disord 10 (1995) 731-740 D.A. Di Monte, et al., Relationship Among Nigrostriatal Denervation, Parkinsonism and Dyskinesias in the MPTP Primate Model, Mov Disord 15 (2000) 459-466 P.J. Blanchet , et al., Dyskinesia and Wearing-Off Following Dopamine DI Agonist Treatment in Drug-Naive l-Methy-4-Phenyl-l,2,3,6-Tetrahydropyridine-Lesioned Primates, MovDisord 11 (1996) 91-94 B. Gomez-Mancilla and P.J. Bedard, Effect of D1 and D2 Agonists and Antagonists on Dyskinesia Produced by L-DOPA in l-Methyl-4-Phenyl-l, 2, 3, 6-Tetrahydropyridine-Treated Monkeys, J Pharmacol Exp Ther 259 (1991) 409-413 P.J. Blanchet,e/ al., Differential Effect of Selective D-l and D-2 Dopamine Receptor Agonists on Levodopa-Induced Dyskinesia in l-Methyl-4-Phenyl-1, 2, 3, 6-Tetrahydropyridine-Exposed Monkeys, J Pharmacol Exp Ther 267 (1993) 275-279 J.R. Taylor, et al., Dihydrexidine, a Full Dopamine DI Agonist, Reduces MPTP-Induced Parkinsonism in Monkeys, Eur J Pharmacol 199 (1991) 380-391 K. Shiosaki, etal., ABT-431: The Diacetyl Prodrug of A-86929, a Potent and Selective Dopamine D1 Receptor Agonist: In Vitro Characterisation and Effects in Animal Models of Parkinson's Disease, J Pharmacol Exp Ther 276 (1996) 150-160 O. Rascol, et al., Induction by Dopamine D1 Receptor Agonist ABT-431 of Dyskinesia Similar to Levodopa in Patients With Parkinson's Disease, Arch Neurol 58 (2001) 249-254 R.K.B. Pearce, et al., The Dopamine Re-uptake Blocker Brasofensine Reverses Akinesia without Dyskinesia in MPTP-Treated and L-DOPA-Primed Common Marmosets, Mov Disord in press (2001) M.J. Hansard, et al., The Antiparkinsonian ability of Bupropion in MPTP-Treated Common Marmosets, Brit J Pharmacol 125 (suppl) 373-80 L. Johnston et al., 5-Ht-1A-D2 receptor interaction in the reversal of Parkinsonism: SLV308, a novel D2 partial agonist with 5-HT-1A agonist activity, has anti-Parkinsonian effects in MPTP-treated common Marmomsets. British Pharmacological Science, Winter meeting, Birmingham, December 2001. R.W. Feenstra et al., SLV306, Drugs of the Future 26 (2001) 128-132 P-A. Loschmann, et al., Stereoselective Reversal of MPTP-Induced Parkinsonism in the Marmoset after Dermal Application of N-0437, Eur J Pharmacol 166 (1989) 373-380 N.M.J. Rupniak, et al., Antiparkinsonian Efficacy of a Novel Transdermal Delivery System for (+)PHNO in MPTP-Treated Squirrel Monkeys, Neurology 39 (1989) 329-335 J.T. Hutton, et al., Transdermal Dopaminergic D2 Receptor Agonist Therapy in Parkinson's Disease With N-0923 TDS: A Double-Blind Placebos-Controlled Study, Mov Disord 16 (2001) 459-463 T. Kanda, et al., Adenosine A2A Antagonist: A Novel Antiparkinsonian Agent that Does Not Provoke Dyskinesia in Parkinsonian Monkeys, Ann Neurol 43 (1998) 507-513

50

[75] [76] [77] [78] [79] [80]

[81] [82] [83]

[84]

[85]

P. Jenner / Experimental Models of Parkinson's Disease

T. Kanda, er al., Combined Use of the Adenosine A2A Antagonist KW-6002 with L-DOPA or with Selective Dl or D2 Dopamine Agonists Increases Antiparkinsonian Activity hut Not Dyskinesia in MPTP-Treated Monkeys, Exp Neurol 162 (2000) 321 -327 P.J. Blanchet, et al., Amantadine Reduces Levodopa-Induced Dyskinesias in Parkinsonian Monkeys. Mov Disord 13 (1998) 798-802 S.M. Papa and T.N. Chase, Levodopa-induced Dyskinesias Improved by a Glutamate Antagonis3t in Parkinsonian Monkeys, Ann Neurol 39 (1996) 574-578 B. Henry, et al., The a2-Adrenergic Receptor Antagonist Idazoxan Reduces Dyskinesia and Enhances Anti-Parkinsonian Actions of L-DOPA in the MPTP-Lesioned Primate Model of Parkinson's Disease. Mov Disord 14 (1999) 744-753 R. Grondin, et al., Noradrenoceptor Antagonism with Idazoxan Improves L-DOPA-Induced Dyskinesias in MPTP-Monkeys, Arch Pharmacol 361 (2000) 181-186 L. Smith, et al., Effect of Dose, Frequency of Administration and of Delivery to the Brain on the Ability of L-DOPA to Prime for Dyskinesia in Drug-naive MPTP-treated Common Marmosets: Effect of Entacapone Pretreatment. Mov Disord in press (2001) P.J. Bedard, et al., Dopamine Agonists as First Line Therapy of Parkinsonism in MPTP Monkeys. In: Beyond the Decade of the Brain. Volume 2, Dopamine Agonists in Early Parkinson's Disease. Wells Medical Limited 5 (1997) 101 -110 R.K.B. Pearce, et al., De Novo Administration of Ropinirole and Bromocriptine Induces Less Dyskinesia Than L-DOPA in the MPTP-Treated Marmoset, Mov Disord 13 (1998) 234-41 R. Grondin, et al., Cabergoline. a Long-Acting Dopamine D2 -like Receptor Agonist. Produces a Sustained Antiparkinsonian Effect with Transient Dyskinesias in Parkinsonian Drug-Naive Primates. Brain Res 735 (1996) 298-306 P. Blanchet, et al., Continuous Administration Decreases and Pulsatile Administration Increases Behavioural Sensitivity to a Novel Dopamine D2 Agonist (U-91356A) in MPTP Exposed Monkeys. J Pharmacol Exp Ther 272 (1995) 854-859 C.W. Olanow, et al., Continuous Dopamine-Receptor Stimulation in Early Parkinson's Disease. Trends Neurosci 23 (2000) (suppl) S117-SI26

Parkinson's Disease E. Ronken and G.J.M. van Scharrenburg (Eds.) IOS Press, 2002

51

SLV308: A Novel Antiparkinsonian Agent with Antidepressant and Anxiolytic Properties Andrew C. McCreary, Eric Ronken, Jan van der Heyden, Arnoud Herremans, Tinka Tuinstra, Steve Long and Guus van Scharrenburg Solvay Research Laboratories, Solvay Pharmaceuticals, Weesp, The Netherlands Key Words: SLV308; Parkinson's Disease; Partial Agonist; D2Receptors; 5-HT1A Receptors; Depression; Anxiety; Treatment

Introduction Parkinson's disease (PD) is a progressive neurodegenerative disorder primarily affecting dopaminergic neurones arising from the substantia nigra pars compacta projecting to the caudate nucleus and putamen. The neurodegeneration of this pathway results in the primary symptoms of PD such as postural rigidity, bradykinesia and a resting tremor [1]. In addition to the motor abnormalities that are seen, many Parkinsonian patients also suffer from secondary symptoms such as depression, sleep disturbances and dementia which can affect quality of life indices more than the primary motor abnormalities [2]. Current therapy for Parkinson's disease focuses upon alleviating the motor symptoms. The use of L-dihydroxypenylalanine (L-DOPA) as the precursor for dopamine synthesis is effective in improving the motor functioning of PD patients and more recently the effects of L-DOPA have been potentiated by the administration of the peripherally acting decarboxylase inhibitors such as carbidopa which reduce the systemic metabolism of L-DOPA. However, the effects of L-DOPA treatment decline with time and patients may experience characteristic fluctuations of treatment efficacy, known as "on-off' effects [3]. During the off-state patients suffer weakness, akinesia and "freezing" [3]; in addition patients may develop other side-effects as a consequence of long-term treatment. Such side effects include the development of psychoses, dyskinesias or dystonia [4]. More recently the ergots bromocriptine and pergolide, and the non-ergots such as ropinirole and pramipexole, have been used for the treatment of PD. However, a net improvement in side effect liability has not per se been described compared to those seen after prolonged LDOPA treatment. Moreover, increased blood pressure, nausea and vomiting have also been seen and can still become treatment limiting [4]. Although slow titration to clinically effective doses may allay some of these side effects. One potential hypothesis for the development of psychoses, dystonia and dyskinesias and the fluctuations in treatment efficacy is perhaps a consequence of the pulsatile plasma concentrations that are observed due to short half lifes of the compounds employed for treatment. It is to be expected that by maintaining steady-state plasma or brain concentrations of a drug there would be a reduction in the development of side effects (for review see Bronstein see this volume). Interestingly, it has been considered that another means to reduce potential sideeffects would be the use of partial dopamine D2 receptor agonists i.e. compounds that do not maximally stimulate dopamine D2 receptors. The use of partial dopamine D2 receptor agonists is expected to result in less adaptive changes as a consequence of treatment, i.e. it would act as an agonist to complete tonic dopamine levels when dopamine tone is low, but

52

A.C. McCreary et al. /A Novel Antiparkinsonian Agent with Antidepressant and Anxiolytic Properties

would also act as an antagonist under conditions of high dopamine tone. With such compounds pharmacotherapy of PD is expected to be improved significantly compared with present therapies, whereas the incidence of side effects such as nausea, vomiting, dyskinesia, dystonia, on-off phenomenon and development of psychoses would be avoided or substantially decreased. Comorbid major depressive disorder is seen in a high percentage of PD patients (ca. 60-70% or higher). In addition to the known dopaminergic neurotransmission related abnormalities that are known to occur in the brains of PD patients central serotonergic and noradrenergic function are also perturbed [5; Brooks this volume]. In preliminary positron emission tomographic (PET) studies PD patients have demonstrated a 20% reduction in binding of the silent 5-HT|A receptor antagonist 11C-WAY 100635 in the median raphe nucleus, a serotonergic somatic area, in patients presenting with or without a history of major depressive disorder. Moreover, postsynaptic 5-HT1A receptors density in the cortex is reduced but in only in patients with current or a history of major depressive disorder [6]. These data therefore suggest that a 5-HT1A receptor agonist will be valuable for the treatment of comorbid major depression in PD. We decided to develop candidate antiparkinsonian compounds combining partial dopamine D2 receptor agonism with full 5-HT1A receptor agonism. A range of compounds have been synthesised and tested in vitro and in animal models predictive of efficacy in Parkinson's disease and for potential antidepressant and anxiolytic-like efficacy. The nonergot 7-[4-methyl-l-piperazinyl]-2(3H)-benzoxazolone, monohydrochloride (SLV308; see figure 1) was selected as a compound sharing potent partial dopamine D2 receptor agonism in combination with weaker full 5-HT|A receptor agonism, ultimately providing an antiparkinsonian, antidepressant and anxiolytic-like profile [7]. In the present chapter we assess some of the preclinical effects of SLV308 and other reference compounds on putative indices of in vitro receptor binding and functional activity and in vivo behavioural effects.

.HCL

Figure 1. SLV308 - 7-[4-methyl-l-piperazinyl]-2(5//)-benzoxazolone.

monohydrochloride

In vitro receptor binding and functional studies In vitro receptor binding assays were performed with SLV308 and a range of reference compounds at several different receptors (table 2). SLV308 displayed appreciable affinity for dopamine D2, D1, D4, 5-HT1A, a 1 and a 2 adrenoceptors. In addition the clinically relevant antiparkinsonian drug lisuride was tested for binding at D2 receptors and was fund to have a pKi value of 8.9.

A. C. McCreary et al. /A Novel Antiparkinsonian Agent with Antidepressant and Anxiolytic Properties

53

Table 1. Affinities of SLV308 and reference compounds in receptor binding assays. Results are expressed as pKi values; spaces indicate data not determined. Compound

D2

D1

D4

a1

a2

5-HT1A

SLV308

7.5

7.5

7.6

7.3

7.1

7.5

Bromocryptine Talipexole Quinpirole Apomorphine

8.2 6.7 5.6 7.3 8.6 6.3 9 5.7

7.2 5.6 5.1 7.9

6.3 5.9 6.7 7.5 8.6 5.5 8.4

8.1

7.3 7.3 5.5 6.5

8.6 5.4 5.3 6.5 8 5.3 9.6 8.6

Terguride 3-PPP/Preclamol Roxindole (±)-8-OH-DPAT

6.2 <5.1

<5.5

<5.3

5.9 7 5.8 5.6

5.7 5.2

The effects of SLV308 were tested upon signal transduction cascade systems, using cell lines stably transfected with the target receptors. SLV308 acted as an agonist at D2 receptors, but in contrast to the full D2 receptor agonists quinpirole and talipexol, exerted approximately 50% agonism, similar to terguride (figure 2). In antagonism studies it was found that SLV308 and terguride also acted as partial antagonists when in the presence of the full agonist quinpirole. Thus, SLV308 acts both as an agonist (pEC50 = 8; a=0.55) and displayed antagonist properties (pA2 = 8.4) and is therefore considered a partial agonist. In addition to this activity at dopamine D2 receptors, SLV308 exerted full (100%) but weak agonist actions at 5-HT1A receptors (pEC50 = 6.3). Data for the effects of SLV308 and reference compounds at dopamine D2 receptors are shown in table 3. N.B. In the present studies 3H-spiperone was used as the radioligand for D2 receptors which tends to underestimate the affinity of full dopamine D2 receptor agonists such as pramipexol, quinpirole, apomorphine and ropinirole.

1E-12

1E-11 1E-10

1E-09

1E-08

1E-07

1E-06

1E-05

agonist concentration (M)

Figure 2. Dopamine D2,L receptor-mediated attenuation of forskolin-stimulated accumulation of cAMP. Forskolin-induced accumulation of cAMP was set to 100% and drug effects are expressed as a percentage. Data are presented as mean ± SEM of at least three independent observations.

54

A.C. McCreary et al. / A Novel Antiparkinsonian Agent with Antidepressant and Anxiolytic Properties

Table 2. The affinities and agonist effects of SLV308 and various reference compounds on dopamaine receptors. Data are expressed as pKi, pEC50 and % of agonism.

Compound

pKi

Dopamine D2 Receptor % Agonism pEC50

SLV308

7.6

8.0

55

Quinpirole Terguride Talipexol Bromocriptine Prairapexol Ropinirole

5.63

7.0 9.4 7.4 9.0 7.3 7

100 52 96 100 86 100

8.6 6.7 8.2 <6 <6

In vivo studies The substantia nigra pars compacta of rats was unilaterally lesioned with 6-hydroxydopamine (6-OHDA; a neurotoxin selective for dopaminergic neurones when it is administered with a noradrenaline reuptake blocker) and after a period of time their turning behaviour was tested after challenge with SLV308 or other reference dopamine receptor agonists. This turning rat model, as it is called, has predictive validity as a model for antiparkinsonian drugs [8]. It was demonstrated that SLV308 induced turning contralateral to the lesion side as expected from a putative antiparkinsonian agent and dopamine D2 receptor agonist. Figure 3 and see also table 5.

Figure 3. The Behavioural effects of SLV308 in the 6-OHDA unilateral turning model. * p<0.05 versus control.

In contrast to the agonist effects (turning contralateral to the lesion side) that were observed in the 6-OHDA turning rat model SLV308 acted as an antagonist at dopamine D2 receptors demonstrated by the antagonism of apomorphine-induced climbing behaviour (ED50 = 0.68

A. C. McCreary et al. /A Novel Antiparkinsonian Agent with Antidepressant and Anxiolytic Properties

55

and 0.46, po and ip, respectively). Terguride was less active than SLV308 in this model (ED50 = 1.7 mg/kg ip). SLV308 was without significant effect upon the reversal of reserpine-induced hypomotility although other D2 receptor agonists reversed this, for example apomorphine (least effective dose (LED) = 0.1 mg/kg sc), quinpirole (LED = 0.3 mg/kg ip) and talipexole (LED = 1 mg/kg ip). SLV308 was also tested for effects upon spontaneous locomotor activity in naive animals and, like the dopamine receptor agonists tested, a reduction in locomotor activity was found. In addition, SLV308 increased lower-lip retraction in rats (LED = 0.34 mg/kg po; extrapolated ED50 = 1.26 mg/kg po) which was antagonised by pretreatment with the silent 5-HT1A receptor antagonist WAY 100635. Table 3. The effects of SLV308 and reference compounds on turning behaviour in unilaterally 6-OHDA lesioned rats, data are expressed as the dose which induced 60 turns contralateral to the lesion side with a 2 standard-deviation confidence interval.

Compound SLV308 ip SLV308 po Apomorphine sc Quinpirole ip Talipexole ip Ropinirole po Pramipexol po Roxindol po

6-OHDA Turning Rat (ED60turns mg/kg) 0.036 (0.0024-0.051) 0.012(0.005-0.019) 6.86(4.24->10) 0.14(0.126-0.17) 2.04 ((1.56-5.92)

Subsequently, SLV308 was tested in the differential rate of low reinforcement 72 sec (DRL72S) and forced swim paradigms, which predict antidepressant efficacy. In the DRL72S procedure SLV308 increased the number of responses and reduced the number of reinforcements (figure 4), an effect consistent with antidepressant efficacy. Similar effects were seen with the antidepressant fluvoxamine (Luvox; data not shown). Talipexole and terguride also had an effect in this model (table 4). SLV308 acted as a putative antidepressant in the forced swim paradigm along with quinpirole and talipexole, but not terguride (table 4). The data are expressed as the dose at which results in an 8 second decrease in mean immobility in the first 3-minute period of testing. SLV308 was tested in the adult ultrasonic vocalisation (AUV) procedure. SLV308, along with the clinically relevant anxiolytics diazepam, chlordiazepoxide and alprazolam (data not shown), and, consistent with the other dopamine receptor agonists tested, reduced the number of ultrasonic calls when rats were introduced into an environment where they had previously received an electric shock. N.B. Presently all doses of SLV308 are represented as the doses of the monohydrochloric acid salt.

56

A.C. McCreary et al. /A Novel Antiparkinsonian Agent with Antidepressant and Anxiolytic Properties

Figure 4. The effects of SLV308 (ip) in the DRL72S paradigm. SLV308 induced an increase in response rate and a reduction in reinforcement rate which is indicative of an antidepressant profile. * represents significance from control.

SLV308(n=12) 150- -

-^-resp40

125- -

~*~"*'*J

15 */

100-

--

12

-9

*/

75-

-6

5025-

-3

0-

-0 I oortr

0

0.1

Q3

1

doses(mg/kgip) Table 4. The effects of SLV308 and reference compounds in models of antipderessant and anxiolytic. DRL72S, data expressed as antidepressant profile; the forced swim test, data expressed as the dose at which results in an 8 second decrease in mean immobility in the first 3-minute period of testing; the adult ultrasonic vocalisation procedure, data expressed as the ED50 to reduce ultrasonic vocalisations. Data in parentheses indicate associated confidence intervals. NE, no effect. Compound

SLV308ip SLV308po Clonidine Apomorphine Quinpirole Talipexole Terguride

Forced Swim (ED8sec mg/kg ip)

Ultrasonic Vocalisations (ED50mg/kg)

Not tested

0.2 (0.04-0.9) 0.2 (0.04-03)

0.0046 (0.001-0.008) 0.008(0-0.017)

0.01 NE to 8.0 NE to 1.0 10.0 1.0

>l >3 sc 0.09 (0.05-0.2) 0.3 (0.2-0.4) >3

0.05 (0.03 -> 0.3) 0.01 (0.005 - 0.02) sc 0.007(0-0.01) 0.01 (0.007 - 0.02) 0.04 ( 0 - > 0.1)

DRL72S Antidepressant Profile (mg/kg ip) 1.0

Summary and Discussion SLV308 displays a unique receptor binding profile with appreciable binding affinities at a number of monoaminergic receptors, but most prominently the dopamine D2 and 5-HT1A receptors. In vitro studies demonstrated that SLV308 is a partial D2 receptor agonist and weak but full 5-HT1A receptor agonist, data which were supported with behavioural tests. SLV308 acted as D2 receptor antagonist when tested for the ability to inhibit apomorphineinduced climbing. Terguride was also found to be active but at doses approximately three times less than those of SLV308 when tested orally. SLV308 also possesses potent dopamine D2 receptor agonist properties in the 6-OHDA lesioned turning rat model (ED60turns = 0.036 mg/kg orally) and was extremely potent compared with the other clinically relevant compounds that were tested. In experiments conducted in the MPTP

A. C. McCreary et al. /A Novel Antiparkinsonian Agent with Antidepressant and Anxiolytic Properties

57

treated marmoset model, a model with predictive validity as a model for PD [9; Jenner this volume] SLV308 was also very potent in reversing the effects of MPTP-ablation [7]. SLV308 was not active in the reserpinised rat model, probably due to partial agonist nature of the SLV308, since, although dopamine tone is low, the system per se is not a supersensitised system like that seen in the turning rat or MPTP-treated marmoset and the intrinsic activity of SLV308 is not high cf. the full agonists tested. SLV308, like the other compounds tested, caused a reduction in spontaneous locomotor activity. More likely this gradation of effects across the different behavioural paradigms is caused by the nature of the reduction in tonic dopamine levels. For example, 6-OHDA and MPTP causes an ablation of dopamine (where pronounced antiparkinsonian-like effects of SLV308 are seen) whereas reserpine pretreatment (where SLV308 was devoid of effects) causes a dopamine depletion of approximately 70% and in naive animals which have normal dopaminergic tone a reduction in locomotor activity was seen with SLV308. This might actually suggest that SLV308 will nicely discriminate against the tonic levels of dopamine in PD patients and can readily treat the early through to the late stages of PD. The D2 receptor partial agonist action of SLV308 have been supported by previous neurochemical findings. SLV308 failed to act as an agonist in superfusion experiments conducted in striatal slices, yet in the presence of quinpirole an antagonist action was revealed. In vivo microdialysis studies, however, suggested that SLV308 was able to reduce dopamine overflow in the nucleus accumbens suggesting an agonist action in this test system [7]. Clearly then, SLV308 is a true partial agonist and under low dopaminergic tone conditions, such as one would expect to find in the nigrostriatal pathway in PD patients, SLV308 will mimic the effects of endogenous dopamine. Adaptations beyond the nigrostriatal system, e.g. those involving the globus pallidus and the subthalamic nucleus, are expected not to take place after prolonged treatment with SLV308. A low propensity for dystonia, dyskinesias and the development of psychoses and a delay in the onset or even abolishment of on-off symptomatologies, is predicted for chronic SLV308 administration. SLV308 is also expected to induce less aberrant motor stimulation which would lead to less restlessness reduce nocturnal wakefulness and reduce daytime sleeping. Of all of the reference compounds tested terguride was the only compound that possessed partial agonist properties at the dopamine D2 receptor. However, it must considered that terguride is an ergot and will possess associated side-effect problems such as nausea, vomiting, sleep disturbances and an increase in blood pressure. Moreover in the present studies terguride was not as potent as SLV308. SLV308 therefore clearly stands out as a premier potential medicament for the treatment of the motor symptomatologies of PD. SLV308 is a weak but full 5-HT1A receptor agonist in vitro and these actions were verified in vivo using the lower lip retraction paradigm. The effects of SLV308 were antagonised by the silent 5-HT1A receptor antagonist WAY 100635. It is more than likely that this property might underly the unique properties of SLV308 in paradigms that model depression and anxiety. In this respect SLV308 has a unique profile insofar as the reference compounds demonstrated a lack of consistent concrete antidepressant effects. Apomorphine failed to show efficacy in either the DRL72S or the forced swim models and while quinpirole was active in the swim test it was devoid of effects in the DRL72S paradigm. Talipexole failed to show good potency in the DRL72S (cf. the forced swim paradigm) and terguride failed to show efficacy in the forced swim paradigm. In conclusion, SLV308 demonstrated superior efficacy with high potency in tests modeling antidepressant-like effects. In the AUV procedure all of the compounds demonstrated putative anxiolytic actions, but what is clear, with the exception of quinpirole, is that, SLV308 demonstrated superior potency in this model. These putative antidepressant effects of SLV308 are extremely important given the consideration that between 60-70% or more of PD patients present with comorbid symptoms of major depressive disorder, which has a pronounced

58

A. C. McCreary et al. /A Novel Antiparltinsonian Agent with Antidepressant and Anxiolytic Properties

effect on quality of life in PD patients [2]. The role of 5-HT1A receptors has been implicated in the depressive symptoms seen in PD (see Introduction; Brooks this volume) and it is well known that 5-HT1A receptor agonists are efficacious in preclinical models of depression, such as those presented here and, moreover, that buspirone (a 5-HT1A receptor agonist) has been widely used in the clinic for the treatment of depressive and anxiety disorders. Taken together the results from preclinical studies suggest that SLV308 is a novel agent possessing a unique binding profile compared with other known PD therapies. In addition to treating movement abnormalities (as predicted from the turning rat and MPTPtreated marmoset models) SLV308 is likely to possess antidepressant and anxiolytic-like effects which will be beneficial to PD patients presenting with such comorbid symptomatologies. SLV308 is expected to have a low propensity to induce psychoses, dystonia, dyskinesias. In addition SLV308 will be devoid of the side-effects commonly seen with ergot compounds.

References

[3] [4] [5]

[6]

[7] (8] [9]

O. Hornykiewicz, Dopamine and brain function. Pharmacol Rev 18 (1966) 925-964. A. Schrag. et al., What contributes to quality of life in Parkinson's disease. J. Neurol. Nerosurg. Psychiatry. 69 (2000)308-312. S. Factor, Dopamine agonists in Parkinson's Disease and Parkinsonian Syndromes: The Medicinal Clinics of North America. (Eds) Ster M. and Hurtig M. 83:2. (1999). Published W.B. Saunders. D. Lambert and C. Waters, Comparitive tolerability of the newer generation antipsychotic agents: a review. Drugs and Aging 16:1, (2000) 55-65. W. Birkmayer and P. Rieder, Biochemical changes in Parkinson's Disease. In Parkinson's Disease Biochemistry, Clinical Pathology, and Treatment. Translated by Reynolds G, Blau G and Reynolds L. Springer-Verlag, New York (1980). M. Doder et al., Brain serotonin HT1A receptors in Parkinson's Disease with and without depression measured by positron emission tomography and 11C-WAY100635. Movement Dis 15: Supp 3. (2000) 213. R.W. Feenstra, et al., SLV308. Drugs of the Future. 26 (2001) 128-132. U. Ungerstedt, Postsynaptic supersensitivity after 6-hydroxy-dopamine-induced degeneration of the nigro-striatal dopamine system, Acta Physiol Scand 82 (Suppl.367): (1971) 69-93. R.S. Burns, et al., A primate model of parkinsonism: selective destruction of dopaminergic neurons in the pars compacta of the substantia nigra by MPTP. Proc Natl Acad Sci USA, 80 (1983) 4546-4550.

Parkinson's Disease E. Ronken and G.J.M. van Scharrenburg (Eds.) IOS Press, 2002

59

Adenosine A2A Receptor Antagonists and Parkinson's Disease Peter J. Richardson Department of Pharmacology, University of Cambridge, Tennis Court Road Cambridge CB2 JQJ, UK Abstract. The adenosine A2A receptor is one of four cloned receptors known to mediate the regulatory actions of adenosine in many tissues. This receptor is expressed at particularly high levels in the striatum, primarily in striatopallidal output neurons and cholinergic interneurons. There are many convergent lines of evidence which suggest that adenosine A2A receptor antagonists could serve as novel, non-dopamine based, therapeutic agents for the amelioration of Parkinson's Disease. These range from the molecular to the behavioural and are briefly reviewed here. Keywords: adenosine A2A receptor, symptomatic relief, striatonigral neurons

1. Introduction Parkinson's disease is caused by the degeneration of dopaminergic nigrostriatal neurons of the basal ganglia, which results in bradykinesia, tremor, and rigidity. Currently symptomatic relief is the mainstay of therapy, since there is no known drug that prevents this degeneration [1]. There is an unmet need for anti-parkinsonian agents that do not exhibit the side effects (on/off effects, dyskinesias etc) associated with L-DOPA, and dopamine receptor agonist, therapy. Since dyskinesias are observed with dopamine based therapies, it seems reasonable to try drugs with different modes of action in order to avoid them. The major non-dopamine therapy currently available is the use of acetylcholine receptor antagonists, but these are restricted by their side effects and their low efficacy at the doses used. Adenosine is a ubiquitous metabolite present in the CSF and plasma at concentrations between 10 and 100 nM [2]. The precise mechanisms controlling extracellular adenosine concentrations are incompletely understood, although it is accepted that breakdown of cellular ATP, perhaps when energy demand exceeds supply, results in high intracellular adenosine. The principal immediate source of adenosine is AMP although hydrolysis of Sadenosyl homocysteine can contribute to adenosine production. This then leaves the cell via nucleoside transporters, resulting in elevated levels of interstitial adenosine (between 1 and 10mM in ischaemic conditions, [3]). Indeed adenosine is often understood to serve as a local homeostatic regulator i.e. in many systems adenosine tends to inhibit cell function (thereby saving energy) while also causing vasodilation so increasing local energy supply [4]. A significant amount of extracellular adenosine is also derived from ATP released from cells. In the central nervous system, this may take the form of co-release with classical transmitters such as dopamine, nor-adrenaline or acetylcholine (ACh) [8]. The released ATP is degraded by a series of ectoenzymes present on the surface of many cells, the rate limiting step in the production of adenosine being the ecto-5' nucleotidase [9]. It has also been widely reported that glutamate receptor stimulation causes the release of adenosine

60

P.J. Richardson / Adenosine A2A Receptor Antagonists and Parkinson's Disease

and ATP from neurons [10-13]. Some specific instances where glutamate stimulated release may be important in normal striatal function are discussed below. In general, the relative importance of adenosine formed intracellularly or extracellularly appears to vary under different conditions. For example it is likely that adenosine derived from released ATP is important under "normal" conditions, whereas that derived from intracellular sources is more important under conditions of stress e.g. hypoxia. The half-life of extracellular adenosine is approximately 10 seconds, implying a highly dynamic control of interstitial adenosine. The major process regulating the concentration of this nucleoside is uptake into surrounding cells via the sodium-dependent and equilibrative nucleoside transporters, which have been recently cloned [5-7]. After uptake, the nucleoside may be phosphorylated by adenosine kinase or broken down to inosine by adenosine deaminase. Some measurements have been made of striatal adenosine levels in models of Parkinson's disease, but there has been no consistent evidence for a detectable change when compared to the normal striatum [14]. This does not however invalidate the potential of adenosine receptors as novel targets for drugs in the therapy of Parkinson's disease since, as outlined below, the dopamine D2 receptor serves to oppose the actions of the A2a receptor in the normal striatum. Thus in the absence of dopamine there is an apparently excessive A2a receptor drive which contributes to the motor abnormalities characteristic of this disease.

2.

Parkinson's Disease and neuronal interactions of the striatum

In the normal striatum the two major output pathways (to the pailidum and substantia nigra, see Figure 1) are in balance. However this balance is lost in the absence of dopamine. This is reflected in an increase in proenkephalin expression in the striatopallidal neurons which suggests an increase in the activity of this pathway, and a reduced expression of Substance P which suggests a fall in the activity of the striatonigral pathway [15,16]. Any attempt to design novel therapies should therefore be based on an understanding of the dominant influences on these two pathways, and the identification of systems open to pharmacological intervention which could restore the balance. In addition changes in expression of these neuropeptides can be used to assay the potential efficacy of new therapeutic compounds in vivo. The major regulators of these two striatal projection systems are an excitatory drive from the cerebral cortex, balanced by inhibition via GABAergic interneurons. Another potential source of GABA in the striatum is the extensive recurrent collaterals of the projection neurons (which together account for approximately 90% of striatal neurons). These recurrent collaterals synapse in the immediate area of the dendritic trees of the neurons from which they arise [17], but their role is unclear. It was expected that a given collateral served to feedback inhibit the cell from which it arises and/or its immediate neighbours. However there is little direct evidence for this [18, 19]. Other regulators of the projection neurons include dopamine from nigrostriatal neurons (which degenerate in Parkinson's disease), acetylcholine from large interneurons as well as a number of other agents including neuropeptides present in some GABAergic interneurons (e.g. somatostatin, NPY) and cannabinoids. There are also significant striatal afferents that release histamine and serotonin. 2.1

Glutamate mediated stimulation of striatal projection neurons

The dominant excitatory input into the striatum is from the cortex [20], mainly layer V. The glutamatergic synapses on the striatal projection neurons are located on the spines of the

P.J. Richardson / Adenosine A2A Receptor Antagonists and Parkinson's Disease

61

dendrites, while other inputs e.g. dopamine and GABA tend to be localised to the spine necks or the dendritic shafts [17]. This glutamatergic input also stimulates the interneurons of the striatum with most such cells expressing a wide variety of ionotropic glutamate receptors [17, 21, 22]. Repetitive activation of corticostriatal neurons can elicit long term potentiation (LTP) or long term depression (LTD) in the striatum [21]. A number of striatal modulators regulate the release of glutamate from the corticostriatal nerve terminals including adenosine (via the adenosine A1 receptor), dopamine (via the dopamine D2 receptor) and ACh via muscarinic receptors [28]. 2.2

Dopamine mediated modulation of striatal projection neurons

The influence of dopamine in the striatum is both excitatory and inhibitory. In situ hybridisation and single cell RT-PCR analyses have revealed the expression of both D1 and D2 receptors in the projection neurons, with most of the striatonigral neurons expressing the D1 receptor, and most of the striatopallidal neurons expressing the D2. There is evidence that most neurons express both Di-like and D2-like receptors [24, 75]. Since dopamine depletion, and dopamine receptor antagonists, cause an increase in the expression of enkephalin in the striatopallidal neurons, and a decrease in expression of Substance P in the striatonigral neurons [15,16], it appears that the predominant influence of the D2 receptor is to inhibit striatopallidal neuronal firing while the D1 receptor facilitates striatonigral firing.

Figure 1. Diagram of the dominant influences on striatal output neurons and subsequent processing in the basal ganglia. MSN: medium spiny neuron, the predominant neuron of the striatum comprising both major output neuron types: the striatopallidal neuron expressing enkephalin (enk) and the D2 and A2A receptors, and the striatonigral neuron expressing Substance P (SP), dynorphin (Dyn) and the D1 and A1 receptors. These

62

P.J. Richardson /Adenosine A2A Receptor Antagonists and Parkinson's Disease

constitute the first stages of the indirect and direct basal ganglia pathways respectively. The indirect pathway via the Gpe (external segment of the globus pallidus), and the STN (subthalamic nucleus) to the SNr (substantia nigra pars reticulata)/Gpi (internal segment of the globus pallidus) and then to the thalamus inhibits movement. In contrast the direct pathway via the SNr/Gpi to the thalamus excites the thalamocortical neurons and thus increases movement. The recurrent collaterals of the striatonigral neurons are hypothesised to release both Substance P, which stimulates cholinergic interneurons. and GABA which is presumed to inhibit them. The corticostriatal excitatory input onto all the cell types shown has been omitted for clarity. Filled arrows: inhibition, empty arrows: excitation, shaded arrow: mixed. Note 1. opposing actions of dopamine and ACh on the MSNs 2. the A2A receptor stimulates the release of ACh 3. the A2A receptor inhibits the release of GABA from recurrent collaterals and/or GABA interneurons which may contribute to the overactivity of the striatopallidal neurons in the absence of dopamine.

In addition these receptors influence the excitatory drive from the corticostriatal neurons. As mentioned above the terminals of these neurons bear inhibitory D2 receptors [25], while the D| receptor increases the responses of projection neurons to NMDA [26]. Combined stimulation of D1 and NMDA receptors also increases extracellular adenosine which in turn reduces further glutamate release (via A1 receptors on the corticostriatal nerve terminals). Thus DI receptors have a dual effect - increasing the sensitivity of neurons to glutamate while also inducing a feedback which reduces the amount of glutamate released from the corticostriatal terminals [11]. In the absence of dopamine, the lack of D2 receptor activation would increase glutamate stimulation of the striatum, and the absence of DI receptor stimulated adenosine release may exacerbate this effect. 2.3

GABA mediated inhibition of striatal projection neurons

The GABAergic interneurons are probably the major inhibitory influence on the projection neurons since, as mentioned above, there is no detectable effect of the other major source of GABA. the recurrent collaterals [18]. There are 4 major classes of interneuron three of which release GABA and can be distinguished on the basis of their expression of somatostatin, parvalbumin and calretinin [27]. Electrophysiological analysis also readily distinguishes two types which have profound influences on striatal neuronal function [19]. Dual recording suggests that many of the GABA interneurons are electronically coupled and could serve as inhibitors of large numbers of projection neurons, with each interneuron potentially innervating (and therefore influencing) one hundred projection neurons. It has been suggested that these interneurons act as feed-forward inhibitors, stimulated by corticostriatal afferents, and thus they may serve to inhibit weakly stimulated projection neurons, allowing only strongly stimulated neurons to fire [19]. This could focus the widespread input from the cortex prior to further processing in other basal ganglia nuclei. Since these authors failed to observe any evidence of GABA mediated recurrent collateral mediated control of the interneurons. the role of the collaterals remains unclear. 2.4

ACh mediated control of striatal projection neurons

Cholinergic interneurons comprise approximately 2% of the total striatal neuron number. they have large soma (diameter of approximately 30mm) and extensive dendritic and axonal arborisations. The precise role of these neurons has been difficult to determine although the ability of dopamine to inhibit ACh release in the striatum (via D2 receptors) lent support to the concept of an ACh/dopamine balance in the normal striatum. However the D1 receptor increases ACh release and under conditions of dopamine depletion there is no overall change in ACh levels in the striatum [28, 29]. Thus it would appear that the dopamine mediated regulation of these neurons is in balance under normal conditions, although it is likely that the fine control of these neurons is disturbed in the absence of dopamine. It is

P.J. Richardson /Adenosine A2A Receptor Antagonists and Parkinson's Disease

63

interesting that the Substance P (NK|) receptor is selectively expressed in these interneurons where it causes depolarisation [31], thus suggesting a link between striatonigral neuronal firing and ACh mediated modulation of projection neurons. Indeed one could hypothesise that Substance P release from recurrent collaterals contributes to the integration of striatal function by excitation of the cholinergic interneurons. The ability of ACh receptor agonists to stimulate c-fos expression in striatopallidal neurons and to inhibit it in striatonigral neurons, suggests that ACh has precisely opposite effects to those of dopamine on these neurons [28]. More evidence that ACh antagonises striatal dopamine was obtained when the cholinergic interneurons were ablated using an immunotoxin [30]. Unilateral destruction of these interneurons caused contralateral turning, with and without stimulation with apomorphine, suggesting a supersensitivity to dopamine receptor agonists on the lesioned side. In addition decreased expression of proenkephalin and increased expression of Substance P was observed suggesting excessive activity of the striatonigral neurons and inhibition of the striatopallidal neurons. In time these effects were compensated for by reduction in dopamine receptor expression, although the mice remained abnormally sensitive to dopamine agonists. This clearly demonstrated the opposing effects of dopamine and ACh in the striatum, particularly on the activity of the projection neurons (see Figure 1). It also suggested that the ACh neuron plays a central role in the phenotype arising from striatal dopamine depletion. Thus any agent capable of reducing striatal ACh neurotransmission may be of use in the treatment of Parkinson's disease. 3.

The adenosine A2A receptor and its actions in the striatum

There are four cloned adenosine receptors designated A1, A2A, A2B and A3, [32] all of which have 7 transmembrane domains and are coupled to G-proteins. They are widely expressed in many tissues [33] and in the nervous system regulate the release of most neurotransmitters, including GABA, glutamate, ACh, nor-adrenaline, and dopamine [3438]. The adenosine A2 receptor was originally identified as a striatal membrane receptor capable of stimulating adenylyl cyclase. In time it became apparent that there was at least two A2 receptors on the basis of antagonist studies, and this subsequently confirmed by cloning when both the A2A and A2B receptors were sequenced [39,40]. Both A2 receptors are linked to Gs and thus the stimulation of adenylyl cyclase, with the A2A receptor corresponding to the receptor previously detected in striatal membrane preparations. This receptor is particularly highly expressed in the striatum, nucleus accumbens and olfactory tubercle as detected by in situ hybridisation and iminunohistochemistry [41, 42]. The more sensitive method of RT-PCR has revealed that this receptor has a much wider distribution than that suggested by in situ hybridisation, being detected in all brain areas tested (including the cortex, cerebellum, thalamus, hippocampus as well as the striatum) and in peripheral tissues such as lung, skin, eye, bladder [33]. In the striatum the receptor mRNA is detected predominantly in the GABA/enkephalin containing striato-pallidal neurons [43]. At present the only other striatal cell type known to express this receptor is the large cholinergic interneuron [33, 44, 45]. In a recent study it was shown that kainate receptor induced suppression of GABAergic transmission in the striatum is mediated by the release of adenosine that acted on presynaptic A2A receptors. Since there is little detectable GABAergic transmission from recurrent collaterals of projection neurons [18], the authors concluded that GABAergic interneurons may also express the A2A receptor [13]. The archetypal adenosine A2A receptor agonist is CGS21680, although others (e.g. APEC) have been developed. The first adenosine receptor antagonists were xanthines (e.g. caffeine and theophylline) which show little selectivity between the A1, A2A and A2B

64

P.J. Richardson /Adenosine A2A Receptor Antagonists and Parkinson's Disease

receptors. Subsequent modifications of the xanthine nucleus led to the development of antagonists with A2A receptor selectivity, such as KF17837 which shows 60-fold selectivity for the A2A receptor over the A1 [46]. More recently, a number of other xanthine and nonxanthine antagonists with A2A receptor selectivity have been developed [47]. The most selective of these antagonists is a non-xanthine ZM241385 [48], but others including KW6002 (a derivative of KF17837) show sufficient selectivity for use in vitro and in vivo. 3.1

Control of ACh release in the striatum

The A2A receptor was shown to be present in the striatal cholinergic interneurons on the basis of its ability to stimulate the release of ACh from synaptosomes [49]. Since this receptor was thought to be expressed solely in the striatum, and because it stimulated ACh release, it was initially suggested that antagonists could serve as striatal specific anticholinergic agents, and thus used in PD therapy [50]. Since these early days this stimulatory effect of the receptor has been replicated in nerve terminals, brain slices, and in vivo by microdialysis [45,51]. In striatal cholinergic nerve terminals the A2A receptor was shown to activate separate but interacting signal transduction pathways, both of which increased the release of ACh. Thus activation of protein kinase A resulted in an increased release due to a stimulation of a P-type calcium channel. However when protein kinase A was inhibited, another A2A receptor stimulated pathway was revealed which operated via protein kinase C and an N-type calcium channel [52]. One aspect of this dual signalling in the cholinergic interneuron is the implication that when cAMP levels are reduced in these nerve terminals (e.g. by dopamine D2, or adenosine A1, receptor stimulation) the A2A receptor would still increase ACh release. In other systems the A2A receptor has also been shown to regulate MAPkinase and KAtp channel activity [53,54]. In vivo the A2A receptor agonist CGS21680 increased extracellular levels of ACh in the rat striatum, in the presence and absence of GABA antagonists, showing that this effect was not secondary to a modulation of GABA release (see below) [51]. In the absence of dopamine the effect of this agonist was greater which is consistent with the A2A and D2 receptors having opposite effects on the cholinergic interneuron. 3.2

Control of striatal GABA release

After the cloning of the A2A receptor the major site of receptor mRNA expression was clearly shown to be the striatopallidal output neuron, which also expresses the dopamine D2 receptor and enkephalin. Analysis of the effect of A2A receptor agonists on GABA release showed that in contrast to the ACh neuron, this receptor inhibited the release of GABA from striatal synaptosomes [55], an effect not mediated by cAMP [56]. Similarly analysis of the effect of A2A receptor agonists on IPSCs in striatal slices also clearly showed that the receptor inhibited the release of GABA onto medium spiny neurons [57], although in these experiments permeable cAMP analogues mimicked the effect of A2A receptor stimulation. It may be that, as in the cholinergic nerve terminal, this receptor operates more than one signalling pathway when regulating GABA release. Given the high levels of expression of the A2A receptor in the striatopallidal neurons it was generally assumed that this was an effect on the recurrent collaterals of these neurons [58], consistent with the observation that these recurrent collaterals synapse within the dendritic fields of striatopallidal neurons [17]. However since it has been difficult to show inhibition by recurrent collaterals of the projection neurons [18. 19], it has been suggested that the GABAergic terminals most affected by the A2A receptor may he those of interneurons rather than the striatopallidal recurrent collaterals [13]. It is probable that the site of action of the A2A receptor will he determined when attempts have been made to determine whether the GABAergic

P.J. Richardson / Adenosine A2A Receptor Antagonists and Parkinson's Disease

65

interneurons express this receptor. In vivo microdialysis has revealed that lesioning of the nigrostriatal tract increases pallidal GABA levels, an effect consistent with the increased activity of the striatopallidal neurons. This increase in pallidal GABA can be ameliorated by striatal infusion of D2 agonists, an effect counteracted by CG521680 [59], illustrating again the antagonistic effects of the A2A and D2 receptors. The A2A antagonist KW6002 reduced the increase in pallidal GABA [60], an effect consistent with the ability of similar antagonists to reduce the activity of the striatopallidal pathway as measured by proenkephalin expression [16, 58]. 3.3

Control of glutamate responses in the striatum

There are conflicting reports as to the ability of A2A receptor agonists to control glutamate release in the striatum. Thus CGS21680 has been reported to increase glutamate concentrations in vivo [61], but most reports describe solely an A1 mediated inhibition of glutamate release at the corticostriatal nerve terminal [62]. However it is interesting that the A2A receptor is expressed on the spines of the striatopallidal neurons [42], and it has been reported that receptor activation can control the responses of these neurons to glutamate, by modulating the opening of NMDA receptor channels. It has also been suggested that blockade of the A2A receptor will affect intracellular signalling pathways of the striatopallidal neurons, and thus the responsiveness of these cells to the glutamate inputs [63]. 3.4

A2A/D2 receptor interactions

One of the early indications that the A2A receptor may be a suitable target for Parkinson's disease therapy was the observation by Ferre et al that A2A receptor agonists reduce the affinity of the dopamine D2 receptor for agonists [64]. This effect was mediated within the membrane, probably by a direct interaction. Subsequently it was shown that A2A receptor stimulation also opposed the action of dopamine D2 receptors at the level of signal transduction, an expected result given their opposite effects on adenylyl cyclase [reviewed in 23]. Besides showing the opposing influences of these two receptors, which suggested that A2A receptor blockade would be pro-dopaminergic and thus anti-Parkinsonian, this led to the proposal that the effect of A2A antagonists in models of PD was primarily due the relief of the antagonistic effect of the A2A receptor on D2 receptor function i.e. A2A receptor blockade would permit residual dopamine to be more effective in stimulating D2 receptors. In a D2 receptor deficient mouse, the elevated expression of proenkephalin, and reduced expression of Substance P, were reversed by the A2A receptor antagonist KW6002. This antagonist also reversed the locomotor deficits in these animals [65]. Therefore although the A2A and D2 receptors have opposing effects, the benefits of A2A receptor blockade can be observed in the absence of D2 receptors. In summary in vitro and in vivo studies are broadly consistent with the A2A receptor regulating the release of ACh and GABA in the striatum. Since ACh opposes the actions of dopamine on the striatal projection neurons, blockade of the A2A receptor may have dopamine-like effects. Similarly since the A2A receptor inhibits the inhibitory effect of GABA onto striatal projection neurons, blockade of the receptor would reduce the activity of these neurons by increasing GABA action. This may be particularly important in reducing the over activity of the striatopallidal projection neurons seen in the absence of dopamine. Finally the antagonistic effects of the A2A and D2 receptors also suggests that blockade of the A2A receptor would have pro-dopaminergic effects.

66

3.5

P.J. Richardson /Adenosine A2A Receptor Antagonists and Parkinson's Disease

Control of GABA release in the globus pallidus

The only other part of the basal ganglia in which the role of the A2a receptor has been studied in any detail is the external segment of the globus pallidus i.e. that area innervated by the striatopallidal neurons which express this receptor. In this area conflicting results have been obtained with one group reporting inhibition of GABA release by the A2a receptor in the pallidum [66] and another stimulation of release [67]. However the preparations and conditions were widely different, which may explain the discrepancy. In order to determine which process is the more physiologically relevant (i.e. inhibition or stimulation) electrophysiological analysis is required. Of particular interest is the possibility that the stimulation of release is the more physiological, which would suggest that two terminals of the same neurons (i.e. the pallidal and striatal terminals of the striatopallidal neurons) may regulate the release of transmitter in opposite directions. In addition under these circumstances blockade of the A2A receptor would reduce GABA release in the pallidum, and thus reduce the influence of the striatopallidal neuron which is overactive in Parkinson's disease.

4.

In vivo effects of A2A receptor ligands

Apomorphine induced turning in unilaterally lesioned rats (6-hydroxydopamine lesioning of the nigrostriatal tract) can be profoundly potentiated by methylxanthines, an effect thought to be due to blockade of striatal A2 receptors [68]. After the cloning of the A2A and A2A receptors, and the development of A2A receptor selective ligands, it was shown that A2A receptor agonists could counteract dopamine receptor agonist induced turning in a manner sensitive to both A2A and ACh receptor antagonists [69, 70, 71]. The administration of A2A receptor agonists, and dopamine receptor antagonists, cause sedation, locomotor depression and catalepsy, effects blocked by the co-administration of A2A receptor antagonists. Interestingly in haloperidol treated mice, A2A antagonists and L-DOPA have synergistic effects in reversing the catalepsy [72]. The opposing actions of adenosine A2A receptors and dopamine receptors on neuronal activity was outlined in earlier sections. These effects are also reflected in behavioural studies, with dopamine agonists and A2A receptor antagonists having similar effects. However the most significant effect of A2A receptor antagonists is seen in MPTP treated marmosets, where KW6002 is able to significantly improve the motor deficits as measured by locomotor activity and disability. In particular, repetitive dosing with KW6002 does not cause dyskinesia, even in animals exhibiting dyskinesia with L-DOPA [73], and the synergistic effect with dopamine agonists was again observed. These data are consistent with a wealth of studies not mentioned here which show synergistic and beneficial effects of A2A receptor antagonists in many models of PD [reviewed in 74].

5.

Conclusion

From the above discussion it can be seen that the adenosine A2A receptor, via a number of mechanisms involving GABA and ACh, antagonises the effects of dopamine receptor agonists at the molecular, cellular and whole animal levels. Furthermore, in animal models of Parkinson's disease, A2A receptor antagonists have beneficial effects and can be used synergistically with current therapies such as L-DOPA. In monkeys with L-DOPA induced dyskinesias. A2A antagonists have beneficial effects on disability and movement without

P.J. Richardson /Adenasine A2A Receptor Antagonists and Parkinson's Disease

67

causing dyskinetic movements. The implication of all these studies is that A2A antagonists should be beneficial in the treatment of Parkinson's disease. References [1] [2] [3] [4] [5] [6] [71 [8] [9] [10]

[11]

[12] [13] [14] [15]

[16] [17] [18] [19] [20] [21] [22] [23] [24] [25]

J. Kopin, Pharmacology of Parkinson's Disease therapy; an update Ann. Rev. Pharmaco.l Toxicol. 32 (1993)467-495 J.P. Headrick, Ischemic preconditioning: bioenergetic arid metabolic changes and the role of endogenous adenosine, J. Mol Cell Cardiol 28 (1996) 1227-1240 B.B. Fredholm, Adenosine receptors in the central nervous system, News Pharmacol Sci 10 (1995) 122-128 A.C. Newby, Adenosine and the concept of "retaliatory metabolites", Trends in Biochem. Sci. 9 (1984) 42-44 J. Wang et al., Na dependent purine nuceloside transport from human kidney: Cloning and functional characterisation, Am. J. Physiol. 273 (1997) F1058-F1065 S.Y. Yao et al., Molecular cloning and functional characterization of NBMPR sensitive (es) and NBMPR (ei) insensitive equilibrative nucleoside transporter proteins (rENT I and rENT2) from rat tissues, J. Biol. Chem. 272 (1997) 28423-28430 C.R. Crawford et al, Cloning of the human equilibrative NBMPR insensitive nucleoside transporter ei by functional expression in a transport deficient cell line, J. Biol. Chem 273 (1998) 5288-5293 P.J. Richardson and S. J. Brown ATP release from affinity purified rat cholinergic nerve terminals, J. Neurochem. 48 (1987) 622-630 H. Zimmermann and H. Braun Ectc-nucleontidases—molecular structures, catalytic properties, and functional roles in the nervous system, Prog Brain Res. 120 (1999) 371-385. C.G. Craig and T.D. White N-methyl-D-aspartate- and non-N-methyl-D-aspartate-evoked adenosine release from rat cortical slices: distinct purinergic sources and mechanisms of release, J Neurochem. 60 (1993) 1073-1080. J. Harvey and M.G. Lacey A postsynaptic interaction between dopamine D1 and NMDA receptors promotes presynap tic inhibition in rat nucleus accumbens via adenosine release, J. Neurosci. 17 (1997)5271-5280 S. Delaney et al., Levels of endogenous adenosine in rat striatum.I.- Regulation by ionotropic receptors, nitric oxide and free radicals, J. Pharmacol. Exp. Ther. 285(1998)561-567 K. Chergui et al., Functional GluR6 kainate receptors in the striatum: indirect down regulation of synaptic transmission, J. Neurosci 20 (2000) 2175-2182 M. Nomoto Neurobiology of adenosine receptors In (H. Kase et al., eds) Adenosine receptors and Parkinsons Disease ISBN: 0 12 400405-9 Academic Press, (2000) pp245-256 C.R. Gerfen et al., Dopamine differentially regulates dynorphin, substance P and enkephalin expression in striatal neurons: in situ hybridization histochemical analysis, J. Neurosci 11 (1991)10161031 A..K. Dixon and P.J. Richardson Control of gene expression in basal ganglia by adenosine receptors In (H. Kase et al., eds) Adenosine receptors and Parkinsons Disease ISBN: 0 12 400405-9 Academic Press, (2000) pp 149-170 C.R. Gerfen and C.J. Wilson The basal ganglia In Handbook of Chemical Neuroanatomy Vol 12 Integrated systems of the CNS (L. W. Swanson et al., eds) Elsevier Science (1996) 371-468 D. Jaeger et al., Surround inhibition among projection neurons is weak or nonexistent in the rat striatum, J. Neurophysiol. 194 (1994)1-4 T. Koos and J.M. Tepper Inhibitory control of neostriatal projection neurons by GABAergic interneurons, Nature Neurosci. 2 (1999) 467-472 H. Kita Glutamatergic and GABAergic postsynaptic responses of striatal spiny neurons to intrastriatal and cortical stimulation recorded in slice preparations, Neurosci 70 (1996) 925-940 P. Calabresi et al, The corticostriatal projection: from synaptic plasticity to dysfunctions of the basal ganglia, Trends in Neurosci 19 (1996)19-24 P.J. Richardson et al, Correlating cell physiology with gene expression in single neurons, - Neurochem 74 (2000)839-846 S. Ferre et al, Adenosine-dopamine receptor-receptor interactions as an integrative mechanism in the basal ganglia, Trends in Neurosci 20(1997) 482-487 D.J. Surmeier et al, Coordinated expression of dopamine receptors in neostriatal medium spiny neurons, J. Neurosci. 16 (1996) 6579-6591 G. Maura et al, Release-regulating D-2 dopamine-receptors are located on striatal glutamatergic nerveterminals, J Pharnacol Exp Ther. 247 (1988) 680-684

68

[26] [271 [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43]

[44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54]

P.J. Richardson /Adenosine A2A Receptor Antagonists and Parkinson's Disease

M.S. Levine et al. Modulatory actions of dopamaine on NMDA receptor mediated responses are reduced in DlA deficient mutant mice, J. Neurosci 16 (1996) 5870-5882 Y. Kawaguchi et al, Striatal interneurons: chelical. Physiological arid morphological characterisation. Trends in Neurosci 18 (1995) 527-535 G.O. Di Chiara et al,. Modulatory functions of neurotransmitters in the striatum: ACh/ dopamine/NMDA interactions, Trends in Neurosci. 17(1994) 228-233 R. Bertorelli et al, Dopamine depletion preferentially impairs Dl-receptor over D2-receptor regulation of striatal in vivo acetylcholine-release, J. Neurochem 59 (1992) 353-357 S. Kaneko et lt, Synaptic integration mediated by striatal cholinergic interneurons in basal ganglia function, Science 289 (2000) 633-637 M.I. Bell et al, Characterisation of the mechanism of action of tachykinins in rat striatal cholinergic interneurones, Neurosci. 87 (1998) 649-658 B.B. Fredholm et al. Nomenclature and classification of purinoceptors. Pharmacol. Rev. 46 (1994) 143-156 A . K . Dixon et al. Tissue distribution of adenosine receptor mRNAs in the rat. fir. J. Pharmacol. 118 (1996) 1461-1468 A.C. Dolphin and S.A. Prestwich Pertussis toxin reverses adenosine inhibition of neuronal glutamate release, Nature 316 (1985) 148-150. H.H. Harms et al., Adenosine modulates depolarization-induced release of 3H-noradrenaline from slices of rat brain cortex, Eur. J. Pharmacol. 49 (1978) 305-308. H. H. Harms et al., Effects of adenosine on depolarization-induced release of various radiolabelled neurotransmitters from slices of rat corpus striatum, Neuropharmacol. 18 (1979)577-580. C. Hollins, and T.W. Stone Adenosine inhibition of g-aminobutyric acid release from slices of rat cerebral cortex, Br, J. Pharmacol. 69 (1980) 107-112. M.L. Michaelis et al., Adenosine modulation of synaptosomal dopamine release. Life Sci. 24 (1979) 2083-2092. C. Maenhaut et al, RDC8 codes for an adenosine A2 receptor with physiological constitutive activity. Biochem Biophys Res Comtnun 173 (1990) 1169-1178 J.H. Stehle et al., Molecular cloning and expression of the cDNA for a novel A2-adenosine receptor subtype, Mol Endocrinol 6 (1992) 384-393 J. Flink et al, Molecular cloning of the rat A2 adenosine receptor: selective coexpression with dopamine D2 receptors in the striatum, Mol. Brain Res. 14 (1992)186-195 D.L. Rosin et al. Immunohistochemical localization of adenosine A2A receptors in the rat central nervous system. J. Comp. Neurol 401(1998) 163-186 S. Schiffmann et al., Striatal restricted adenosine A2 receptor (RDC8) is expressed by enkephalin but not Substance P neurons: an in situ hybridisation histochemistry study. J. Neurochem. 57 (1991)10621067 P.J. Richardson et al., Correlating cell physiology with gene expression in single neurons, it Neurochem 74 (2000) 839-846 Z. Preston et al., Adenosine receptor expression and function in rat striatal cholinergic interneurones. fir. J. Pharmacol 130 (2000) 886-890. J. Shimada et al, (EO-l,3-Diqalkyl-7-methyl-8-(3,4.5-trimethoxystyryl)xanthines: potent and selective adenosine A2 antagonists, J. Med. Chem. 35 (1992) 2342-2345 E. Ongini and B. B. Fredholm Pharmacology of adenosine A2A receptors. Trends in Pharmacol. Sci. 17(1996)364-372 S.M. Poulcher et al., The in vitro pharmacology of ZM241383. a potent, non xanthine. A2A selective adenosine receptor antagonist. Br J. Pharmaco.l 115(1995)1096-1102 S.J. Brown et al., Both Al and A2A purine receptors regulate striatal acetylcholine release. J. Neurochem 55(1990)31 -38 S.J. Brown Purinergic modulation of cholinergic synaptosomes. Ph. D. Thesis. (1988) University of Cambridge M. Kurokawa et al., Adenosine A2A receptor modulation of striatal acetyicholine release in vivo. J Neurochem. 66( 1996) 1882-1888 A.K. Gubitz et al., Dual signalling by the adenosine A2a receptor involves activation of both N- and Ptype calcium channels by different G proteins and protein kinases in the same striatal nerve terminals. J. Neurochem. 67 (1996) 374-381 T. Kleppisch and M.T. Nelson Adenosine activates ATP sensitive potassium channels in arterial myocytes via A2 adenosine receptors and cAMP dependent protein kinase, Proc. Natl. Acad. Sci USA 92 12441 - 12445 V. SexI et al., Stimulation of mitogen activated MAP kinase via the A2a adenosine receptor in primary human endothelial cells. J. Biol Chem. 272 (1997) 5792-5799

P.J. Richardson / Adenosine A2A Receptor Antagonists and Parkinson's Disease

69

I.P. Kirk and P.J. Richardson Adenosine A2a receptor mediated modulation of striatal [3H]-GABA and [3H]-ACh release, J. Neurochem. 62 (1994) 960-966 I.P. Kirk and P.J. Richardson Inhibition of striatal GABA release by the adenosine A2a receptor is not mediated by increases in cyclic AMP, J. Neurochem. 64(1995) 2801-2809T. A. Mori et al.,The role of adenosine A2A receptors in regulating GABAergic synaptic transmission in striatal medium spiny neurons, J. Neurosci 16 (1996) 605-611 P.J. Richardson et al., Adenosine A2A receptor antagonists as new agents for the treatment of Parkinson's disease, Trends in Pharmacol Sci 18 (1997) 338-344 S. Ferre e al.t, The striopallidal neuron: a main locus for adenosine-dopamine interactions in the brain, 7. Neurosci 13 (1993)5402-5406 M. Ochi et al., Systemic administration of adenosine A2A receptor antagonist reverses increased GABA release in the globus pallidus of unilateral 6-hydroxydopamine-lesioned rats: a microdialysis study, Neurosci 100 (2000) 53-62 P. Popoli et al., Adenosine A2A receptor stimulation enhances striatal extracellular glutamate levels in rats, Eur. J. Pharmacol 287(1995) 215-219 I. Flagmeyer et al., Adenosine A1 receptor mediated depression of corticostriatal and thialamostriatal glutamatergic synaptic potentials in vitro Brain Res 778(1997) 178-185 T.N. Chase, Novel approaches to the palliation of Parkinson's disease, Movement Disorder 13 M39 S. Ferre et al., Proc Natl Acad Sci USA 88 (1991) 7327-7241 S. Aoyama et al., Rescue of locomotor impairment in dopamine D2 receptor deficient mice by an adenosine A2A receptor antagonist, J. Neurosci 20(2000) 5848-5858 M. Kurokawa et al., Inhibition by KF17837 of adenosine A2A receptor-mediated modulation of striatal GABA and ACh release, Br. J. Pharmacol. 113 (1994) 43-48 R.D. Mayfield et al., Adenosine A2A receptor modulation of electrically evoked endogenous GABA release from slices of rat globus pallidus, J. Neurochem 60(1993)2334-2337 K. Fuxe and U. Ungerstedt Action of caffeine on supersenitive dopamine receptors: considerable enhancement of receptor response to treatment with DOPA and dopamine agonists, Med Biol 52 (1974)48-54 S.J. Brown et al., Striatal A2 receptor regulates apomorphine induced turning in rats with unilateral dopamine denervation, Psychopharmacol. 103(1991) 78-82. S.V. Vellucci et al., Adenosine A2 receptor regulation of apomorphine induced turning in rats with unilateral striatal dopamine denervation, Psychopharmacol. 111(1993) 383-388 M. Morelli et al., Adenosine A2 receptors interact negatively with dopamine Dl and D2 receptors in unilaterally 6-hydroxydopamine lesioned rats, Eur. J. Pharmacol. 251 (1994) 21-25 T. Kanda et al., KF17837: a novel selective adenosine A2A receptor antagonist with anticataleptic activity, Eur. J. Pharmacol. 256 (1994), 264-268 T. Kanda et al., Combined use of the adenosine A2A antagonist KW-6002 with L-DOPA or with selective Dl or D2 dopamine agonists increases antiparkinsonian activity but not dyskinesia in MPTPtreated monkeys, Exp Neurol 162 (2000) 321-327 A.H. Tahar et al., Selective adenosine A2A receptor antagonism as an alternative therapy for Parkinson's disease In (H. Kase et al., eds) Adenosine receptors and Parkinson's Disease, ISBN: 0 12 400405-9 Academic Press, (2000) pp 229-244 O. Aizman et al., Anatomical and physiological evidence for Dl and D2 dopamine receptor colocalization in neostriatal neurons, Nature Neurosci. 3 (2000) 226-230

This page intentionally left blank

Part II Understanding and Influencing Neuro-Degeneration

This page intentionally left blank

Parkinson's Disease E. Ronken and G.J.M. van Scharrenburg (Eds.) IOS Press, 2002

73

Bioenergetics in Neurodegeneration M. Flint Beal, M.D. Department of Neurology and Neuroscience, Weill Medical College of Cornell University and the New York Hospital-Cornell Medical Center New York, NY, USA Abstract. There is increasing evidence linking both oxidative damage as well as energy dysfunction in neurodegenerative diseases. Mitochondria play a key role in both apoptotic and necrotic cell death. We have found that oxidative damage appears to play a key role in substantia nigra cell death mediated by the toxin MPTP. Mice, who overexpress manganese superoxide dismutase, are resistant. Mice, who have a deficiency in manganese superoxide dismutase or glutathione peroxidase, show increased vulnerability. There is also evidence for involvement of apoptotic cell death pathways. Mice overexpressing Bcl-2 or which have a dominant negative inhibitor of interleukin converting enzyme show resistance to MPTP toxicity. We have investigated whether treatments, which can modulate energy metabolism, may show neuroprotective effects. We have found that creatine administration is effective in blocking neuronal cell death in two transgenic mouse models of neurodegenerative diseases as well as against MPTP toxicity. These findings suggest a novel therapeutic strategy for the treatment of neurodegenerative diseases. Keywords: oxidative stress, mitochondrian, MPTP superoxide dismutase, ICE, bioenergetics

There is marked and increasing evidence that implicate mitochondria as playing a critical role in both necrotic and apoptotic cell death. Apoptotic and necrotic cell death are distinct forms of cell death as defined morphologically however under most circumstances they appear to represent a continuum [1]. Under conditions of severe energetic insults necrotic cell death predominates whereas when cellular energy reserves are preserved apoptosis predominates. A requirement for mitochondrial calcium uptake in glutamate mediated excitotoxicity has been demonstrated [2, 3]. On the other hand mitochondria are essential in controlling specific pathways involved in apoptotic cell death [4]. The mechanisms by which they exert this function include release of caspase activators such as cytochiome c, caspase 9 and apoptosis inducing factor. The redistribution of cytochrome c during apoptosis can be prevented by overproduction of the antiapoptotic protein Bcl-2 whereas the overexpression of Bax induces the release of cytochrome c [5]. A crucial regulator in cell death appears to be the mitochondrial permeability transition pore ( PTP) . The activation of the permeability transition pore leads to a nonspecific increase in membrane permeability to solvents with a molecular mass up to 1.5 kd. A number of components of the permeability transition pore include the inner mitochondrial membrane adenine nucleotide transporter, which then interacts with cyclophilin D, and the voltage dependent anion channel in the outer membrane. There also appear to be interactions with the peripheral benzodiazepine receptor in the outer membrane, as well as the mitochondrial creatine kinase. Bax has been shown to interact with the voltage dependent anion channel to accelerate opening of the permeability transition pore, which may then contribute, to cytochrome c release. Several agents are known to activate the PTP. These include elevations in calcium as well as oxidizing agents. Closure of the pore is favored by protons ( low matrix pH) as well as adenine nucleotides

74

M. F. Beal / Bioenergetics in Neurodegeneration

(ADP) . Cyclosporin A prevents the interaction of cyclophilin with the adenine nucleotide transporter. Cyclosporin A has been shown to prevent both necrotic cell death and excitotoxicity in vitro as well as apoptotic cell death. It was recently shown however that bongkrekic acid, which interacts with the adenine nucleotide transporter can block apoptotic cell death induced by NMDA but not that induced by staurosporin [6]. There is substantial evidence implicating mitochondrial dysfunction in Parkinson's disease. In idiopathic Parkinson's disease there is a 30-40% decrease in complex I activity in the substantia nigra as well as reduced staining for complex I subunits in the substantia nigra [7-10]. Strong support for mitochondrial DNA encoded defect comes from two studies showing the cybrids made from individuals with Parkinson's disease show reductions in complex I activity [11, 12]. These defects are associated with increased free radical production as well as increased susceptibility to the MPTP metabolite, MPP+ [13]. There is also impaired mitochondrial calcium buffering. We recently reported a family with multisystem degeneration with Parkinsonism associated with the 11778 mitochondrial DNA mutation, which produces a complex I defect [14]. A larger study however of mitochondrial DNA complex I subunits and transferase RNA failed to show homoplasmic mutations, suggesting that the etiology of the complex I defects in Parkinson's disease is either due to heteroplasmic mutations, or may involve both genetic and environmental interactions [15]. In support of interactions between the environment as well as genetic alterations we found that several genetic mutations result in increased susceptibility to MPTP toxicity. MPTP ( l-methyl-4-phenyl-l,2,3,6-tetrahydropyridine) first came to light as a contaminant of synthetic opiates that led to an outbreak of Parkinsonism in young individuals [16]. MPTP is metabolized to MPP+ ( l-methyl-4-phenylpyridinium) which inhibits complex I of the electron transport chain, leading to reductions in mitochondrial ATP production [17]. We investigated whether MPTP toxicity is influenced by either overexpression or deficits in free radical scavenging enzymes. We initially studied whether overexpression of manganese superoxide dismutase could attenuate MPTP toxicity [18]. We found that overexpression of human manganese SOD in transgenic mice resulted in increased manganese SOD localized to mitochondria in neurons, and a 50% increase in total manganese SOD activity in brain homogenates. In these mice who overexpress manganese SOD there was a significant reduction in MPTP toxicity. Following MPTP these mice showed a 3-fold greater dopamine level than those in controls. There were no alterations in MPP+ levels suggesting that the effects were not due to an alteration in metabolism of MPTP. There was a significant increase in 3-nitrotyrosine levels in the littermate controls but not in the transgenic mice overexpressing human manganese SOD. This suggests that toxicity mediated by peroxynitrite was markedly attenuated in these mice. These results provide evidence that manganese SOD plays a significant role in MPTP toxicity. We recently performed a series of experiments, which examined whether a deficiency of manganese superoxide dismutase resulted in increased vulnerability to several mitochondrial toxins [19]. As noted above the major free radical scavenging enzyme in mitochondria is manganese superoxide dismutase. A 100% knockout of this enzyme results in neonatal lethality [20, 21]. Mice who express a 50% decrease however show no evidence of neuropathological or behavioral abnormalities at 2-4 months of age [22]. Nevertheless crossing these mice into transgenic ALS mice with SOD-1 mutations significantly exacerbates the onset of symptoms and decreases survival [23]. We examined the susceptibility of the heterozygous manganese superoxide dismutase knockout mice to MPTP, malonate and 3-nitroproprionic acid ( 3-NP) neurotoxicity ( Andreassen et al., 2000. in press) . Compared to littermate wild-type mice the mice with partial deficiency of manganese superoxide dismutase showed increased vulnerability to dopamine depletion after systemic administration of MPTP. They also showed significantly larger striatal

M.F. Beal/Bioenergetics in Neurodegeneration

75

lesions produced by both 3-NP and malonate. The mice with partial deficiency of manganese superoxide dismutase also showed an increase production of hydroxyl radicals after malonate injection as measured with the salicylate hydroxyl radical trapping method. These findings therefore provide further evidence for a central role of defective mitochondrial energy production and a resulting increase in free radical formation in MPTP neurotoxicity. We carried out an analogous series of experiments in mice deficient in glutathione peroxidase [24]. In this instance we examined mice which had a complete knockout of glutathione peroxidase as compared to the heterozygote controls, which have a 50% decrease as compared to normal mice. Interestingly mice that have a complete knockout of glutathione peroxidase do not show any phenotypic, behavioral or neuropathologic abnormalities [25]. This is somewhat surprising since glutathione peroxidase is the major detoxifying enzyme for hydrogen peroxide within mitochondria. Furthermore recent evidence suggest that glutathione peroxidase also plays a major role in detoxifying peroxynitrite [26]. Both malonate and 3-nitropropionic acid are inhibitors of succinate dehydrogenase, which model Huntington's disease [27, 28]. MPTP has been extensively used to replicate the dopaminergic neuronal loss occurring in Parkinson's disease [16]. We found that glutathione peroxidase knockout mice show a 2-fold increase in lesion volume following intrastriatal injections of malonate. Malonate induced increases in the conversion of salicylate to 2,3 and 2,5-dihydroxybenzoic acid, indices of hydroxyl radical generation, were significantly greater in homozygote glutathione peroxidase knockout mice as compared with both heterozygote glutathione peroxidase knockout and wild-type control mice. Administration of MPTP resulted in significantly greater depletions of dopamine, 3,4dihydroxyphenylacetic acid ( DOPAC) and homovanillic acid ( HVA) in glutathione peroxidase knockout mice than those seen in wild-type control mice. The striatal 3nitrotyrosine concentrations after MPTP were significantly increased in glutathione peroxidase knockout mice as compared with wild-type control mice. Systemic 3-NP administration resulted in significantly greater striatal damage, and increases in 3nitrotyrosine, in the glutathione peroxidase knockout mice as compared with wild-type control mice. These results provide marked evidence for an involvement of free radicals in MPTP toxicity. They demonstrate that overexpression of manganese SOD that attenuates MPTP toxicity or underexpression, which exacerbates MPTP toxicity, are associated with altered free radical generation. We also found that MPTP toxicity is exacerbated in mice that are deficient in glutathione peroxidase. The heterozygous manganese SOD knockout mice and the glutathione peroxidase homozygous knockout mice do not show any phenotypic abnormalities at baseline. Nevertheless following MPTP toxicity they show a marked exacerbation of the dopaminergic depletion. This therefore provides further evidence for an interaction between genetic defects and environmental toxins, which could play a role in the pathogenesis of Parkinson's disease. There is increasing evidence, which has implicated apoptotic cell death in the pathogenesis of Parkinson's disease. As mentioned above this could be a consequence of energy defects at the mitochondrial level. The evidence for apoptosis in Parkinson's disease however remains somewhat controversial. One study using in situ end-labeling demonstrated DNA damage whereas another did not [29, 30]. A recent study using electron microscopy found morphologic evidence of apoptosis in the substantia nigra of a Parkinson's disease patient [31]. Furthermore an increase in Bcl-2 protein was found in the caudate nucleus and putamen of Parkinson disease patients [32]. A study quantifying DNA aggregation has been recently published supporting apoptotic cell death [33]. The protooncogene Bcl-2 was initially characterized because of its ability to inhibit apoptosis. Bcl-2 is widely expressed in the nervous system and is localized to the outer mitochondrial

76

M. F. Beal / Bioenergetics in Neurodegeneration

membrane, and the plasma reticulum and nuclear membrane [34]. Bcl-2 expression inhibits apoptosis in neural cells induced by a variety of stimuli [35]. It also inhibits necrotic cell death in some paradigms, such as oxidative neuronal cell death induced by depletion of glutathione [36]. It has been demonstrated to protect cells from the lethal effects of hydrogen peroxide or tert-butyl hydroperoxide in a dose-dependent manner [37, 38]. It also protects neural cells from cyanide aglycemia induced lipid peroxidation, as well as compromised mitochondrial respiration [39]. It protects from AMPA induced toxicity in cortical cultures [40]. Bcl-2 increases the capacity of neural cell mitochondria to accumulate calcium [41]. It may play a critical role in regulation of membrane potential and volume homeostasis in mitochondria in response to apoptotic and necrotic stimuli [42]. Bcl2 has been shown to maintain the mitochondrial membrane potential and to enhance proton efflux after treatment with either calcium or tert-butyl hydroperoxide, and it prevents activation of the mitochondrial permeability transition [43]. Because of the proposed involvement of both mitochondrial dysfunction and oxidative injury in the pathogenesis of MPTP toxicity we investigated whether MPTP toxicity is attenuated in Bcl-2 overexpressing mice. The mice we examined have a neuron specific enolase promotor fused to a human Bcl-2 cDNA [44]. These mice overexpress Bcl2 in multiple tissues including the substantia nigra pars compacta. We examined the effects of both acute and chronic daily dosing regimen of MPTP in the Bcl-2 overexpressing mice as compared to littermate controls [45]. Chronic ( daily administration over 5 days) administration of MPTP has been shown to induce apoptotic cell death in the substantia nigra pars compacta of mice [46]. There was no evidence of apoptosis found with a more acute dosing regimen [47]. We initially suspected that neuroprotection in Bcl-2 overexpressing mice would be more profound with a chronic dosing regimen. Much to our surprise there was almost complete protection against MPTP neurotoxicity induced by an acute dosing regimen whereas there was only partial protection with a chronic dosing regimen [45]. This was demonstrated by direct measurement of dopamine and its metabolites as well as of tritiated mazindol binding. We investigated the mechanism of neuroprotection in the Bcl-2 overexpressing mice. We found that these mice showed a significant decrease in the increases in 3-nitrotyrosine induced by MPTP. We also found that caspase activation was inhibited. These results therefore indicate that Bcl-2 overexpression protects against MPTP toxicity by mechanisms involving both antioxidant activity as well as inhibition of apoptotic pathways. We more recently extended these studies to an examination of MPTP toxicity in transgenic mice expressing a domimnt negative mutant of interleukin-1 beta converting enzyme [48]. These mice have been shown to extend survival when crossed into both transgenic mouse models of amyotrophic lateral sclerosis and Huntington's disease [49]. A critical role of interleukin converting enzyme ( ICE-like proteases) has been well established in apoptosis [50]. ICE knockout mutant mice generated by gene targeting techniques were found to be only partially defective in apoptosis induced by anti-fas antibody [51]. The neurons from knockout mice however showed resistance to trophic factor mediated apoptosis [49]. Furthermore these mice showed resistance to middle cerebral artery occlusion mediated infarcts. Other studies showed elevated levels of mature interleukin-1 beta following apoptotic cell death indicating activation of ICE in apoptosis [52,53]. The mice, which we utilized to examine MPTP toxicity, were generated by inserting a cysteine for the g]ycine in the active site of ICE [49]. The cysteine residue in the active site is required for interleukin-1 beta convertase activity. We demonstrated the transgenic mice expressing the dominant negative ICE inhibitor are almost completely resistant to MPTP neurotoxicity [48]. Following administration of MPTP there were significant reductions in dopamine. DOPAC and HVA in the control mice, whereas there is a complete inhibition of

M.F. Beal / Bioenergetics in Neurodegeneration

77

this in the transgenic mutant mice. There was also a complete protection of loss of tyrosine hydroxylase immunoreactive neurons in the substantia nigra. We demonstrated that this was not due to an alteration in processing of MPTP to MPP+. These results provide evidence implicating ICE mediated apoptotic cell death in MPTP toxicity. If similar mechanisms occur Parkinson's disease then agents that might inhibit caspase mediated apoptotic cell death may prove useful. Another potential strategy to make the two modulate cell death in Parkinson's disease is to improve cellular bioenergetics. Two strategies, which we have examined to study this, are the administration of coenzyme Q10 or creatine. Coenzyme Q10 is both an antioxidant, which is particularly important in lipophilic membranes, as well as a coenzyme of the electron transport chain. We demonstrated that platelet levels of coenzyme Q10 are significantly decreased in Parkinson's disease patients [54]. We also found a direct linear correlation between coenzyme Q10 levels and reductions in complex I activity of the electron transport chain in Parkinson's disease patients [55]. We found that coenzyme Q10 supplementation could significantly increase plasma levels of coenzyme Q10 in Parkinson's disease patients [56]. Furthermore coenzyme Q10 in older mice significantly attenuated MPTP toxicity [57]. We also demonstrated that coenzyme Q10 could significantly increase mitochondrial concentrations in older mice [58]. These observations have led to a clinical trial to determine whether coenzyme Q10 can delay the need for L-dopa and attenuate progress on the unified Parkinson's disease rating scale ( UPDRS) in Parkinson's disease patients. Another strategy to potentially improve brain energy metabolism is to administer creatine, the substrate of creatine kinase. Creatine kinase is a key enzyme involved in regulating energy metabolism in cells with inherently high and fluctuating energy requirements including the brain [59]. The enzyme catalyzes the reversal of transfer of the phosphoryl group from phosphocreatine to ADP. This therefore generates ATP. Several cytoplasmic and mitochondrial isoforms have been identified, and along with the substrates creatine and phosphocreatine, they constitute an integral cellular energy buffering and transport system connecting sites of energy production with sites of energy consumption [60]. The mitochondrial isoform of creatine kinase is located at contact sites between the inner and outer mitochondrial membranes, where it is associated with porin [59, 61]. Mitochondrial CK can directly convert intramitochondrial produced ATP to phosphocreatine, which then gets transported to sites of energy consumption. There is a large body of evidence to suggest that the creatine kinase system is important in regulating energy homeostasis in the brain. Creatine has been shown to effectively stimulate mitochondrial stage 3 respiration leading to net production of phosphocreatine [62, 63]. Furthermore both creatine and its analogue cyclocreatine modulate rates of ATP production through the creatine kinase system.As noted above there is evidence that the mechanism of MPTP toxicity involves the conversion of MPTP to MPP+ by monoamine oxidase B. MPP+ is then taken up by the dopamine transporter into neurons, where it is accumulated within mitochondria, and then inhibits complex I activity of the electron transport chain [64]. Oral administration of creatine has previously been demonstrated by us to increase brain levels of phosphocreatine [58]. This may therefore provide an energy buffer against ATP depletion. We therefore examined whether oral supplementation with either creatine or cyclocreatine could produce significant protection against MPTP induced dopamine depletions in mice [65]. We found that doses of 0.25, 0.5 and 1% creatine exerted dosedependent significant neuroprotection effects that disappeared at 2 and 3% creatine, consistent with a U shaped dose response curve. Cyclocreatine exerted significant protection against dopamine depletions at doses of 0.5 and 1% cyclocreatine in the diet. The effects on the dopamine metabolites HVA and DOPAC parallel those seen with dopamine. Cyclocreatine exerted similar effects on HVA

78

M.F. Beal / Bioenergetics in Neurodegenerat ion

and DOPAC. We subsequently examined the effects of 1% creatine and 1% cyclocreatine on MPTP at 15 mg/kg i.p. for 6 doses. With this dosing regimen there was a 70% depletion of dopamine as compared with controls. Administration of both creatine and cyclocreatine produced marked significant protection against MPTP induced depletions of dopamine, DOPAC and HVA. Histological examination of serially cut, Nissl stained and tyrosine hydroxylase positive immunostained sections through the midbrains of control, creatine-MPTP treated and MPTP treated mice showed that there was marked protection against the loss of dopaminergic neurons with creatine administration. These findings therefore demonstrate that two different strategies to improve brain energy metabolism, administration of either coenzyme Q10 or creatine produce marked neuroprotective effects against MPTP toxicity. Several lines of evidence implicate defective energy metabolism in the pathogenesis of Parkinson's disease. Strategies to improve mitochondrial function might therefore be useful] in the treatment of Parkinson's disease. If both MPTP neurotoxicity and Parkinson's disease involve similar pathogenic mechanisms, both coenzyme Q10 and creatine administration may prove to be novel therapeutic strategies in attempting to slow the neurodegenerative process.

The secretarial assistance of Sharon Melanson is gratefully acknowledged. This work was supported by the Department of Defense, NINDS and the Parkinson's Disease Foundation. References [1J [2] [3] [4] |5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16]

M. Ankarcrona, el al., Glutamate-induced neuronal death: a succession of necrosis or apoptosis depending on mitochondrial function. Neuron 15 ( 1995) 961-973 S.L. Budd and D.C. Nicholls, Mitochondria, calcium regulation, and acute glutarnate excitotoxicity in cultured cerebellar granule cells, Journal of Neurochemistry 67 ( 1996) 2282-2291. A.K. Stout.et al., Glutamate-induced neuron death requires mitochondrial calcium uptake. Nature Neuroscience 1 ( 1998) 366-373. D.R. Green and J.C. Reed, Mitochondria and apoptosis, Science 281 ( 1998) 1309-1312. S. Desagher, et al., Bid-induced conformational change of Bax is responsible for mitochondrial cytochromec release during apoptosis, Journal of Cellular Biology 144 ( 1999) 891-901. S.L. Budd. et al., Mitochondrial and extramitochondrial apoptotic signaling pathways neurons. Proceedings of the National Academy of Sciences USA 97 ( 2000) 6161-6166. N. Rattori, et al., Immunohistochemical studies on complexes I. II, III and Parkinson's disease. Annals of Neurology 30 ( 1991) 563-571. B. Janetzky. et al., Unaltered aconitase activity, but decreased complex I activity in substnatia nigra pars compacta of patients with Parkinson's disease, Neuroscience Letters 169 ( 1994) 126-128 V.M. Mann, et al., Brain, skeletal muscle and platelet homogenate mitochondrial disease. Brain 115 ( 1992)333-342. A.H.V. Schapira. et al., Mitochondrial complex I deficiency in Parkinson's disease. Journal of Neurochemistry 54 ( 1990) 823-827. M. Gu. et al., Mitochondrial DNA transmission of the mitochondrial defect in Parkinson's disease. Annals of Neurology 44 ( 1998) 177-186. R.H. Swerdlow, et al., Origin and functional consequences of the complex 1 defect in Parkinson's disease, Annals of Neurology 40 ( 1996) 663-671. J.P. Sheehan, et al., Altered calcium homeostasis in cells transformed by mitochondria from individuals with Parkinson's disease, Journal of Neurochemistry 68 ( 1997) 1221-1233. D. K. Simon, et al., Familial multisystem degeneration with parkinsonism associated with the 11778 mitochondrial DNA mutation., Neurology 53 ( 1999) 1787-1793. D.K. Simon, et al., Mitochondrial DNA mutations in complex I and tRNA genes in Parkinson's disease, Neurology 54 ( 2000) 703-709. B.R Bloem. et al., The MPTP model: versatile contributions to the treatment of idiopathic Parkinson's disease. Journal of Neurological Sciences 97 ( 1990) 273-293.

M.F. Beal / Bioenergetics in Neurodegeneration

[17] [18] [19] [20] [21] [22]

[23] [24]

[25] [26] [27] [28] [29] [30] [31] [32] [33] [34]

[35] [36] [37] 38] [39] [40] [41] [42] [43] [44] [45]

79

T.P. Singer, et al., Biochemical events in the development of parkinsonism induced by MPTP, Journal of Neurochemistry 49 ( 1987) 1-8. P. Klivenyi, et al., Manganese superoxide dismutase overexpression attenuates MPTP toxicity, Neurobiology of Disease 5 ( 1998) 253-258. O.A. Andreassen, et al, Effects of an inhibitor of poly ( ADP-ribose) polymerase, desmethylselegiline, trientine, and lipoic acid in transgenic ALS mice, Experimental Neurology ( 2000a) in press. R.M. Lebovitz, et al., Neurodegeneration, myocardial injury, and perinatal death in mitochondrial superoxide dismutase-deficient mice, Proceedings of the National Academy of Sciences USA 93 ( 1996)9782-9787. Y. Li, et al., Dilated cardiomyopathy and neonatal lethality in mutant mice lacking manganese superoxide dismutase, Nature Genetics 11 ( 1995) 376-381. K. Murakami, et al., Mitochondrial susceptibility to oxidative stress exacerbates cerebral infarction that follows permanent focal cerebral ischemia in mutant mice with manganese superoxide dismutase deficiency, Journal of Neuroscience 18 ( 1998) 205-213. O.A. Andreassen, et al., Partial deficiency of manganese superoxide dismutase exacerbates a transgenic mouse model of ALS, Annals of Neurology 47 ( 2000b) 447-455. P. Klivenyi, et al., Mice deficient in cellular glutathione peroxidase show increased vulnerability to malonate, 3-nitropropionic acid, and l-methyl-4-phenyl-l,2,5,6-tetrahydropyridine, Journal of Neuroscience 20 ( 2000) 1-7. Y.S. Ho, et al., Mice deficient in cellular glutathione peroxidase develop normally and show no increased sensitivity to hyperoxia, Journal of Biological Chemistry 272 ( 1997) 16644-16651. H. Sies, et al., Glutathione peroxidase protects against peroxynitrite-mediated oxidations, Journal of Biological Chemistry 272 ( 1997) 27812-27817 M.F. Beal, et al., Neurochemical and histologic characterization of excitotoxic lesions produced by the mitochondrial toxin 3-nitropropionic acid, Journal of Neuroscience 13 ( 1993a) 4181-4191. M.F. Beat, et al., Age-dependent striatal excitotoxic lesions produced by the endogenous mitochondrial inhibitor malonate, Journal of Neurochemistry 61 ( 1993b) 1147-1150. S. Koset, et al., On the question of apoptosis in the parkinsonian substantia nigra, Acta Neuropathologica 93 ( 1997) 105-108. H. Mochizuki, et al., Histochemical detection of apoptosis in Parkinson's disease, Journal of Neurological Sciences 137 ( 1996) 120-123. P. Anglade, et al., Apoptosis and autophagy in nigral neurons of patients with Parkinson's disease Histology and Histopathology 12 (1997) 25-31. M. Mogi, et al., BcI-2 Protein is increased in the brain from parkinsonian patients, Neuroscience Letters 215 ( 1996) 137-139. N.A. Tatton, Increased caspase 3 and Bax immunoreactivity accompany nuclear GAPDH translocation and neuronal apoptosis in Parkinson's disease, Experimental Neurology 166 ( 2000) 29-43. S. Krajewski, et al., Investigation of the subcellular distribution of the bcl-2 oncoprotein: residence in the nuclear envelope, endoplasmic reticulum, and outer mitochondrial membranes, Cancer Research 53 (1993)4701-4714. D.E. Bredesen, Neural apoptosis, Annals of Neurology 38(1995) 839-351. D.J. Kane, et al., Expression of bcl-2 inhibits necrotic neural cell death, Journal of Neuroscience Research 40 ( 1995)269-275. D.M. Hockenbery, et al., Bcl-2 functions in an antioxidant pathway to prevent apoptosis, Cell 75 ( 1993)241-251. D.J. Kane, et al., Bcl-2 inhibition of neural death: decreased generation of reactive oxygen species, Science 262 ( 1993) 1274-1277. K.M. Myers, et al, Bcl-2 protects neural cells from cyanide/aglycemia-induced lipid oxidation, mitochondrial injury, and loss of viability, Journal of Neurochemistry 65 ( 1995) 2432-2440. M.J. White, et al., A Bcl-2 antisense oligonucleotide increases a-amino-3-hydroxy-5-methylisoxazole4-propionic acid (AMPA) toxicity in cortical cultures, Annals of Neurology 42 ( 1997) 580-587. A.N. Murphy, et al., Bcl-2 potentiates the maximal calcium uptake capacity of neural cell mitochondria, Proceedings of the National Academy of Sciences USA 93 (1996) 9893-9898. M.G. Vander Heiden, et al., Bcl-xL regulates the membrane potential and volume homeostasis of mitochondria, Cell ( 1997) 627-637. S. Shimizu, et al., Bcl-2 prevents apoptotic mitochondrial dysfunction by regulating proton flux, Proceedings of the National Academy of Sciences USA 95 ( 1998) 1455-1459. J.-C. Martinou, et al., Overexpression of BCL-2 in transgenic mice protects neurons from naturally occurring cell death and experimental ischemia, Neuron 13 ( 1994) 1017-1030. L. Yang, et al., l-Methyl-4-phenyl-l,2,3,6-tetrahydropyride neurotoxicity neurotoxicity is attenuated in mice overexpressing Bcl-2, Journal of Neuroscience 18( 1998) 8145-8152.

80

M. F. Beal / Bioenergetics in Neurodegeneration

N.A. Tatton and S.J. Kish, In situ detection of apoptotic nuclei in the substantia nigra compacta of 1methyl-4-phenyl- 1,2,3,6-tetrahydropyridine-treated mice using terminal deoxynucleotidyl transferase labelling and acridine orange staining, Neuroscience 77 ( 1997) 1037-1048. V. Jackson-Lewis, et al.. Time course and morphology of dopaminergic neuronal death caused by the neurotoxin I -methyl-4-phenyl-l,2,3.6-tetrahydropyridine, Neurodegeneration 4 (1995)257-269. P. Klivenyi, et al., Neuroprotective effects of creatine in a transgenic animal model of amyolrophic lateral sclerosis. Nature Medicine 5 (1999) 347-350. R.M. Friedlander, et al., Expression of a dominant negative mutant of interleukin-1 b convening enzyme in transgenic mice prevents neuronal cell death induced by trophic factor withdrawal and ischemic brain injury, Journal of Experimental Medicine 185 (1997) 933-940. R.M. Friedlander and J. Yuan. ICE, neuronal apoptosis and neurodegeneration. Cell Death Differentiation 5 (1998) in press. K. Kuida, Caspase-9, International Journal of Biochemistry and Cellular Biology 32( 2000) 121 -124. A.H. Hogquist, et al., Interleukin I is processed and released during apoptosis. Proceedings of the National Academy of Sciences USA 88 ( 1991) 8485-8489. M. Miura, et al., Tumor necrosis factor-induced apoptosis is mediated by a CrmA-sensitive cell death pathway, Proceedings of the National Academy of Sciences USA 92 (1995)8318-8322. C.W. Shults, et al., Coenzyme Q10 is reduced in mitochondria from parkinsonian patients. Annals of Neurology 42 ( 1997a) 261 -264. C.W. Shults, et al., Coenzyme Q10 levels correlate with the activities of complexes I and 'II/III in mitochondria from parkinsonian and nonparkinsonian subjects. Annals of Neurology 42 ( 1997b) 261264. C.W. Shults, et al., Absorption, tolerability, and effects on mitochondrial activity of oral coenzyrne Q10 in parkinsonian patients. Neurology 50 ( 1998) 793-795 M.F. Beal, et al., Coenzyme Q10 attenuates the l-methyl-4-phenyl-l,2,3,tetrahydropyridine ( MPTP) induced loss of striatal dopamine and dopaminergic axons in aged mice. Brain Research 783 (1998) 109-114. R.T. Matthews, et al., Coenzyme Q10 administration increases brain mitochondrial concentrations and exerts neuroprotective effects. Proceedings of the National Academv of Sciences USA 95 (1998) 88928897. T. Wallimann, et al., Intraceilular compartmentation, structure and function of creatine kinase isoenzymes in tissues with high and fluctuating energy demands: the 'phosphocreatine circuit' for cellular energy homeostasis, Biochemical Journal 281 ( 1992) 21-40. W Hemmer and T. Wallimann. Functional aspects of creatine kinase in brain. Developmental Neuroscience 15 ( 1993)249-260. D. Brdiczka, et al., In vitro complex formation between the octamer of mitochondrial creatine kinase and porin, Journal of Biological Chemistry 269 ( 1994) 27640-27644. F. Kernec, et al., Phosphocreatine synthesis by isolated rat skeletal muscle mitochondria is not dependent upon external ADP: a 31P NMR study. Biochemical and Biophysical Research Communications 225 (1996) 819-825. E. O'Gorman. et al., Differential effects of creatine depletion on the regulation of enzyme activities and on creatine-stimulated mitochondrial respiration in skeletal muscle, heart, and brain. Biochimica Et Biophysica Acta 1276 (1996) 161-170. M.R. Gluck, et al., Characterization of the inhibitory mechanism of 1-methyl-4-phenylpyridinium and 4-phenylpyridine analogs in inner membrane preparation. Journal of Biological Chemistry 269 (1994) 3167-3174. R.T. Matthews, et al., Creatine and cyclocreatine attenuate MPTP neuroioxicity. Experimental Neurology 157 ( 1999) 142-149.

Parkinson's Disease E. Ronken and G.J.M. van Scharrenburg (Eds.) IOS Press, 2002

81

Neuronal cell death and apoptosis in neurodegenerative diseases Thomas Klockgether and Ullrich Wullner Department of Neurology, University of Bonn, Sigmund-Freud-Strasse 25, 53105 Bonn, Germany Phone +49-228-287-5726, Fax +49-228-287-5024, e-mail: [email protected] Abstract Apoptosis is a morphologically defined type of cell death that is distinguished from necrosis. Recently, the hypothesis has been put forward that the progressive death of specific populations of neurons in neurodegenerative diseases occurs by apoptosis. We here review the available data concerning possible apoptotic death of nigral dopaminergic neurons in Parkinson's disease (PD). The presence of apoptotic neurons in the substantia nigra of PD has not been demonstrated unequivocally. This may have technical reasons, but may also be due to the absence of apoptosis. More important than the final mode of cell death, however, is the question whether neurons in neurodegenerative disorders can regain full or partial functional capabilitities. Recent observations in a transgenic mouse model of Huntington's disease showing reversibility of the neurological phenotype by blockade of transgene expression indeed question the principal importance of neuronal cell death for neurodegeneration. These observations rather suggest that at least in early disease stages - neurodegeneration is a reversible process of neuronal cell dysfunction. In this scenario, neuronal cell death does not cause the neurological phenotype, but rather serves to remove non-functional neurons.

The different types of neuronal cell death: apoptosis versus necrosis Cell death in the nervous system can occur through one of several mechanisms. Morphologically, two major types of cell death are distinguished: apoptosis and necrosis. Apoptotic neurons undergo a number of well defined morphological changes affecting cytoplasm, cell membranes and nucleus. Characteristic features include cell shrinkage, membrane blebbing, chromatin condensation and fragmentation. Typically, cell organelles retain their normal appearance during apoptosis. Finally, neurons break down into apoptotic bodies which undergo phagocytosis by neighboring cells. Apoptosis does not cause a major inflammatory reaction of the surrounding tissue [1]. Necrosis, on the other hand, is characterized by rapid cell swelling, lysis of the cell membrane and a major inflammatory reaction (Figure 1). While necrosis is a pathological form of cell death resulting from acute cellular injury apoptosis plays central role in the development and homeostasis of multicellular organisms. During development, more than half of the neurons undergo a physiological or programmed cell death that typically occurs via apoptosis. The intracellular pathways mediating apoptosis have been extensively studied, and it is well established that apoptosis is controlled by a number of evolutionarily highly conserved pro- and antiapoptotic genes. These genes have been initially identified in C.elegans: CED-3 and CED-4 promote apoptosis, while CED-9 is an antiapoptotic gene. Upon certain stimuli, these genes orchestrate an internally coded suicide program that results in the self-removal of the apoptotic cell. Principally, two converging pathways mediating apoptosis are distinguished.

82

T. Klockgether and U. Wullner / Neuronal Cell Death and Apoptosis in Neurodegenerative Disease

The receptor-mediated pathway is initiated via death receptors on the cell surface that belong to the tumor necrosis factor (TNF) receptor supergene family. The best characterized death receptor is CD95, also called Fas or Apol [2]. Binding of CD95 ligand to CD95 results in the activation of an intracellular protease, caspase-8, which activates downstream caspases including caspase-3 which serve to execute apoptosis. Caspases are the mammalian homologue of CED-3 [3].

Figure 1. Morphology of apoptotic and necrotic cell death. Apoptosis is characterized by cell shrinkage (2,3). chromatin condensation and frgamentation (2,3), and formation of apoptotic bodies (3,4) which then undergo phagocytosis (4-6). Cell organelles, in particular mitochondria, retain their normal appearance (2-4). Necrosis, on the other hand, is characterized by rapid eel! swelling (7), and lysis of the cell membrane (8).

A second pathway to activate downstream caspases involves the mitochondria. Essential steps of this pathway are the mitochondrial release of cytochrome C and formation of a complex involving Apaf-1. the mammalian homologue of CED-4. The

T. Klockgether and U. Wullner / Neuronal Cell Death and Apoptosis in Neurodegenerative Disease

83

mitochondrial release of cyrochrome C is controlled by members of the bcl-2 family, the homologue of CED-9 [4;5]. The bcl-2 family of evolutionary conserved genes comprises inhibitors (bcl-2, bcl-x1 )and promoters of apoptosis (bax, bcl-xs) [6]. The levels and balance of these two categories of the bcl-2 family members determine the vulnerability of an individual cell towards pro-apoptotic stimuli. Numerous protein targets that are cleaved by caspases have been identified. Among these are cell structure proteins such as actin, kinases, DNA repair enzymes, and signalling molecules such as ICAD (inhibitor of caspase activated desoxyribonuclease). Cleavage of ICAD results in activation of a specific deoxyribonuclease which degrades nuclear DNA in oligosomal size 180-200 bp fragments. Several methods are used to identify and demonstrate apoptosis. The most powerful morphological method is ultrastructural analysis using electron microscopy. Only electron microscopy is capable of showing the characteristic morphological features of apoptosis. Although DNA cleavage does not appear to be an essential step of apoptosis endonucleolytic DNA breakdown has attracted much attention because demonstration of 180-200 bp fragments is taken as biochemical evidence of apoptosis. On a gel electrophoresis DNA fragments typically appear as a DNA ladder. While demonstration of a DNA fragmentation by electrophoresis is quite specific for apoptosis, light microscopical detection of DNA fragments in situ using the terminal deoxyribonucleotidyl transferasemediated DNA end labeling (TUNEL) may result in false-positive findings [7;8]. Apoptosis and necrosis are phenomena which often are observed simultaneously in tissues or cell cultures exposed to the same stimulus. There is increasing evidence that mechanisms causing cell death (apoptosis or necrosis) may be overlapping with integrated events among the components interacting and contributing to a final pathway for neuron death. Importantly, apoptosis is an energy-consuming process; thus, several studies investigated the relation of the cellular ATP content and the induction of apoptosis. An ATP depletion of more than 50% was sufficient to direct the mode of death from apoptosis to necrosis. Higher ATP contents favored the apoptotic process [9]. As a consequence, antiapoptotic treatment may well prevent apoptosis, but simply by driving a cell into necrosis [10]. Is there apoptotic cell death of dopaminergic neurons in Parkinson's disease? Neurodegenerative disorders such as Parkinson's disease (PD), Alzheimer's disease, Huntingtons's disease or spinocerebellar ataxias are characterized by a gradual loss of specific sets of neurons. Most neurodegenerative diseases are apparently sporadic, and their aetiology remains a matter of speculation. In some diseases, such as Huntington's disease and the spinocerebellar ataxias, the disease is due to the mutation of a single gene. There are also rare monogenic variants of the more common disorders like Alzheimer's disease and PD. Several pathogenic mechanisms have been postulated to contribute to the progressive neuronal cell loss in neurodegenerative diseases. These potential disease mechanisms include oxidative stress, calcium overload, excitotoxicity and deficiency of survival factors. It should be noted, however, that this knowledge is mainly derived from cell culture experiments and toxin-induced animal models. Interestingly, all these mechanisms are capable of inducing apoptosis, raising the question whether neuronal cell death in neurodegenerative diseases is apoptotic in nature. Another argument put forward in favour of an apoptotic nature of neuronal cell death in neurodegenerative disorders is the absence of an inflammatory response to cell death. Although inflammation in neurodegenerative diseases is much weaker than in immune-mediated diseases of the central nervous system

84

T. Klockgether and U. Wiillner / Neuronal Cell Death and Apoptosis in Neurodegenerative Disease

T. Klockgether and U. Wullner /Neuronal Cell Death and Apoptosis in Neurodegenerative Disease

85

dopaminergic cells at the time of disease onset, a number that might be estimated from the quantitative studies of Bogerts et al. [25] and McGeer et al. [21]. Paulus and Jellinger, on the basis of an extensive pathoanatomical study found a 58.7 % and 65.8 % loss of melanin-containing neurons in the medial and the lateral substantia nigra of PD patients, respectively, compared to age-matched controls [26]. Therefore, at a given timepoint of one day, no more than approximately 0.1 - 0.2 % dying cells should be present in the substantia nigra of a PD patient. These consideration show that if dopaminergic neurons die by apoptosis it would be extremely difficult to demonstrate the presence of apoptotic neurons in autopsy material. Taken together, the available evidence does not allow a firm conclusion whether dopaminergic neurons of the substantia nigra in PD die from apoptosis or not. The principal difficulty to detect apoptosis in human brain material does not allow to exclude the possibility that apoptosis occurs in PD. The recent data of Hartmann et al. showing an increased percentage of nigral neurons in PD that express activated caspase-3 strongly suggest an involvement of caspases in the pathogenesis of PD [20]. However, execution of apoptosis does not appear to be the only action of caspases. Activation of caspases may also damage neurons in a more chronic way by cleaving proteins essential for normal cell function and survival or by forming cleavage products that have deleterious functions. Thus, Goldberg et al. showed that huntingtin is cleaved in a CAG repeat length-dependent manner by caspase-3 [27]. Obviously, the intracellular actions of caspases that are not directly related to the execution of apoptosis require further studies. Neuronal dysfunction preceding cell death in neurodegenerative disease In the final section of this article we wish to discuss the question how important neuronal cell death is for neurodegeneration and whether treatment strategies designed to interfere with cell death mechanisms - whether apoptotic or not - are really promising. As outlined above, neurodegenerative disease is closely associated with neuronal cell death. However, the demonstration of neuronal cell death in post mortem brain tissue of patients with neurodegenerative diseases does not prove that death and loss of neurons causes the progressive neurological symptoms in the diverse neurodegenerative disorders. Alternatively, the clinical symptoms in neurodegenerative disorders might be due to neuronal dysfunction, while neuronal cell loss might be a secondary phenomenon that serves to remove non-functional neurons. The removal of non-functional neurons would be accomplished by apoptosis in a particularly effective way since apoptosis causes little damage of surrounding tissue. In this case, it might even be harmful to inhibit this process. Apoptosis may thus be only the last step in a chain of events and not a feasible therapeutic target. At present, the question whether neuronal cell death really causes the symptoms of neurodegenerative diseases or whether it is rather a secondary phenomenon cannot be definitely answered. For principal reasons, studies of human post mortem brain tissue are not helpful to answer this question. Most animal models of neurodegenerative diseases do not properly reflect the disease process that leads to neurodegeneration. In particular, toxininduced dopaminergic cell death does not adaequately model disease pathogenesis of sporadic PD. Interestingly, a recent study showed that antiapoptotic treatment with a caspase inhibitor protected dopaminergic nigral neurons from 6-hydroxydopamine-induced cell death but did not prevent nerve terminal loss and the functional impairment [28]. More appropriate animal models are available for monogenic neurodegenerative disorders such as Huntington's disease or the spinocerebellar ataxias [29]. Detailed studies in transgenic models of these disorders showed that animals have neurological impairment without evidence of neuronal cell death or weeks before first sign of neuronal cell death can be

86

T. Klockgether and U. Wullner / Neuronal Cell Death and Apoptosis in Neurodegenerative Disease

detected [30-33]. Table 1 summarizes the evolution of pathogenetic events, behavioural signs and morphological alterations in mice overexpressing an expanded allele of the spinocerebellar ataxia type 1 (SCA1) gene showing that neuronal cell death is a rather late event in the development of the disease. Yamamoto et al. recently described a conditional mouse model of Huntington's disease. Blockade of transgene expression led to an amelioration of the behavioral phenotype supporting the view that irreversible changes that commit neurons to death have not necessarily taken place at a time when neurological signs are already apparent [34]. Although these remarkable finding have been made in a rodent model they raise hope that - at least at early disease stages - neurological symptoms in neurodegenerative diseases are caused by reversible neuronal dysfunction and not by irreversible cell death. Table 1. Temporal evolution of molecular, behavioral and morphological signs in mice overexpressing an expanded allele of the human spinocerebellar ataxia 1 (SCA1) gene [30-32]

References [1]

J.F. Kerr et al., Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics, Br.J.Cancer, 26, (1972) 239-257. [2] A. Ashkenazi and V.M. Dixit. Death receptors: signaling and modulation. Science. 281, (1998) 13051308. [3] N.A. Thornberry and Y. Lazebnik. Caspases: enemies within. Science, 281, (1998) 1312-1316. [4] H. Zou et al., Apaf-1, a human protein homologous to C-elegans CED-4. participates in cytochrome cdependent activation of caspase-3. Cell, 90. (1997) 405-413. D.R. Green and J.C. Reed. Mitochondria and apoptosis. Science. 281. (1998) 1309-1312. J.C. Reed. Bcl-2 family proteins, Oncogene, 17, (1998) 3225-3236. Y. Gavrieli et al., Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J.Cell Biol., 119. (1992) 493-501. [8] C. Charriaut-Marlangue and Y. Ben An. A cautionary note on the use of the TUNEL stain to determine apoptosis, Neuroreport, 7, (1995) 61-64. [9] P. Nicotera et al., Energy requirement for caspase activation and neuronal cell death. Brain Pathol.. 10. (2000)276-282. [ 101 A. Hartmann et al., Caspase-8 is an effector in apoptotic death of dopaminergic neurons in Parkinson's disease, but pathway inhibition results in neuronal necrosis. J.Neurosci., 21. (2001) 2247-2255. [11] Y. Vodovotz et al., Inducible nitric oxide synthase in tangle-bearing neurons of patients with Alzheimer's disease, J.Exp.Med., 184, (1996) 1425-1433. [12] C.B. Thompson. Apoptosis in the pathogenesis and treatment of disease. Science. 267, (1995) 14561462. [ 1 3 ] P. Anglade et al., Apoptosis and autophagy in nigral neurons of patients with Parkinson's disease. Histol.Histopathol. 12. (1997) 25-31.

T. Klockgether and U. Wullner / Neuronal Cell Death and Apoptosis in Neurodegenerative Disease

[14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31]

[32] [33] [34]

87

M.M. Tompkins et al., Apoptotic-like changes in Lewy-body-associated disorders and normal aging in substantia nigral neurons, Am.J.Pathol., 150, (1997) 119-131. H. Mochizuki et al., Histochemical detection of apoptosis in Parkinson's disease, J.Neurol.Sci., 137, (1996) 120-123. S. Kosel et al., On the question of apoptosis in the parkinsonian substantia nigra, Acta Neuropathol.(Berl), 93, (1997) 105-108. U. Wullner et al., Cell death and apoptosis regulating proteins in Parkinson's disease—a cautionary note, Acta Neuropathol.(Berl), 97, (1999) 408-412. S. Vyas et al., Expression of Bcl-2 in adult human brain regions with special reference to neurodegenerative disorders, J.Neurochem., 69, (1997) 223-231. A. Tortosa et al., Bcl-2 and Bax proteins in Lewy bodies from patients with Parkinson's disease and Diffuse Lewy body disease, Neurosci.Lett., 238, (1997) 78-80. A. Hartmann et al., Caspase-3: A vulnerability factor and final effector in apoptotic death of dopaminergic neurons in Parkinson's disease, Proc.Natl.Acad.Sci.U.S.A, 97, (2000) 2875-2880. P.L. McGeer et al., Rate of cell death in parkinsonism indicates active neuropathological process, Ann.Neurol., 24, (1988) 574-576. J.M. Fearnley and A.J. Lees. Ageing and Parkinson's disease: substantia nigra regional selectivity, Brain, 114, (1991)2283-2301. P.K. Morrish et al., An [18F]dopa-PET and clinical study of the rate of progression in Parkinson's disease, Brain, 119, (1996) 585-591. F.J. Vingerhoets et al., Longitudinal fluorodopa positron emission tomographic studies of the evolution of idiopathic parkinsonism, Ann.Neurol., 36, (1994) 759-764. B. Bogerts et al., A morphometric study of the dopamine-containing cell groups in the mesencephalon of normals, Parkinson patients, and schizophrenics, Biol.Psychiatry, 18, (1983) 951-969. W. Paulus and K.Jellinger. The neuropathologic basis of different clinical subgroups of Parkinson's disease, J.Neuropathol.Exp.Neurol., 50, (1991) 743-755. Y.P. Goldberg et al., Cleavage of huntingtin by apopain, a proapoptotic cysteine protease, is modulated by the polyglutamine tract, Nature Genet., 13, (1996) 442-449. R. von Coelln et al., Rescue from death but not from functional impairment: caspase inhibition protects dopaminergic cells against 6-hydroxydopamine-induced apoptosis but not against the loss of their terminals, J.Neurochem., 77, (2001) 263-273. E.N. Burright et al., Mouse models of human CAG repeat disorders, Brain Pathol., 7, (1997) 965-977. E.N. Burright et al., SCA1 transgenic mice: A model for neurodegeneration caused by an expanded CAG trinucleotide repeat, Cell, 82, (1995) 937-948. H.B. Clark et al., Purkinje cell expression of a mutant allele of SCA1 in transgenic mice leads to disparate effects on motor behaviors, followed by a progressive cerebellar dysfunction and histological alterations, J.Neurosci., 17, (1997) 7385-7395. X. Lin et al., Polyglutamine expansion down-regulates specific neuronal genes before pathologic changes in SCA1, Nat.Neurosci., 3, (2000) 157-163. L. Mangiarini et al., Exon 1 of the HD gene with an expanded CAG repeat is sufficient to cause a progressive neurological phenotype in transgenic mice, Cell, 87, (1996) 493-506. A. Yamamoto et al., Reversal of neuropathology and motor dysfunction in a conditional model of Huntington's disease, Cell, 101, (2000) 57-66.

88

Parkinson's Disease E. Ronken and G.J.M. van Scharrenburg (Eds.) IOS Press, 2002

Animal models of neurodegeneration in Parkinson's disease Christopher Earl and Jiirgen Sautter Klinikum der Philipps-Universitdt, Klinik fur Neurologic, Rudolf-Bultmann-Strasse 35033 Marburg, Germany Abstract. Parkinson's disease is demonstrated as a neurodegeneration of the doparnine producing cells of the substantia nigra. This leads to a dopamine depletion in the neostriatum and to motor abnormalities. In order to investigate the principal pathological processes involved in the disease and to design new therapeutic strategies, animal models are developed which mimic the events seen in the progression of the human disease. There are mainly two classes of models available. First, those applying a transient pharmacological impairment of the dopaminergic system and secondly, those that apply neurotoxins resulting in the actual destruction of the nigrostriatal system. This review will focus on rodent models of Parkinson's disease applying either the toxin 6-hydroxydopamine or 1-methyl-4-phenyl-1,2,3,6tetrahydropyridine, summarizing their merits and limitations and their use in neuroprotection studies. Keywords: reserpine, 6-hydroxydopamine, MPTP. retrograde neurotoxic lesioning

Idiopathic Parkinson's disease (PD), first described by James Parkinson in 1817, is a common neurodegenerative disorder leading to a triad of symptoms comprising rigidity, resting tremor, and akinesia. These cardinal symptoms, which are not always equally strongly manifested, may be accompanied by postural abnormalities, vegetative symptoms and psycho-organic disturbances. Although defined over a century and a half ago, its aetiology remains unknown. PD is associated with the progressive dysfunction and destruction of pigmented dopaminergic neurones in the substantia nigra pars compacta (SNc) that project to the neostriatum. The nigrostriatal dopamine (DA) system constitutes one of the major afferents of the putamen and the caudate, the former being more directly involved in motor function. Thus, lesions of the DA system show a profound influence on motor activity. Although there is no spontaneous occurrence of PD in animals, several models have been developed to simulate clinical features of PD in animals. These experimental models induce parkinsonian symptoms either by transient pharmacological impairment of dopaminergic neurotransmission or by partial or near complete neurotoxic destruction of the dopaminergic neurones of the SNc. A well suited animal model of PD should allow first, to screen reliably for new potential therapeutic drugs; secondly to elucidate their mechanisms of action; and thirdly to study the pathophysiological mechanisms underlying the human disease. Transient pharmacological impairment In 1954 Plummer et al. [1] observed that reserpine produced marked sedation and a sharp decrease in motor activity with resultant hypokinesia, akinesia and catalepsy in rats. It was demonstrated in 1957 that reserpine depleted the brain of DA and that the behavioural syndrome could be reversed by L-dopa [2]. Reserpine-treated animals also present other

C. Earl and J. Sautter / Animal Models of Neurodegeneration

89

symptoms which resemble those observed in PD, the most frequent being rigidity of skeletal muscles, tremor and postural flexion. The pharmacological disruption of dopaminergic function has since been studied using a number of other drugs including some neuroleptics such as chlorpromazine, haloperidol and a-methyl-p-tyrosine. The majority of these DA antagonists produce, in both humans and animals, a syndrome of three neurological symptoms: akinesia, which predominates, rigidity and resting tremor [3]. Although DA antagonists induce extrapyramidal symptoms, their relevance for animal models of PD is limited. For example, the induced behavioural impairments are acute rather than chronic. Moreover, most DA antagonists leave the presynaptic terminals and nigral cell bodies intact, whereas PD involves the degeneration of dopaminergic neurones of the SNc. The acute 6-hydroxydopamine model The development of chemical agents that selectively destroy DA neurones in laboratory animals initiated a new era of research in PD. The first such neurotoxin was 6hydroxydopamine (6-OHDA), the isomer of noradrenaline [4]. Bilateral injection of 6-OHDA into the SNc or close to the SN into the medial forebrain bundle in rats leads to a near complete loss of dopaminergic neurones in the SN which causes severe neurological impairments which are suggestive of some symptoms of PD. Thus, rats become akinetic and show a period of aphagia, adipsia and severe sensory neglect [5]. Following unilateral injection of 6-OHDA behavioural deficits are much less pronounced but systemic administration of the DA releasing drug amphetamine or the DA agonist apomorphine amplify the induced motor asymmetry resulting in body rotations [6]. This rotational behaviour reflects the degree of DA depletion and damage to the nigrostriatal DA pathway. Thus, amphetamine and apomorphine induce intense (ipsilateral or contralateral, respectively) body rotations provided that the striatal DA depletion exceeds 90% [7,8]. Whereas a minimum DA depletion of about 60% is required for amphetamine to exert any rotational behaviour, a DA depletion between 75-90% is associated with reliable circling behaviour [9]. A decrease in the number of amphetamine induced ipsilateral rotations, e.g. after treatment with a therapeutic drug or after grafting of dopaminergic neurones, would therefore point to restored DA levels on the previously damaged side. This toxin induced syndrome with its characteristic behavioural response to dopaminergic drugs is an invaluable pharmacological tool and a standard model of PD. However, a major difference to PD concerns the temporal progression which underlies the human neuropathology compared to the rapid (acute) changes induced by injection of 6-OHDA into the SN or close to it [10]. The subacute (intrastriatal) 6-hydroxydopamine model Recent studies using intrastriatal 6-OHDA application suggest that this type of lesion can lead to partial damage of DA terminals and to a delayed and progressive loss of nigral DA neurones with a concomitant neuronal atrophy [11,12,13,14]. A progressive loss of nigral DA neurones pre-labeled with the retrograde fluorescent tracer fluorogold has been observed for up to 16 weeks following a single intrastriatal injection of 6-OHDA [14]. The intrastriatal 6-OHDA lesion model thus may better resemble the slowly developing human disease at the cellular nigral level. Partial depletion of striatal DA as well as amphetamine induced turning behaviour has also been described after intrastriatal 6-OHDA [12,15]. Although increasing doses of intrastriatally applied 6-OHDA led to an increasing DA depletion as indicated by the loss of striatal DA uptake sites and to increasing rotations [15,16], behavioural symptoms are less pronounced in this model. However, a recent study

90

C. Earl and J. Sautter / Animal Models of Neurodegeneration

showed that modifying the dose of 6-OHDA and the number of injection sites within the striatum can produce specific and more prominent behavioural deficits [17]. Mechanisms of 6-OHDA-toxicity Several hypotheses concerning the mechanisms of action of 6-OHDA have been put forward [18]. 6-OHDA can generate highly cytotoxic free radicals (i.e. hydrogen peroxide and hydroxyl radical) [19]. Peroxidative membrane damage probably due to free radical generation has been observed after intrastriatal injection of 6-OHDA [20]. Recently it has been shown that 6-OHDA, similar to 1-methyl-4-phenylpyridinium (MPP+), the presumed neurotoxic metabolite MPTP, is a potent inhibitor of respiratory chain enzymes suggesting a neurotoxic pathway for 6-OHDA not primarily involving oxidation products [21]. In addition, disturbance of intracellular calcium homeostasis by excessive entry of calcium and/or release from intracellular storage sites either may be a causal factor for neurotoxin mediated cell death or may represent a common final pathway involving other mechanisms [22]. Specifically, it is not clear whether intrastriatal 6-OHDA acts directly on the nigral cell bodies (following uptake and retrograde transport) or whether the toxin mediates a rapid destruction of striatal DA-terminals and the progressive nigral cell degeneration is caused by another mechanism (e.g. lack of a target derived neurotrophic factor). The neurotoxicity of both 6-OHDA and MPTP may largely depend on their incorporation into DA neurones via uptake by the dopamine transporter (DAT) [23,24]. DAT is preferentially localized to distal dendritic and axonal processes of nigral dopaminergic neurones [25] thereby facilitating reuptake of DA but also uptake of neurotoxins. If, as suggested, endogenous or environmental toxins play a role in the aetiology of PD [26], the striatal DA terminals may be the site where the degeneration of nigral neurones begins. On the other side there is evidence that mainly pigmented cell bodies in the SN are affected in PD [27]. These neurones contain neuromelanin, an autoxidation byproduct of catecholamines and autoxidation of DA can lead to the formation of cytotoxic quinones and hydroxyl radicals suggesting that the degeneration of DA neurones may begin in the cell body [27,28]. An example of a successful neuroprotective compound tested for its efficacy in this animal model is glial cell line-derived neurotrophic factor (GDNF) which, intracerebrally infused to the SN, completely protected the nigral DA neurones from 6-OHDA induced degeneration [29]. This model has also been very helpful in gene therapeutic studies in which GDNF also proved neuroprotective when applied via adenoviral vectors [30,31] and even when vector delivery was delayed by one week following 6-OHDA [32]. In contrast, nimodipine, a calcium entry blocker, was not effective in protecting the nigral neurones in this animal model [33]. The 1-methyl-4-phenyl-l,2,3,6-tetrahydropyridine model 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) is a potent parkinsonian agent [34] and its systemic application to mice or non-human primates represents widely used and well characterized models of PD [35]. The neurotoxic action of MPTP is mediated by its MAO-mediated metabolite 1-methyl-4-phenyl-pyridinium ion (MPP+) [36]. MPP+ is taken up via the DA uptake system [37] and accumulates within the mitochondria of nigral dopaminergic neurones where it selectively inhibits complex I of the respiratory chain leading to cellular energy failure and generation of cytotoxic free oxygen radicals [38]. The systemic injection of MPTP leads to a bilateral degeneration of the SN. In primates, the toxin induced deficits mirror the neurological abnormalities manifested in human patients such as bradykinesia, rigidity of the limbs, flexed posture, poor movement

C Earl and J. Sautter / Animal Models of Neurodegeneration

91

initiation, masked face, eye-lid closure, impaired chewing movements, drooling, loss of vocalisation and in some cases postural tremor [39,40]. The various parkinsonian symptoms can be evaluated by the use of clinical rating scales and specially designed automated apparatus to measure locomotor activity. Ethical concerns for using non-human primates as well as their limited availability and the costs for such studies, however, put a limit to the use of the primate MPTP-model of PD. Most rodents, including rats, show no or only a limited sensitivity towards MPTP. However, a specific strain, the C57 black mouse, is sensitive towards the toxin and thus commonly used with the MPTP mouse model [41]. Although these mice, when lesioned with MPTP, show massive damage to the dopaminergic system, such as loss of nigral neurones and striatal dopamine depletion, they can recover from behavioural symptoms relatively fast. Thus, initially observable acute symptoms such as mydriasis, piloerection, hypersalivation and clonic seizures wear off approximately thirty minutes after MPTP exposure [42]. Hypokinesia, visible as decreased spontaneous motor activity, can occur but disappears about seven days after MPTP exposure despite a more than 90% loss of striatal DA [43]. On average, a loss of 40% of tyrosine hydroxylase (TH) positive dopaminergic cells can be observed in this model [44,45]. However, it has been speculated that there occurs some recovery of "lost" or functionally suppressed TH-positive nigral cells in this model, especially with low doses of MPTP. This could, of course, also explain some of the functional recovery as mentioned above. The MPTP model of PD has been used in a number of studies to assess the potential neuroprotective effects of a variety of different compounds. Thus, the NMDA antagonists CPP and MK-801 [46], the L-type calcium channel blocker nimodipine [45], and the neurotrophic factors basic fibroblast growth factor (bFGF) and GDNF [47,48] have been shown to exert neuroprotective effects on nigral dopaminergic cells and/or striatal dopamine levels. More recently, various types of radical scavengers, e.g. sodium salicylate [49], a-phenyl-N-tert-butyl nitrone (PBN) [50] and 4-hydroxy-2,2,6,6-tetramethylpiperidine-1-oxyl (tempol) [51] have been successful neuroprotectors against MPTPinduced damage of the nigrostriatal system. To summarize, there are currently two models of PD which best fulfil the criteria required for useful and practical in vivo testing systems for potentially neuroprotective drugs, the subacute (intrastriatal) rat model and the MPTP mouse model. Both models simulate specific features of the human disease to various degrees and the two toxins 6-OHDA and MPTP likely involve several different mechanisms of dopaminergic toxicity. Compared to the relatively simple systemic application of MPTP in mice, intracerebral stereotactic injection of 6-OHDA is more difficult and time consuming. In contrast, the subacute intrastriatal 6-OHDA model may better simulate the progressive mode of the nigral cell loss compared to the more acute deficits produced by MPTP. Specific tests for assessing behavioural (motor) symptoms are available or may be established in both models. We therefore conclude that both models should not be used alone but rather in combination for investigating potential drug effects and the underlying mechanisms. References [I] [2] [3]

A. Plummer et al., Pharmacology of rauwolfia alkaloids, including reserpine, Ann. N. Y. Acad. Sci. 59 (1954)8-21. A. Carlsson et al., 3,4-Dihydroxyphenylalanine and 5-hydroxytryptophan as reserpine antagonists, Nature 180 (1957) 1200. O. Hornykiewicz, Biochemical and pharmacological aspects of akinesia. In: J. Siegfried (ed.), Parkinson's Disease. Huber, Bern, 1972, pp. 127-149.

92

[4] [5] [6J [7] [8] (9] [10] [11] [12] [13] [14]

[15] [16] [17]

C. Earl and J. Sautter / Animal Models of Neurodegeneration

H. Thoenen and J.P. Tranzer, The pharmacology of 6-hydroxydopamine. Ann. Rev. Pharmacol. 13 (1973) 169-180. M.J. Zigmond et al., Compensations after lesions of central dopaminergic neurons: some clinical and basic implications. Trends Neurosci. 13 (1990) 290-296. U. Ungerstedt and G.W. Arbuthnott, Quantitative recording of rotational behaviour in rats after 6hydroxydopamine lesions of the nigrostriatal dopamine system. Brain Res. 24 (1970) 485-493. A. Dravid et al., Spontaneous recovery from motor asymmetry in adult rats with 6-hydroxydopamineinduced partial lesions of the nigrostriatal pathway. Brain Res. 311 (1984) 361-365. M.J. Zigmond et al., Neurochemical compensation after nigrostriatal bundle injury in an animal model of preclinical parkinsonism. Arch. Neural. 41 (1984) 856-861. J.L. Hudson et al., Correlation of apomorphine- and amphetamine-induced turning with nigrostriatal dopamine content in unilateral 6-hydroxydopamine lesioned rats, Brain Res. 626 (1993) 167-174. B.S. Jeon et al., 6-Hydroxydopamine lesion of the rat substantia nigra: Time course and morphology of cell death, Neurodegeneration 4 (1995) 131 -137. K. Berger et al., Retrograde degeneration of nigrostriatal neurons induced by intrasiriatal 6hydroxydopamine injection in rats. Brain Res. Bull. 26 (1991) 301-307. J.L. Cadet et al., Long-term behavioral and biochemical effects of 6-hydroxydopamine injections in rat caudate-putamen, Brain Res. Bull. 26 (1991) 707-713. Y. Ichitani et al., Degeneration of the nigral dopamine neurons after 6-hydroxydopamine injection into the rat striatum, Brain Res. 549 (1991) 350-353. H. Sauer and W.H. Oertel, Progressive degeneration of nigrostriatal dopamine neurons following intrastriatal terminal lesions with 6-Hydroxydopamine: A combined retrograde tracing and immunocytochemical study in the rat, Neuroscience 59 (1994) 401-415. B.E. Jones et al., A continuous striatal infusion of 6-hydroxydopamine produces a terminal axotomy and delayed behavioral effects, Brain Res. 709 (1996) 275-284. S. Przedborski et al., Dose-dependent lesions of the dopaminergic nigrostriatal pathway induced by intrastriatal injection of 6-Hydroxydopamine, Neuroscience 67 (1995) 631 -647. D. Kirik et al., Characterization of behavioural and neurodegenerative changes following partial lesions of the nigrostriatal dopamine system induced by intrastriatal 6-hydroxydopamine in the rat. Exp. Neurol. 152 (1998) 259-277. G. Jonsson, Chemical neurotoxins as denervation tools in neurobiology, Ann. Rev. Neuroscience 3 (1980) 169-187. R.E. Heikkila and G. Cohen, Inhibition of biogenic amine uptake by hydrogen peroxide: A mechanism tor toxic effects of 6-hydroxydopamine, Science 172 (1971) 1257-1258. R. Kumar et al., Free radical-generated neurotoxicity of 6-hydroxydopamine. J. Neurochem. 64 (1995) 1703-1707. Y. Glinka and M.B.H. Youdim, Inhibition of mitochondrial complexes I and IV by 6-hydroxydopamine, Eur. J. Pharmacol.-Environm. Toxicol. Pharmacol. Sect. 295(1995) 329-332. M. Gerlach et al., Molecular mechanisms for neurodegeneration - Synergism between reactive oxygen species, calcium, and excitotoxic ammo acids. Adv. Neurol. 69 (1996) 177-194. J.A. Javitch et al., Parkinsonism-inducing neurotoxin, N-methyl-4-phenyl-1.2,3,6-tetrahydropyridine: uptake of the metabolite N-methyl-4-phenylpyridine by dopamine neurons explains selective toxicity. Proc. Natl. Acad. Sci. USA 82 (1995) 2173-2177. E. Sundstrom et al., Uptake inhibition protects nigro-striatal dopamine neurons from the neurotoxicity of 1-methyl-4-phenylpyridine (MPP+) in mice. Eur. J. Pharmacol. 131 (1986) 289-292. M.J. Nirenberg et al., The dopamine transporter is localized to dendritic and axonal plasma membranes of nigrostriatal dopaminergic neurons, J. Neurosci. 15 (1996) 436-447. P. Jenner et al., New insights into the cause of Parkinson's disease. Neurology 42 (1992) 2241-2250. E. Hirsch, Why are nigral catecholaminergic neurons more vulnerable than other cells in Parkinson's disease, Ann. Neurol. 32 (1992) 88-93. B. Fornstedt, Role of catechol autoxidation in the degeneration of dopamine neurons. Acta. Neurol. Scand. Suppl. 129(1990) 12-4. H. Sauer et al., GDNF but not TGFß prevents delayed degeneration of nigral dopaminergic neurons following striatal 6-hydroxydopamine-lesion, Proc. Natl. Acad. Sci. USA 92 (1995) 8935-8939. R.J Mandel et al., Midbrain injection of recombinant adeno-associated virus encoding rat glial cell line-derived neurotrophic factor protects nigral neurons in a progressive 6-hydroxydopamine-induced degeneration model of Parkinson's disease in rats, PNAS 94 (1997) 14083-14088. D. L. Choi-Lundberg et al., Dopaminergic neurons protected from degeneration by GDNF gene therapy. Science 275 (1997) 838-841. D.A. Kozlowski et al., Delivery of a GDNF gene into the substantia nigra after a progressive 6-OHDA lesion maintains functional nigrostriatal connections. Exp. Neurol. 166(2000) 1-15.

C. Earl and J. Sautter / Animal Models of Neurodegeneration

[33]

[34] [35] [36] [37]

[38] [39]

[40] [41] [42] [43] [44] [45] [46] [47] [48] [49]

[50] [51]

93

J. Sautter et al.,Degeneration of pre-labelled nigral neurons induced by intrastriatal 6-hydroxydopamine in the rat: Behavioral and biochemical changes and pretreatment with the calcium entry blocker nimodipine, Exp. Brain Res. 117 (1997) 111-119. J.W. Langston, MPTP and Parkinson's Disease. Trends Neurosci. 8 (1985) 79-83. M. Gerlach and P. Riederer, Animal models of Parkinson's disease: an empirical comparison with the phenomenology of the disease in man, J. Neural Transm. 103 (1996) 987-1041. S.P. Markey et al., Intraneuronal generation of a pyridinium metabolite may cause drug-induced parkinsonism, Nature 311 (1994) 464-467. R. Drucker-Colin and F. Garcfa-Harnandez, Can we treat Parkinson's disease using adrenal transplants? In: C. Hachinski, (ed.), Challenges in Neurology. F.A. Davis Co., Philadelphia, 1992, pp. 165182. M.F. Beal et al., Do defects in mitochondrial energy metabolism underlie the pathology of neurodegenerative diseases? Trends Neurosci. 16 (1993) 125-131. R.S. Burns et al.,, A primate model of Parkinsonism: selective destruction of dopaminergic neurons in the pars compacta of the substantia nigra by N-methyl-4-phenyl-l,2,3,6-tetrahydropyridine, Proc. Natl. Acad. Sci. USA 80 (1983) 4546-4550. K.S. Bankiewicz, MPTP-induced parkinsonism in nonhuman primates. In: P.M. Conn (ed.), Methods in Neurosciences, Vol. 7, Lesions and Transplantation. Academic Press, London, 1991, pp. 168-182. E. Sundstrom et al., Studies on the effect of l-methyl-4-phenyl-l,2,3,6-terahydropyridine (MPTP) on central catecholamine neurons in C57BL/6 mice. Comparison with three other strains of mice, Brain Res. 405(1987)26-38. E. Sundstrom et al., Chronic neurochemical and behavioral changes in MPTP-lesioned C57BL/6 mice: A model for Parkinson's disease, Brain Res. 528 (1990) 181-188. A. Zuddas et al., Selective lesion of the nigrostriatal dopaminergic pathway by MPTP and acetaldehyde or diethyldithiocarbamate, Neurochem. Int. (Suppl.) 20 (1992) 287S-293S. A. Kupsch et al., Do NMDA receptor antagonists protect against MPTP-toxicity? Biochemical and immunocytochemical analyses in black mice, Brain Res. 592 (1992) 74-83. A. Kupsch et al.,Pretreatment with nimodipine prevents MPTP-induced neurotoxicity at the nigral, but not at the striatal level in mice, NeuroReport 6 (1995) 621-625. L. Turski et al., Protection of substantia nigra from MPP+ neurotoxcity by N-methyl-D-asparate antagonists, Nature 349 (1991) 414-418. D. Otto and K. Unsicker, Basic FGF reverses chemical and morphological deficits in the nigrostriatal system of MPTP-treated mice, J. Neurosci. 10 (1990) 1912-1921. A. Tomac et al., Protection and repair of the nigrostriatal dopaminergic system by GDNF in vivo, Nature 373 (1995) 335-339. B. Ferger et al., Salicylate protects against MPTP-induced impairments in dopaminergic neurotransmission at the striatal and nigral level in mice, Naunyn Schmiedeberg's Arch. Pharmacol. 360 (1999) 256-261. B. Ferger et al., The protective effects of PBN against MPTP toxicity are independent of hydroxyl radical trapping, Pharmacol. Biochem. Behav. 65 (2000) 425-431. R.T. Matthews et al., Novel free radical spin traps protect against malonate and MPTP neurotoxicity, Exp. Neurol. 157 (1999) 120-126.

94

Parkinson's Disease E. Ronken and G.J.M van Scharrenburg (Eds.) IOS Press, 2002

Parkinsonian Features of Ataxia-Telangiectasia and of the Atm-Deficient Mouse Howard T. J. Mount 1 , Yingji Wu 1 , Paul Fluit1, Qiuna Bi' Thomas O. Crawford2 and Allen S. Mandir 1

Centre for Research in Neurodegenerative Diseases and Division of Neurology, Department of Medicine, University of Toronto, 6 Queen's Park Crescent West, Toronto, Ontario M5S 3H2 Canada, 2 Department of Neurology, Johns Hopkins Hospital, Baltimore, MD, U.S.A

Abstract. Ataxia-telangiectasia (A-T) is an autosomal recessive disorder that is caused by mutations in the ATM gene and is associated with progressive cerebellar ataxia and extrapyramidal deficits. Mice deficient in the Ann gene display phenotypes that recapitulate many features of the human disease, including mil., progressive sensorimotor impairment. Although the Atm-knockout (Atm-KO) mice do not exhibit frank neuropathology, an emerging body of evidence suggests that behavioral anomalies in the mice may be accompanied by prodromal degenerative changes, some of which resemble those associated with Parkinson's disease. In particular, Atm-KO mice express markers of oxidative damage, constitutive elevation of the transcription factor NF-kB, increased numbers of lysosomes, loss of tyrosine hydroxylase-immunoreactivity in striatum and substantia nigra and behavioral hypersensitivity to amphetamine. The apparent involvement of extrapyramidal dysfunction in both A-T patients and Atm-KO mice, suggests that ATM deficiency may provide an useful model for understanding aspects of the parkinsonian syndrome. Keywords: cerebellum, purkinje cell, NF-kB, Atm, p7rNTR

Human neurodegenerative diseases rarely have naturally-ocurring equivalents in laboratory animals. Consequently, though they can be described by neurologists, they cannot be studied in an invasive manner. Genetically engineered mouse models circumvent this shortfall and provide an opportunity to gain insights into the process of disease pathogenesis. Recent attempts to engineer a genetic mouse model of Parkinson's disease (PD) have focussed on a-synuclein, a presynaptic phosphoprotein that forms self-aggregating fibrils in the Lewy bodies that characterize idiopathic PD. Two missense mutations of asynuclein (A53T and A30P) have been linked to rare, inherited forms of early-onset familial PD. Moreover, transgenic mice that express very high levels of the wild type human transgene under control of the mouse PDGF-B promoter, exhibit amorphous accumulation of a-synuclein, as well as modest reduction in tyrosine hydroxylase (TH) activity and immunolabeling in the striatum and substantia nigra [1]. The relevance of these anomalies to PD remains unclear. Transgenic mice in which either wild type, or A53T mutant human a-synuclein are expressed at somewhat lower levels, under control of Thy-1 regulatory sequences, exhibit neuronal a-synucleinopathies and Lewy pathology, without an obvious dopaminergic disturbance [2. 3].

H.T.J. Mount et al. /Parkinsonian Features of Ataxia-Telangiectasia

95

Animals with mutations that model other diseases may sometimes afford opportunitites to gain insight into the pathogenesis of PD. Mouse models of the neurodegenerative disorder, ataxia-telangiectasia (A-T) have been found to exhibit multiple cellular, neurochemical and behavioral anomalies that resemble changes found in the PD brain. Strikingly, one laboratory has even reported evidence that the A-T mouse exhibits a marked, selective loss of nigrostriatal dopamine neurons and a behavioral disturbance that can be abrogated by treatment with L-DOPA [4]. Whilst at first blush surprising, the notion of dopaminergic degeneration in the A-T mouse is consistent with a clinical literature that suggests an extrapyramidal component to the neurological manifestions of the disease. In this paper, we review clinical evidence for extrapyramidal involvement in A-T. We then briefly discuss the biology of utaxia-telangiectasia mutated (ATM), the protein that is deficient in A-T patients. Finally, we survey the work from several laboratories, including our own, that suggest potentially broad relevance of neurochemical and cellular changes in the A-T mouse to human neurodegenerative disorders.

1.

Neurological deficits and Degeneration in Ataxia-telangiectasia

A-T is an autosomal recessive disease that involves progressive deterioration of the nervous, immune and endocrine systems. Cardinal features of A-T include cerebellar ataxia, immunodeficiency, thymic degeneration, growth retardation, infertility, premature senescence, chromosomal instability, predisposition to lymphoreticular malignancies and radiation sensitivity [reviewed in 5]. Cerebellar ataxia is most commonly associated with the neurologic aspects of A-T, but often underemphasized is the prevalence of extrapyramidal signs not related to cerebellar pathophysiology. The onset of symptoms and overt signs in A-T is usually within the first 2 years of life. Life span is normally severely restricted in the disease and death often occurs by the third decade. However, case reports demonstrate a wide range of longevity including survival into the sixth decade. Likely this variability results from differing gene mutants and environmental exposure - no treatments have been found to alter the course of this disease. 1.1

Clinical Presentation

Although telangiectasias (dilated blood vessels) in the eye, the back of the knee and the bridge of the nose are frequently seen, they are typically not the first presentations of this disease. The full-fledged clinical scenario is frequently not apparent until later in life, as such a correct diagnosis is often delayed. Most commonly, children begin to walk at an appropriate age, but the instability that is normally present when learning to walk, fails to improve. Though overall motor function may be static or slowly progressive at first, these symptoms are not static and progress, unlike cerebral palsy and other childhood cerebellar ataxias, eventually leading to requirement of a wheelchair during late childhood. The signs that present later in life are remarkably distinct to A-T, yet quite variable among patients [6]. Thus, a limited number of systems appear to be involved in the disease, but patients manifest them at different times. 1.2

Nature of Neurologic Features

In addition to the early appearance of ataxia, neurologic features of A-T include oculomotor abnormalities, neuropathy and a myriad of extrapyramidal features. The co-appearance of these characteristics may give rise to distinctive features of the disease. The ataxic gait of children with A-T has unique characteristics of being narrow based with surprisingly fewer

96

H. T.J. Mount et al. / Parkinsonian Features of Ataxia-Telangiectasia

falls than expected. Dysarthria often presents with cerebellar-type pathology, but also with signs characteristic of extrapyramidal pathology. Despite the cerebellar involvement, hypotonia is not prominent and often there is increased tone with cross activation of the limbs. Both hyperkinetic and hypokinetic features accompany A-T (Table 1). Hyperkinetic features of the disease include athetosis, chorea and myoclonus. Athetosis is prevalent even early in the disease and often progresses to choreiform movements. Despite the presence of a mixed sensory neuropathy, these are true athetoid movements rather than pseudoathethosis as sensory testing is normal to all modalities at time of presentation. Athetosis is present at rest and even in a supine position. Some patients further progress and demonstrate myoclonus. We note in clinic, as have others, dystonic postures [7], often involving the feet during gait and the hands during activation. Outside these hyperkinetic aspects of A-T, other hypokinetic extrapyramidal features abound. Masked faces are common as the disease advances, and sialorrhea may present at an early age. Tone is often increased with contralateral activation and in a proportion of patients, bradykinesia, akinesia and reduced eye blink rate are noted. As in the extrapyramidal presentations of Alzheimer's disease, rest tremor is never noted in these patients [8]. All of the extrapyramidal features of A-T tend to present symmetrically, which we interpret to suggest a pharmacologic rather than structural etiology. 1.3

Progression of Extrapyramidal Signs and Symptoms

Although substantial variability exists in the phenotype of A-T, there is a high correlation between increasing score of the AT scale with age. Thus, although the areas of brain seemingly affected in A-T are likely different among individuals, a characteristic rate of overall disease progression exists [6]. The correlation between presence of non-cerebellar extrapyramidal signs and age is not as high as the correlation between these signs and overall AT score. This further suggests that functional domains of the brain are affected at different times across patients, and seem progressive within the domain. Gait deteriorates so as to require use of a wheelchair, usually by the first decade. Dysarthria tends to progress as well, with prominent slowing. Cerebellar ataxia also progresses over time with axial as well as appendicular involvement. What is most demonstrative of a severe cerebellar ("rubral") tremor may be seen in a proportion of patients as the disease progresses. Facial expression diminishes from normally expressive in the youngest of patients, to diminishment of subtle facial expression and then in some to a mask-like appearance. Other parkinsonian signs such as bradykinesia are prominent in a proportion of patients as the disease progresses. 1.4

Neuropathology

Degeneration of cerebellar Purkinje cells, granule cell loss and diffuse fibrillary gliosis are the hallmarks of A-T. Despite the prevalence of extrapyramidal signs, past reports have not demonstrated consistent changes within the basal ganglia. However, sporadic reports indicate that extra-cerebellar areas of the brain are involved in AT. These include the striatum [9] and substantia nigra pars compacta (SNpc) [10]. More specifically, some report Lewy bodies within the SNpc in A-T. The areas of brain implicated by the clinical and neuropathologic features of A-T suggest that it is reflective of spinocerebellar ataxia (SCA). Classically SCA's are trinucleotide repeat disorders, however this is not the case in A-T. Current theories suggest that expression of polyglutamine proteins encoded by trinucleotide repeats are causative of SCA. However, the finding that some types of SCA the trinucleotide repeats are not in noncoding regions suggests other pathways must be important in neuronal death. Recent neuropathologic findings in our laboratory further

H. T.J. Mount et al. /Parkinsonian Features of Ataxia-Telangiectasia

97

support a link to the SCAs, and may reveal insight into mechanisms of neuronal degeneration. Table 1. Extrapyramidal manifestations associated with A-T Ataxia and cerebellar tremor Choreoathetosis Dystonia Bradykinesia Masked facies Rigidity Lewy bodies within substantia nigra pars compacta (SNpc)

2.

ATM: protein structure and function

Discovery of the gene "mutated in ataxia-telangiectasia" (ATM, 11) fundamentally changed research on A-T. Prior to the discovery, it had been established that the disorder was characterized by multiple "complementation groups". For this reason, it was surmised that the disease was likely polygenic. The realization that mutations in a single gene on 11 q2223 [12] could be identified in all A-T complementation groups challenged this notion. Identification of the gene confirmed the monogenic nature of A-T, and suggested that the ATM protein may play multiple roles within the cell. The ATM gene contains a phosphatidylinositol 3-kinase (PI 3-kinase) domain as well as a leucine zipper motif, which indicates that its function likely requires interaction with DNA. Related molecules from a variety of organisms participate in signal transduction, cell cycle checkpoint control and DNA-repair [13]. The determination that the 370 kDa ATM protein is a nuclear phosphoprotein, devoid of lipid kinase activity, but capable of binding to DNA and of phosphorylating proteins implicated in DNA damage repair, vindicates this notion [reviewed in 14, 15]. Evidence that ATM is found largely outside the nuclei of neurons [16, 17] further suggests the possibility of extranuclear functions of ATM within the CNS. An involvement of the ATM protein in response to DNA damage was initially predicted from the cellular phenotype of A-T cells. Cell cycle checkpoints at the G1, S and G2 phases, that normally arrest cell cycle progression in response to radiation damage, are defective in A-T cells [reviewed in 18]. Hence, semi-conservative DNA synthesis is less inhibited by ionizing radiation or radiomimetic drug treatment in A-T cells than in normal cells ("radioresistant DNA synthesis"). The post-irradiation induction of p53 protein, a substrate of ATM-mediated phosphorylation [19], is delayed and attenuated. The radiosensitivity of A-T cells results from increased apoptosis in response to low radiation doses [18]. Conversely, radioresistance of brain tissue in a mouse model of A-T [20] is consistent with a p53-dependence of radiation-induced cell death in post-mitotic neurons. Radiation sensitivity of A-T cell lines has also been linked to dysregulated, constitutive activation of the transcription factor.,NF-kB [21]. Activation of NF-kB involves phosphorylation and proteolysis of I-kBa, which otherwise complexes to NF-kB, preventing it from migrating to the nucleus and activating transcription. It has been suggested that ATM may play a direct, or indirect role in phosphorylating I-kBa [22], thereby explaining dysregulation of NF-kB in A-T. This finding lends further support to a model that suggests a central role for ATM in management of the cellular response to oxidative stress and DNA damage. The pleiotropic clinical manifestations of A-T and the demonstrated involvement

98

H. T.J. Mount et al. / Parkinsonian Features of Ataxia-Telangiectasia

of ATM in initiating signal transduction pathways that protect cells from injury, suggest an important role for this protein of potential relevance to a variety of degenerative disorders. 3.

The Atm-deficient mouse

At least five laboratories have developed genetic murine models of A-T, by ablating Atm. the mouse homolog of the human ATM gene [23-25, 20, 26]. In large measure, the cellular phenotypes of these mice recapitulate those of A-T patients. However, with the exception of one report [27], that was not confirmed upon reexamination of this and other lines of Atm-knockout (Atm-KO) mice [17], no groups have discerned evidence of frank pathologic changes in the Atm-deficient brain. It has been suggested that absence of CNS degeneration could be explained by the fact that most Atm-KO mice succumb to thymomas by six months of age, and quite simply may not live long enough to express cumulative degenerative changes [23, 28]. This explanation is not entirely satisfactory, as even the rare mice that live as long as 10 months have shown no degenerative changes in the cerebellum. Moreover, one recently derived strain of Atm-KO mice, that was bred on a 129Sv/C57B6 background and maintained in a pathogen-free environment, failed to develop thymic lymphomas, despite living much longer (13-15 months) than previous strains [26]. No gross degenerative changes were observed in the cerebella of these animals. Initially, the absence of profound degenerative changes in the Atm-KO mouse brain was viewed with disappointment, as it indicated that the A-T mouse would not fully model the significant pathologic changes that occur in human CNS. However, an emerging body of evidence indicates a variety of more subtle anomalies that point to progressive behavioral impairment, oxidative damage, trophic dysregulation and neurochemical deficits.

3.1

Subtle behavioral defects

Atm-deficient mice exhibit subtle anomalies in sensorimotor function. Two reports suggest that Atm-KO mice have a mild gait asymmetry [23, 4], although we [P.F and H.T.J.M.] and others [26] have been unable to replicate these findings. A modest reduction in open field activity [23, 26] has also proven difficult to reproduce [4. P.F and H.T.J.M.]. In contrast, two laboratories, in addition to our own, have found anomalies in the performance of various rotarod tasks [23, 26], suggesting that such tests may provide more robust indices of the behavioral impairment. In a longitudinal study that involved daily testing on an accelerating rotarod, we found the rotarod performance deficit of Atm-KO mice to emerge only after the 14lh postnatal week, thereby indicating that the behavioral deficit may be progressive. Strain-specific anomalies in development of the corpus callosum could potentially contribute to differences in the behavioral observations recorded by various laboratories. Atm-deficient mouse strains have been developed on 129/SvEv and 129/J and, in one instance, may have been crossed into Balb/c backgrounds. Some lines of these substrains display a high incidence of callosal defects [29, 30]. As callosal defects are associated with motor impairment, it will be important to test the contribution of genetic background to the behavioral anomalies of Atm-KO mice. 3.2

Degeneration of dopaminergic neurons?

In the absence of obvious, discrete degenerative changes, it can be difficult to ascribe deficits in murine motor performance to a specific anatomic locus, or motor system. However, mice with neurochemical lesions to the nigrostriatal dopaminergic system exhibit gait anomalies that resemble putative gait anomalies of Atm-KO mice. As deficits in

H.T.J. Mount et al. /Parkinsonian Features of Ataxia-Telangiectasia

99

dopaminergic transmission can be offset by L-DOPA treatment, Segal and coworkers administered L-DOPA to Atm-KO 129SvEv/B16 mice and reported that the gait disruption was abrogated by L-DOPA treatment [4]. Moreover, the Atm-KO mice displayed marked hyperactivity in response to a challenge dose of amphetamine, a finding that is interpreted as indicating upregulation of postsynaptic striatal dopamine receptors. The behavioral evidence of a dopaminergic phenotype was reinforced by immunocytochemical findings. Atm-KO mice displayed a marked reduction in immunostaining for tyrosine hydroxylase (TH) in the striatum and mesencephalon. Striatal immunolabeling with antisera to the dopamine transporter was also dramatically reduced. This suggests that the loss of TH immunolabeling may be explained by degeneration of dopaminergic fibers. Subsequent retrograde labeling studies confirmed a reduction in the number of TH-immunopositive cell bodies in the substantia nigra [31]. The apparent loss of TH immunoreactivity was restricted to dopaminergic neurons, as sections through the locus coeruleus showed no obvious loss of noradrenergic cell bodies. In addition, the deficit appeared to be progressive. Whilst subtle changes in TH immunoreactivity were detected at 2 months, significant (50%) cell loss was only observed in 4 month old mice. To date, the hypodopaminergic phenotype has been observed in one of the five lines of Atm-KO mice. It will be important to establish whether the phenotype is peculiar to this line, or a bonafide consequence of Atm-deficiency. To address this issue, we are currently completing an examination of catecholamine and metabolite levels in age-matched groups of Atm-KO animals with an inbred 129SvEv-Tac background [23]. 3.3

Oxidative Stress and Damage

The absence of any profound degenerative changes in Atm-deficient brain prompted several groups, including our own, to look for indicators of cellular stress and prodromal degeneration. Published neurochemical anomalies in the Atm-KO brain include elevated expression and activity of the enzyme hemoxygenase-2, as well as increased levels of nitrosylated proteins in the whole brain [28]. Isoprostanes, indicators of oxidative damage to lipid, were elevated in the testes, though they were not significantly affected in the brain. Conversion of peripherally administered salicylate to the nonenzymatic hydroxylation product, 2,3-dihydroxybenzoic acid, revealed basal levels of reactive oxygen species to be elevated 7.5-fold in the cerebellum and 3.5-fold in basal ganglia, but unaffected in the cortex of Atm-KO mice [32]. Collectively, these neurochemical markers provide strong evidence of oxidative damage in the Atm-KO brain. This interpretation is supported by electron micrographs of the 12 week-old Atm-deficient CNS, that reveal a dramatic increase in the number of lysosomes [17], a correlate of degenerative changes that is also observed in afflicted regions of the Alzheimer brain [33]. Transcription factors associated with a cellular stress response may also be affected in the Atm-KO brain. We have found constitutive kB-like binding activity to be elevated in nuclear extracts from Atm-deficient cerebella of mice aged 17 to 90 days, in comparison to levels of binding activity in wild type littermates [Fig 1]. A similar enhancement of nuclear KB binding activity was observed in immortalized peripheral cell lines from A-T patients [21]. Intriguingly, normalization of constitutive KB binding activity in these patient cell lines was achieved by expression of a cDNA that encoded a truncated form of the protein, IkBa. The truncated IkBa, like full length LcBa, complexes to NF-kB. However, the truncated form does not undergo phosphorylation-induced protelysis and hence functions as an inhibitor of NFC-IB activation. A role for dysregulated NF-kB signaling in the A-T phenotype was inferred from the observation that truncated IkBa also abrogated the characteristic radiosensitivity of A-T patient cells.

100

H.T.J. Mount et al. /Parkinsonian Features of Ataxia-Telangiectasia

The significance of altered nuclear NF-kB binding activity in the Atm-KO brain is unclear. It is consistent with the notion of a protective response to cellular stress in affected brain areas and with the observation of increased nuclear translocation of transcriptionally active NF-kB in post-mortem parkinsonian brain [34]. Elevated constitutive NF-kB activity could interfere with induction of genes involved in cytoprotective response to cellular stress, and could also plausibly affect NF-kB-mediated transcription of the tyrosine hydroxylase gene [35], thereby contributing to the hypodopaminergic murine phenotype. Dysregulation of the enzyme, poly(ADP-ribose) polymerase (PARP) might also contribute to the Atm-KO hypodopaminergic phenotype. PARP is a nuclear enzyme responsible for transferring the ADP-ribose moiety of NAD+ to nuclear protein acceptors, in response to stimuli that elicit DNA damage. In response to DNA damage, A-T patient cells fail to exhibit increased levels of (ADP-ribose)n polymers [36]. The absence of a poly(ADP-ribosyl)ation response in the A-T cells may be related to the abnormally high levels of constitutive poly(ADP-ribose) synthesis reported in some A-T patient cells [37]. Activation of PARP by methamphetamine, or MPTP, has been shown to elicit rapid depletion of mouse brain NAD+ (a cofactor in dopamine biosynthsis), decreases in catecholamine levels and the eventual demise of nigral dopamine neurons [38-42]. In AtmKO mice, one group reported no significant alteration of PARP activity in the spleen or thymus [43]. The status of PARP activity in the Atm-KO brain remains to be investigated.

Figure 1. Electrophoretic gel mobility shift assays reveal a delay in the ontogeny of constitutive NF-kBbinding in nuclear extracts from Atm-KO cerebella (KO), relative to samples from wild type littermates. However, beyond postnatal day 7 (7 d) and into maturity (3 mos), constitutive kB binding activity is elevated in Atm-KO cerebella relative to age-matched wild type mice. • indicates the position of specific KB binding. NS indicates non-specific binding in the presence of an excess unlabeled kB probe sequence. Specific binding was not displaced by an excess of mkB, a mutant KB oligonucleotide.

3.4

Developmental Anomalies and Defects in Trophic Factor Responsiveness: Is the p75NTR signaling involved in neuronal degeneration ?

A profound loss of cerebellar Purkinje neurons is the neuropathologic hallmark of A-T. In A-T patients, most Purkinje cells appear to migrate into the Purkinje cell layer and develop morphologically before dying [44]. However, a high proportion of surviving Purkinje neurons have smooth neurites, display abnormal patterns of dendritic arborization and are displaced in the middle and superficial strata of the molecular layer. It has been argued that the presence of such ectopic, poorly differentiated Purkinje cells suggests an important role for ATM early in cerebellar development [45]. In the rodent nervous system, nerve growth factor (NGF). a member of the neurotrophin family of trophic factors, regulates survival and

H.T.J. Mount et al. /Parkinsonian Features of Ataxia-Telangiectasia

101

maturation of Purkinje cells over this period of development [46-48]. As a first step toward understanding how A-T cerebellar dysmorphogenesis might occur, we examined effects of NGF on Purkinje cells isolated from the Atm-KO brain. Neurotrophins mediate effects via p75NTR, the original "NGF receptor" and by way of higher affinity receptors that are protein products of the trk proto-oncogne family [reviewed in 49]. The Trk family of receptors are tyrosine kinases that discriminate in their binding with neurotrophins. TrkA binds NGF; TrkB preferentially interacts with brainderived neurotrophic factor (BDNF) and neurotrophin-4/5; TrkC binds neurotrophin-3. In contrast, p75NTR binds all neurotrophins, but induces NF-kB selectively upon binding NGF [50]. For most responsive cell types, functional neurotrophin receptors are dimeric complexes of Trk, with or without p75NTR. In this context, p75NTR is thought to serve a ligand presentation function. However, p75NTR can also elicit trophic and regressive effects on neurons through signaling mechanisms that are independent of Trks. As the Purkinje cell expresses abundant p75NTR in the absence of TrkA, this cell type provides an opportunity to examine such p75NTR-mediated responses. We previously reported that p75NTR serves a dual survival/death function [51]. We demonstrated NGF-elicited improvement in cultured Purkinje cell survival to be p75NTR-dependent. In addition, we found that reagents capable of blocking p75NTR, in the absence of exogenous NGF, dramatically improved Purkinje cell survival. These findings are consistent with an earlier report of cell death associated with unliganded p75NTR expression [52]. We have recently found that dissociated Purkinje cell cultures from Atm-KO mice survive less well than do cells from wild type littermates, when grown in the absence of exogenous trophic factor. As expected, addition of NGF improved survival of the wild type Purkinje cells. In contrast, NGF killed Atm-KO cells. The cytotoxic effect of NGF was blocked by inhibition of p75NTR:NGF interaction. As p75NTR was expressed at comparable levels in Atm-KO and wild type cerebellum, differences downstream of the receptor.ligand interaction were implicated. We interpret the data as indicating that Atmdeficiency selectively abolishes p75NTR-mediated trophic signaling, without impairing the cytotoxic effects of receptor activation. The reason for a selective deficit in p75NTR-mediated trophism is unclear. One possibility is that NGF fails to elicit a trophic action because of disruption in NF-kB signaling. NGF binding to p75NTR stimulates sphingomyelin hydrolysis, leading to the generation of ceramide [53] and the activation of NF-kB [50]. A delayed ontogeny of constitutive NF-kB activity in the perinatal Atm-KO cerebella [Fig. 1] might indicate delayed expression of NF-kB subunits. However, both p65 and p50 subunits NF-kB were expressed at comparable levels in Atm-/- and Atm+/+ brain and supershift experiments with antibodies to these proteins indicate that they are major components of the cerebellar kB oligonucleotide binding complexes. A second possibility is that Atm-deficiency causes an altered response to the NGFp75NTR-stimulated generation of ceramide. NGF interaction with p75NTR increases ceramide by rapidly activating neutral sphingomyelinase-mediated hydrolysis of phopholipid sphingomyelin. Exogenous ceramide mimics the Purkinje cell survivalpromoting action of NGF, consistent with the notion that ceramide is a downstream element in this p75NTR-triggered pathway. However, ceramide is also an inducer of apoptosis [54] and elevated constitutive levels of ceramide, or over-expression of p75NTR, cause cellular apoptosis. Under conditions of genotoxic stress, ceramide is generated by the enzyme ceramide synthase. ATM has been shown to play an important role in constraining ceramide generation through this pathway [55]. Thus, in Atm-KO neurons, absence of a constraint on ceramide synthase could cause high constitutive levels of ceramide, that render cells vulnerable to a cytotoxic effect of further ceramide, or NGF treatment.

102

H. T.J. Mount et al. /Parkinsonian Features of Alaxia-Telangiectasia

Primary cultures of rat dopaminergic neurons also exhibit ceramide-induced apoptosis and this toxicity is preceded by a transient generation of a reactive oxygen species and nuclear translocation of NF-kB [34]. The involvement of reactive oxygen species in the pathogenesis of PD has long been suspected, based on markers of oxidative damage such as lipid peroxidation [56], superoxide dismutase activity [57], decreased levels of reduced glutathione [58], and increased iron content in the end-stage of the disease [59, 60]. Involvement of p75NTR in dopaminergic cell fate has not been investigated, although NGF-p75NTR-mediated generation of ceramide has been shown to regulate dopamine release from nigral neurons [61]. The parkinsonian substantia nigra does express elevated levels of the cytokine TNF-a, which like NGF elicits both regressive and protective effects through a receptor-mediated activation of sphingomyelinase and nuclear translocation of NF-kB. Receptors for TNF-a are present on the dopaminergic neurons [62] and inappropriate ceramide generation may very well contribute to the pathogenesis of PD [63. 64]. Summary Both A-T patients and the Atm-KO mouse exhibit phenotypes that overlap those associated with Parkinson's disease. The Atm-KO mouse displays a progressive sensorimotor disturbance that may involve dopaminergic neurons. In addition, the mouse shows evidence of oxidative damage and constitutive activation of cellular repair mechanisms. An emerging literature suggests that these same processes are affected in other degenerative diseases, including PD. Thus, insights derived from studying the Atm-KO mouse may illuminate pathogenic mechanisms that are broadly relevant to human neurodegenerative conditions.

Acknowledgement The authors thank Drs. Carrolee Barlow and Anthony Wynshaw-Boris for providing breeding pairs of Atmheterozygotic mice. This work was assisted by grants to H. Mount from the A-T Children's Project (Boca Raton. FL, U.S.A.), the Medical Research Council of Canada and the March of Dimes Birth Defects Foundation (White Plains. NY. U.S.A.). References [ 1 ] E. Masliah et al., Dopaminergic loss and inclusion body formation in a-synuclein mice: Implications for neurodegenerative disorders, Science 287 (2000) 1265-1269. [2] H. van der Putten et al., Neuropathology in mice expressing human a-synuclein. J. Neurosci. 20 (2000) 6021-6029. [3] P.J. Kahle et al., Subcellular localization of wild-type and Parkinson's disease-associated mutant asynuclein in human and transgenic mouse brain, J. Neurosci. 20 (2000) 6365-6373. [4] R. Eilam et al., Selective loss of dopaminergic nigro-striatal neurons in brains of Atm-deficient mice. Proc. Natl. Acad. Sci(USA) 95 (1998) 12653-12656. [5] R.P. Sedgwick and E. Boder, Ataxia-Telangiectasia. In: J.M.B.V. de Jong (ed.), Handbook of Clinical Neurology. Vol 16(60): Hereditary Neuropathies and Spinocerebellar Atrophies. Elsevier, Amsterdam. 1991,pp/347-423. |6] T.O. Crawford et al., Quantitative neurologic assessment of ataxia-telangiectasia. Neurology 54 (2000) 1505-1509. [7] J.B. Bodensteiner et al., Progressive dystonia masking ataxia in ataxia-telangiectasia. Arch. Neurol. 37 (1980)464-465. [8] U. Kischka et al., Electrophysiologic detection of extrapyramidal motor signs in Alzheimer's disease. Neurology 43 (1993) 500-505. [9] M. Koepp et al., Dystonia in ataxia telangeiectasia: Report of a case with putaminal lesions and decreased striaial [ 123 I]iodobenzamide binding. Mov Disord 9 (1994) 455-459

H. T.J. Mount et al. / Parkinsonian Features of Ataxia-Telangiectasia

103

[10] D. P. Agamanolis and J.I. Greenstein, Ataxia-telangiectasia. Report of a case with Lewy bodies and vascular abnormalities within cerebral tissue, J. Neuropathol. Exp. Neurol. 38 (1979) 475-489. [11] K. Savitsky et al., A single ataxia telangiectasia gene with a product similar to PI-3 kinase. Science 268 (1995) 1749-1753. [12] R.A. Gatti et al., Localization of an ataxia-telangiectasia gene to chromosome 1 lq22-23, Nature 336 (1988)577-580. [13] V.A. Zakian, ATM-related genes: what do they tell us about functions of the human gene? Cell 82 (1995)685-7. [14] G. Rotman and Y. Shiloh, ATM: a mediator of multiple responses to genotoxic stress, Oncogene 18 (1999)6135-6144. [15] M.F. Lavin and K.K. Khanna, Review: ATM: the protein encoded by the gene mutated in the radiosensitive syndrome ataxia-telangiectasia, Int. J. Radiat. Biol. 75 (1999) 1201-1214. [16] R.O. Kuljis et al., ATM immunolocalization in mouse neuronal endosomes: implications for ataxiatelangiectasia, Brain Res. 842 (1997) 351-358. [17] C. Barlow et al., ATM is a cytoplasmic protein in mouse brain required to prevent lysosomal accumulation, Proc. Natl. Acad. Sci. (USA) 97 (2000) 871-876. [18] M.S. Meyn, Ataxia-telangiectasia and cellular responses to DNA damage, Cancer Res. 55 (1995) 59916001. [19] K.K. Khanna et al., ATM associates with and phosphorylates p53: mapping the region of interaction, Nature Genet. 20 (1998) 398-400. [20] K.-H. Herzog et al., Requirement for Atm in ionizing radiation-induced cell death in the developing central nervous system, Science 280 (1998) 1089-1091. [21] M. Jung et al., Correction of radiation sensitivity in ataxia telangiectasia cells by a truncated Ik-B-a, Science 268 (1995) 1619-1621. [22] M. Jung et al., ATM gene product phosphorylates I kappa B -alpha, Cancer Res. 57(1997)24-27. [23] C. Barlow et al., Atm-deficient mice: a paradigm of ataxia telangiectasia. Cell 86 (1996) 159-171. [24] A. Elson et al., Pleiotropic defects in ataxia-telangiectasia protein-deficient mice, Proc. Natl. Acad. Sci. (USA)93(1996) 13084-13089. [25] Y. Xu et al., Targeted dirsruption of ATM leads to growth retardation, chromosomal fragmentation during meiosis, immune defects, and thymic lymphoma. Genes Dev. 10 (1996) 2411-2422. [26] P.R. Borghesani et al., Abnormal development of Purkinje cells and lymphocytes in Atm mutant mice, Proc. Natl. Acad. Sci. (USA) 97 (2000) 3336-3341. [27] R.O. Kuljis et al., Degeneration of neurons, synapses, and neuropil and glial activation in a murine Atm knockout model of ataxia-telangiectasia, Proc. Natl. Acad. Sci. (USA) 94 (1997) 12688-12693. [28] C. Barlow et al., Loss of the ataxia-telangiectasia gene product causes oxidative damage in target organs, Proc. Natl. Acad Sci. (USA) 96 (1999) 9915-9919. [29] F. Magara et al., Genetic background changes the pattern of forebrain commissure defects in transgenic mice underexpressing the ß-amyloid-precursor protein, Proc. Natl. Acad. Sci. (USA) 96 (1999) 4656-61. [30] D. Wahlsten, Deficiency of the corpus callosum: incomplete penetrance and substrain differentiation in BALB/c mice, J. Neurogenet. 5 (1989) 61-76. [31] R. Eilam et al.. Progressive degeneration of dopaminergic nigro-striatal neurons in brains of ATMdeficient mice, Soc Neurosci. Abstr. (1999) 25:377. [32] K.L. Quick et al., Basal levels of free radicals are elevated in basal ganglia and cerebellum, but not cortex, in ataxia-telangiectasia transgenic mice, Soc. Neurosci. Abstr. 25 (1999) 487. [33] A.M. Cataldo et al., Lysosomal hydrolases of different classes are abnormally distributed in brains of patients with Alzheimer disease, Proc. Natl. Acad. Sci. (USA) 88 (1991) 1098-11002. [34] S. Hunot et al., Nuclear translocation of NF-kappaB is increased in dopaminergic neurons of patients with Parkinson Disease, Proc. Natl. Acad. Sci. (USA) 94 (1997) 7531-7536. [35] P. Jensen and K. O'Malley, Potential role of NF-kB transcription factor in the regulation of the tyrosine hydroxylase gene, Soc. Neurosci. Abstr. 24 (1998) 1067. [36] M.J. Edwards and A.M.R. Taylor, Unusual levels of (ADP-ribose)n and DNA synthesis in ataxia telangiectasia cells following g-ray irradiation, Nature 287 (1980) 745-747. [37] L.A. Zwelling et al., Ataxia-telangiectasia cells are not uniformly deficient in poly(ADP-ribose) synthesis following X-irradiation, Mutat. Res. 120 (1983) 69-78. [38] C. Cosi and M. Marien, Decreases in mouse brain NAD+ and ATP induced by l-methyl-4-phenyl1,2,3,6-tetrahydropyridine (MPTP): prevention by the poly(ADP-ribose) polymerase inhibitor, benzamide, Brain Res. 809 (1998) 58-67.

104

H.T.J.

Mount et al. / Parkinsonian Features of A taxia -Telangiectasia

[39] C. Cosi and M. Marien, Implication of poly(ADP-ribose) polymerase (PARP) in neurodegeneralion and brain energy metabolism. Decreases in mouse brain NAD+ and ATP caused by MPTP are prevented by the PARP inhbitor benzamide. Ann N.Y. Acad. Sci. 890 (1999) 227-239. [40] C. Cosi et al., Benzamide, an inhibitor of poly(ADP-ribose) polymerase, attenuates methamphetamineinduced dopamine neurotoxicity in the C57BI/6N mouse, Brain Res. 735 (1996) 343-348. [41] C. Cosi et al., Poly(ADP-ribose) polymerase inhibitors protect against MPTP-induced depletions of striatal dopamine and cortical noradrenaline in C57B1/6 mice. Brain Res. 729 (1996) 264-269. [42] A.S. Mandir et al., Poly(ADP-ribose) polymerase activation mediates 1-methyl-4-phenyl-l,2,3,6tetrahydropyridine (MPTP)-induced parkinsonism, Proc. Natl. Acad. Sci. (USA) 96 (1999) 5774-5779. [43] F. Dantzer et al., Poly(ADP-ribose) polymerase activity is not affected in ataxia telangiectasia cells and knockout mice, Carcinogenesis 20 (1999) 177-180. [44] R.A. Gatti and H.V. Vinters, Cerebellar pathology in ataxia telangiectasia: the significance of basket cells, Kroc Foundation Series 19 (1985) 225-232. [45] H.V. Vinters et al., Sequence of cellular events in cerebellar ontogeny relevant to expression of neuronal abnormalities in ataxia-telangiectasia, Kroc Foundation Series 19 (1985) 233-255. [46] C. Confort et al.. Nerve growth factor enhances DNA synthesis in cultured cerebellar neuroblasts. NeuroReport 2 (1991) 566-568. [47] L. Aloe and E. Vigneti, In vivo and in vitro NGF studies on developing cerebellar cells, NeuroReport 3 (1992)279-282. [48] H.T.J. Mount et al., Purkinje cell survival is differentially regulated by metabotropic and ionotropic excitatory amino acid receptors, 7. Neurosci. 13 (1993) 3173-3179. [49] W.J. Friedman and L.A. Greene, Neurotrophin signaling via Trks and p75, Exper. Cell Res. 253 (1999) 131-142. [50] B.D. Carter et al., Selective activation of NF-k-B by nerve growth factor through the neurotrophin receptor p75. Science 272 (1996) 542-545. [51] H.T.J. Mount et al., Differential involvement of metabotropic and p75 neurotrophin receptors in effects of nerve growth factor and neurotrophin-3 on cultured Purkinje cell survival, J Neurochem. 70 (1998) 1045-1053. [52] S. Rabizadeh et al., Induction of apoptosis by the low-affinity NGF receptor. Science 261 (1993) 345348. [53] A. Ito and K. Horigome, Ceramide prevents neuronal programed cell death induced by nerve growth factor deprivation. J. Neurochem. 65 (1995) 463-466. [54] L.M. Obeid et al., Programmed cell death induced by ceramide. Science 259 (1993) 1769-1771. [55] W.C. Liao et al., Ataxia telangiectasia-mutated gene product inhibits DNA damage-induced apoptosis via ceramide synthase, J. Biol. Chem. 274(1999) 17908-17917. [56] D.T. Dexter et al., Basal lipid peroxidation in substantia nigra is increased in Parkinson's disease. 7. Neurochem. 52 (1989) 381-389. [57] H. Saggu et al., A selective increase in paniculate superoxide dismutase activity in parkmsonian substantia nigra, 7. Neurochem. 53 (1989) 692-697. [58] J. Sian et al., Alterations in glutathione levels in Parkinson's disease and other neurodegenerative disorders affecting basal ganglia, Ann. Neurol. 36 (1994) 348-355. [59] D.T Dexter et al.. Increased nigral iron content in parkinsonian brain, Lancet 2(8569) (1987) 1219-1220. [60] D.T Dexter et al., Increased nigral iron content and alterations in other metal ions occurring in brain in Parkinson's disease, 7. Neurochem. 52(1989) 1830-1836. [61] A. Blochl and C. Sirrenberg, Neurotrophins stimulate the release of dopamine from rat mesencephalic neurons via Trk and p75Lntr receptors, 7. Biol. Chem. 271 (1996) 21100-21107. [62] G. Boka et al., Receptor-mediated activation of tumor necrosis factor in Parkinson's disease? Neurosci. Lett. 172(1994) 151-154. [63] V. France-Lanord et al., Mitochondrial free radical signal in ceramide-dependent apoptosis: a putative mechanism for neuronal death in Parkinson's disease, 7. Neurochem. 69 (1997) 1612-1621. [64] P.P. Michel et at., Neuropharmacologic aspects of apoptosis: significance for neurodegenerative diseases. Clin. Neuropharmacol. 22(1999) 137-150.

Parkinson's Disease E. Ronken and G.J.M. van Scharrenburg (Eds.) IOS Press. 2002

105

Neuroprotection in Parkinson's Disease Through Selective Gene Induction B. Drukarch, J. Flier, F.L. van Muiswinkel Section of Experimental Neurology, Department of Neurology, Research Institute Neurosciences Vrije Universiteit, vd. Boechorststraat 7, 1081 BT Amsterdam, The Netherlands, e-mail: [email protected]

Abstract Parkinson's disease (PD) is a brain disorder for which currently no cure is available. Therapy consists mainly of amelioration of symptoms through suppletion of dopamine (DA) with dopaminomimetics. Pathologically, PD is characterized by degeneration of DA-producing neurons in the substantia nigra (SN). DA autooxidation, a pathway of DA metabolism in the SN that comprises a chain of oxidation reactions in which highly neurotoxic DA-quinones are formed, has been implicated in PD pathogenesis. In fact, the level of DA-quinones has been documented to be increased in the Parkinsonian SN. Hence, stimulation of cellular mechanisms responsible for quinone detoxication in the brain may provide neuroprotection in PD. DA-quinones are efficiently inactivated by the concerted action of so-called phase II biotransformation enzymes, in particular NAD(P)H:quinone oxidoreductase (NQO) and glutathione transferase(s), both of which are present in the human SN. The activity of these protective enzymes can be upregulated by a large variety of structurally unrelated compounds. These compounds, of which the group of sulfur-containing dithiolethiones may be of particular interest, have been shown to act via stimulation of an antioxidant response element (ARE) present in the promotor region not only of phase II biotransformation genes but also in the promoter region of several other genes encoding detoxifying proteins. In this manner, phase II enzyme inducers, such as the dithiolethiones, elicit a highly coordinated, multi-faceted protective response against DA-quinone toxicity, which warrants their full evaluation as neuroprotectants in PD. Keywords: sulfarlem, NQO, phase II, biotransformation, glutathione

1.

Introduction

Parkinson's disease (PD) is a progressive neurodegenerative disorder characterized clinically by the classical motor system trials of tremor, hypokinesia and rigidity, in addition to which signs of autonomic nervous system failure and psychiatric dysfunction may occur. At the cellular level, loss of dopamine(DA)-producing neurons located in a brain area known as the substantia nigra (SN) is the most prominent feature of the disease. As a consequence of this degeneration DA levels are reduced not only in the SN, but also in the corpus striatum to which the nigral DAergic neurons project. It is this striatal DA depletion which directly underlies the motor symptoms of PD. PD affects approximately 0.5 % of the population over the age of 50 years, thereby, considering the fast growing number of the elderly, forming an increasing economic burden for society. Thus far, mere amelioration of PD symptoms with "classical" dopaminomimetics, in the form of the DA precursor L-Dopa and/or DA D2 receptor agonists, has been the mainstay of pharmacotherapy for this disease [1]. However, treatment with these dopaminomimetics does not mitigate progression of the disease process responsible for PD, a factor that is thought to be causatively involved in the declining efficacy and the occurrence of disabling

106

Drukarch el al. / Neuroprotection in Parkinson's Disease Through Selective Gene Induction

side-effects noted upon continued use of such compounds [2,3]- Also, for later additions to the range of pharmacotherapeutics in PD, in particular glutamate receptor antagonists, neurotrophic factors and the newest generation of DA agonists, despite claims based on the outcome of animal experiments, a neuroprotective effect in humans awaits confirmation [3,4]. Thus, to optimally manage PD, there is (still) a strong requirement for drugs capable of retarding or, ideally, halting the ongoing degeneration of DAergic neurons in the Parkinsonian SN. The requirement for such clinically effective neuroprotectants is the more pressing as recent advances in diagnostic procedures, in particular PET/SPECT brain imaging, allow for very early, even subclinical, diagnosis of PD, thereby considerably enlarging the "therapeutic window" for these compounds [5]. Based on current insights into PD pathogenesis to be described below, we here advocate investigation of so-called phase II biotransformation enzyme inducers on their neuroprotective potential in PD. 2.

Quinone Detoxication as a Target for Neuroprotection in Parkinson's Disease

Although environmental factors should not be ruled out, the large body of evidence gathered in the last decade favors endogenous dysfunction at the level of the SN as the primary cause of DAergic degeneration in PD. Thus, apart from disturbances in DA metabolism, the biochemical pathology of PD is characterized by an early reduction in the level of the endogenous antioxidant glutathione (GSH) in the SN, later on accompanied by impairment of mitochondrial respiration, increased free iron content and upregulation of anti-oxidant enzyme activity, in particular superoxide dismutase [6,7]. Despite that the exact mechanisms which lead to these events are still to be elucidated, they are compatible with an ongoing state of so-called oxidative stress in the Parkinsonian SN [8]. This conclusion is underscored by the finding of elevated levels of oxidized lipids, proteins and DNA in the SN of PD patients. In fact, from investigations using the 6-hydroxydopamine and MPTP animal models for PD, strong evidence has been obtained that oxidative stress may be directly responsible for DAergic cell loss in the nigro-striatal system [9]. Considering this presumed causal link between oxidative stress and DAergic neurodegeneration in PD, it is obvious that identification of likely sources of oxidant (over)production in the SN will provide targets for development of effective neuroprotectants for PD. Over the years the issue of oxidative stress in PD has been covered by a large number of excellent reviews. In this context, we recently drew particular attention to the role of highly toxic chemical entities known as DA-derived quinones [10]. These quinones are formed in a chain of reactions commonly known as DA autooxidation, which functions as one of the pathways of DA breakdown in the human brain. Intererestingly, as DA autooxidation is facilitated by the presence of non-protein bound iron and considering the increase in free iron content, there is reason to suspect that the actual rate of DA autooxidation, and thus the production of quinones, is enhanced in the Parkinsonian SN (see below) [11]. Like other quinones of catecholaminergic origin, DA-derived quinones in general are highly reactive, electron-deficient species that may readily bind in a covalent fashion to cellular nucleophiles such as DNA and reduced sulphydryl groups present in protein cysteinyl residues and the thiol anti-oxidant GSH [11]. It is therefore noteworthy, that the level of cysteinyl conjugates of DA, formed as result of the interaction between DA-quinones and reduced sulphydryl groups, has been reported to be increased in the SN of PD patients [12]. Thus, besides toxicity due to inactivation of essential cellular constituents, nucleophilic attack by quinones is hazardous to cellular survival since it ultimately leads to lowering of GSH levels and consequent reduction of cellular antioxidant capacity, as observed in the Parkinsonian brain. Moreover, albeit thus far only under experimental conditions, in its turn the glutathionyl adduct resulting from the direct

Drukarch et al. /Neuroprotection in Parkinson's Disease Through Selective Gene Induction

107

interaction between GSH and DA-quinone has been shown to be prone to oxidation into a neurotoxic benzothiazine derivative [13]. More important, however, for the mechanism underlying the, oxidative stressinduced, progressive neurodegeneration observed in PD may be the propensity of DAderived quinones for so-called redox-cycling [11]. This redox-cycling occurs when, at the expense of cellular reducing equivalents in the form of NADPH, DA-derived quinones undergo a one-electron reduction resulting in the formation of semi-quinones. These semiquinones are highly reactive and unstable DA intermediates that not only bind to cellular nucleophiles but also readily re-oxidize into quinones with the concomitant production of toxic superoxide free radicals and hydrogen peroxide. In this way, NADPH-driven reduction of DA-derived quinones elicits a cascade of reversible oxidation and reduction reactions, a redox-cycle, that is accompanied by excessive release of oxidants and depletion of NADPH, thereby initiating a variety of detrimental events culminating in DAergic cell death [10]. These include not only the aforementioned lipid and protein oxidation, oxidative damage to nucleic acid, and deprivation of cellular energy production, but also aberrant redox-sensitive gene transcription (e.g. via NF-kappaB) and apoptosis, phenomena which have all been observed to occur in the PD brain [7], Taken together, there is sufficient evidence to believe that quinone toxicity plays a major role in PD pathology and as such provides an excellent focus for neuroprotective strategies. 3.

Quinone Detoxication in the Brain

As outlined above, it is evident that neuroprotective strategies for PD should try and address the issue of DA-derived quinone toxicity. More in particular, to be optimally effective, neuroprotective agents should be capable of preventing both electrophilic attack on cellular macromolecules as well as redox-cyling of DA-derived quinones. As described recently by Rescigno and coworkers, two-electron reduction of quinones, leading to formation of their corresponding diphenols or substituted diphenols, provides the first and crucial step in the scavenging of these highly reactive species [14]. This line of defense may involve a direct, non-enzymatic interaction of DA-quinones with low molecular weight reductants, in the first place GSH. It should be noted however, that GSH synthesis is strongly dependent on the supply of precursor amino acids, the rate of which may not be sufficient to keep up with the demand of quinone detoxication [15]. From the viewpoint of protective options, it is therefore plausible to assume that quinone reactivity can be controlled in a more efficient manner by specific enzymes dedicated to the task of quinone inactivation. In this context, the action of at least two enzymes, both belonging to the group of phase II biotransformation enzymes, is of note. In general, phase II biotransformation enzymes have been shown to protect against xenobiotics and endogenous toxins by catalyzing the transformation of reactive electrophiles such as DA-quinones into more stable, hydrophilic conjugates that are prone to be excreted from the cell [16,17]. Originally, the enzymatic detoxication of DA-derived quinones was thought to rely on the phase II enzyme NAD(P)H:quinone oxidoreductase (NQO; EC 1.6.99.2) [11,14]. NQO, also referred to as DT-diaphorase, is a flavoenzyme that, in contrast to ubiquitous quinone reductases such as NADPH-cytochrome P450 reductase and NO-synthase, catalyzes an unique, direct two-electron reduction of (DA) quinones yielding chemical entities known as hydroquinones, thereby averting formation of the extremely toxic, free semi-quinones [18,19]. Subsequently, further detoxication of hydroquinones is ensured by other phase II enzymes such as UDP glucuronosyltransferases or sulfotransferases [20]. Recently, apart from NQO, Mu class GSH transferases (EC 2.5.1.18) have been identified as an additional protective mechanism against the toxicity of DA-quinones. More specifically, the Mu 2-2 subtype of GSH transferase was found to catalyze the reductive conjugation of GSH to the DA autooxidation product cyclic DA-quinone. In this reaction, 4-S-glutathionyl-5,6-

108

Drukarch et al. / Neuroprotection in Parkinson's Disease Through Selective Gene Induction

dihydroxyindole is formed, a hydrophilic entity that is resistant to redox-cycling and easily amenable to further detoxication either by other phase II enzymes or via direct cellular excretion [21,22]. Taken together, it is safe to conclude that the combined action of phase II biotransformation enzymes, most notably NQO and GSH transferase, has the potential to provide a powerful cellular defense mechanism against the oxidative stress inherent to the process of DA autooxidation. From the viewpoint of therapeutic targets for neuroprotection in PD it is noteworthy therefore, that recently we were able to demonstrate the presence of NQO immunoreactivity in both DAergic neurons and glial cells in the human SN [10], thereby extending earlier data obtained by others using rat brain sections [23]. Moreover, abundant expression of Mu class GSH transferases. including the Mu 2-2 form, also has been reported in the human SN [21 ]. 4.

Induction of Quinone Reductase Activity as a Means of Neuroprotection in Parkinson's Disease

Given their presence in relevant brain areas, the pivotal role of phase II biotransformation enzymes in the inactivation of DA-derived cytotoxic quinones makes these antioxidant proteins attractive targets for development of innovative neuroprotective agents to treat PD. For this particular purpose, drugs capable of simultaneously stimulating the activity of NQO and GSH transferase(s), i.e. enzymes with a direct quinone detoxifying function, would be expected to be of special therapeutic interest [10]. In this context it is of note, that besides their capability to neutralize quinones, NQO and GSH transferase(s) share their inducability not only by these toxicants, but also by a large array of other, structurally diverse chemicals, including phenolic antioxidants (e.g. butylated hydroxyanisole), aromatic isothiocyanates (e.g. sulforaphane) and michael reaction acceptors (e.g. dimethylfumarate) [10,24,25]. In fact, pretreatment of neuronal cells with such phase II enzyme inducers has been demonstrated to markedly reduce the toxicity of glutamate. hydrogen peroxide as well as DA via a coordinated up-regulation of the expression of phase II enzymes, in particular NQO and GSH transferase [26,27]. These findings thus strongly support the idea that the joint efforts of this army of defense enzymes can be protective against neuronal cell death such as observed in PD. The chemistry of phase II enzyme inducers has been comprehensively reviewed. The astonishing variety of structures capable of this action makes it unlikely that one single, specific, complementary receptor is involved in the inducer signal transduction pathway. Rather, the unifying feature among these inducers appears to be their electrophilicity and, hence, reactivity with sulphydryl groups either by nucleophilic substitution or in redox reactions [28]. Moreover, inducer activity has been suggested to be dependent on activation of a so-called antioxidant response element (ARE) present in the promotor region of several phase II enzymes. Nucleotide sequence analysis has revealed the ARE to consist in general of AP-l/AP-1 like binding elements arranged as inverse or direct repeats separated by either three or eight nucleotides, followed by a GC box [29]. Interestingly, although characterization of the nature of the cytosolic signal transduction cascade is far from complete, ARE's of various phase II genes have been found to bind to a complex of nuclear proteins. Indeed, analysis of ARE-nuclear protein complexes has identified a number of nuclear transcription factors, including c-Jun, Jun-B, Jun-d, c-Fos, Fral, Nrfl. Nrf2, Ah (aryl hydrocarbon) receptor, and the estrogen receptor, capable of binding to the ARE's of phase II genes [30]. Amongst these transcription factors, Nrfl and Nrf2 are known to positively regulate the ARE-mediated expression and induction of phase II genes, in particular that of NQO [31]. In addition to genes of phase II biotransformation enzymes, genes encoding several other proteins that participate in cellular defense against oxidative stress also contain ARE-like sequences. Of these, in the context of PD pathogenesis and neuroprotective treatment, the genes for gamma-glutamylcysteine

Drukarch et al. /Neuroprotection in Parkinson's Disease Through Selective Gene Induction

109

synthetase, i.e. the rate-limiting enzyme in GSH synthesis, and the ferritin L gene coding for the light chain of ferritin, i.e. the major intracellular iron-binding protein, merit special attention [32]. Taken together, the conservation of the ARE in the genes of many detoxifying enzymes and other protective proteins indicates that the expression of these genes may be coordinately regulated by a single mechanism involving the ARE. In this manner, the common transcriptional regulation of such genes provides cells not only with a versatile, yet highly orchestrated capability to withstand oxidative stress, but also with an attractive neuroprotective drug target.

5.

Gene Induction and Neuroprotection in Parkinson's Disease: Focus on Dithiolethiones

In the context of phase II gene induction as described in the previous section, a particularly interesting class of compounds is the dithiolethiones, a cyclic, sulfur-containing group of agents, originally described as constituents of cruciferous vegetables [33]. The dithiolethiones, of which oltipraz and anethole dithiolethione are currently available for use in humans, not only increase the activity of phase II biotransformation enzymes in various cellular preparations in vitro, but are active also in vivo in both animals and humans with only minor side-effects reported [10]. Concomitant with their stimulatory effect on phase II biotransformation, dithiolethiones are especially attractive as potential neuroprotectants in PD in that they also boost general cellular antioxidant capacity by acting as regular oxidant scavengers [34,35], by upregulating the activity of oxidant scavenging enzymes such as superoxide dismutase [36], by modulating redox-sensitive gene transcription (e.g. inhibition of peroxide-mediated activation of NF-kappaB) [37], by increasing the expression of metalbinding proteins such as ferritin [38], and by stimulation of those enzymes responsible for the maintenance of GSH pools, in particular gamma-glutamylcysteine synthetase, glutathione disulfide-reductase, and glucose-6-phosphate dehydrogenase [33,35,39,40]. Indeed, such dual characteristics, i.e. induction of specific quinone reductase activity together with a more broad antioxidant profile, have been shown recently to be instrumental to the capacity of phase II enzyme inducers to protect against DA neurotoxicity, at least in vitro [26]. In contrast to other organ systems, thus far only very limited data are available on the action of dithiolethiones in nervous tissue. Recently, we were able to show that treatment with anethole dithiolethione (trade name Sulfarlem®) increases GSH levels in cultured astrocytes, a cell type known to mediate neuroprotective responses in the brain [39,40]. Moreover, in this brain cell preparation anethole dithiolethione provided marked protection against hydrogen peroxide toxicity under low intracellular GSH levels, such as observed in the Parkinsonian SN [40]. Here, we extend these data by demonstrating that anethole dithiolethione is capable of protecting astroglial cells also against glutamate toxicity, both upon simultaneous treatment as well as following pretreatment with anethole dithiolethione (Fig. 1). Intriguingly, while maintaining astroglial viability under these experimental conditions, anethole dithiolethione was unable to prevent glutamate-induced GSH depletion (Fig. 1), thus suggesting a GSH-independent protective mechanism of action requiring future elucidation. Together, these results obtained by us and others provide sufficient support to conclude that dithiolethiones, and in particular anethole dithiolethione, are multifaceted protective compounds which deserve full evaluation of their neuroprotective potential in PD.

110

Drukarch et al. / Neuroprotection in Parkinson's Disease Through Selective Gene Induction

Fig. 1. (A) Rat C6 astroglioma cells were treated for 24 hr with sodium glutamate (Na-Glu, 20mM) or Sulfarlem (10mM) alone or the combination thereof. Lactate dehydrogenase (LDH) release, as a measure of cellular viability, and intracellular glutathione (GSH) content were measured. Data are depicted as means ± SD (n=3). Statistical analysis was performed using ANOVA followed by Newman-Keuls post-hoc test. (*p<0.0l, **p<0.001 in comparison with control #p<0.01, ##p<0.001 in comparison with Na-Glu treatment alone). (B) Rat C6 astroglioma cells were incubated with Sulfarlem (10mM) for 24 hr. Subsequently, culture medium was changed and the cells were incubated with Na-Glu (20mM) for an additional 24 hr. Thereafter. LDH release and GSH content were measured. Data are depicted as means ± SD (n=3). Statistical analysis was performed using ANOVA followed by Newman-Keuls post-hoc test. (**p<0.001 in comparison with control ##p<0.001 in comparison with Na-Glu treatment alone).

References [1] [2] [3] [41 [5] [6] [7] [8] [9] [10] [11] [12] [13]

[14] 115] [16]

J.J. Hagan et al., Parkinson's Disease: Prospects for Improved Drug Therapy. Trends Pharmacoi Sci. 18(1997) 156-163. Y, Agid, Levodopa: Is Toxicity a Myth?, Neurology 50 (1998) 858-863. S.B. Dunnett and A. Bjorklund, Prospects for New Restorative and Neuroprotective Treatments in Parkinson's Disease, Nature 399(Suppl) (1999) A32-A39. C.W. Olanow et al., Dopamine Agonists and Neuroprotection in Parkinson's Disease. Ann. Neurol. 44(Suppl) (1998) S167-S174. i.C. Stoof et al., Leads for the Development of Neuroprotective Treatment in Parkinson's Disease and Brain Imaging Methods for Estimating Treatment Efficacy, Eur. J. Pharmacol. 375 (1999) 75-86. J. Sian et al., Parkinson's disease: A Major Hypokinetic Basal Ganglia Disorder. J. Neural Transm 106(1999)443-476. P. Jenner and C.W. Olanow. Understanding Cell Death in Parkinson's Disease. Ann. Neurol. 44(Suppl)(1998)S72-S84. S. Fahn and G. Cohen, The Oxidant Stress Hypothesis in Parkinson's Disease: Evidence Supporting It. Ann. Neurol. 32 (1992) 804-812. E. Grunblatt et al., MPTP and 6-Hydroxydopamine-Induced Neurodegeneration as Models for Parkinson's Disease: Neuroprotective Strategies, J. Neurol. 247(Suppl 2) (2000) II/95-II/102. B. Drukarch and F.L. van Muiswinkel, Drug Treatment of Parkinson's Disease: Time for Phase II. Biochem. Pharmacol. 59(2000) 1032-1031. A. Bindoli et al., Biochemical and Toxicological Properties of the Oxidation Products of Catecholamines. Free Radic. Biol. Med. 13 (1992) 391-405. P.E. Spencer et al., Conjugates of Catecholamines with Cysteine and GSH in Parkinson's Disease: Possible Mechanisms of Formation Involving Reactive Oxygen Species. J. Neurochem. 71 (1998) 2112-2122. H. Li et al., Brain Mitochondria Catalyze the Oxidation of 7-(2-Aminoethyl)-3,4-Dihydro-5Hydroxy-2H-1.4-Benzothiazine-3-Carboxylic Acid (DHBT-1) to Intermediates that Irreversibly Inhibit Complex I and Scavenge Glutathione: Potential Relevance to the Pathogenesis of Parkinson's Disease. J. Neurochem. 71 (1998) 2049-2062. A. Rescigno et al., Some Aspects of Tyrosine Secondary Metabolism. Biochem. Pharmacol. 56 (1998) 1089-1096. A. Meister and M.E. Anderson. Glutathione. Annu. Rev. Biochem 52 (1983) 711-760. U.A. Meyer. Overview of Enzymes of Drug metabolism. J. Pharmacokinet. Biopharm. 24 (1996) 449-459.

Drukarch et al. /Neuroprotection in Parkinson's Disease Through Selective Gene Induction

[17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32]

[33] [34] [35] [36] [37] [38] [39] [40]

111

J. Wilkinson IV and M.L. Clapper, Detoxication Enzymes and Chemoprevention, Proc. Soc. Exp. Biol. Med. 216(1997) 192-200. C. Lind et al., DT-Diaphorase as a Quinone Reductase: A Cellular Control Device Against Semiquinone and Superoxide Radical Formation, Arch. Biochem. Biophys. 216 (1982) 178-185. J. Segura-Aguilar and C. Lind, On the Mechanism of the Mn3+-Induced Neurotoxicity of Dopamine: Prevention of Quinone-Derived Oxygen Toxicity by DT Diaphorase and Superoxide Dismutase, Chem. Biol. Interact. 72 (1989) 309-324. E. Cadenas et al., Pro- and Antioxidant Functions of Quinones and Quinone Reductases in Mammalian Cells, Adv. Enzymol. 65 (1992) 97-146. S. Baez et al., Glutathione Transferases Catalyze the Detoxication of Oxidized Metabolites oQuinones) of Catecholamines and May Serve as an Antioxidant System Preventing Degenerative Cellular Processes, Biochem. J. 324 (1997) 25-28. J. Segura-Aguilar et al., Human Class Mu Glutathione Transferases, in Particular Isoenzyme M2-2, Catalyze Detoxication of the Dopamine Metabolite Aminochrome, J. Biol. Chem. 272 (1997) 57275731. M. Schultzberg et al., Distribution of DT diaphorase in the Rat Brain: Biochemical and Immunohistochemical Studies, Neuroscience 27 (1988) 763-776. F.L. van Muiswinkel et al., L-Dopa Stimulates Expression of the Antioxidant Enzyme NAD(P)H:Quinone Oxidoreductase (NQO) in Cultured Astroglial Cells, Free Radic. Biol. Med. 29 (2000) 442-453. T. Primiano et al., Antioxidant-Inducible Genes, Adv. Pharmacol. 38 (1997) 293-328. S. Duffy et al., Activation of Endogenous Antioxidant Defenses in Neuronal Cells Prevents Free Radical-Mediated Damage, J. Neurochem. 71 (1998) 69-77. T.H. Murphy et al., Enhanced NAD(P)H:Quinone Reductase Activity Prevents Glutamate Toxicity Produced by Oxidative Stress, J. Neurochem. 56 (1991) 990-995. P. Talalay et al., Identification of a Common Chemical Signal Regulating the Induction of Enzymes that Protect Against Chemical Carcinogenesis, Proc. Natl. Acad. Sci. USA 85 (1988) 8261-8265. A.K. Jaiswal, Antioxidant Response Element, Biochem. Pharmacol. 48 (1994) 439-444. A.K. Jaiswal, Regulation of Genes Encoding NAD(P)H:Quinone Oxidoreductases, Free Radic. Biol. Med. 29 (2000) 254-262. R. Venugopal and A.K. Jaiswal, Nrfl and Nrf2 Positively and c-Fos and Fral Negatively Regulate the Human Antioxidant Response Element-Mediated Expression of NAD(P)H:Quinone Oxidoreductase 1 Gene, Proc. Natl. Acad. Sci. USA 39 (1996) 14960-14965. A.T. Dinkova-Kostova and P. Talalay, Persuasive Evidence that Quinone Reductase Type 1 (DT Diaphorase) Protects Cells Against the Toxicity of Electrophiles and Reactive Forms of Oxygen, Free Radic. Biol. Med. 29 (2000) 231-240. S.S. Ansher et al., Biochemical Effects of Dithiolthiones, Food Chem. Toxicol. 24 (1986) 405-415. M-O. Christen, Anethole Dithiolethione: Biochemical Considerations, Methods Enzymol. 252 (1995) 316-323. S. Khanna et al., Protective Effects of Anethole Dithiolethione Against Oxidative Stress-Induced Cytotoxicity in Human Jurkat T Cells, Biochem. Pharmacol. 56 (1998) 61-69. E. Antras-Ferry et al., Oltipraz Stimulates the Transcription of the Manganese Superoxide Dismutase Gene in Rat Hepatocytes, Carcinogenesis 18 (1997) 2113-2117. C.K. Sen et al., Inhibition of NF-kappaB Activation in Human T-Cell Lines by Anetholdithiolthione, Biochem. Biophys. Res. Commun. 218 (1996) 148-153. T. Primiano et al., Induction of Hepatic Heme Oxygenase-1 and Ferritin in Rats by Cancer Chemopreventive Dithiolethiones, Carcinogenesis 17 (1996) 2291-2296. R. Dringen et al., Anethole Dithiolethione, a Putative Neuroprotectant, Increases Intracellular and Extracellular Glutathione Levels During Starvation of Cultured Astroglial Cells, NaunynSchmiedeberg's Arch. Pharmacol. 358 (1998) 616-622. B. Drukarch et al., Anethole Dithiolethione Prevents Oxidative Damage in Glutathione-Depleted Astrocytes, Eur. J. Pharmacol. 329 (1997) 259-262.

This page intentionally left blank

Part III Towards Neuroprotective Efficacy

This page intentionally left blank

Parkinson's Disease E. Ronken and G.J.M. van Scharrenburg (Eds.) IOS Press, 2002

115

The RET-Dependent Neuroprotective Effects of GDNF are Mediated by Activation of Proteinkinase B (PKB)/AKT Jiirgen Schlegel1, Frauke Neff1 and KarlaEggert2 Neuropathology, Institute of Pathology, Munich Technical University, Germany 2 Department of Neurology, University of Marburg, Germany Abstract. Glial cell line derived neurotrophic factor (GDNF) is a potent and specific neuroprotective factor which protected dopaminergic neurons from lesion-induced degeneration in several in vivo and in vitro models of Parkinson's disease. Here we show, that GDNF activates the endogenous phosphatidylinositol-3' kinase (PI3K)/proteinkinase B/Akt (PKB/Akt) signalling pathway and the mitogen activated protein kinase (MAPK) cascade in human neuroblastoma cell line SHSY5Y. In comparison to EGF treatment which delivers a strong MAPK signal, GDNF showed a predominant and prolonged activation of PI3K/Akt. Consistent with this finding we also found a PI3K-dependent binding of phosphorylated Bad protein to 14-3-3 proteins. Our data support the hypothesis that GDNF exerts its neuroprotective effects by activation of the anti-apoptotic PI3K/Akt signalling pathway and subsequent inactivation of Bad.

1. Introduction Glial cell line derived neurotrophic factor (GDNF) has attracted much attention as a potential therapeutic agent for the treatment of Parkinson's disease [1]. It has been shown previously that GDNF exerts its effects by functional Ret receptor complexes [2, 3]. The RET gene was originally cloned as an oncogene in human papillary thyroid carcinomas. It has then been shown that activating mutations in the RET gene are also the cause for two familial cancer syndromes, multiple endocrine neoplasia (MEN2) and familial medullary thyroid carcinomas (FMTC). In contrast, inactivating mutations of the RET gene are the cause of Hirschsprung's disease [4]. The RET encoded protein belongs to the TGFB superfamily of transmembrane receptors. GDNF signals involving the Ret receptor are transduced by binding of GDNF or one of the other GDNF ligands including neurturin (NTN), artemin (ART) and persephin (PSP) to a glycosyl phosphatidylinositol anchored ligand binding protein (GDNF receptor, GFR1 through GFR4). The dimeric GDNF/GFR1 complex then binds to Ret and the signal is transduced by phosphorylation of the intracellular tyrosine kinase domain of Ret. The intracellular signal transduction of GDNF is mediated by different pathways including the mitogen activated protein kinase (MAPK) cascade, the phosphatidylinositol-3-kinase (PI3K)/Akt and the stress activated protein kinase (SAPK)/Jun N-terminal kinase (JNK) pathway [5]. The PI3K/Akt pathway is thought to be a key regulator of neurotrophin-dependent neuronal survival since Akt (also called proteinkinase B, PKB, or related to A and C type kinases, RAC, hereafter referred to as PKB/Akt) mediates its anti-apoptotic effect by regulating several downstream targets including the bcl-2 family factor Bad, the forkhead transcription factor FKHR-1 and NFkB [6].

116

J. Schlegel et al. / The RET-Dependem Neuroprotective Effects of GDNF

2. Material and Methods Human neuroblastoma cells SH-SY5Y were cultivated in DMEM supplemented with 15% FCS, 2 mM L-Glutamine, penicillin (100 U/ml) and streptomycin (100 mg/ml) under standard cell culture conditions at 37°C and 5%CO2. Cells were grown without serum for 24 hrs prior to experiments. Subsequently, rhGDNF (ICN, Cosa Mesa, USA) was added to serum-free medium to a final concentration of 10 ng/ml, or, as control, rhEGF (10ng/ml) for 15min. and 5min, respectively. Wortmannin (100 nM) or PD98059 (50mM) were used as specific inhibitors of the PI3K or MAPK signalling pathways. Protein-extraction was performed using standard methods. For immuneprecipitation, equal amounts of protein were dissolved in lysis Buffer, 3mg of PY20-anti-pTyr- antibody (Santa Cruz) or anti-14-3-3 antibody (Upstate biotechnology) was added to each sample and incubated at 4°C overnight. Subsequently, 60ml Protein A-sepharose (50% beads) was added and incubated at 4°C for 2h. Beads were centrifuged at 13.000rpm for 60sec. and washed twice with lysis buffer. The pellet was resuspended in l00ml Sample Buffer (62,5mM Tris-Cl pH 6,8, 2% SDS, 10% Glycerol, 50mM DTT, 0,1% Bromphenol blue) and boiled for 10 minutes. The samples were centifuged for 60 sec. at 13.000 rpm and 20ml of each sample were loaded to SDS-gels for Western analysis. Equal amounts of protein were separated by sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to Immobilon membranes (Millipore. Bedford, MA). Unspecific binding sites were subsequently blocked by incubation with 5% (w/v) nonfat dry milk in TTBS (0.1% Triton X-100, 20mM Tris, 136mM NaCl at pH 7.5) for 1 hr. Membranes were incubated with rabbit polyclonal primary antibodies (antiphospho-Akt, anti-Akt, anti-phospho-Erkl/2, anti-Erkl/2, anti-phospho-Bad, and anti-Bad from New England Biolabs, anti-GFRl, anti-Ret, and anti-EGF receptor from Santa Cruz) diluted 1:1000 in TTBS for 1 2 - 1 4 hrs at 4°C. HRP-conjugated anti-rabbit immunoglobulins (diluted 1:2000 in 5% nonfat dry milk/TTBS) served as secondary antibodies, with which incubation was performed for 1 hr at room temperature, before a 60 incubation sec with ECL reagent (Amersham Pharmacia). Immunoreactivity was visualized by exposure to X-OMAT MA-films (Eastman Kodak). 3. Results Here we investigated the GDNF mediated signal transduction in human SH-SY5Y neuroblastoma cells. By Western analysis, SH-SY5Y cells showed protein expression of GFR1 and Ret (Fig. 1). They also expressed the EGF receptor, PI3K, PKB/Akt, MAPK and Bad. After treatment with rhGDNF or rhEGF neuroblastoma cells exhibited rapid phosphorylation of PI3K detected by immune precipitation and subsequent Western analysis using specific anti-p85 and anti-phospho-tyrosine antibodies. Using anti-phosphoAkt and anti-phospho-ERKl/2 antibodies a rapid induction of PKB/Akt- and MAPKphosphorylation could be detected. The activation of PKB/Akt was comparable in GDNFand EGF-treated cells, in contrast, the MAPK phosphorylation was much stronger in response to EGF than to GDNF (Fig. 1). There were no differences in total PKB/Akt or MAPK protein levels. The phosphorylation of PKB/Akt but not MAPK-phosphorylation could be completely blocked by Wortmannin treatment. Activation of MAPK but not Akt could be decreased by PD98059 (Fig. 1). Using a combined immune precipitation and Western analysis approach Bad protein bound to 14-3-3 protein was detected in SH-SY5Y cells indicating phosphorylation and inactivation of Bad through 14-3-3 binding. There was an increase of 14-3-3 bound Bad in

J. Schlegel et al. / The RET-Dependent Neuroprotective Effects of GDNF

117

response to GDNF treatment (Fig. 1). This increase could be completely blocked by Wortmannin treatment but not with PD98059. The amount of 14-3-3 bound Bad also increased after EGF treatment.

Figure 1. Western Analysis of protein extracts from SH-SY5Y cells using antibodies directed against GFR1, Ret, phospho-Erkl/2, phospho-Akt, and Bad after immunoprecipitation using an anti-14-3-3 antibody. Proteins were extracted after 24 hrs in serum free medium, and after treatment with EGF or GDNF either alone or with specific inhibitors of PI3K or MAPK. Western analysis showed an induction of phospho-Erkl/2 after stimulation with EGF and a weaker phosphorylation after treatment with GDNF, which could be blocked by PD98059 as an inhibitor of MAPK. The phosphorylation of PKB/Akt after treatment with EGF or with GDNF was at comparable levels and could be blocked by Wortmannin as an inhibitor of PI3K. Consequently, Western analysis of 14-3-3 bound Bad showed an increase after stimulation with EGF and GDNF which could be blocked by Wortmannin. There were no differences of the expression levels of GFR 1 or Ret with or without treatment.

4. Discussion Glial cell line derived neurotrophic factor (GDNF) is a potent and specific neuroprotective factor which protected dopaminergic neurons from lesion-induced degeneration in several in vivo and in vitro models of Parkinson's Disease [7-10]. The mechanism by which GDNF exerts its neuroprotective function, however, is poorly understood. It has been shown previously, that the neurotrophic effects of GDNF are mediated by activation of cellular signalling pathways via functional Ret receptor complexes [11, 12]. Using cell lines or primary neurons specific effects could be attributed to individual pathway. Whereas MAPK

118

J. Schlegel et al. / The RET-Dependent Neuroprotective Effects of GDNF

signaling appears to be necessary for neurite outgrowth [13], the activation of the PI3K/Akt pathway seems to be required for neurotrophin-dependent cellular survival [14-16]. Most studies, however, focussed on only one signal transduction pathway. Recent data indicate, however, that complex interactions between multiple activated pathways are responsible for specific responses of neuronal cells to an individual factor [17]. This is particularly true for neurorestorative or neuroprotective responses where both the cellular reactions to the lesion and the effects of the protective factor accumulate to a specific response. In accordance with published data we detected simultaneous activation of the MAPK and the PI3K/Akt pathway after GDNF treatment [5]. However, the activation of MAPK by EGF was much more pronounced than after GDNF treatment whereas PKB/Akt phosphorylation by EGF and GDNF was at comparable levels. The predominant activation of PKB/Akt by GDNF appears to be responsible for the neurotrophic effect and in the context of neurodegeneration for survival signalling. The PKB/Akt signalling pathway has attracted much attention because PKB/Akt delivers an anti-apoptotic signal. Consequently, PI3K dependent activation of PKB/Akt is also a major candidate for signalling the neuroprotective response to GDNF. Several factors involved in apoptosis and cellular survival have been shown to be regulated by PKB/Akt [18]. The Bcl-2 family member Bad was first identified because of its ability to bind Bcl-2 thereby blocking the anti-apoptotic effect of Bcl-2. Phosphorylation of Bad by PKB/Akt results in binding of phospho-Bad to 14-3-3 proteins leading to its inactivation [19]. Here we show that Bad is phosphorylated and inactivated in response to GDNF stimulation. We found a Wortmannin-sensitive binding of phosphorylated Bad to 14-3-3 protein after stimulation of SH-SY5Y cells with GDNF. In conclusion, our data indicate that the neuroprotective effect of GDNF in human neuroblastoma cells seems be mediated by a predominant PI3K-dependent phosphorylation of PKB/Akt and a subsequent inactivation of Bad by phosphorylation and binding to 14-3-3 proteins.

Acknowledgements Supported by a grant of the Deutsche Forschungsgemeinschaft (Forschergruppe "Neuroprotektion: Wachstumsfaktoren und Sauerstoffradikalfanger", Projekt E) References [1] [2] [31 [4] [5] [6] [7] [8] [9]

D.M. Gash et al., Neuroprotective and neurorestorative properties of GDNF. Ann Neurol 44 (1998) S121-125. J.J. Treanor et al.,. Characterization of a multicomponent receptor for GDNF. Nature 382 (1996) 80-83. M. Trupp et al., Functional receptor for GDNF encoded by the c-ret proto-oncogene. Nature 381 (1996)785-788. B. Pasini et al., RET mutations in human disease. Trends Genet 12 (1996) 138-144. M.S. Airaksinen et al.,. GDNF family neurotrophic factor signaling: four masters, one servant? Mol Celt Neurosci 13 (1999) 313-325. E.S. Kandel, and N. Hay, The regulation and activities of the multifunctional Serine/Threonine kinase Akt/PKB. Exp Cell Rex 253 (1999) 210-29. K. Eggert et al., Glial cell line-derived neurotrophic factor protects dopaminergic neurons from 6hydroxydopamine toxicity in vitro. Neurosci Lett 269 (1999) 178-182. A.C.Granholm et al., Glial cell line-derived neurotrophic factor improves survival of ventral mesencephalic grafts to the 6-hydroxydopamine lesioned striatum. Exp Brain Res 116 (1997) 29-38. R.J Mandel et al., Midhrain injection of recombinant adeno-associated virus encoding rat glial cell line-derived neurotrophic factor protects nigral neurons in a progressive 6-hydroxydopamine-induced degeneration model of Parkinson's disease in rats. Proc Natl Acad Sci U S A 94 (1997) 14083-14088

J. Schlegel et al. / The RET-Dependent Neuroprotective Effects of GDNF

[10] [11] [12] [13] [14] [15] [16]

[17] [18] [19]

119

A. Tomac et al, Protection and repair of the nigrostriatal dopaminergic system by GDNF in vivo. Nature 373 (1995) 335-339. K. Robertson, and I. Mason, The GDNF-RET signalling partnership. Trends Genet 13 (1997) 1-3. M. Trupp et al., Ret-dependent and -independent mechanisms of glial cell line-derived neurotrophic factor signaling in neuronal cells. J Biol Chem 274 (1999)., 20885-20894. D.H. van Weering, J.L. and Bos, Glial cell line-derived neurotrophic factor induces Ret-mediated lamellipodia formation. J Biol Chem 272 (1997) 249-254. H. Dudek et al., Regulation of neuronal survival by the serine-threonine protein kinase Akt. Science 275(1997)661-665. E.M. Eves et al., Akt, a target of phosphatidylinositol 3-kinase, inhibits apoptosis in a differentiating neuronal cell line. Mol Cell Biol 18 (1998) 2143-2152. R.M. Soler et al., Receptors of the glial cell line-derived neurotrophic factor family of neurotrophic factors signal cell survival through the phosphatidylinositol 3-kinase pathway in spinal cord motoneurons. J Neurosci 19 (1999) 9160-9169. T. Pawson, and T.M.Saxton, Signaling networks—do all roads lead to the same genes?. Cell 97 (1999) 675-678. S.G. Kennedy et al., The PI 3-kinase/Akt signaling pathway delivers an anti-apoptotic signal. Genes Dev 11(1997) 701-713. J. Downward, How BAD phosphorylation is good for survival. Nat Cell Biol 1 (1999)., E33-35.

120

Parkinson's Disease E. Ronken and G.J.M. van Scharrenburg (Eds.) IOS Press. 2002

An Immortalized Neuronal Cell Line Derived from the Substantia Nigra of an Adult Rat: Application to Cell Transplant Therapy. Christian Arriagada1, Julio Salaza1, Takeshi Shimahara3, Raul Caviedes1 and Pablo Caviedes1,2* Program of Molecular & Clinical Pharmacology, JCBM Faculty of Medicine, University of Chile, Santiago, Chile; 2Sansum Medical Research Institute, Santa Barbara, CA, USA; 3 NBCMCNRS. Gif-sur-Yvette, France Abstract Immortalized cell lines offer great advantages over other in vitro preparations. Indeed, cell lines can provide large amounts of genetically homogeneous tissue, and are readily accessible and easy to manipulate. However, cell lines must deal with problems such as lack of stability, which may result in loss of differentiated properties and death. In our lab, we have successfully produced immortalized cells lines of various tissues from adult mammalian donors (rat, mouse, bovine, human), using an original transformation protocol. Briefly, primary cultures are kept in the presence of media conditioned by the UCHT1 rat thyroid tumoral cell line, which induces transformation after variable periods of time. In this presentation, we discuss the characteristics of the neuronal cell RCSN, derived from the substantia nigra of an adult rat. By immunohystochemistry, the cell line shows positive neuronai markers (NSE, synaptophysin, MAP-2, neurofilament) and lack glial traits (GFAP, SI00). RCSN cells possess catecholaminergic traits: presence of TH, melanin and DAT. This cell line also line exhibits intracellular Ca2+ increments in response to excitatory neurotransmitter agonists. Finally, using stereotaxic surgery, we implanted suspensions of this cell line (500.000 cells in 4 mL) in the striatum of rats previously lesioned in the nigrostriatal pathway with 6 OH dopamine. The implanted rats show a steady decrease in rotations, leveling off at 75% of the initial rotation values after 16 weeks post implant. Sham operated rats, where only the vehicle was implanted, show no improvement. The results suggest that the RCSN cell line appears as a model for the study of cell mechanisms related to Parkinson's like neurodegeneration. The effect of the implanted RCSN cell line in the improvement of the rotational behavior in 6 OH Dopamine lesioned rats is encouraging in our quest to establish a human substantia nigra cell line with the UCHT1 protocol, and generate a model that could be applied towards cell transplant therapy in humans. Keywords: dopaminergic cell line, xenotransplantation, xenografting

* Corresponding author: Pablo Caviedes. Program of Molecular & Clinical Pharmacology, ICBM, Faculty of Medicine, University of Chile. Casilla 7000, Correo 7, Santiago, Chile. Phone: (562) 678-6075 Fax: (562) 7372783 E-mail: [email protected]. Funded by Fondecyt grants #1980906. 7980058 & 1990622 (Chile), and INSERM(France)/Conicyt(Chile) Exchange Program

C. A rriagada et al. / Immortalized Neuronal Cell Line

121

Parkinson's disease is a neurodegenerative disorder affecting an estimated one million patients in the United States alone [1], and it is determined by the degeneration of the dopaminergic neurons of the substantia nigra pars compacta (SN) which project to the striatum. The cardinal features of Parkinson's disease are: tremor, mainly at rest; muscular rigidity, which leads to difficulties in walking, writing, speaking and masking of facial expression; bradykinesia, a slowness in initiating and executing movements; and stooped posture and instability [2]. In later stages of the disease, patients undergo cognitive deterioration. The degeneration of the SN determines an imbalance in the regulation of the circuitry of the basal ganglia, whose GABAergic output to the thalamus is increased, therefore determining an inhibition of this projection nucleus [3]. Hence, messages relayed from the thalamus to the cortex, especially those coding for voluntary movement, are slowed down. Many strategies are being pursued to develop new therapies for Parkinsonian patients, oriented to prevention of the damage of the nigrostriatal system and to the replacement of lost neurons. These techniques range from the use of dopaminotrophic factors [4] and viral vectors [5], to the transplantation of primary xenogenetic tissue [6]. Cell transplantation with fetal tissue [26], xenograft from porcine tissue [7,8] and immortalized cell lines [9,10,11] appear as clinically promising experimental treatments in late stage Parkinson's disease, where tissue replacement is essential. Indeed, considering that Parkinson's patients develop symptoms when approximately 50-70% of the dopaminergic neurons in the SN are lost [3], the development of cell transplant models appears as a much needed step in the cure for the disease. Further, the discrete anatomical localization of the degenerating tissue make it a likely candidate for cell transplant therapy with minimally invasive stereotaxic procedures. More than two hundred patients have received transplants worldwide [12]. Clinical improvement has been confirmed by functional studies using positron emission tomography of striatal fluorodopa (F-DOPA) uptake after transplantation [12,13] and was correlated to good graft survival and innervation of the host striatum in postmortem studies of transplanted patients [1]. However, and in spite of these promising findings, neural transplantation remains a controversial procedure, facing ethical dilemmas in procurement of tissue (human, human fetal), and in obtaining adequate amounts of cells. The development of neuronal, dopaminergic cell lines from the SN would greatly contribute to overcome these difficulties, and would provide a limitless supply of cells that could be easily manipulated and modulated in vitro prior to transplantation. At present, successful attempts to establish such lines have been made in rat SN, using SV4O virus transfection [9,10,11]. However, cell lines so established tend to exhibit problems in their stability and/or viability [14]. Our group has successfully established a continuously growing cell line from the SN of an adult Fisher 344 rat [15,16], named RCSN-3, using an original protocol developed in our laboratory which has yielded continuously growing cell lines from mammalian tissue of diverse origins that stably retain differentiated traits [16-21]. In the present work, we show that the RCSN-3 cell lines presents neuronal markers by immunohystochemistry and lack glial traits. We also transplanted suspensions of this cell line in the striatum of 6 OH dopamine lesioned rats, achieving up to 75% recovery in the rotational behavior of the animals. Part of this work has been presented in abstract form [22].

122

C. Arriagada el al. /Immortalized Neuronal Cell Line

MATERIAL AND METHODS Cell culture a) Establishment of cell lines. The RCSN-3 cell line was derived from the substantia nigra of 4 month old normal Fisher 344 rat. The cell material used to establish primary cultures was then transformed to a permanent cell line by exposing them to media conditioned by UCHT 1 cells (Figure 1), a process that induces transformation in cell cultures [16-22]. For standard culture conditions, the cells were kept in feeding medium consisting of DMEM/ Ham F12 nutrient mixture ( 1 : 1 ) (Sigma Chemical Co., Saint Louis. MO, USA) modified to contain 6 g/1 glucose, 10% bovine serum, 2.5% fetal bovine serum. 100 U/ml penicillin, 100 (mg/ml streptomycin (Sigma) supplemented with 10% (v/v) with UCHT1 conditioned medium. The cultures were maintained in an incubator at 37°C with 100% humidity and an atmosphere of 10% C02 and were monitored routinely for the appearance of transformation foci or morphological changes. After 10 weeks in culture, transformation foci were evident. The cultures were expanded and part were criopreserved in liquid nitrogen. The cell line was cloned by dilutional culturing, giving rise to the clonal line RCSN-3. Cells are passaged at confluence with trypsinization (1% trypsin. Gibco. Grand Island, NY. USA). b) Culture of cell lines. For standard growth conditions, RCSN-3 CNh cells were cultured in feeding medium. Media was renewed completely twice a week. For differentiation, the cell were kept in a media consisten of DMEM/Ham F12 nutrient mixture, supplemented with 2% adult bovine serum and 1% (v/v) of N3 supplement as previously described [18] and 1% (v/v) Site+3 supplement (Sigma). Cell were allowed to differentiate for 1 week.

C. Arriagada et al. /Immortalized Neuronal Cell Line

123

Figure 1. UCHT1 transformation protocol. Media conditioned by the rat thyroid UCHTI cell line for 48 hrs. is freeze-thawed 3 times in the absence of cryopreservants. The media is filtered through 0.2 mm filters to yield a cell free conditioned media. Primary cultures of mammalian rigin are kept in the presence of 10-20% (v/v) for the time range noted. Successful transformation is assessed by the generation of transformation foci in the culture.

Morphology Cells were fixed in formaldehyde 4% in phosphate buffer pH 7,4. a) Citology. Cytochemical reactions included: Hematoxilin - eosin staining, ferrous ion capture to demonstrate the presence of melanin in the form of neuromelanin, paraformaldehyde-glyoxylate staining to demonstrate the presence of catecholamines. b) Immunohystochemistry. Fixed cells were permeabilized in an ascending/descending alcohol battery ranging from 50 to 96%. The blocking reaction was carried out using BSA 1% in phosphate buffer. The antibodies utilized here were the next: /. neuronal markers: NSE (pre-diluted, Biogenex), Synaptophysin (pre-diluted, Biogenex) and MAP-2 (1:1000, Sigma); 2. glial markers: GFAP (pre-diluted, Biogenex) and S-100 (pre-diluted, Biogenex); 3. functional markers: TH (1:1000 1:1500, Sigma). The incubation with the primary antibodies was carried out overnight and an ABC detection kit (Biogenex) was used to

124

C. Arriagada et al. /Immortalized Neuronal Cell Line

develop the reaction and utilizing DAB as chromogen. Specific primary antibodies, fluorescein labeled secondary antibodies and tetanus toxin (kind gifts. Dr. Lautaro Perez. ICBM) were used to evaluate the presence of neurofilament 200 kD and tetanus toxin receptor. Intracellular Ca 2+ measurements For intracellular Ca2+ measurements, the cells were palled onto 35 mm culture dishes. The variations of intracellular Ca2+ were assessed by Ca2+ imaging techniques using Fluo-3. The cells were incubated at 37°C for 40-60 mm with the indicator. The dishes were visualized in an Olympus BH2 microscope equipped with epifluorescence (halogen lamp). The microscope was connected to a Cooled Extended Isis digital camera (Photonic Science Ltd, Robensbridge, UK) connected to a dedicated PC equipped with a Axon Digidata 2000 digitizing board (Axon Instruments, Foster City, CA). Images were acquired at 12 bit resolution and 1 Hz using a customized software Axon Imaging Workbench 2.1.80 (Axon), and stored in the computer hard disk for later analysis. The compositions of the normal extracellular solutions were (in mM): 135 or 145 NaCl, 5 KC1, 2 MgCl 2 , 1.5 or 2.5 CaCI2, 10 4-(2-hidroxyethil) piperazine-1-ethanesulfonic acid (HEPES)-NaOH, 10 Dextrose pH=7.4). Surgical procedures, behavioral testing. 4 adult male Fisher 344 rats (20-250 g) were lesioned by unilateral injection of 6hydroxydopamine bromide at two sites along the medial forebrain bundle. Assessment of apomorphine induced rotational behavior (i.p injection of 5 mg apomorfine per kg body weight. National Health Service. Chile) was carried out visually twice, once per week. before transplantation. Only rats with more than 160 rotations every 30 mm were utilized and three times after transplantation (days 30, 55 and 80). For transplant, confluent cultures were washed in PBS and dissociated with 1% trypsin. 500.000 cells in a volume of 4 mL were implanted through blunt Hamilton syringe and deposited at AP +1.0 min. ML -2.5 mm and V -4.7 mm (coordinates relative to bregma), toothbar set at -2.5. Rotational behaviour was assessed visually every two weeks after transplantation. RESULTS The RCSN-3 cell line grows on monolayers, with a doubling time of 52 hrs, a plating efficiency of 21% and a saturation density of 410.000 cells/cm2, when kept in feeding medium. Figure 2 shows that undifferentiated RCSN cells tend to exhibit a epithelial like morphology, with short or no processes and a more acidophylic cytoplasm. After differentiation, cell proliferation is greatly reduced, and RCSN cells develop processes and establish contact with neighboring cells. The presence of melanin was evidenced with the ferrous ion capture technique, demonstrating a homogeneous distribution of the pigment in the cytoplasm, with faint labelling in undifferentiated stages and a substantial increase upon differentiation.

C. Arriagada et al. /Immortalized Neuronal Cell Line

125

Figure 2. The top panels represent phase contrast microscopy images. Note the decrease in cell number and the increased length and number of dendrites. H-E: Hematoxilin-eosin.

Inmunohistochemical characterization demonstrates that RCSN cells express neuronal traits, evidenced by the positive inmunolabelling for NSE, synaptophysin and MAP-2. NSE and synaptophysin show a finely granular pattern, evenly distributed in the cytoplasm (Figure 3). Synaptophysin is specially intense at the zone of cell-cell interaction. MAP-2 shows a fibrillary pattern of labelling, surrounding vacuole-like cytoplasmic structures. Microfilament 200 kD labels differentiated cells homogeneously, and tetanus toxin is present in the cell membrane in a patch-like distribution.

126

C. Arriagada et al. /Immortalized Neuronal Cell Line

Figure 3. Immunohystochemistry for neuronal markers. NSE: Neuron specific enolase. SNP: Svnaptophysin. MAP-2: microtubular associated protein-2. The bottom panels represent images taken under epifluorescence microscopy conditions, taken in differentiated cells.

C. Arriagada et al. /Immortalized Neuronal Cell Line

127

Figure 4. The top panels represent immunohystochemistry for tirosine hydroxylase (TH) in RCSN cells, under control and differentiated conditions. The bottom panels Micrograph of RCSN cells using the ferrous ion capture technique, where fluorescent areas represent catecholamine deposits.

Functional neuronal markers are shown in figure 4, which presents immunohystochemi-cal staning for tyrosine hydroxylase. The labeling is slightly less intense in non-differentiated cells, and the label is distributed in the entire cytoplasm following a granular pattern. The presence of catecholamines is also clear from the micrographs presented in figure 4, with a cytoplasmic distribution. Glial markers GFAP and S100 were negative in both control and differentiated conditions (data not shown). When differentiated, up to 40% of fluo-3 loaded RCSN cells respond with intense increase in intracellular Ca2+ when stimulated externally with 200 mM glutamate, and even more intensely when using simultaneous depolarizing conditions (70 mM K+), a situation depicted in figure 5. Of 16 cells explored, the Ca2+ signal peaks after 1 sec. of stimulation, and returns to basal level between 30-40 sec. after the peak.

128

C. A rriagada et al. / Immortalized Neuronal Cell Line

Figure 5. Ca2+ signals in fluo-3 loaded RCSN lines. The image shows cells 3 sec after being stimulated with the addition of 200 uM glutamate, and even more intensely when using simultaneous depolarizing conditions (70 mM K + ). Fluorescence intensity is depicted in a pseudo color scale, which in ascending order is black blue - green - yellow - orange - red.

Finally, we transplanted suspensions of RCSN cells in the striatum of rats with 6OH dopamine- induced lesions of the substantia nigra. Figure 6 shows typical patterns in the evolution of the rotational behaviour after transplantation, characterized by either a smooth, decreasing exponential type curve which level off after 12 weeks at approximately 25% of the initial rotation rate. Another pattern involves a greater drop in rotations 2 weeks after transplantation, followed by an increase and later a sustained decrease in the rate of rotation to again plateau after 12 weeks. At 16 weeks, the rats were sacrificed, and section of the striatu were taken and immunohystochemically stained with TH and DOPA decarboxilase antibodies. As shown in figure 5, cells staining positively for both markers are present in the striatum, showing intense labeling and neurite formation.

C. Arriagada el al. / Immortalized Neuronal Cell Line

129

Figure 6. Transplant of RCSN-3 cells in the striatum of 6OH dopamine lesioned rats. The graphs at the left represent the decrease in the rate of rotation after transplant. The microphotographs on the right represent tirosine hydroxilase (TH) and DORA decarboxilase (DOPA-DC) immunostaining in striatal sections of two rats sacrificed 16 weeks after transplantation.

DISCUSSION The use of immortalized cell lines to study basic phenomena at the cellular and molecular levels is highly desirable. Ideally, such cell models should derive from the differentiated tissue, and retain the properties of the originating tissue in vitro. However, current transformation protocols have yet to overcome problems involving viability and/or stability [14]. Stem cells are a promising possibility, but finding the adequate in vitro conditions to attain the state of differentiation desired in such cells can be cumbersome and costly. Our UCHT1 transformation protocol seems to overcome these drawbacks to the extent that all our cell lines so generated retain differentiated properties for extended periods of time [16-21]. Further, we have even immortalized pathological tissue, such as the cerebral cortex of a mouse bearing an extra copy of chromosome 16 [17,18], an animal model of Down's syndrome, which retains the same neurotransmitter dysfunction previously observed in primary cultures of the same origin [23]. The present study shows that the RCSN-3 clonal cell line retains general properties of neuronal tissue, and possesses specific characteristics of the SN, such as the presence of tyrosine hydroxylase, DOPA

130

C. Arriagada et al. /Immortalized Neuronal Cell Line

decarboxilase and catecholamines. Previous studies have shown that this cell line exhibits neurotoxicity in response to the external application of salsolinol [22], a dopamine-derived isoquinoline which reportedly acts as an endogenous dopaminergic neurotoxin, inducing selective neuronal cell death and eliciting symptoms almost identical to idiopathic Parkinson's disease in humans [24]. Further, salsolinol has been reported to induce behavioral changes similar to those observed in Parkinson's disease [25], making it an interesting candidate for the study of dopamine induced toxicity. If we also consider that RCSN-3 cells respond to glutamate with increases in intracellular Ca2+, the cell line appears as a likely candidate for the study of excitotoxic mechanisms involved in neurodegeneration, as well as more specific mechanisms related to dopamine derived neurotoxicity. However, a very exciting application arises from our present study. The RCSN-3 cell line appears as a model for cell transplant therapy, inducing a sustained and progressive reduction in the rotational behavior of 6 OH dopamine lesioned rats. Interestingly, no previous in vitro differentiation was utilized in our transplantation experiments, which may prove to be a practical asset, as the cells either have enough dopaminergic function at the time of the inoculation, or the in vivo microenvironment in the striatum may be enough to sustain or induce a differentiated phenotype in the RCSN-3 line. Cell grafting in Parkinsonian models has been the subject if intense study [26]. Procedures have involved the use of primary mesencephalic cells [12,27], genetically modified cells that do not differentiate into neurons in the host striatum [28], and dopaminergic neurons derived from precursors expanded in vitro [1]. The latter may reflect the situation of RCSN-3 cells more closely, as it offers many of the advantages of expanded precursors. Indeed, the expansion of cells in culture followed by a differentiation period or genetic manipulation may reduce the amount of tissue needed and provide optimal cell material. However, it must be pointed out that non-expanded neuronal precursor have shown differentiation in vivo when transplanted [29-32], in vitro expanded cells have been less efficient in developing into neurons once transplanted [33,34]. It appears that in the latter case a differentiation period in culture, rather than just an expansion of the cells, is necessary [1]. Previous work with primary fetal mesencephalic tissue suggests that behavioral recovery be directly correlated to the number of surviving dopaminergic neurons in the host striatum [35] We cannot adequately address this issue in our study, since we have used a reduced number of animals, although the survival rate reported n successful grafts of dopaminergic neurons is rather low (3-5%) [12]. From the study in the striatal sections in our transplanted animals, our RCSN-3 cell line appears well in these figures, with the added advantage of not requiring prior differentiation in vitro. Our present study presents the RCSN-3 cell line as a model to study Parkinson's related mechanisms at the cellular level, and presents the UCHT1 protocol as a method for generating limitless cellular material for cell transplant. It is tempting to speculate that this cell line may also be successful in transplantation in higher order of mammal, such as primates. Indeed, recent reports of successful xenografting of porcine cells in primates [36,37] is encouraging, and we therefore intend to transplant the RCSN-3 cells in MPTP primates shortly. Undoubtedly, the success of xenografting would expand our potential sources of tissue, and provide more opportunities of generating adequate immortalized cell lines to propose for transplantation in humans.

C. Arriagada et al. /Immortalized Neuronal Cell Line

131

REFERENCES [ 1 ] L. Studer et al., Transplantation of expanded mesencephalic precursors leads to recovery in parkinsonian rats. Nature Neurosci. 1 (1998) (4):290-295. [2] J. Sian, et al., Neurotransmitters and Disorders of the Basal Ganglia. In Basic Neurochemistry: Molecular, cellular and medical aspects. Siegel, GJ (Ed), 1999, 6* Edition. [3] A.M. Lozano et al., New developments in understanding the etiology of Parkinson's disease and in its treatment, Current Opinion in Neurobiology, 8 (1998):783-790. [4] H. Takayama, et al. Basic fibroblast growth factor increases dopaminergic graft survival and function in a rat model of Parkinson's disease. Nat. Med. 1 (1995) 53-58. [5] D. L. Choi-Lundberg, et al., Dopaminergic neurons protected from degeneration by GDNF gene therapy Science 275(1997) 838-841. [6] T. Deacon, et al. Histological evidence of fetal pig neural cell survival after transplantation into a patient with Parkinson's disease. Nat. Med. 3 (1997) 350-353 [7] O. Isacson et al., Transplanted xenogenic neural cells in neurodegenerative disease models exhibit remarkable axonai target specificity and distinct growth patterns of glial and axonal fibres. Nat Med., 1 (1995) 1189-1194. [8] J.M. Schumacher and O. Isacson Neuronal xenotransplantation in Parkinson's disease. Nat Med, 3 (1997) 474-475. [9] E.D. Clarson et al., Immortalized dopamine neurons: A model to study neurotoxicity and neuroprotection. Proc Soc Exp Biol Med., Nov 222(2) (1999) 157-63. [10] F.G. La Rosa et al., Inhibition of proliferation and expression of T-antigen in SV4O large T-antigen gene-induced immortalized cells following transplantations. Cancer Lett., 26;1 13(1-2) (1997) 55-60. [11] K.N. Prasad et al., Efficacy of grafted immortalized dopamine neurons in an animal model of parkinsonism: a review. Mol Genet Metab 1998 Sep;65(l) (1998) 1-9 [12] C. W. Olanow et al., Fetal nigral transplantation as a therapy for Parkinson's disease. Trends Neurosci. 19(1996) 102-109 [13] O.K. Wenning, et al. Short- and long-term survival and function of unilateral intrastriatal dopaminergic grafts in Parkinson's disease. Ann. Neurol. 42 (1997) 95-107 [14] F. Tiemarm, and W.J. Deppert, Virol 68 (1994) 2869-2878 [15] C. Arriagada et al., Establishment and characterization of a neuronal cell line derived from the substantia nigra of the adult rat. Soc. Neurosci. Absti., 24(1) (1998) 93.13 [16] A. Dagnino-Subiabre et al., Angiotensin receptor II is present in dopaminergic cell line of rat substantia nigra and it is down regulated by aminochrome. Molec. & Cell Bioch em. (2000), in press. [17] D.D. Allen et al., Impaired cholinergic function in cell lines derived from the cerebral cortex of normal and trisomy 16 mice. Fur J Neurosci. 12 (2000) 3259 - 3264. [ 18] A.M. Cardenas et al., Intracellular calcium signals in immortal cell lines derived from the cerebral cortex of normal and trisomy 16 fetal mice, an animal model of human trisomy 21 (Down syndrome). Neuroreport, 10(2) (1999), 363-369. [19] J.L. Liberona et al., Ion channels during differentiation of a human skeletal muscle cell line. J. of Muscle Res. & CellMotil., 18(5) (1997), 587-598. [20] P. Caviedes et al., Ion currents in a skeletal muscle cell line from a Duchenne muscular dystrophy patient. Muscle & Nerve. 17(1994), 1021-1028. [21] P. Caviedes et al., Calcium fluxes, ion currents and dihydropyridine receptors in a newly established cell line from rat heart muscle. J. Molec. & Cell Cardiol. 25 (1993), 829-845. [22] J. Salazar et al., The RCSN cell line, derived from the substantia nigra of an adult rat, as a model for pathophysiological studies and cell transplant therapy in Parkinson's disease. Symposium: Neurodegenerative Mechanisms in Parkinson's disease. XX Congress of the Latinamerican Association for Physiological Sciences. 3-6 de September 3-6, 2000. Cancun, Mexico. [23] J. Fiedler et al., Altered cholinergic function in cultured neurons from the trisomy 16 mouse fetus, a model for Down Syndrome. Brain Res., 658 (1994), 27-32. [24] M. Naoi et al.,. N-methyl-(R)-salsolinol as a dopaminergic neurotoxin: from animal model to an early marker of Parkinson's disease. J. Neural Transm [Suppl.J 50 (1997) 89-105. [25] M. Naoi et al., Dopamine-derived endogenous 1 (R),2(N)-dimethyl-6,7-dihydroxy- 1,2,3,4tetrahydroisoquinoline, N-rnethyl(R)-salsolinol, induce Parkinsonism in rat: biochemical, pathological and behavioral studies. Brain Res. 709 (1996) 285-295. [26] T. Alexi et al., Neuroprotective strategies for basal ganglia degeneration: Parkinson's and Huntington's diseases Progress in Neurobiology. 60 (2000) 409-470 [27] O. Lindvall et al., Grafts of fetal dopamine neurons survive and improve motor function in Parkinson's disease. Science 247 (1990) 574-577

132

C. Arriagada et al. / Immortalized Neuronal Cell Line

[28] L.S. Shihabuddin et al., The search for neural progenitor cells: prospects for the therapy of neurodegenerative disease. Molecular Medicine Today. 5 (1999) 474-480. [29] O Briistle et al., Host guided migration allows targeted introduction of neurons into the embryonic brain. Neuron, 15(1995) 1275-1285 [30] K. Campbell et al., Regional incorporation and site specific differentiation of striatal precursors transplanted to the embryonic forebrain ventricle. Neuron, 15 (1995), 1259-1273. [31] G. Fishell, Striatal precursors adopt cortical identities in response to local cues. Development, 121 (1995) 803-812 [32] C. Vicario-Abejon et al., Cerebellar precursors transplanted to the neonatal dentate gyms express fatures characteristic of hippocampal neurons J. Neurosci. 15 (1995),6361-6363. [33] C.N. Svendsen et al., Survival differentiation of rat and human epithelial growth factor-responsive precursor cells following grafting into the lesioned adult central nervous system. Exp. Neurol, 137 (1996),376-388. [34] C.N. Svendsen et al., Long-term survival of human central nervous system progenitor cells transplanted into a rat model of Parkinson's disease. Exp Neurol, 148 (1997): 135-146. [35] P. Brundin et al. Human fetal dopamine neurons grafted in a rat model of Parkinson's disease: immunological aspects, spontaneous and drug-induced behavior, and dopamine release. Exp. Brain Res 70(1988) 192-208. [36] O. Isacson et al., Transplanted xenogeneic neural cells in neurodegenerative disease models exhibit remarkable axonal target specificity and distinct growth patterns of glial and axonal fibres. Nat Med. 1 (1995): 1189-1194. [37] J.M. Schumacher and O. Isacson, Neuronal xenotransplantation in Parkinson's disease. Nat Med. 3 (1997)474-475.

Parkinson's Disease E. Ronken and G.J.M. van Scharrenburg (Eds.) IOS Press, 2002

133

Lentiviral Vector Delivery of GDNF in Primate Models of Parkinson's Disease Prevents Neurodegeneration Jeffrey H. Kordowera,*, Marina E. Emborga, Jocelyne Blochb, Shuang Y. Maa, Yaping Chua, Liza Leventhala, Jodi McBridea, Er-Yun Chena, Stephane Palfia, Ben Zion Roitberga, James E. Holdenc,d, W. Douglas Brownd, Robert Pyzalskid, Michael D. Taylorc, Paul Carveye, ZaoDung Linge, Didier Tronof, Phillippe Hantrayeg, Nicole Deglonb and Patrick Aebischerb,h a

Department of Neurological Sciences, Rush Presbyterian-St. Luke's Medical Center, Chicago, IL 60612, USA b Division of Surgical Research and Gene Therapy Center, Lausanne University Medical School, Lausanne, Switzerland c Department of Medical Physics, University of Wisconsin, Madison, WI53706, USA Department of Radiology, University of Wisconsin, Madison, WI 53706, USA e D e p a r t m e n t o f Pharmacology, Rush Presbyterian-St f

Department of Genetics and Microbiology, Faculty of Medicine, University of Geneva, Geneva, Switzerland 7 Commissariat a I'Energie Atomique (CEA), CNRS, Unite de Recherche Associe (URA), 2210 Service Hospitalier Frederic Joliot, CEA, Direction des Sciences du Vivant (DSV), Departement de Recherche Medicale (DRM), Orsay cedex, France h Swiss Federal Institute of Technology, EPFL, Lausanne, Switzerland. Abstract: Lentiviral delivery of glial cell line-derived neurotrophic factor (lentiGDNF) was tested for its trophic effects upon degenerating nigrostriatal neurons in nonhuman primate models of Parkinson's disease (PD). lenti-GDNF was injected into the striatum and substantia nigra of nonlesioned aged rhesus monkeys or young adult rhesus monkeys treated 1 week prior with 1 methyl-4-pheny 1-1,2,3,6tetrahydropyridine (MPTP). Extensive GDNF expression with anterograde and retrograde transport was seen in all animals. In aged monkeys, lenti-GDNF augmented dopaminergic function. In MPTP-treated monkeys, lenti-GDNF reversed functional deficits and prevented nigrostriatal degeneration. Additionally, lentiGDNF injections to intact rhesus monkeys revealed long-term gene expression (8 months). In MPTP-treated monkeys, lenti-GDNF treatment reversed motor deficits in a hand-reach task. These data indicate that GDNF delivery using a lentiviral vector system can prevent nigrostriatal degeneration and induce regeneration in primate models of PD and might be a viable therapeutic strategy for PD patients. Reprinted and adapted with permission from Science 290: 767-773 (2000). © 2000 American Association for the Advancement of Science.

Corresponding author.; E-mail: [email protected].

134

J. H. Kordower et al. / Lentiviral Vector Delivery of GDNF

Introduction Parkinson's disease is a progressive disorder resulting from degeneration of dopaminergic neurons within the substantia nigra. Surgical therapies aimed at replacing lost dopaminergic neurons or disrupting aberrant basal ganglia circuitry have recently been tested [1]. However, these clinical trials have focused on patients with advanced disease, and the primary goal of forestalling disease progression in newly diagnosed patients has yet to be realized. Glial cell line-derived neurotrophic factor (GDNF) has potent trophic effects on dopaminergic nigral neurons [2], suggesting that this factor could provide neuroprotection in patients with early PD. We have shown that intraventricular administration of GDNF failed to improve clinical function or prevent nigrostriatal degeneration in a patient with PD, and this failure resulted from an ineffective delivery method [3]. Gene therapy is a powerful means to deliver trophic molecules to the central nervous system in a site-specific manner. Robust transfer of marker and therapeutic genes has recently been demonstrated in the rodent and nonhuman primate brain with the use of a lentiviral vector [4]. The transgene expression is long-term and nontoxic. Using two different nonhuman primate models of PD, we examined whether lentiviral-mediated delivery of GDNF could reverse the cellular and behavioral changes associated with nigrostriatal degeneration in primates. For the first model, we chose nonlesioned aged monkeys that displayed a slow progressive loss of dopamine within the striatum and tyrosine hydroxylase (TH) within the substantia niura without frank cellular degeneration [5]. These aged monkeys demonstrate changes within the nigrostriatal system that model some of the cellular changes seen in early PD [6]. In the second model, young adult monkeys received unilateral intracarotid injections of 1-methyl4-phenyl-l,2,3,6-tetrahydropyridine (MPTP) to induce extensive nigrostriatal degeneration, resulting in a behavioral syndrome characterized by robust motor deficits. Methods and Results Eight aged (approximately 25 years old) female rhesus monkeys received injections of lentiviral vectors encoding ß-galactosidase (lenti-ßGal; n = 4) or GDNF (lenti-GDNF; n = 4) targeted for the striatum and substantia nigra and were killed 3 months later. Postmortem, all GDNF injections were localized to the caudate nucleus, putamen, and supranigral regions, as revealed by standard staining procedures. The cDNA coding for a nuclear-localized ß-galactosidase (LacZ) and the human GDNF containing a Kozak consensus sequence (a 636-bp fragment position 1 to 151 and 1 to 485; GenBank accession numbers L19062 and L19063) were cloned in the SIN-W-PGK transfer vector [7]. The packaging construct and vesicular stomatis virus G protein (VSV-G) envelope used in this study were the PCMVDR8.92, PRSV-Rev, and the PMD.6 plasmids described previously [8,9]. The viral particles were produced in 293T cells as previously described [10]. The titers (3 to 5 x 108 TU/ml) of the concentrated LacZ-expressing viruses (200,000 and 250,000 ng p24/ml in experiment 1 and 450,000 ng p24/ml in experiment 2) were determined on 293T cells. The GDNF-expressing viral stocks were normalized for viral particles content using p24 antigen measurement. All experimentation was performed in accordance with NIH guidelines and institutional animal care approval. Level II Biosafety procedures were used. Under MRI guidance, each monkey received six stereotaxic injections of lent-ßGal or lenti-GDNF bilaterally into the caudate nucleus, putamen, and substantia nigra. Injections were made into the head of the caudate nucleus (10 ml), body of the caudate nucleus (5 ul), anterior putamen (10 ul), commissural putamen (10 m,l), postcommissural putarnen (5 ml). and substantia nigra (5 ml). Injections were made through a 10

J.H. Kordower et al. / Lentiviral Vector Delivery of GDNF

135

ml Hamilton syringe connected to a pump at a rate of 0.5 ml/min. During the injection, the needle was raised 1 to 2 mm to better disperse the lentivirus through the intended target. The needle was left in place for an additional 3 min to allow the injectate to diffuse from the needle tip. The left side was injected 6 weeks before the right. During the first surgical session, there was a technical failure with the virus aggregating in the needle, which prevented its injection into the brain. This was con-firmed at postmortem examination using GDNF-immunohistochemistry and ßGal histochemistry. Thus, the left side served as an additional control for the right side. GDNF immunohistochemistry was performed with a commerdally available antibody (R&D Systems, Minneapolis, MN; 1:250), using the ABC method and nickel intensification. Deletion or substitution for the primary antibody served as controls. Under control conditions, no staining was observed. All aged monkeys receiving lenti-GDNF displayed robust GDNF immunoreactivity within the right striatum and substantia nigra. In contrast, no monkeys receiving lenti-ßGal displayed specific GDNF immuno-reactivity in the right striatum. Rather, these monkeys displayed robust expression of ßGal similar to that reported previously [10]. In lentiGDNF-treated animals, GDNF immunoreactivity within the striatum was extremely dense and distributed throughout the neuropil. When the primary antibody concentration was decreased to one-tenth of the standard, the intense striatal neuropil staining was diminished, and GDNF-immunoreactive perikarya were easily seen. Numerous GDNF-immunoreactive perikarya were also seen within the substantia nigra of lenti-GDNF-injected monkeys. Within the striatum and substantia nigra, Nissl-stained sections revealed normal striatal cytoarchitecture without significant cytotoxicity. Macrophages were occasionally observed within the needle tracts. Gliosis was similar across treatment groups and was principally confined to the regions immediately surrounding the needle tracts. Lenti-GDNF injections resulted in marked anterograde transport of the trophic factor. Intense GDNF immunoreactivity was observed within fibers of the globus pallidus and substantia nigra pars reticulata after striatal injections. GDNF-containing fibers emanating from putaminal injection sites were seen coursing medially toward and into the globus pallidus. These staining patterns were clearly distinct from the injection site and respected the boundaries of the striatal target structures. In contrast, anterograde transport of ßGal was not observed in lenti-ßGal monkeys. This suggests that secreted GDNF, and not the virus per se, was anterogradely transported. Aged monkeys underwent fluorodopa (FD) positron emission tomography (PET) before surgery and again just before being killed. All procedures followed an overnight fast. After sedation with ketamine (10-15 mg/kg), the animal was intubated, and femoral angiocatheters were placed for tracer injection and blood sampling. Anesthesia was then maintained by 1 to 2% isofluorane for the remainder of the procedure. Carbidopa (2-3 mg/kg i.v.) was administered 30 min before the FD study. The animal was placed in a stereotaxic head holder constructed of materials compatible with PET scanning, and a transmission scan was acquired for correction of the emission data for attenuation. FD (185 MBq) was administered over 30 s and a 90-min three-dimensional dynamic emission scan started. The scan included 22 frames with durations increasing from 1 min initially to 5 min at the end. The bed was moved cyclically by the interplane distance between each pair of 5min scans to give a net coronal sampling interval of 2.125 mm. Regions of interest (ROI) were placed on the caudate nucleus, putamen and occipital cortex in individual morphometic MR images coregistered with the FD image data. Cortical time courses were used as input functions to generate functional maps of the uptake rate constant Ki, by the modified graphical method [11]. Striatal ROIs were transferred to the functional maps, and the Ki values were evaluated as the ROI means for each structure. Before treatment, all monkeys displayed symmetrical FD uptake in the caudate and putamen bilaterally (ratio: 1.02 ± 0.02). Similarly, there was symmetrical (4% difference)

136

J.H. Kordower el al. /Lentiviral Vector Delivery of GDNF

FD uptake in all lenti-ßGal-treated monkeys after lentivirus injections. In contrast, FD uptake was significantly asymmetrical (27%) in lenti-GDNF-treated monkeys with greater uptake on the side of the GDNF expression (P < 0.007). With respect to absolute values, lenti-ßGal animals displayed a trend toward reduced FD uptake after treatment relative to baseline levels (P = 0.06). Qualitatively, three of four lenti-GDNF-treated monkeys displayed clear increases in FD uptake on the treated side. This increase in uptake (K, value) between the groups just failed to reach statistical significance (P = 0.06). Within the striatum, lentiviral delivery of GDNF increased a number of markers of dopaminergic function. All monkeys were perfused with saline. The brain was removed, immersed in ice-cold saline for 10 min, and slabbed on a monkey brain slicer. Slabs through the head of the caudate and putamen were punched bilaterally with a 1 mm brain punch. These punches were processed for HPLC [12]. The tissue slabs were immersed in Zamboni's fixative. Stereological counts and volumes of TH-immunoreactive neurons were performed with NeuroZoom software using the optical dissector method for cell counting and the nucleator method for measuring neuronal volume. Optical density measurements were performed to assess the relative intensity of TH staining within the caudate nucleus and putamen. On the left side where there was no lentiGDNF expression, the intensity of TH immunoreactivity within the caudate nucleus and putamen was similar between groups. In contrast, significant increases in optical density measures of TH immunoreactivity were seen in the right striatum of lenti-GDNF-infused monkeys relative to lenti-ßGal-treated animals or the contralateral side. In this regard, there was a 44.1% and a 38.9% increase in optical density measures of TH immunoreactivity within the caudate nucleus and putamen, respectively. At the time of death, tissue punches were taken throughout the caudate nucleus and putamen of all monkeys. Relative to lenti-ßGal-treated animals, measurement of dopamine (DA) and homovanillic acid (HVA) revealed significant increases in the right caudate nucleus (140% DA, P < 0.001; 207% HVA, P <0.001) and putamen (47.2% DA, P <0.05: 128% HVA, P < 0.01) in lenti-GDNF-treated aged monkeys. Lentiviral delivery of GDNF to aged monkeys resulted in an increase in the number of TH-immunoreactive neurons within the substantia nigra. Regardless of the extent of GDNF immunoreactivity within the midbrain, the organization of TH-immuno-reactive neurons was similar in all animals, and these neurons were not observed in ectopic locations within this locus. Stereological counts revealed an 85% increase in the number of THimmunoreactive nigral neurons on the side receiving lentivirally delivered GDNF (Fig. 4A) relative to lenti-ßGal-treated animals. On the side (left) that did not display GDNF immunoreactivity, lenti-GDNF-treated animals contained 76,929 ±4918 TH-immunoreactive neurons. This is similar to what was seen in lenti-ßGal-infused animals (68,543 ± 5519). Whereas lenti-ßGal-infused monkeys contained 63,738 ± 6094 TH-immunoreactive nigral neurons in the right side, lenti-GDNF-treated monkeys contained 118,170 ± 8631 THimmunoreactive nigral neurons in this hemisphere (P <0.001). A similar pattern was seen when the volume of TH-immunoreactive substantia nigra neurons was quantified. TH-immunoreactive neurons from lenti-ßGal- and lenti-GDNFtreated monkeys were similar in size in the left nigra where there was no GDNF expression (11,147.5 ± 351 mm3 and 11,458.7 ± 379 mm3, respectively). In contrast, a 35% increase in neuronal volume was seen on the GDNF-rich right side in lenti GDNF-injected aged monkeys (lenti-ßGal 10,707.5 ± 333 mm3; lenti-GDNF 16,653.7 4 ± 1240 mrn3; P<0.001). Although Stereological counts of TH mRNA-containing neurons were not performed, there was an obvious increase in the number of TH mRNA-containig neuron, within the right substantia nigra in lenti-GDNF-treated monkeys compared with lenti-ßGal-containing animals. With regard to the relative levels of TH mRNA expression within individual nigral

J. H. Kordower et al. / Lentiviral Vector Delivery of GDNF

137

neurons, the pattern of results was similar to that observed with TH-immunoreactive neurona] number and volume. The TH riboprobe was prepared as previously described [3]. The probe was conjugated to 2 mM biotin-14-CTP (Gibco BRL/Life Technologies, Rockville, MD). 1 mg Pvu 1-linearized pBS-TH31, 5 mM DTT, 50 U RNasin, 4 U T3RNA polymerase, 0.5 mM CTP, and 0.25 mM of ATP, GTP, and UTP. Tissue was processed for immunohistochemistry by the ABC method using this probe as the primary antibody. Optical density measurements were performed using Nffl Image. On the left side, the optical density of TH mRNA within nigral neurons was similar between lenti-ßGal- and lenti-GDNF-treated monkeys (78.28 ± 2.78 and 80.58 ± 2.5, respectively). In contrast, there was a significant (21.5%) increase in the optical density for TH mRNA in lenti-GDNFtreated monkeys (98.3 ± 1.5) relative to lenti-ßGal-treated monkeys (77.2 ± 2.3) on the right side (P< 0.01). In the second experiment, 20 young adult rhesus monkeys were initially trained 3 days per week until asymptotic performance was achieved on a hand-reach task in which the time to pick up food treats out of recessed wells was measured [5,12]. Each experimental day, monkeys received 10 trials per hand. Once per week, monkeys were also evaluated on a modified parkinsonian clinical rating scale (CRS). All monkeys then received an injection of 3 mg MPTP-HC1 into the right carotid processed artery, initiating a parkinsonian state. One week later, monkeys were evaluated on the CRS. Only monkeys displaying severe hemiparkinsonism with the classic crooked arm posture and dragging leg on the left side continued in the study (n = 10). It is our experience that monkeys with this behavioral phenotype display the most severe lesions neuroanatomically and do not display spontaneous recovery behaviorally [12]. On the basis of CRS scores, monkeys were matched into two groups of five monkeys, which received on that day lenti-ßGal or lenti-GDNF treatment. Using magnetic resonance imaging (MRI) guidance, we gave all monkeys lentivirus injections into the caudate nucleus (n = 2), putamen (n = 3), and substantia nigra (n = 1) on the right side using the same injection parameters as in experiment 1. One week later, monkeys began retesting on the hand-reach task three times per week for 3 weeks per month. Testing was performed during weeks 2 to 4 for month 1, and weeks 1 to 3 for months 2 and 3. Monkeys were not tested for the first week in month 1 to allow them time to recover from surgery. Testing was not performed for the final week of months 2 and 3 to allow for routine veterinary care (month 2) and transportation to the University of Wisconsin for PET scans (month 3). For statistical analyses, the times for an individual week were combined into a single score. During the weeks of hand-reach testing, monkeys were also scored once per week on the CRS. Individuals blinded to the experimental treatment between performed all behavioral assessments. Three months after lentivirus treatment, monkeys received a FD PET scan and were killed 24 to 48 hours later, and tissues were histologically processed as before. Within 1 week after the lentivirus injections, one monkey from each group died. Necropsies from these animals revealed only the presence of mild necrosis from multifocal random hepatocellular coagulaion. On account of these deeths, all remaining monkeys underwent detailed necropssies after death, and no significant abnormalities in any organs were seen. Before MPTP treatment, all young adult monkeys scored 0 on the CRS. After MPTP, but before lentivirus injection, monkeys in the lenti-GDNF and lenti-ßGal groups averaged 10.4 ± 0.07 and 10.6 ± 0.6, respectively, on the CRS (P >0.05). After lentivirus treatment, significant differences in CRS scores were seen between the two groups (KolmogorovSmirnov test, P <0.0001). CRS scores of monkeys receiving lenti-ßGal did not change over the 3-month period after treatment. Scores began to decrease in the first month after lenit-

138

J.H. Kordower el al. /Lentiviral Vector Delivery of GDNF

GDNF treatment. However, statistically significant differences between lenti-GDNF and lenti-ßGal were only discerned at posttreatment observations 6, 7, 8 and 9 (KolmogorovSmirnov test, P <0.04 for each comparison). Lenti-GDNF-treated animals also improved performance on the operant hand-reach task. Under the conditions before MPTP administration, animals in both groups performed this task with similar speed. For the "unaffected" right hand, no differences in motor function were discerned for either group relative to performance before MPTP administration or to each other (P > 0.05). In contrast, performance with the left hand was significantly improved in lenti-GDNF-treated animals relative to controls (P < 0.05). After MPTP, all lenti-ßGal-treated animals were severely impaired, with monkeys often not performing at all, or requiring more than the maximally allowed 30 s. In contrast, three of the four lentiGDNF monkeys performed the task with the left hand at near-normal levels, whereas one lenti-GDNF-treated monkey was impaired and performed this task in a manner similar to the lenti-ßGal-treated animals. Between the groups, significant differences in performance were discerned on posttreatment tests 4, 6, 7, 8, and 9 (P < 0.05 for each comparison). Just before being killed, all monkeys underwent FD PET scans. Qualitatively, all lentißGal-treated monkeys displayed pronounced FD uptake in the left striatum and a comprehensive loss of FD uptake on the right side (Fig. 2C). In contrast, two of four lentiGDNF-treated animals displayed robust and symmetrical FD uptake on both sides. The remaining two lenti-GDNF monkeys displayed reduced FD uptake on the right side, but with K, values 50 to 100% greater than lenti-pGal controls. Quantitatively, no differences in FD uptake were observed between groups within the left striatum (P > 0.05). In contrast, there was a significant (>300%) increase in FD uptake in lenti-GDNF-treated animals in the right striatum relative to lenti-ßGal-treated animals (P < 0.05). When the right striatum was subdivided, significant increases in FD uptake were only seen within the putamen of lentiGDNF-treated animals (P < 0.05). After death, a strong GDNF-immunoreactive signal was seen in the caudate nucleus, putamen, and substantia nigra of all lenti-GDNF-treated, but none of the lenti-ßGal-treated animals. The intensity and distribution of GDNF immunoreactivity was indistinguishable from what we observed in aged monkeys. All lenti-ßGal-treated monkeys displayed a comprehensive loss of TH immunoreactivity within the striatum on the side ipsilateral to the MPTP injection. In contrast, all lenti-GDNF-treated monkeys displayed enhanced striatal TH immunoreactivity relative to ßGal controls. However, there was variability in the degree of striatal TH immunoreactivity in lenti-GDNF-treated animals and that variability was associated with the degree of functional recovery seen on the hand-reach task. Two lentiGDNF-treated monkeys displayed dense TH immunoreactivity throughout the rostrocaudal extent of the striatum. In these monkeys, the intensity of the TH immunoreactivity was greater than that observed on the intact side. These two animals displayed the best functional recovery. A third lenti-GDNF-treated monkey also displayed robust functional recovery on the hand-reach task. However, the enhanced striatal TH immunoreactivity in this animal was limited to the post-commissural putamen. The fourth lenti-GDNF-treated monkey did not recover on the hand-reach task. Although putaminal TH immunoreactivity in this animal was still greater than controls, the degree of innervation was sparse and restricted to the medial postcommissural putamen. Lenti-GDNF treatment enhanced the expression of THimmunoreactive fibers throughout the nigrostriatal pathway. Unlike what was observed in aged monkeys, however, some TH-immunoreactive fibers in the striatum displayed a morphology characteristic of both degenerating and regenerating fibers. Large, thickened fibers could be seen coursing in an irregular fashion in these animals. Rostrally, these fibers appeared disorganized at times, with a more normal organization seen more caudally. TH-

J.H. Kordower et al. /Lentiviral Vector Delivery of GDNF

139

immunoreactive sprouting was also seen in the globus pallidus, substantia innominata, and lateral septum. These novel staining patterns were not immunoreactive for dopamine ßhydroxylase confirming the dopaminergic phenotype of this response. Quantitatively, lentißGal-treated monkeys displayed significant decreases in the optical density of THimmnunoreactive fibers within the right caudate nucleus (71.5%; P < 0.006) and putamen (74.3% P < 0.0007) relative to the intact side. When analyzed as a group, TH optical density in the right caudate nucleus and putamen of lenti-GDNF-treated monkeys was significantly greater than that seen in lenti-ßGal-treated monkeys (P < 0.001 for both) and was similar to that seen on the intact side of these animals (P > 0.05 for both). All lenti-ßGal-treated monkeys displayed a dramatic loss of TH-immunoreactive neurons within the substantia nigra on the side ipsilateral to the MPTP injection. In contrast, the nigra from all four of the lenti-GDNF-treated displayed complete neuroprotection, regardless of the degree of functional recovery. In lenti-ßGal-treated monkeys, intracarotid injections of MPTP resulted in an 89% decrease in the number, and an 81.6% decrease in the density, of TH-immunoreactive nigral neurons on the side ipsilateral to the toxin injection (P < 0.001). In contrast, lenti-GDNF-treated monkeys displayed 32% more THimmunoreactive nigral neurons (P < 0.001) and an 11% increase in TH-immunoreactive neuronal density (P < 0.05) relative to the intact side. In lenti-ßGal-treated animals, MPTP significantly reduced (32%) the volume of residual TH-immunoreactive nigral neurons on the lesion side relative to the intact side (P < 0.001). In contrast, the volume of THimmunoreactive neurons in lenti-GDNF-treated animals was significantly larger (44.3%) on the lesioned side relative to the intact side (P < 0.001). Finally, the optical density of TH mRNA was quantified bilaterally in all animals. In lenti-ßGal-treated animals, there was a significant decrease (24.0%) in the relative optical density of TH mRNA within residual neurons on the MPTP-lesioned side relative to the intact side (P < 0.03). In contrast, lentiGDNF-treated animals displayed a significant increase (41.7%) in relative optical density of TH mRNA relative to the intact side or lenti-ßGal-treated animals (P < 0.001). Sections from all monkeys were stained for CD45, CD3, and CD8 markers to assess the immune response after lentiviral vector injection. These antibodies are markers for activated microglia, T cells, and leukocytes including lymphocytes, monocytes, granulocytes, eosinophils, and thymocytes. Staining for these immune markers was weak, and often absent, in these animals. Mild staling for CD45 and CD8 was seen in two animals. Some CD45-immunoreactive cells displayed a microglial morphology. Other monkeys displayed virtually no immunoreactivity even in sections containing needle tracts. Two additional intact young adult rhesus monkeys received lenti-GDNF injections into the right caudate and putamen and the left substantia nigra using the same injection protocol. These animals were killed 8 months later and were evaluated by immunohistochemistry and enzyme-linked immunosorbent assay (ELISA) for long-term gene expression. Brain punches were homogenized in 150:1 buffer I [0.1 M tris-buffered saline, pH 8.1, containing 1 mM EDTA 1% aprotinin, 10 mg/ml leupeptin. 14 mg/ml pepstatin, 4 mM phenylmethylsulfonyl fluoride (PMSF)] for 30 s in the ice slurry. An equal amount of buffer 11 (0.1 M tris-buffered saline, pH 8.1, containing ImM EDTA, 1% aprotinin 10 g/ml leupeptin, 14 mg/ml pepstatin, 4 mM PMSF, and 0.5% NP-40) was then added. The tubes were shaken for 2 hours. The supenatant was collected for ELISA and protein measurements. The ELISA reaction was completed in 96-well plate (Dynatech Chantilly, VA) according to the ELISA manufacturer's instructions (GDNF Emax - ImmunoAssay Systems Kit G3520, Promega, Madison, WI). The optical densities were recorded in ELISA 31 plate reader (at 450 nm wave length; Dynatech). Some lysates were diluted to ensure all the optical densities were within the standard curve. The concentrations or GDNF were calculated against six-point standard curve and then adjusted to picograms of GDNF per milligram of total protein. The total protein in each tissue lysate was measured using Bio-

140

J.H. Kordower et al. /Lentiviral Vector Delivery of GDNF

Rad protein assay kit (Bio-Rad Richmond, CA). Robust GDNF immunoreactivity was seen in the right caudate, right putamen, and left ventral midbrain in both animals. In the right substantia nigra, many GDNFimmunoreactive neurons were seen. This labeling represents retrograde transport of GDNF after injections of lenti-GDNF into the right striatum. Further, dense GDNF-immunoreactive fiber staining, representing anterograde transport of the trophic factor, was seen within the right substantia nigra pars reticulata. Tissue punches taken at the time of death revealed significant levels of GDNF produced by striatal cells 8 months after lentiGDNF injections. On the side without a striatal injection, 0.130 ± 0.062 and 0.131 ± 0.060 ng/mg protein of GDNF were seen in the caudate nucleus and putamen, respectively. Significantly higher GDNF levels were observed within the caudate nucleus (2.25 ±0.312 ng/mg protein; P < 0.001) and putamen (3.5 ± 0.582 ng/mg protein; P < 0.001) on the lentiGDNF-injected side. Discussion Our study demonstrates that delivery of GDNF cDNA into the nigrostriatal system using a lentiviral vector system can potently reverse the structural and functional effects of dopamine insufficiency in nonhuman primate models of aging and early Parkinson's disease. Most critically, lenti-GDNF delivery prevented the motor deficits that normally occur after MPTP administration. In this regard, functional disability was prevented on both a subjective clinical rating scale modeled after the Unified Parkinson's Disease Rating Scale and an objective operant motor test. Consistent expression of GDNF was observed in aged and lesioned monkeys with significant and biologically relevant levels of GDNF observed for up to 8 months after lentivirus injection. Indeed, the 2.5 to 3.5 ng/mg protein of GDNF produced after lenti-GDNF injections compares very favorably to the 50 to 152 pg/mg protein of striatal GDNF produced after intrastriatal adenovirus injections in monkeys [13]. This consistent gene expression occurred without significant toxicity to aged monkeys, and minor toxicity in two of the MPTP-treated monkeys, supporting our previous observations [10]. Still, the death of two monkeys needs to be addressed. Pathological analyses revealed only a mild necrosis from multifocal random hepatocellular coagulation in these animals, and this was not deemed to be the cause of death. No other young adult or aged monkeys from this or a previous study [10] displayed morbidity or mortality after lentivirus injections. Further, detailed necropsies from the remaining MPTP-treated animals failed to reveal any relevant pathology. Although the absolute cause of death remains elusive, we hypothesize that the death of these two monkeys relates to the impact of the surgical procedure 1 week after the MPTP injections and is unrelated to the lentivirus injection. In aged monkeys, lentiviral delivery of GDNF augmented host nigrostriatal function as determined by a variety of morphological, physiological, and neurochemical dependent measures. In this regard, lenti-GDNF increased the size and number of TH-immunoreactive neurons within the substantia nigra; increased the expression of TH mRNA within these neurons; increased the levels of dopamine, dopaminergic metabolites, and dopaminergic markers in the striatum; and in-creased FD uptake within the striatum as determined by PET scan. Enhanced nigrostriatal dopamine function was consistently associated with the expression of lentivirally delivered GDNF, as enhanced nigrostriatal function was only seen on the side with robust gene expression. We used aged monkeys to model specific cellular changes that occur in aging and the earliest aspects of PD. Phenotypic down-regulation of the TH gene and protein are among

J.H. Kordower et al. / Lentiviral Vector Delivery of GDNF

141

the earliest pathological events seen within the substantia nigra in PD [6], and analogous changes are seen in aged rhesus monkeys [5]. The number of TH-immunoreactive nigral neurons seen in lenti-ßGal-injected animals was similar to that previously reported for aged rhesus monkeys [5]. In contrast, lenti-GDNF-treated aged monkeys displayed nigral neurons in numbers similar to those seen in young adult animals. The possibility that the lenti-GDNF spurred neurogenesis of dopaminergic nigral neurons cannot be ruled out. However, the delivery of lenti-GDNF to the nigral region resulted in transgene expression throughout the midbrain. Yet, TH-immunoreactive neurons were observed only within established catecholaminergic nuclei and not in ectopic midbrain locations. A more parsimonious explanation is that GDNF up-regulated TH-immunoreactivity in aged nigral neurons that had previously down-regulated TH expression below detectable levels. The enhanced TH mRNA expression seen within nigral neurons after lenti-GDNF treatment supports this interpretation. Lenti-GDNF also prevented the behavioral and neuroanatomical effects of MPTP-induced nigrostriatal degeneration. It is notable that, unlike many other neuroprotection paradigms, the lenti-GDNF injections were performed after the parkinsonian state was initiated, thus better modeling what can be attempted in PD patients. The exact mechanism by which lenti-GDNF exerted its effects requires further elucidation. It is clear that neuroprotection was achieved within the substantia nigra, as these neurons do not degenerate within a week of MPTP treatment [14]. However, striatal fibers can degenerate during this time, and whether the GDNF is preventing degeneration or inducing sprouting of degenerating fibers still needs to be established. Indeed, there is evidence for both mechanisms as some animals displayed fiber morphology and topography indicative of regeneration. A critical question is whether preservation of striatal innervation, nigral perikarya, or both, is required for functional recovery. Although the number of animals in our study is too small to provide a definitive answer, it is notable that all lenti-GDNF-treated monkeys had complete preservation of nigral perikarya. Yet, functional recovery on the hand-reach task was absent only in the one monkey with sparsest striatal reinnervation. Thus it appears that GDNF-mediated striatal reinnervation is critical for functional recovery in nonhuman primates, a concept supported by recent studies performed in rodents [15,16]. The failure to potently protect dopaminergic innervation in the one monkey may be due to variability in the speed by which nigrostriatal fibers are lost after MPTP. At the time of the lenti-GDNF injections, dopaminergic fibers in this monkey may have regressed to a level where access to the GDNF was limited, and regrowth to the striatum was impossible. Not only was lenti-GDNF capable of preventing the degeneration of nigrostriatal neurons in MPTP-treated monkeys, it augmented many of the morphological parameters relative to the "intact" side. It is likely that the unilateral 3 mg MPTP dose induced a small loss of TH-immunoreactive neurons on the contralateral side. Thus the increased numbers of TH-immunoreactive neurons may reflect complete neuroprotection on the side of GDNF expression contrasted with a small loss of TH-immunoreactive neurons on the side not injected. Conclusions and outlook We injected lentivirus into both the striatum and substantia nigra in order to maximize the chance for an effect. For lenti-GDNF therapy to be a practical clinical approach, studies determining the regions of GDNF delivery critical to reverse progressive nigrosttiatal degeneration are needed. The importance of related biological events such as anterograde transport of GDNF from injection sites to target regions also needs to be established.

142

J.H. Kordower et al. / Lentiviral Vector Delivery of GDNF

Finally, potential adverse events resulting from lenti-GDNF inducing supranonnal levels of striatal dopamine needs to be evaluated. Toward this end, vectors with built-in inducible systems that can modulate gene expression in cases of dose-limiting side effects need to be developed. Still, the reversal of slowly progressive cellular phenotypic changes seen in aged monkeys, combined the structural and functional neuroprotection and regeneration seen in MPTP-treated monkeys, indicates that lentiviral delivery of GDNF may provide potent clinical benefits for patients with PD. Acknowledgements This research was supported by a grant from the Department of Defense, The Charles Shapiro Foundation, NS40578, and by the Swiss National Science Foundation and the Swiss National Program in Neurological Diseases. We thank T. Kladis and J. Stansell for expert histological assistance, F. Pidoux and M. Rey for the technical assistance in the production of the lentiviral vectors, K. Gibbons for assistance with PET scans, and J. Sladek jr. for photographic assistance. References [ I]

C. Honey et al., New developments in the surgery for Parkinson's disease. Can. J. Neurol 2 (1999) S45-52. [2] A. Bjorklund et al., Studies on neuroprotective and regenerative effects of GDNF in a partial lesion model of Parkinson's disease. Neurobiol. Dis. 4 (1997) 186- 200. [3] J.H. Kordower et al., Clinicopathological findings following intraventricular glial-denved neurotrophic factor treatment in a patient with Parkinson's disease. Ann. Neurol. 46 (1999) 419-424. [4] L. Naldini et al., In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science, 272 (1996) 263-267. [5] ME. Emborg et al., Age-related declines in nigral neuronal function correlate with motor impairments in rhesus monkeys. J. Comp. Neurol. 401 (1998) 253-265. [6] A. Kastner et al., Tyrosine hydroxylase protein and messenger RNA in the dopaminergic nigral neurons of patients with Parkinson's disease. Brain Res. 606 (1993) 341-345. [7] R. Zufferey et al., Woodchuck hepatitis virus posttranscriptional regulatory element enhances expression of transgenes delivered by retroviral vectors. J. Virol. 73 (1999) 2886-2892. [8] R. Zufferey et al., Multiply attenuated lentiviral vector achieves efficient gene delivery in vivo.Nature Biotech. 15(1997)871-875. [9] A.F. Hottinger et al., Complete and long-term rescue of lesioned adult motoneurons by lentiviralmediated expression of glial cell line-derived neurotrophic factor in the facial nucleus. J. Neurosci. 20 (2000) 5587-5593. [10] J.H. Kordower et al., Lentiviral gene transfer to the nonhuman primate brain. Exp. Neurol. 160 (1999) 1-16. [11] C.S. Patlak and R.G. Blasberg, Graphical evaluation of blood-to-brain transfer constants from multiple-time uptake data. Generalizations. J. Cereb. Blood Flow Metab. 5 (1985) 584-590. [12] J.H. Kordower et al., Encapsulated PC 12 cell transplants into hemiparkinsoman monkeys: a behavioral, neuroanatomical, and neurochemical analysis. Cell Transplant. 4(1995) 155-171 [ 13] D.A. Kozlowski et al. ASNTR. Abstr. 7 (2000) 25. [ 14] J.L. Eberling et al., Dopamine transporter loss and clinical changes in MPTP-lesioned primates.Brain Res. 832(1999) 184-187. [15] B. Connor et al., Differential effects of glial cell line-derived neurotrophic factor (GDNF) in the striatum and substantia nigra of the aged Parkinsonian rat. Gene Ther. 6 (1999) 1936-1951. [16] C. Rosenblad et al., Sequential administration of GDNF into the substantia nigra and striatum promotes dopamine neuron survival and axonal sprouting but not striatal reinnervation or functional recovery in the partial 6-OHDA lesion model. Exp. Neurol. 16 (2000) 503-516.

Parkinson's Disease E. Ronken and G.J.M. van Scharrenburg (Eds.) IOS Press, 2002

143

Design of Clinical Studies for Neuroprotective Efficacy J.M. Bronstein M.D., Ph.D. UCLA School of Medicine, Department of Areurology, 710 Westwood Plaza, Los Angeles, CA 90095, USA Abstract. Parkinson's disease (PD) is a neurological disorder characterized by progressive loss of dopaminergic neurons and clinical disability. Current therapy for the treatment of PD improves many of the symptoms but none have been shown to he neuroprotective. Previous trials evaluating potential neuroprotective agents have utilized both clinical measures and functional imaging. The advantages and shortcomings of these measures of disease progression are discussed. It is recommended that future trials include either PET or SPECT scanning in addition to clinical measures of disease severity to monitor the effectiveness of new agents in slowing neuronal loss in PD. Keywords: clinical diagnosis, neuroprotection, neuroimaging (SPECT, PET, MRI), DATATop

Great advances have been made in the past 30 years in the symptomatic treatment of Parkinson's disease (PD) but strategies for delaying disease progression have been largely unsuccessful. Although modern therapy for PD has resulted in an almost normal life-span, many patients will eventually become disabled by complications of therapy (i.e. motor fluctuations and drug-induced psychosis), postural instability, and cognitive decline. These disabilities result in a lower quality of life for patients and their families in addition to posing a huge financial burden on society. Delaying the onset of these disabling symptoms by even 25-50% would therefore result in a marked improvement in patients' quality of life. Thus, one major goal of future therapies is to alter the natural progression of the disease. The natural progression can be measured both clinically and pathologically although they are clearly related. Using both of these measures in studying potential inventions that alter the progression of the disease hold distinct advantages over relying on only one. 1.

Parkinson's disease

PD is a progressive neurodegenerative disorder characterized clinically by tremor, rigidity, and bradykinesia early in the disease. As the disease advances, many patients suffer from postural instability and cognitive decline. The motor symptoms experienced by patients with PD reflect at least in part, the deafferentation of the striatum due to cell loss in the substantia nigra. This cell loss leads to a deficiency in dopamine that clearly causes much of the disability, especially early in the disease. This is exemplified by the dramatic improvement in the clinical status of a patient after dopamine repleting therapy. Disability seen later in the disease likely reflects additional extra-nigral neuronal degeneration in such areas as the locus coeruleus and nucleus basalis of Meynert. Decreases in neurotransmitters other than dopamine have also been demonstrated. For example, serotonin levels are reduced in PD and are believed to lead to depression, which is commonly associated with the disorder.

144

2.

J.M. Bronstein / Design of Clinical Studies for Neuroprotective

Efficacy

Clinical measures of disease progression

There are several studies that have measured the effect of various treatments on motor disability in PD but few that have documented the natural progression without therapy. Hoehn and Yahr's classic description of the progression of clinical features in PD led used a now well-established rating scale [1]. Using this scale (HY I-V), they documented progression to advanced disease in 37% of patients within 5 years and 63% within 10 years (HY stage III to IV). Disease progression using this measure is not linear since it took an average of 14 years to reach the most advanced level (HY V). Other limitations of the HY scale include the large increments between stages and the fact that it does not measure other aspects of disability including cognition, complications of therapy, and quality of life. Thus, the HY scale is a simple measure of motor disability in PD but it is far from comprehensive enough to be used as a primary measure of progression. In an effort to develop a more comprehensive measure of clinical progression in PD, the United Parkinson's Disease Rating Scale (UPDRS) was created [2]. Disease progression early in PD was well documented using the UPDRS in the Deprenyl and Tocopherol Antioxidative Therapy of Parkinsonism (DATATOP) study [3]. In this study, 800 patients were enrolled with early (HY I or II) and untreated PD of a duration of less then 5 years and followed for an average of 21 months. The motor component of the UPDRS deteriorated by 8-9% per year in the untreated group although a smaller subset progressed at a much slower rate of 34% per year. A slower rate of decline similar to that of the subset in the DAT ATOP study was described in two subsequent studies [4,5]. Some of the patients in the later studies were receiving treatment and clinical measures were taken after a relatively short period of medication withdrawal. The use of symptomatic medication adds potential confounds in the interpretation of disease progression. Symptomatic effects are difficult to wash out given their relatively long duration of effects and the symptomatic treatments themselves may have an effect on disease progression. Indeed, a long duration of action of levodopa has been well described as has a long duration of the symptomatic effect of selegeline. These studies exemplify some of the challenges in using clinical measures in monitoring disease progression in PD. There was variability between studies of 2-3 fold, measures were influenced by symptomatic effects of treatments, and it is unknown if the nonlinear nature of deterioration measured reflected the progression of the underlying disease process or flaws in the rating scales used. 3.

Nonclinical measures of disease progression

Cell loss in the substantia nigra has been estimated from pathological studies in normal aging and in PD [6]. Decline in neuronal counts increased exponentially with disease duration. Postmortem studies obviously can not be practically used to determine the effect of potential therapies on disease progression. Functional imaging studies offer a reasonable alternative in approximating substantia nigra cell loss. Fluorodopa uptake measures using positron emission tomography (PET) directly correlates with pathological measures [7]. More recently, single-photon emission computed tomography (SPECT) using radiolabeled dopamine transporter analogues have also been successfully used to measure the rate of loss of dopaminergic function in PD [8]. Using PET and putamenal measures of [1xF]dopa influx, Morrish et al, estimated the rate of decline in PD to be 12.5% per year [9]. They also noted a more rapid rate of progression in advanced patients but this accelerated rate did not reach statistical significance. SPECT studies have also demonstrated the ability of functional imaging's ability to measure disease progression in PD and estimates an annual decline in dopamine transporter uptake of 3 to 12% [8]. Thus, PET and SPECT offer powerful and objective measures of disease progression that reflect the rate of loss of

J. M. Bronstein / Design of Clinical Studies for Neuroprotective Efficacy

145

dopaminergic neurons. One can make predictions of how many patients need to be studied in order to detect a statistically significant neuroprotective effect using PET and SPECT. Based on the estimates of Morrish et al, the rate of loss of dopamine uptake in putamen, expressed as Ki min/year (mean +/- SD) in new PD patients is 0.0008+/- 0.0008 [9]. The number of subjects needed per group, assuming a power of 0.80 using a two tailed t test (p value of 0.05) and a two year scanning interval, are as follows: 100% neuroprotection, 9 per group; 75% neuroprotection, 15 per group; 50% neuroprotection, 32 per group; and 25 % neuroprotection, 126 per group. Marek predicts that the follow sample size would be required using SPECT and a two year scan interval: 100% neuroprotection, 16-20 per group; 50% neuroprotection, 57-68 per group; and 25 % neuroprotection, 208-260 per group [8]. Despite these advantages over clinical measures, there are some important limitations when using these measures. It is still unclear whether progression is linear throughout the disease process and there is considerable variability in measurements of progression between individual subjects and studies. Although it is likely that putamenal measures of dopaminergic function are the most sensitive measure of PD progression, there may be variations in the rate of progression between the more and less effected putamen. Finally, PET and SPECT measurements do not necessarily reflect disability which is an important, if not the most important goal of any therapeutic intervention [9]. This may be due to the fact that these techniques are only measuring dopaminergic function in the striatum and not looking at other regions and transmitter systems that are involved in PD. 4.

Previous attempts at neuroprotection in PD

There have been few controlled clinical trials evaluating neuroprotective therapies in PD. Ideally, a neuroprotective treatment would intervene in a specific aspect of the disease process. The lack of potential protective agents reflects the general lack of insight we have into the etiology of the disease. Oxidant stress has been a predominant theory in the pathogenesis of PD based on pathological findings and knowledge of dopamine metabolism. This theory suggests that the normal metabolism of dopamine results in the generation of hydroxyl radicals leading to cell death. Early attempts at neuroprotective therapy have therefore been focused on anti-oxidant therapy, blocking dopamine metabolism, and reducing the amount of levodopa used and thus secondarily reducing dopamine metabolism. More recently, neurotrophic agents have been tested with the hope that they will slow disease progression regardless of the cause. The DAT ATOP Study was the 1st large study to evaluate the possible neuroprotective effect of selegeline (a monoamine oxidase type B inhibitor, MAO-B) and tocopherol (vitamin E) in PD (3). The hypothesis was that MAO-B inhibition would lead to a decrease in free radical formation with dopamine degradation and tocopherol would act as a free radical scavenger and decrease free radical-induced cell damage. Eight hundred early, untreated PD patients were enrolled and the primary endpoint of the study was time interval in which disability reached a point requiring the initiation of levodopa therapy. Tocopherol was found to be no better than placebo in slowing the onset of clinical disability but there was a significant benefit in the selegeline group. Post hoc analysis revealed that there was an unanticipated symptomatic effect of selegeline making it difficult to interpret potential neuroprotective effects. Follow-up analysis on the DATATOP patients revealed that early use of selegeline did not alter the progression of disability in the patients never on levodopa [10] nor did it alter the occurrence of levodopa-induced adverse events [11]1. Thus, after enrolling 800 patients and following many of them for 3 years, neither selegeline nor tocopherol demonstrated any long-term disease altering actions in PD. Dopamine agonists have been used as a symptomatic therapy in PD for several years

146

J. M. Bronstein / Design of Clinical Studies for Neuroprotective

Efficacy

but have been recently studied for their potential neuroprotective properties. The later was driven by their ability to lower the amount of levodopa needed by PD patients and therefore may secondarily reduce oxidative stress. Dopamine agonists have also been shown to have free radical scavenging ability and may stimulate growth factor release. There have been several studies demonstrating dopamine agonists' ability to delay the onset of motor fluctuations in patients in early PD but it has remained unclear whether the benefits were due to purely symptomatic effects of the drugs or whether they alter the disease course [12]. In these studies, patients were randomized to initial treatment with levodopa or a dopamine agonist and they were followed for the development of motor fluctuations. Some of these studies also followed progression using PET and SPECT. The studies investigating the effect of dopamine agonists on disease progression using functional imaging revealed a trend for neuroprotection using pramipexole, pergolide, or ropinirole but none apparently reached statistical significance (all of these investigations have yet to be published). Taken together, dopamine agonists can delay clinical measures of disease progression but have small or no effect on the rate of decline of dopaminergic cell loss. These studies exemplify the power of performing both clinical and non-clinical measures of progression in PD. Glial derived neurotrophic factor (GDNF) has potent neuroprotective actions in vitro and in animal models of PD and therefore was an excellent candidate as a disease altering compound in Parkinson patients. Because it is a peptide, it must be delivered directly into brain tissue or into the cerebral spinal fluid (CSF) for it to reach target cells. A multicenter placebo controlled trial was performed in advanced patients with PD. Originally, the study was designed to determine the safety of GDNF in patients and clinical measures of disease severity were used as secondary endpoints (e.g. UPDRS. HY, timed tasks, need for symptomatic medications). Patients were required to have a reservoir surgically inserted in order to deliver study material to the CSF (intraventicularly). There was no apparent benefit when comparing active drug versus placebo but there were significant side effects associated with GDNF administration. In my opinion, there were serious design flaws that limited interpretation of these data. Subjects were followed for less than one year and decline in clinical measures cannot be readily detected in such a short time period. Functional imaging studies were not performed so there was no data obtained on the effect GDNF had on the integrity of the dopaminergic system. Finally, the patients tested in this study had advanced disease and the neuroprotective actions of GDNF had only been demonstrated in a smaller loss of dopaminergic neurons in animal models. 5. Lessons learned from studies in multiple sclerosis There are several important lessons to be learned when investigating disease altering treatments by studying the progress made in the field of multiple sclerosis (MS). For several years, clinical measures were used to evaluate potential therapies with little success. The course of the disease varies considerably between and within patients over time. There was a confounding placebo effect in all therapies tested making interpretation of data difficult. These factors required investigators to use large groups of patients studied over long periods of time. This increased the expense of the studies and reduced the number of potential agents that could be tested. It was recently discovered that disease activity could be monitored using magnetic resonance imaging (MRI) to determine the number of brain lesions and therefore allowing for an objective measure of disease progression. The inclusion of MRI evaluation in clinical MS trials soon resulted in the conformation that three different drugs were effective in altering the course of MS. More recently, it was found that the measuring the extent of brain atrophy in addition to the number of lesions added further information in the predicting the progression of clinical disability. Thus, the use of objective imaging techniques has markedly improved the ability to evaluate new

J.M. Bronstein / Design of Clinical Studies for Neuroprotective Efficacy

147

candidate disease altering therapies and that more than one measure may offer additional information. 6. Conclusion Current therapies for the treatment of PD offer great symptomatic relief but there is a need for disease modifying therapies to reduce long-term disability. Clinical measures of motor performance, quality of life, and cognition are necessary and important since these are the aspects of patient's lives we are trying to improve but these measures have clear limitations when studying potential neuroprotective agents. Prolonged symptomatic effects confound clinical rating scales as does placebo effects. Marked variability of clinical measures requires the enrollment of large numbers of patients to be studied. Functional imaging such as PET and SPECT complement clinical measures and offers a biological substrate for observed effects of study drug. I feel that PET or SPECT therefore should be included in all studies evaluating neuroprotection in PD. In addition to measuring dopaminergic function in the striatum, imaging other transmitter systems such as serotonin may provide further power to studying disease-modifying agents. References. [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [ ll] [12]

M.M. Hoehn and M.D. Yahr, Parkinsonism. Onset, progression and mortality. Neurology.17 (1967) 427-432 S. Fatin and R.L. Elton, and members of the UPDRS Development Committee. United Parkinson's Disease Rating Scale. In: FaHn S, Marsden CD, Goldstein M, Calne CD, eds. Recent Developments in Parkinson's Disease Volume IL Florham Park, Ni: Macmillian; (1987) 153-163. Group TPS. Effect of Deprenyl on the Progression of Disability in Early Parkinson's Disease. N Engl J Med. 321(1989) 1364-71. C.S. Lee et al. Clinical Observations on the Rate of Progression of Idiopathic Parkinsonism. Brain. 117 (1994)501-507. C.W. Olanow et al. The effect of deprenyl and levodopa on the progression of Parkinson's disease. Ann Neurol 38 (1995) 771-777. I.M. Fearnley and A.J. Lees. Ageing and Parkinson's disease: substantia nigra regional selectivity. Brain. 114(1991)2283-2301. H.I. Snow et al. Correlations in humans between premortem PET [18F] fluorodopa uptake, postmortem dopaminergic cell counts, and striatal dopaminergic levels. Ann Neurol. 34 (1993) 324330. K. Marek, Dopaminergic Dysfunction in Parkinsonism: New Lession from Imaging. The Neuroscientist. 5 (1999) 333-340. P.K. Morrish, G.V. Sawle ,D.J. Brooks. An [I8F]dopa-PET and clinical study of the rate of progression in Parkinson's disease. Brain. 119(Pt 2) (1996) 585-91. Group PS. Impact of Deprenyl and Tocopherol Treatment on Parkinson's Disease in DATATOP Subjects Not Requiring Levodopa. Ann Neurol. 39 (1996) 29-36. Group PS. Impact of Deprenyl and Tocopherol Treatment on Parkinson's Disease in DATATOP Patients Requiring Levodopa. Ann Neurol. 39 (1996) 37-45. O. Rascol et al. A five-year study of the incidence of dyskinesia in patients with early Parkinson's disease who were treated with ropinirole or levodopa. N Engl J Med 342 (2000) 1484-91.

148

Parkinson's Disease E. Ronken and G.J.M. van Scharrenburg (Eds.) /OS Press. 2002

Regulation of Glial cell line-Derived Neurotrophic Factor (GDNF) for dopamine conservation in models for Parkinson's disease Eric Ronken, Damian McCrossan and Jaap Venema Solvay Research Laboratories, Solvay Pharmaceuticals BV, C.J. van Houtenlaan 36, 1381 CP Weesp. The Netherlands Abstract. Glial cell line-Derived Neurotrophic Factor (GDNF) has been established as a powerful trophic factor for midbrain dopaminergic neurons. Thus, it facilitates dendritic outgrowth and striatal dopamine turnover in normal nigral neurons and increases resistance to dopaminergic neurotoxins such as MPTP or 6hydroxydopamine. Moreover, dopaminergic neurons that have been lesioned by these neurotoxins can be partially recovered by administration of GDNF in vivo. To apply these powerful effects of GDNF in patients with Parkinson's disease, it may be worthwhile to identify the mechanisms by which GDNF expression can be regulated locally in the caudate nucleus, the putamen and the substantia nigra. This approach would circumvent delivery problems, associated with large proteins such as trophic factors and would enforce local physiological ceiling levels to dosing thereby precluding the clear risk for overdosing. To do this, we investigated expression of GDNF in rat brain and studied signal transduction mechanisms by which GDNF expression can be controlled. Keywords: GDNF, neuroprotection, controlled expression

Introduction Parkinson's disease (PD) is a slow, but progressive neurodegenerative disorder, primarily affecting midbrain dopaminergic neurons with their cell bodies in the substantia nigra and projecting onto the caudate nucleus and putamen [1,2,3,4]. Current drug therapy aims for symptomatic relief, using dopamine agonists as a proven principle with clear clinical efficacy. However, long-term strategies to attenuate further degeneration of dopaminergic neurons are still not realized. One of those strategies that is currently under evaluation is the potential use of trophic factors in neurodegenerative disorders. Of these trophic factors, Glial cell line-Derived Neurotrophic Factor (GDNF) stands out for its potent support of dopaminergic midbrain neurons and may serve as the prototypic trophic factor to design long-term therapeutic strategies that will slow down the degeneration process in PD. GDNF Glial cell line-Derived Neurotrophic Factor (GDNF) was identified in a program dedicated to find trophic factors that promote dopaminergic neuronal survival and differentiation[5,6]. This program was initiated because of the hypothesis 1) that trophic factors play a major role in developmental organization and maintenance of brain's cellular architecture, and 2) that during aging, the brain may become deficient of certain (amounts of) growth factors.

E. Ronken et al. /Regulation of GDNF for Dopamine Conversation

149

The latter may result in degeneration of specific brain circuits, e.g. loss of nigrostriatal dopamine (DA) neurons in Parkinsonian patients [7]. The consistent presence of trophic factors that promoted DA uptake in primary mesencephalic neuronal cell cultures in conditioned medium of the rat glioma cell line B49 was previously reported [8]. Using this bioassay, the trophic factor was purified to apparent homogeneity, followed by partial protein sequencing of the aminoterminus, using Edman degradation [6]. A cDNA library, prepared from B49 cells was screened subsequently by an oligo probe, based on the obtained amino acid sequence and yielded a full length clone, encoding GDNF. Like the purified GDNF, recombinant GDNF was found to promote the survival of dopaminergic neurons with extreme potency. Incubation of primary mesencephalic dopaminergic cell cultures with GDNF increased the number of surviving tyrosine hydroxylase-positive neurites and increased specific uptake of tritiated DA in a concentration-dependent manner [6]. Meanwhile, a number of GDNF congeners have been identified using homology cloning. These ligands have various degrees of preference for the GDNF receptor family that has also emerged subsequently, namely, GFRa-1, -2, -3 and -4. These include neurturin [9], persephin [10], Artemin [11] and Enovin [12]. It is believed that any given ligand forms a complex with a subunit which then signals through a common receptor kinase, cRET. Neuroprotective effects of GDNF The neurotoxins 6-OHDA (6-hydroxydopamine) and MPTP (1-methyl 4-phenyl 1,2,3,6tetrahydropyridine) are known to inactivate dopaminergic cell functions [13,14,15]. In addition to the survival promoting effects on mescencephalic cells in primary culture, GDNF was able to offer to these cells neuroprotection against toxic insults by these compounds. It was discovered that the DA phenotype of the striatum or the substantia nigra (tyrosine hydroxylaseGFRaimmunoreactivityand possessed the ability to synthesize dopamine) can be "renewed" or restored by intraventricular or intranigral administration of GDNF following its elimination by 6-OHDA or MPTP [16,17,18,19,39]. Furthermore, the rotational behavior and reduction in activity induced by such lesions were also shown to have been alleviated, indicating a possible restoration of motor circuitry. In culture, the restorative properties of GDNF on mesencephalic dopaminergic neurons have also been examined. Not only has GDNF shown to promote restoration of neurites after exposure to MPTP, it has demonstrated stimulant effects on neuronal differentiation [20,21]. Neuroprotection offered by GDNF is not the sole property of the dopaminergic mesencephalon and striatum of the central nervous system. For example, protection and neuronal survival is granted through GDNF to motor neurones following spinal rhizotomy [22], retinal ganglion cells following axotomy [23] and to the photoreceptors of the degenerating rd/rd mouse retina [24]. These data, therefore, warranted the evaluation of GDNF in a clinical scenario. Clinical studies with GDNF GDNF has been evaluated clinically in a placebo-controlled multi-center trial for Parkinson's disease in advanced patients [42]. Administration was achieved using a GDNF reservoir, implanted surgically, into the ventricular system. Patients were monitored primarily for safety and tolerability and secondly, for potential efficacy using the UPDRS and Hoehn and Yahr rating scales. Also need for additional symptomatic medication was

150

E. Ronken el al. /Regulation of GDNF for Dopamine Conversation

checked. Despite promising data from tissue and animal models, initial results from this trial failed to show benefit to patients and was discontinued in April 1999. It may be argued that the relatively high amounts of trophic factors that were used, were unwarranted and may explain the high drop out rate due to lack of efficacy [42]. Moreover, the inclusion of advanced Parkinson patients may well preclude a positive outcome of such trials, suggesting that trophic factor therapy should be further evaluated in early Parkinson patients or as an adjunct to symptomatic therapy [42]. Delivery of GDNF Preclinical studies that would provide an alternative strategy of GDNF delivery include gene therapy with viral vectors either of endogenous brain areas or of cultured cells destined for transplantation. Successful transfection of the striatum with a human GDNF transgene has been accomplished using adenoviral and lentiviral vectors [24,,26,39,40,41]. Injection of lentiviral GDNF into the striatum of aged rhesus monkeys protected THpositive neurons from age-related degeneration, doubled the amount of dopamine in caudate and putamen and increases dopamine turnover threefold. Moreover, lentiviral GDNF completely protected nigral dopamine against MPTP-neurotoxic lesions and improved neurotoxin-induced symptoms in these animals. These observations were comparable to those observed in rat exposed to 6-OHDA and treated subsequently with GDNF delivered by an adenoviral vector system [25]. Transplantation of polymer capsules containing hamster kidney cells engineered to secrete GDNF were also shown to be beneficial to the DA containing cells within the substantia nigra prior to lesion of the medial forebrain bundle [27]. In a study also examining investigating cellular delivery of GDNF through transfected transplant tissue, a titer of GDNF was achieved that was adequate to protect striatal neurones against a quinolate-mediated lesion [28]. Transgene mediated local delivery of GDNF is proving to be a promising alternative to bolus injection in animal models, however, questions over safety and dosage remain unaddressed. Taking into account current understanding of local synthesis and secretion of GDNF in culture and in the brain, it is feasible that a strategy to deliver GDNF may be receptor-mediated in the Parkinsonian brain. GDNF regulation The promoter organization of rat and human GDNF genes have been sequenced providing binding sites for various transcription factors mapped out. Therefore, convergence on the promoter via signal transduction pathways can therefore be studied [29,30,31]. We and others have identified a number of key pathways that induce or repress GDNF expression in rat and human cell lines or in primary astrocyte cultures [32,33,34]. In rat C6 glioma cells, GDNF expression can be stimulated by stimulating protein kinase C and measured by ELISA (Promega, Madison WI, USA [32]) (see fig. 1) but was found insensitive to cAMPdriven pathways. Phorbol ester (PMA)-stimulated expression of GDNF could be blocked completely by the protein synthesis inhibitor cycloheximide and by the mRNA synthesis inhibitor actinomycin D, indicating de novo synthesis of immunoreactive GDNF. Moreover, all synthesized GDNF was secreted; no vesicular storage is apparent in C6 cells. In contrast to C6 glioma cells, the human astrocytoma cell line SK-N-AS, GDNF can be stimulated by increasing cAMP through forskolin or use of dibutyryl-cAMP. Moreover. this can be mimicked by prostaglandins PGE2 and PGI2 [35]. Interestingly, basal GDNF expression could be attenuated by activating PKC by PMA or by stimulating cells by TNFa or Interleukin-lß.

E. Ronken et al. /Regulation of GDNF for Dopamine Conversation

151

Fig. 1: Regulation of GDNF expression in C6 glioma cells by the phorbol ester PMA (A) and SK-N-AS astrocytoma cells by the cAMP-promoting agent forskolin (B).

These results indicate that in SK-N-AS cells appropriate stimulation of receptors that are positively linked to cAMP production, acting through protein kinase A, may facilitate GDNF synthesis and release. Thus, at least three completely distinct signal transduction pathways can be further discriminated that converge onto GDNF gene regulation. We have corroborated differences in the ability to stimulate GDNF expression that are cell line specific (e.g. human vs. rat, or glioma vs. astrocytoma; table 1 [32,35]. Therefore, by identifying receptors that are present on CNS compartments that express GDNF and are linked to the relevant and cell-specific signal transduction pathways, it may be possible to upregulate endogenous levels of GDNF to a concentration that would contribute to maintaining or even restoring dopaminergic innervation of the caudate nucleus and the putamen as established both in vitro and in vivo in animal models. Validity of receptor targeted, GDNF mediated neuroprotection was presented in a model of ischemic brain injury. Here, Wang et al., have demonstrated that the vitamin D mediated reduction in the extend of cortical infarction was concurrent with an increase in cortical GDNF [36].

152

E. Ronken et al. /Regulation of GDNF for Dopamine Conversation

Cell line C6

Compound

pEC50 PMA PIC50

SK-N-AS

Staurosporine Cycloheximide Actinomycine D

7.46 7.20 8.66

Forskoline

pEC50 6.11

PMA TNF-a IL-1ß

PIC50 5.52 6.89 8.72

LPS

IC50 (ng/ml) 25,1

Table 1: summary of concentration-response relationships between incubation conditions and GDNF expression in two ceil lines. pEC50 values are the -log concentrations at which 50% of maximal achievable effect of that compound is obtained; pIC50 values denote the -log concentration at which the effects of 10 -7M PMA are attenuated by 50% (in C6 cells) or by l0-6M forskolin (SK-N-AS).

Identification of more receptors that are coupled to GDNF expression is currently ongoing. These efforts involve assessment of several receptors that are known to be present on striatal GABAergic interneurons (see below). In addition, an active orphan receptor program for identifying novel G-protein coupled receptors by homology cloning has yielded at least one new orphan G-protein coupled receptor that is located in the CNS of rat and man. Moreover, in the human brain it is localized exclusively in the caudate nucleus and the putamen. Obviously, these receptors are key candidate receptors to design specific pharmacotherapy that acts exclusively in the striatum. There, they potentially regulate expression of trophic factors that allow maintenance of dopaminergic innervation or facilitate reinnervation of dopaminergic axons. Alternatively, due to its exclusive localization, this and other receptors may also be of relevance for addressing motor disorders. GDNF distribution Cells expressing GDNF mRNA have been localized in rat brain using in situ hybridisation [37,38]. We investigated the localization of GDNF immunoreactivity by immunofluorescence on rat brain vibratome sections (Primary antibodies to GDNF; R&D systems), primary polyclonal anti-GAD67 was supplied by Chemicon used at 1:1000 and anti-TH was supplied by MP Products and used at 1:500; secondary antibody (1:1000) raised in donkey and streptavidin conjugated with a fluorochrome (1:5000) were supplied by Jackson). GDNF immunoreactivity was found to be present throughout the rat brain, including cortex, hippocampus, striatum and substantia nigra. Notably, in the striatum, GDNF immunoreactivity was found predominantly in the matrix and was colocalized with GAD67 immunoreactivity, a marker for GABAergic neurons. In the substantia nigra, GDNF was found colocalized with TH immunoreactivity. the marker for dopaminergic

E. Ronken et al. /Regulation of GDNF for Dopamine Conversation

\ 53

neurons (fig. 2). These data corroborate findings from in situ hybridisation published recently [38].

Fig. 2: Immunofluorescence staining of the rat substantia nigra (left) and the rat striatum (right). Fluorescein conjugated streptavidin was employed for GDNF staining whilst rhodamine conjugated secondary antibodies were used for the colocalising antibody. The top two panels are colocalisation experiments for tyrosine hydroxylase (TH) and the lower two panels are for glutamate decarboxylase (GAD67).

The role of striatal production of GDNF may provide target-derived neurotrophic communication with afferent nuclei. Neutralisation of this signal and its effect on nigrostriatal integrity has not been attempted. However, when exogenous GDNF was injected into the striatum, it underwent retrograde transport to dopaminergic cell bodies after receptor-mediated internalization [17,18,39]. Moreover, when lentiviral GDNF is expressed in striatum, it increases striatal dopamine content and turnover and increases the volume of TH expressing neurons in the substantia nigra [16,25,39]. We can therefore postulate that a retrograde neurotrophic signal, deriving from the striatum, is crucial for TH expression in the substantia nigra. Current studies address the colocalization with calcium-binding proteins to identify the subset of striatal GDNF-containing GABAergic neurons. Moreover, investigations into signal transduction pathways in striatal tissue, specifically GABAergic neurons in the matrix, are ongoing. Identification of neurotransmitter and neuropeptide receptors, linked to these pathways, will allow vitro evaluation of receptor-mediated manipulation of GDNF expression.

154

E. Ronken et al. /Regulation of GDNF for Dopamine Conversation

In the substantia nigra, GDNF may have an autocrine function. Thus, increase of GDNF expression in this region is expected to have neuroprotective effects as well, similar to what has been observed upon GDNF manipulations within the striatum. Indeed, successful transfection of the substantia nigra with GDNF vector plasmids has proved beneficial in relieving the effects of basal ganglia lesions. However, the lack of re-establishment of striatal DA in these studies, raises questions on the mechanism of action. Discussion GDNF has been found to support dopaminergic neurons in vitro and in vivo. Moreover, GDNF expresses profound neuroprotective properties in animal models for Parkinson's disease. Moreover, GDNF can partially revert neurotoxin-induced damage in these animal models. Disappointingly, while GDNF has shown consistent and potent effects in treating Parkinsonian symptoms in animal models, it was not found effective in actual Parkinson patients when studied in clinical trials. Although the exact cause is unknown, one possible cause is the delivery of the trophic factors, resulting in safety problems. Several delivery strategies are under consideration, including encapsulated cell lines that secrete biologically active GDNF [27] or by using gene therapy with viral vectors, encoding GDNF [25,41]. Based on the neuroprotective effects of GDNF and its potential use in Parkinson's disease, we set out to investigate the options to regulate expression of GDNF locally in the brain. By locally regulating GDNF production, a physiological ceiling will be imposed on the system, thereby avoiding the risk of receptor overstimulation and potential damage to DA or other cell types that were to be protected in the first place. GDNF was detected in relevant rat brain areas for studying the nigrostriatal dopamine system. Target-derived neurotrophic support by retrograde transport of upregulated striatal GDNF will exert neuroprotective effects in animal lesion models for Parkinson's disease, whereas nigral GDNF production will to a large part depend on the continued presence of dopaminergic neurons. Localization of GDNF immunoreactivity needs to be confirmed in human brains as well to evaluate applicability in normal and aging brains as well as in Parkinsonian patients. Using cell lines, we and other established that GDNF is subject to regulation through signal transduction pathways. This is currently being evaluated for brain tissue as well and will allow identification of suitable receptor systems to assess acute and chronic exposure of agonists or antagonists on the status of dopaminergic neurons, i.e. complexity. DA turnover and resistance to neurotoxins. Together, these approaches may be used as novel approaches to design causal treatment for Parkinson's disease and in addition the concept may be extended to other neurodegenerative processes such as Alzheimer's disease, Huntington's disease and that induced by traumatic brain injury and stroke. References [ I] [2] [3] [4] [5]

R.D.Adams and M. Victor, Principles of Neurology. 5th edition. McGraw-Hill Inc. 1993. E. Hirsch et al., Melanized dopaminergic neurons are differentially susceptible to degeneration in Parkinson's disease. Nature, 334 (1988) 345-348. H Bernheimer et al., Brain dopamine and the syndromes of Parkinson and Huntington. Clinical, morphological and neurochemical correlations. J Neurol Sci 20 (1973) 415-455. P. Riederer and S. Wuketich, Time course of nigrostriatal degeneration in Parkinson's disease. A detailed study of influential factors in human brain amine analysis. J Neural Transm. 38 (1976) 277 301. L.-F.H. Lin et al., GSNF: a glial cell line-derived neurotrophic factor for midbrain dopaminergic neurons. Science 260 (1993) 1130-1132.

E. Ronken et al. /Regulation of GDNF for Dopamine Conversation

[6] [7] [8] [9] [10] [11] [12]

[13] [14] [15] [16] [17] [18] [19]

[20]

[21]

[22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33]

155

L.-F.H. Lin et al., Purification and initial characterization of rat B49 glial cell line-derived neurotrophic factor. J Neurochem 63 (1994) 758-768. D. Casper, Growth factors and dopaminergic neurons. In: CNS neurotransmitters and neuromodulators: Dopamine, T.W. Stone, Ed. CRC press, chapter 9: pp. 131-163 (1996). J. Engele et al., Conditioned media derived from glial cell lines promote survival and differentiation of dopaminergic neurons in vitro: role of mesencephalic glia. J Neurosci Res 30 (1991) 359-371. P.T. Kotzbauer et al., Neurturin, a relative of glial cell line-derived neurotrophic factor. Nature 384 (1996)467-470. J. Milbrandt et al., Persephin, a novel neurotrophic factor related to GDNF and neurturin. Neuron 20 (1998)245-253. R.H Baloh et al, Artemin, a novel member of the GDNF family, supports peripheral and central neurons and signals through the GFRalpha3-RET receptor complex. Neuron 21 (1998) 1291-1302. S. Masure et al., Enovin, a member of the glial cell line-derived neurotrophic factor (GDNF) family with growth promoting activity on neuronal cells. Existence of tissue-specific expression of different splice variants. Eur J Biochem 266 (1999) 892-902. U. Ungerstedt, 6-Hydroxydopamine induced degeneration of central monoamine neurons. Eur J Pharmacol 5 (1968) 107-110. J.W. Langston et al., Chronic parkinsonism in humans due to a product of meperidine-analog synthesis. Science 219 (1983) 979-980. M. Gerlach et al., MPTP mechanisms of neurotoxicity and their implications for Parkinson's disease. Eur J Pharmacol 208 (1991) 273-286. B.J. Hoffer et al., Glial cell line-derived neurotrophic factor reverses toxin-induced injury to midbrain dopaminergic neurons in vivo. Neurosci Lett 182 (1994)107-111. A. Tomac et al., Protection and repair of the nigrostriatal dopaminergic system by GDNF in vivo. Nature 373 (1995) 335-339. A. Tomac et al, Retrograde transport of glial cell line-derived neurotrophic factor in the adult nigrostriatal system suggests a trophic role in the adult. Proc NatlAcadSci 92 (1995) 8274-8278. K.E. Bowenkamp et al, Intracerebroventricular glial cell line-derived neurotrophic factor improves motor function and supports nigrostriatal dopamine neurons in bilaterally 6-hydroxydopamine lesioned rats. Exp Neurol 145 (1997)104-17. J.G. Hou et al., Glial cell line-derived neurotrophic factor (GDNF) exerts neurotrophic effects on dopaminergic neurons in vitro and promotes their survival and regrowth after damage by l-methyl-4phenylpiperidinium (MPP+). J Neurochem 66 (1996) 74-82. H.R. Widmer et al, Glial cell-derived neurotrophic factor stimulates the morphological differentiation of cultured ventral mesencephalic calbindin- and calretinin-containing neurones. Exp Neurol 164 (2000)71-81. R.W. Oppenheim et al, Developing motor neurones rescued from programmed cell death and axotomy-induced cell death by GDNF. Nature 373 (1995) 344-346. N. Klocker et al, In vivo neurotrophic effects of GDNF on axotomized retinal ganglion cells. Neuroreport 8 (1997) 3439-42. M. Frasson et al., Glial cell line-derived neurotrophic factor induces histologic and functional protection of rod photoreceptors in the rd/rd mouse. Invest Ophthalmol Vis Sci 40 (1999) 2724-2734. J.H. Kordower et al, Neurodegeneration prevented by lentiviral vector delivery of GDNF in primate models of Parkinson's disease. Science 290 (2000) 767-73. J.H. Kordower JH and P. Aebischer, Gene therapy to the rescue in Parkinson's disease. Trends Pharmacol Sci. 22 (2001) 105-106. J.L. Tseng et al, GDNF reduces drug-induced rotational behavior after medial forebrain bundle transection by a mechanism not involving striatal dopamine. J Neurosci 17 (1997) 325-333. E. Perez-Navarro et al., Glial cell-derived neurotrophic factor protects striatal calbindinimmunoreactive neurons from excitotoxic damage. Neuroscience 75 (1996) 345-352. P.A. Baecker et al, Characterization of a promoter for the human glial cell line-derived neurotrophic factor gene. Mol Brain Res 69 (1999) 209-222. L. Grimm et al, Analysis of the human GDNF gene reveals an inducible promoter, three exons, a triplet repeat within the 3'-UTR and alternative splice products. Hum Molec Genet 7 (1998) 18731886. N. Matsushita et al, Cloning and structural organization of the gene encoding the mouse glial cell linederived neurotrophic factor, GDNF. Gene 203 (1997) 149-157. A.N. Verity et al, Regulation of glial cell line-derived neurotrophic factor release from rat C6 glioblastoma cells. J Neurochem 70 (1998) 531-539. E. Appelet al., Regulation of GDNF in cultured astrocytes by inflammatory stimuli. Neuroreport 8 (1997)3309-3312.

156

[34] [35] [36] [37]

[38] [39] [40] [41] [42]

E. Ronken et al. /Regulation of GDNF for Dopamine Conversation

B. Wiesenhofer et al., Glial cell line-derived neurotrophic factor (GDNF) and its receptor (GFR-al) are strongly expressed in human gliomas. Acta Neuropathol 99 (2000) 131 -137. A.N. Verity et al., Differential regulation of 3glial cell line-derived neurotrophic factor (GDNF) expression in human neuroblastoma and glioblastoma cell lines. J Neurosci Res 55 (1999) 187-197. Y. Wang et al., Vitamin D3 attenuates cortical infarction induced by middle cerebral arterial ligation in rats. Neuropharmacology 39 (2000) 873-880. M. Trupp et al., Complementary and overlapping expression of Glial cell-derived neurotrophic factor, c-Ret protooncogene and GDNF receptor-a indicates multiple mechanisms of trophic actions in the adult rat CNS. J Neurosci 17 (1997) 3554-3567. J.L. Bizon et al., Subpopulations of striata! interneurones can be distinguished on the basis of neurotrophic factor expression JCN 408 (1999) 283-298. H. Sauer et al., Glial cell line-derived neurotrophic factor but not transforming growth factor ß3 prevents delayed degeneration of nigral dopaminergic neurons following striatal 6-hydroxydopamine lesion. Proc NatlAcadSci 92 (1995) 8935-8939. D.M. Gash et al., Functional recovery in Parkinsonian monkeys treated with GDNF. Nature 380 (1996)252-255. P.A. Lapchak et al., Adenoviral vector-mediated GDNF gene therapy in a rodent lesion model of late stage Parkinson's disease. Brain Res 777 (1997) 153-160. Drug Report, IDdb abstract 322075: Report on discontinuation of GDNF trials in PD patients. 1999.

157

Concluding Remarks The 1st Solvay Pharmaceuticals Conference (25-27 October 2000, Como, Italy) was devoted to Parkinson's disease, aiming to update current views with recently emerged opportunities both from the clinic and from preclinical science. Dr. Brooks provided the keynote lecture, giving an indepth view on the problems with classification of movement disorders and the use of imaging techniques to assist in classifying such disorders as well as its potential to identify "surrogate" markers for neuroprotection. This was extended further by Dr Bronstein, who discussed a number of examples of clinical trials that try as best as is allowed to use existing classification schemes to identify neuroprotective activity, e.g. using trophic factors. However, as Dr Brandtstatter and Dr Oertel illustrated, the quality of life of patients, from Parkinson's disease is in a significant part caused by comorbid depression. As such, more than 60% of the patients seem to suffer from Major Depressive Disorder, which can be treated by currently available antidepressants. To study potentially effective symptomatic therapy, predictive animal models need to be available. Fortunately, research on Parkinson's disease has several different animal models at its disposal, all of them predictive for measuring efficacy of dopaminergic agonists, as Dr Jenner presented in his contribution. Rodent models, lesioned with 6hydroxydopamine or MPTP, provide models with good face validity, construct validity as well as predictive validity. Moreover, a primate model, such as the MPTP-lesioned common Marmoset offers additional clinical signs - e.g. disability scores, gait, and posture - that can be used in a predictive manner for evaluating prospective drug therapy. Interestingly, the MPTP-lesioned Marmoset was found very sensitive also to partial dopamine D2 receptor agonists, indicating that therapy does not need to involve full agonism with concomitant systemic burden, to achieve clinical efficacy. Dr Earl took the rodent models one step further in employing these animal models for assessing neuroprotective or neurotrophic strategies, thereby making the jump from symptom management to disclosing potential causal/neuroprotective therapy. The contributions on symptom management were completed by Dr Richardson, who provided a completely different target, the adenosine A2a receptor, which, if blocked by suitable receptor antagonists, can mimic the activities of the archetypal dopamine D2 receptor agonists. Thus, employing new insights regarding neuronal networks, previously described by Dr Standaert, other targets for future therapy can be identified and found equally effective in animal models, previously described by Drs Jenner and Earl. Finally, these animal models were found to be of use by Dr Caviedes to study the potential beneficial use of xenografting of cells, prepared from primary dopaminergic cells. Thus, implantation of such cells, immortalized using an original transformation protocol, effectively counteracted the contralateral turning behavior of rats that were previously lesioned on one side with 6-OHDA. However challenging symptomatic therapy is, the approaches seem not to be directed at conserving remaining dopaminergic neurons. The progressive decline of available dopaminergic neurons in the substantia nigra will ultimately render patients in an unacceptable condition, in which pharmacotherapy will be ineffective. Fortunately, a number of new promising approaches aiming at halting the neurodegeneration are pursued. By assuming that dopaminergic neurons in Parkinsonian patients suffer from oxidative stress, mechanisms to counteract reactive oxygen species have been investigated.

158

Concluding Remarks

Dr Drukarch and co-workers provided a case for induction of phase II biotransformation enzymes that allow cells to generate more glutathione, thereby decreasing their vulnerability to repeated oxidative challenges. Moreover, this may be done by activating specific and conserved promoter elements, known to respond to antioxidant stress, thereby allowing completely novel modes of long-term neuroprotective therapy. These issues were elegantly supported by Dr Beal, providing compelling evidence for bioenergetic malfunction and its consequences for neuronal function when energy supply is suboptimal. A significant fraction of Parkinsonian patients suffer from mutations in complex I of the mitochondrial respiratory chain. Thus, systems that challenge this system will contribute to neurotoxin-induced degeneration, whereas approaches that are designed to circumvent or supplement the identified defects, offer better resistance against the same neurotoxins. In this line of thinking, Dr Mount and co-workers have been active in another neurodegenerative disorder, ataxia-telangiectasia (AT), a disorder that is caused by a point mutation in the AT gene product (atm). This mutation affects signal transduction of NGF trophic factor receptors, yielding specific degeneration of cerebellar Purkinje cells. Although Parkinson's disease does not seem to have a specific genetic cause, manipulations of trophic factor expression and signalling seem to offer opportunities for the treatment of neurodegenerative disorders. Dr Schlegel and co-workers, in their presentation, focused on the signal transduction of one particular trophic factor, glial cell line-derived neurotrophic factor (GDNF), which was recently identified to convey potent and selective neurotrophic and neuroprotective effects at dopaminergic cells in vitro and in vivo in animal models. Dr Zurn and Dr Aebischer build on the potential of GDNF by constructing encapsulated cell lines that are capable of continuously secreting low levels of GDNF in vivo after implantation. Impressive results were presented, which convincingly indicated that GDNF approaches are neurorestorative in animal models for Parkinson's disease. However, in preliminary patient studies, Dr Bronstein discussed the problems associated with exogenous administration of GDNF to human Parkinson patients, indicating that GDNF concentrations surpassing the physiological ceiling levels induce adverse events causing a massive drop-out rates in clinical trials. This and other evidence led Solvay scientists to another approach to increase GDNF levels in the brain, namely by activating local cell sources that are present in relevant brain areas. This can be achieved by conventional small molecule based pharmacological approaches, which trigger the endogenous production of GDNF or other neuroprotective factors. In summary, the meeting was focused on the preclinical and clinical sides of both symptom alleviation as well as on evaluating options for future causal therapy for Parkinson's disease, aiming at neuroprotection. Several discussions pointed out the importance of the use of animal models and transgenics, control of trophic factors and the need for better (surrogate) markers for measuring neuroprotection in early patients. Furthermore, rethinking of maintaining cellular resistance to noxious stimuli either by inducing genes that support this requirement or alternatively find ways to improve the bioenergetic equilibrium of specific systems was encouraged. As such, the outcome of this meeting is valid for other neurological disorders, such as Huntington's disease, Alzheimer's disease and multiple sclerosis. E. Ronken and G.J.M. van Scharrenburg

159

Author Index Aebischer, P. Arriagada, C. Beal, M.F. Bi,Q. Bloch, J. Brandstadter, D. Bronstein, J.M. Brooks, D.J. Brown, W.D. Carvey, P. Cautreels, W. Caviedes, P. Caviedes, R. Chen, E.-Y. Chu, Y. Crawford, T.O. Deglon, N. Drukarch, B. Earl, C. Eggert, K. Emborg, M.E. Flier, J. Fluit, P. Hantraye, P. Herremans, A. Holden, J.E. Jenner, P. Klockgether, T. Kordower, J.H. Leventhal, L. Ling, Z.D.

133 120 73 94 133 29 143 5 133 133 v 120 120 133 133 94 133 105 88 115 133 105 94 133 51 133 39 81 133 133 133

Long, S. Ma,S.Y. Mandir, A.S. McBride,J. McCreary, A.C. McCrossan, D. Mount, H.T.J. Neff.F. Oertel,W.H. Palfi,S. Pyzalski, R. Richardson, P.J. Roitberg, B.Z. Ronken, E. Salazar,J. Sautter, J. Schlegel,J. Shimahara, T. Standaert, D.G. Steinborn, C. Taylor, M.D. Trono,D. Tuinstra, T. Turski, L. van der Heyden, J. van Muiswinkel, F.L. van Scharrenburg, G.J.M. Venema, J. Wu, Y. Wullner, U.

51 133 94 133 51 148 94 115 29 133 133 59 133 3,51,148,157 120 88 115 120 17 v 133 133 51 v 51 105 3,51,157 148 94 81