nature neuroscience november 2005

www.nature.com/natureneuroscience

EDITORIAL OFFICE [email protected] Park Avenue South, New York, NY 10010-1707Tel: (212) 726 9319, Fax: (212) 696 0978Editor: Sandra Aamodt Senior Editor: Kalyani Narasimhan Associate Editors: Cara Allen, I-han Chou, Annette Markus, Charvy NarainCopy Editor: Dorothy Moore, Lavanya ReddySenior Production Editor: Nicole D. FournierProduction Editor: Jessica IannuzziCover Design: Kimberly CaesarEditorial Assistant: Jessica Chen

MANAGEMENT OFFICESNPG New York345 Park Avenue South, New York, NY 10010-1707Tel: (212) 726 9200, Fax: (212) 696 9006Publisher: Beatrice RenaultExecutive Editor: Linda MillerChief Technology Officer: Howard RatnerGlobal Head of Advertising: Fabien SavenayHead of Nature Research & Reviews Marketing: Sara GirardAssistant Marketing Manager: Amy MaurerProduction Coordinator: Diane TempranoAssociate Director, Web Publishing: Timo HannayNew Technology Director: Greg SuprockHead of Web Services: Anthony BarreraWeb Production Editor: Edwina Hay

NPG LondonThe Macmillan Building, 4 Crinan Street, London N1 9XWTel: 44 207 833 4000, Fax: 44 207 843 4996Managing Director: Annette ThomasPublishing Director: Peter CollinsEditor-in-Chief, Nature Publications: Philip CampbellMarketing Director: Della Sar

NPG TokyoChiyoda Building 2-37 Ichigayatamachi, Shinjuku-ku, Tokyo 162-0843Tel: 81 3 3267 8751, Fax: 81 3 3267 8746Asia-Pacific Publisher: Antoine E. BocquetManager: Koichi NakamuraSenior Marketing Manager: Peter YoshiharaAsia-Pacific Sales Director: Kate YoneyamaAsia-Pacific Sales Manager: Rinoko Asami

DISPLAY ADVERTISING [email protected] (US/Canada) [email protected] (Europe) [email protected] (Japan)US Head of Display Advertising: Stephen Schwartz, Tel: (212) 726 9256, Fax: (212) 696 9481Global Head of Display Advertising Sales: John Michael, Tel: 44 207 843 4960, Fax: 44 207 843 4996Head of Display AdvertisingEurope: Gerard Preston, Tel: 44 207 843 4960, Fax: 44 207 843 4996Business Development Manager: Claire Hines, Tel: 44 207 843 4960, Fax: 44 207 843 4996Asia-Pacific Sales Manager: Rinoko Asami, Tel: 81 3 3267 8751, Fax: 81 3 3267 8746Western Region Sales Manager: George Lui, Tel: (415) 781 3804, Fax: (415) 781 3805Sales Executives:New England: Sheila Reardon, Tel: (617) 399 4098, Fax: (617) 426 3717New York, Mid-Atlantic, Southeast: Jim Breault, Tel: (212) 726 9334, Fax: (212) 696 9481Midwest: Mike Rossi, Tel: (212) 726 9255, Fax: (212) 696 9481Northwest: Mathieu DesJardins, Tel: (415) 781 6422, Fax: (415) 781 3805Eastern England/Scotland/Italy/Spain/Israel: Matthew Wilkinson, Tel: 44 207 843 4960, Fax: 44 207 843 4749South/West United Kingdom/Scandinavia/Holland: Marianne Boulakas, Tel: 44 207 843 4969, Fax: 44 207 843 4749Northern Germany: Gerard Preston, Tel: 44 207 843 4960, Fax: 44 207 843 4749Southern Germany/Austria/Switzerland/France/Belgium: Sabine Hugi-Frst, Tel: 41 52761 3386, Fax: 41 52761 3419

NATUREJOBS [email protected] (US/Canada) [email protected] (Europe) [email protected] (Japan)Publisher: Ben Crowe, Tel: (212) 726 9245, Fax: (212) 696 9482US Sales Manager: Peter Bless, Tel: (212) 726 9248, Fax: (212) 696 9482Japan Sales Manager: Rinoko Asami, Tel: 81 3 3267 8751, Fax: 81 3 3267 8746naturejobs Sales Director: Nevin Bayoumi, Tel: 44 207 843 4961, Fax: 44 207 843 4996

SITE LICENSE BUSINESS UNITAmericas: Tel: (888) 331 6288 [email protected]/Pacific: Tel: 81 3 3267 8751 [email protected]/New Zealand: Tel: 61 3 9825 1160 [email protected]: Tel: 00 91 11 324 4186 [email protected]: Tel: 44 207 843 4609 [email protected]

CUSTOMER SERVICE www.nature.com/helpSenior Global Customer Service Manager: Gerald CoppinFor all print and online assistance, please visit www.nature.com/helpPurchase subscriptions:Americas: Nature Neuroscience, Subscription Dept., 303 Park Avenue South #1280, New York, NY 10010-3601. Tel: (866) 363 7860, Fax: (212) 689 9108Europe/ROW: Nature Neuroscience, Subscription Dept., Macmillan Magazines Ltd., Brunel Road, Houndmills, Basingstoke RG21 6XS, United Kingdom. Tel: 44 1256 329 242, Fax: 44 1256 812 358Japan: Nature Neuroscience, Nature Japan K.K., MG Ichigaya Bldg. 5F, 19-1 Haraikatamachi, Shinjuku-ku, Tokyo 162-0841. Tel: 81 3 3267 8751, Fax: 81 3 3267 8746India: Harpal Singh Gill, Macmillan Magazines Ltd, 5A/12 Ansari Road, Darya Ganj, New Delhi, 110 002 India. Tel: 00 91 11 324 4186, Tel/Fax: 00 91 11 327 2010

REPRINTS [email protected] Neuroscience Reprint Department, Nature Publishing Group, 345 Park Avenue South, New York, NY 10010-1707, USA.For commercial reprint orders of 600 or more, please contact:UK Reprints Sales Executive: Christine Fothergill, Tel: 44 207 843 4967, Fax: 44 207 843 4749US Reprints Sales Executive: Sharda Tulsie, Tel: (212) 726 9631, Fax: (212) 679 0843

20

05 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce

i

VOLUME 8 NUMBER 11 NOVEMBER 2005

Nature Neuroscience (ISSN 1097-6256) is published monthly by Nature Publishing Group, a trading name of Nature America Inc. located at 345 Park Avenue South, New York, NY 10010-1707. Periodicals postage paid at New York, NY and additional mailing post offices. Editorial Office: 345 Park Avenue South, New York, NY 10010-1707. Tel: (212) 726 9319, Fax: (212) 696 0978. Annual subscription rates: USA/Canada: US$199 (personal), US$1,809 (institution). Canada add 7% GST #104911595RT001; Euro-zone: 271 (personal), 1,558 (institution); Rest of world (excluding China, Japan, Korea): 175 (personal), 1,005 (institution); Japan: Contact Nature Japan K.K., MG Ichigaya Building 5F, 19-1 Haraikatamachi, Shinjuku-ku, Tokyo 162-0841. Tel: 81 (03) 3267 8751, Fax: 81 (03) 3267 8746. POSTMASTER: Send address changes to Nature Neuroscience, Subscriptions Department, 303 Park Avenue South #1280, New York, NY 10010-3601. Authorization to photocopy material for internal or personal use, or internal or personal use of specific clients, is granted by Nature Publishing Group to libraries and others registered with the Copyright Clearance Center (CCC) Transactional Reporting Service, provided the relevant copyright fee is paid direct to CCC, 222 Rosewood Drive, Danvers, MA 01923, USA. Identification code for Nature Neuroscience: 1097-6256/04. Back issues: US$45, Canada add 7% for GST. CPC PUB AGREEMENT #40032744. Printed by Publishers Press, Inc., Lebanon Junction, KY, USA. Copyright 2005 Nature Publishing Group. Printed in USA.

E D I TO R I A L1413 Taking addiction research into the clinic

B O O K R E V I E W1415 The War of the Soups and the Sparks

By Elliot S ValensteinReviewed by Nicholas C Spitzer

N E W S A N D V I E W S1417 How neurons keep in touch

Fekrije Selimi & Nathaniel Heintz see also p 1534

1418 Synaptic plasticity and self-organization in the hippocampusGyrgy Buzski & James J Chrobak see also p 1560

1420 Glial cells under remote controlKlaus-Armin Nave & Markus H Schwab

1422 Senseless makes sense for spinocerebellar ataxia-1Vikram Khurana, Tudor A Fulga & Mel B Feany

1424 How the brain recovers following damageYalin Abdullaev & Michael I Posner see also p 1603

1425 Time to smell the rosesCara Allen see also p 1568

I N T R O D U C T I O N : N E U R O B I O LO G Y O F A D D I C T I O N1427 Neurobiology of addiction

I-han Chou & Kalyani Narasimhan

S P O N S O R S F O R E WO R D : N E U R O B I O LO G Y O F A D D I C T I O N1429 The neuroscience of addiction

Nora Volkow & Ting-Kai Li

Semaphorin controls cerebellar granule cell migration

(p 1516)

Addiction is pervasive, affecting millions of people around the world.

The progression from recreational drug use to drug dependence and

addiction is influenced by many factors, including the nature of the

drug, the personality of the user, and environmental stressors. In this issue, we present reviews and opinion pieces

on the neurobiology of drug abuse, decision making and habit formation,

as well as a commentary on how the neuroscience of addiction should

guide public policy and treatment. This special focus is sponsored by the National Institute on Drug Abuse and

the National Institute on Alcohol Abuse and Alcoholism.

(pp 14271489)

20

05 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce

iii


NATURE NEUROSCIENCE

COM M E N TA R I E S : N E U R O B I O LO G Y O F A D D I C T I O N1431 Neurobiology of addiction: treatment and public policy ramifications

Charles Dackis & Charles OBrien

1437 The role of neuroadaptations in relapse to drug seekingYavin Shaham and Bruce T Hope

1440 How do we determine which drug-induced neuroplastic changes are important?Peter W Kalivas

1442 Plasticity of reward neurocircuitry and the dark side of drug addictionGeorge F Koob & Michel Le Moal

P E R S P E C T I V E S : N E U R O B I O LO G Y O F A D D I C T I O N1445 Is there a common molecular pathway for addiction?

Eric J Nestler

1450 Genetic influences on impulsivity, risk taking, stress responsivity and vulnerability to drug abuse and addictionMary Jeanne Kreek, David A Nielsen, Eduardo R Butelman & K Steven LaForge

1458 Decision making, impulse control and loss of willpower to resist drugs: a neurocognitive perspectiveAntoine Bechara

R E V I E W S : N E U R O B I O LO G Y O F A D D I C T I O N1465 Nicotine addiction and comorbidity with alcohol abuse and mental illness

John A Dani and R Adron Harris

1471 Laboratory models of alcoholism: treatment target identification and insight into mechanismsDavid M Lovinger & John C Crabbe

1481 Neural systems of reinforcement for drug addiction: from actions to habits to compulsionBarry J Everitt & Trevor W Robbins

B R I E F COM M U N I C AT I O N S1491 The cerebellum communicates with the basal ganglia

E Hoshi, L Tremblay, J Fger, P L Carras & P L Strick

1494 Shift of activity from attention to motor-related brain areas during visual learningS Pollmann & M Maertens

1497 The essential role of stimulus temporal patterning in enabling perceptual learningS-G Kuai, J-Y Zhang, S A Klein, D M Levi & C Yu

1500 COMT genotype predicts longitudinal cognitive decline and psychosis in 22q11.2 deletion syndromeD Gothelf, S Eliez, T Thompson, C Hinard, L Penniman, C Feinstein, H Kwon, S Jin, B Jo, S E Antonarakis, M A Morris & A L Reiss

A R T I C L E S1503 MPS-1 is a K+ channel -subunit and a serine/threonine kinase

S-Q Cai, L Hernandez, Y Wang, K H Park & F Sesti

20

05 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce

v


NATURE NEUROSCIENCE

Activity-dependent decrease in global excitability

(p 1542)

Binocular rivalry in the lateral geniculate nucleus

(p 1595)

1510 Lbx1 and Tlx3 are opposing switches in determining GABAergic versus glutamatergic transmitter phenotypesL Cheng, O Abdel Samad, Y Xu, R Mizuguchi, P Luo, S Shirasawa, M Goulding & Q Ma

1516 The transmembrane semaphorin Sema6A controls cerebellar granule cell migrationG Kerjan, J Dolan, C Haumaitre, S Schneider-Maunoury, H Fujisawa, K J Mitchell & A Chdotal

1525 TARP -8 controls hippocampal AMPA receptor number, distribution and synaptic plasticityN Rouach, K Byrd, R S Petralia, G M Elias, H Adesnik, S Tomita, S Karimzadegan, C Kealey, D S Bredt & R A Nicoll

1534 Cbln1 is essential for synaptic integrity and plasticity in the cerebellumH Hirai, Z Pang, D Bao, T Miyazaki, L Li, E Miura, J Parris, Y Rong, M Watanabe, M Yuzaki & J I Morgan see also p 1417

1542 Activity-dependent decrease of excitability in rat hippocampal neurons through increases in IhY Fan, D Fricker, D H Brager, X Chen, H-C Lu, R A Chitwood & D Johnston

1552 Fine-scale specificity of cortical networks depends on inhibitory cell type and connectivityY Yoshimura & E M Callaway

1560 Induction of sharp waveripple complexes in vitro and reorganization of hippocampal networksC J Behrens, L P van den Boom, L de Hoz, A Friedman & U Heinemann see also p 1418

1568 Encoding a temporally structured stimulus with a temporally structured neural representationS L Brown, J Joseph & M Stopfer see also p 1425

1577 Drosophila melanogaster homolog of Down syndrome critical region 1 is critical for mitochondrial functionK T Chang & K-T Min

1586 Transcriptional and behavioral interaction between 22q11.2 orthologs modulates schizophrenia-related phenotypes in miceM Paterlini, S S Zakharenko, W-S Lai, J Qin, H Zhang, J Mukai, K G C Westphal, B Olivier, D Sulzer, P Pavlidis, S A Siegelbaum, M Karayiorgou & J A Gogos

1595 Neural correlates of binocular rivalry in the human lateral geniculate nucleusK Wunderlich, K A Schneider & S Kastner

1603 Neural basis and recovery of spatial attention deficits in spatial neglect M Corbetta, M J Kincade, C Lewis, A Z Snyder & A Sapir see also p 1424

1611 Perceptions of moral character modulate the neural systems of reward during the trust game M R Delgado, R H Frank & E A Phelps

T E C H N I C A L R E P O R T1619 A hybrid approach to measuring electrical activity in genetically specified neurons

B Chanda, R Blunck, L C Faria, F E Schweizer, I Mody & F Bezanilla

N AT U R E N E U R O S C I E N C E C L A S S I F I E D See back pages

Down Syndrome Critical Region 1 and mitochondrial function

(p 1577)

20

05 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce

NATURE NEUROSCIENCE VOLUME 8 | NUMBER 11 | NOVEMBER 2005 1413

E D I TO R I A L

Taking addiction research into the clinic

As the focus articles in this issue demon-strate, our understanding of the neu-robiology of addiction has progressed substantially over the last decade, largely through studies in rodents. However, despite having some of the best animal models in neu-roscience, researchers have been less success-ful in translating this knowledge into effective therapies. To solve this problem, we need to remove the roadblocks to development and testing of new treatments.

Some medications are available. For alcohol-ism, the opioid receptor antagonist naltrexone and a newer drug, acamprosate, are approved in the US and Europe. (Acamprosate affects GABA and glutamate receptors, but its mecha-nism of action is not fully understood.) Each drug moderately reduces the amount and fre-quency of drinking in clinical trials, and com-bination therapy with both drugs seems to be more effective than either alone. For cocaine addiction, modafinil (an atypical stimulant that is approved for other uses) seems to increase abstinence during treatment1. For heroin addicts, the -opioid receptor partial agonist buprenorphine is available in Europe and the US. Buprenorphine is as effective as methadone (a full agonist that is tightly regu-lated), and has no abuse potential or overdose liability. However, in the US, federal regula-tions prohibit physicians from treating more than 30 patients. Such restrictions are common for treatments that target addiction to illegal drugs, and they unfairly reduce the availability of help for patients who need it most.

Other therapies affect not only drug addiction, but also the response to natural rewards, such as food. This is a potentially serious problem for drug design, as the brain circuits that respond to natural rewards partially overlap with those that respond to addictive drugs. For example, topira-mate, an anticonvulsant that acts at AMPA/kain-ate glutamate receptors, is used for alcoholism and cocaine addiction, but it is also effective against binge eating, suggesting that it may affect responses to food. Rimonabant is a cannabinoid receptor 1 antagonist that reduces the expression of several addictions in animal models, and is

now in clinical trials in humans. However, the compound is also in clinical trials as a treatment for obesity, and seems to promote modest weight loss, so it too is likely to affect eating behavior.

Preliminary results from clinical trials sug-gest that other approaches may be promising. Vaccines against cocaine2 or nicotine seem to reduce drug intake in patients who produce enough antibodies, which then intercept the drug before it reaches receptors in the brain. N-acetylcysteine (a food supplement available over the counter) reduces cocaine addiction without affecting the response to food in rats, and is now being tested in people.

Some animal models of addiction mimic the diagnostic characteristics of the human disor-der very closely, but unfortunately they are too complicated for use in large-scale drug screen-ing. For example, in one model of drug seeking and relapse, rats are trained first to press a lever for access to a drug, and then to work for an intermediate cue that predicts the drugs future availability (as money might do for human addicts). If this behavior fails to produce the drug, the animal eventually will stop press-ing the lever, but later exposure to stress, the addictive drug, or cues associated with the drug can cause the drug-seeking behavior to start again, mimicking relapse in human addicts. This entire assay takes 10 to 12 weeks.

Such models are time-consuming and expen-sive, so investigators typically begin screening for drug-induced neural changes with simpler tests. Promising hypotheses can then be studied with more complex behaviors. The disadvan-tage of this approach is that it does not allow the identification of neural adaptations that are specific to compulsive drug seeking and relapseand such neural changes might be the most directly relevant to addiction treat-ment. For example, drug self-administration by rodents produces distinct effects in the brain that are not seen when the same amount of drug is simply injected by researchers.

Beyond the practical problems involved in identifying targets, pharmaceutical compa-nies lack incentives to develop medicines for addiction. Treating addiction to illegal drugs

raises legal issues, and the common view of addiction as a character defect rather than a neurobiological disorder3 creates a public relations problem. In addition, the addicts who most need treatment are least likely to have jobs or medical insurance. Insurance companies often restrict the availability of treatment, even for the many alcoholics and nicotine addicts who have coverage.

For these reasons, pharmaceutical com-panies are often reluctant to undertake full-scale development of new drugs, or even to release the compounds that they already have for testing in addiction. Instead clinical tri-als for addiction typically involve drugs that are already approved for other uses, which reduces the costs of bringing them to market. In contrast, most drug targets proposed by basic researchers have not been tested prop-erly in the clinic, either because drugs to target these mechanisms do not exist or because the companies that own the compounds have not made them available to researchers.

What can be done to improve the situation? Governments and charitable foundations could provide better incentives for drug develop-ment, such as promising to purchase a certain amount of any effective drug that is developed, or they could purchase candidate compounds from pharmaceutical companies for testing in their own trials. Doctors and insurance com-panies should begin to think of drug addiction as a chronic disease that must be treated over the long term, despite difficulties with patient compliance, like schizophrenia or hypertension. Behavioral choices contribute to many health problemsdiet and exercise to heart disease, for example, or smoking to lung cancerbut we do not refuse these patients medical atten-tion because they are culpable for their illness. A similar attitude toward addicted people would go a long way toward improving their care.

1. Dackis, C.A., Kampman, K.M., Lynch, K.G., Pettinati, H.M. & OBrien, C.P. Neuropsychopharmacology 30, 205211 (2005).

2. Martell, B.A., Mitchell, E., Poling, J., Gonsai, K. & Kosten, T.R. Biol. Psychiatry 58, 158164 (2005).

3. Dackis, C. & OBrien, C. Nat. Neurosci. 8, 14311436 (2005).

20

05 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce


N E W S A N D V I E W S

Figure 1 A model for Cbln1 action in Purkinje cellparallel fiber synapses. Cbln1 is secreted from the parallel fiber bouton, where it acts in combination with Grid2 and other unknown molecules (X) to stabilize the PF-PC interaction. This process involves as-yet unidentified intracellular signaling cascades and possibly retrograde signals to the PF bouton (dashed arrow). The actions of Cbln1 are restricted to the immediate vicinity of the presynaptic terminal owing to its properties as a large, poorly diffusible glycoprotein complex. This counteracts the destabilizing effect of Grid2 activation by a small, diffusible ligand (presumably glutamate (blue arrows)). At sites distal to the actions of Cbln1, activation of Grid2 has a destabilizing effect on postsynaptic structures, perhaps through its ability to control autophagy locally. In this way, the actions of Cbln1 and Grid2 in Purkinje cell spines can control both synapse stabilization and refinement.

How neurons keep in touchFekrije Selimi & Nathaniel Heintz

Cerebellin 1 is abundant in the cerebellum, but its function remains a mystery. Hirai et al. now show that this gene is required to maintain parallel fiber-Purkinje cell synapses, via the orphan glutamate receptor subunit Grid2. These findings provide further evidence that there is a molecular pathway devoted to maintenance of synapses.

Neurons face a classic issue: it is not enough to find the right partner; they also need to work hard to keep it. Forming and maintain-ing the proper synaptic contacts between specific types of neurons is the last step in establishing functional neuronal networks in the brain. Any dysfunction during this processwhich includes formation, matura-tion and elimination of synapsescan lead to cognitive disorders1. Moreover, learning and memory formation requires continuous synaptic plasticity at both the functional and structural level2. Thus, regulating the stability of synapses must be a dynamic process that responds to bidirectional signaling between pre- and postsynaptic elements, altering syn-aptic properties and stability in response to specific cues. Whereas the role of activity in the maintenance of synapses seems clear3, very little is known about the identity of the mole-cules that mediate this particular transsynaptic action. In this issue, Hirai et al.4 demonstrate that the transsynaptic function of the secreted glycoprotein Cbln1 is essential for the stability and plasticity of a specific synapse: the parallel fiberPurkinje cell (PF-PC) synapse.

Cerebellin, discovered two decades ago, is a 16-mer peptide associated with Purkinje cells5. Subsequent studies have focused on the rela-tionship of the full-length cerebellin protein Cbln1 to the C1Q/TNF family of secreted proteins and on its biochemical properties6 but have not yielded significant insight into its role in the cerebellum. In this new study, Hirai

and colleagues describe in detail the phenotype of a mouse lacking Cbln1, providing new and important insight into Cbln1 function. The gross phenotype of Cbln1 knockout mice is unremarkable. They are ataxic, although no obvious abnormality of cerebellar histology is evident in the light microscope. However, when the authors examined the Cbln1 knock-out cerebellum at the electron microscope level, a striking phenotype emerged: 78% of the distal spines that are the site of contact for parallel fibers on Purkinje cells lacked a presynaptic partner. Moreover, detailed analysis of serial sections by electron microscopy showed that a significant fraction of the remaining synapses had mismatched pre- and postsynaptic special-izations. Electrophysiological studies confirmed this result, demonstrating a deficit of transmis-sion between parallel fibers and Purkinje cells.

This phenotype is a pleasant surprise, because it is similar to that of the glutamate receptor 2 (Grid2) knockout mouse. The function of the Grid2 gene, which is specifi-cally expressed in cerebellar Purkinje cells, is not yet fully understood. However, strong evi-

dence that Grid2 is involved in stabilization of PF-PC synapses has accumulated from detailed analysis of the Grid2-null phenotype during development7 and from the observation that inactivation of the Grid2 gene in adult animals also leads to detachment of the parallel fibers from Purkinje cell spines and mismatching of the pre- and postsynaptic specializations8. These results provided strong evidence for the existence of a specific pathway dedicated to governing the structural stability of synapses in the brain, rather than their formation.

Given the phenotypic similarities between Cbln1 and Grid2 knockout mice, the authors produced and analyzed Cbln1/Grid2 double-knockout animals to test the involvement of these two genes in a common pathway. The phenotype of the double mutant animals was identical to that of the Grid2 knockout, demonstrating that these two genes do indeed function as part of the same signaling pathway. Taken together, these data strongly implicate Cbln1 in synapse stabilization in the cerebel-lum, although an additional role for Cbln1 in the development of synapses seems probable.

Fekrije Selimi and Nathaniel Heintz are at the

Laboratory for Molecular Biology, and the Howard

Hughes Medical Institute, The Rockefeller

University, New York, New York 10021, USA.

Fekrije Selimi is also at the Centre National de la

Recherche Scientifique, UMR7102, Paris, France.

e-mail: [email protected]

Reb

ecca

Hen

retta

X

Cbln1

Grid2XGrid2

Stabilization

Glutamatespillover?

Destabilization of PSD

PSD

+

20

05 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce

1418 VOLUME 8 | NUMBER 11 | NOVEMBER 2005 NATURE NEUROSCIENCE


Resolution of this important issue must await further experimentation.

The extensive similarities between the Cbln1 and Grid2 knockout mice, including deficits in cerebellar long-term depression4,9 and genetic evidence that these two genes act in the same pathway, suggested that Cbln1 acts with Grid2 in the postsynaptic density of PC spines. However, Hirai and colleagues showusing a combination of in vitro studies, hybridization in situ and a transgenic approachthat Cbln1 is a glycoprotein secreted by granule cells, and not by Purkinje cells. Thus, Cbln1 and Grid2 are part of a transsynaptic pathway controlling the stability and plasticity of the PF-PC synapse.

The transsynaptic action of Cbln1 is particu-larly interesting given the biochemical proper-ties of other C1q/TNF family proteins10. These proteins form large complexes, which interact with a variety of molecules through their globu-lar C1q domains. They often are important in tissue remodeling. In some cases, this action requires associated protease activitiescer-tainly a handy feature if one hopes to remodel extracellular structures that promote synapse stability. Furthermore, the demonstration that Grid2 is tethered to a molecular complex regu-lating autophagy11 and that activation of Grid2 receptors in Lurcher mice results in stimula-tion of this important catabolic pathway11,12 suggests an intracellular mechanism through which the Cbln1/Grid2 pathway could effect changes in PF-PC synaptic structure.

A very interesting parallel can be made with the neuromuscular junction. Agrin is a glyco-protein secreted by motoneurons at the neuro-

muscular junction13. It was thought to promote synapse formation by clustering acetylcholine receptors. Indeed, agrin knockout mice have very few acetylcholine receptor clusters and lack neuromuscular junctions. However, this phenotype can be rescued by inactivating the gene encoding choline acetyltransferase in those same mice, thus preventing acetylcholine receptor activation. These results, together with in vitro studies, have indicated that the role of agrin is to inhibit the destabilizing effect of the activation of acetylcholine receptors14,15.

Given the results of Hirai et al., and previous studies of Grid2, a similar model (Fig. 1) can be proposed for the roles of the Cbln1/Grid2 signaling pathway at the PF-PC synapse. Grid2 seems to have a dual function. Its regulation of autophagy suggests a destabilizing role for this receptor in PF-PC synapses, whereas the phe-notype of the Grid2 knockout suggests that it is necessary for synapse stabilization and main-tenance. One role for Cbln1 secretion from PF boutons may be to locally inhibit the destabiliz-ing action of Grid2 receptors, as agrin inhibits the destabilizing effects of nAChR activity at the neuromuscular junction. This would pro-mote stabilization of PC-PF contacts immedi-ately adjacent to the active zone. Outside of the active zone, which is presumably free of Cbln1 owing to its limited ability to diffuse out of the cleft, Grid2 would retain its destabilizing actions. This would explain the phenotypes observed by Hirai et al. as well as the role of this new pathway in promoting both stabiliza-tion of the synaptic contact and matching of the pre- and postsynaptic specializations.

Although a great deal of additional infor-mation will be required to understand in detail the mechanisms regulating the struc-tural integrity of central synapses, Hirai et al. have made an important step in identi-fying a mechanism for synapse stabilization that operates transsynaptically in the brain. Their studies reveal a common logic for this important process at the neuromuscular junction and at central synapses, and sug-gest that the transsynaptic actions of large, secreted glycoproteins on neurotransmitter receptors may provide a key function for structural remodeling of these critical CNS structures.

1. Zoghbi, H.Y. Science 302, 826830 (2003).2. Chklovskii, D.B., Mel, B.W. & Svoboda, K. Nature 431,

782788 (2004).3. Hua, J.Y. & Smith, S.J. Nat. Neurosci. 7, 327332

(2004).4. Hirai, H. et al. Nat. Neurosci. 8, 15341541

(2005).5. Slemmon, J.R., Danho, W., Hempstead, J.L. & Morgan,

J.I. Proc. Natl. Acad. Sci. USA 82, 71457148 (1985).

6. Bao, D., Pang, Z. & Morgan, J.I. J. Neurochem. 95, 618629 (2005).

7. Kurihara, H. et al. J. Neurosci. 17, 96139623 (1997).

8. Takeuchi, T. et al. J. Neurosci. 25, 21462156 (2005).

9. Kashiwabuchi, N. et al. Cell 81, 245252 (1995).10. Kishore, U. et al. Trends Immunol. 25, 551561

(2004).11. Yue, Z. et al. Neuron 35, 921933 (2002).12. Selimi, F. et al. Neuron 37, 813829 (2003).13. McMahan, U.J. et al. Curr. Opin. Cell Biol. 4, 869874

(1992).14. Misgeld, T., Kummer, T.T., Lichtman, J.W. & Sanes,

J.R. Proc. Natl. Acad. Sci. USA 102, 1108811093 (2005).

15. Lin, W. et al. Neuron 46, 569579 (2005).

Synaptic plasticity and self-organization in the hippocampusGyrgy Buzski & James J Chrobak

A new paper reports that long-term potentiation in the hippocampus, a model of learning and memory, can induce sharp wave-ripple complexes, which are thought to be critical for the stabilization of memory traces in cortex.

After buying a new cell phone, we quickly transfer our phone book to the new gadget

Gyrgy Buzski is at the Center for Molecular

and Behavioral Neuroscience, Rutgers, the State

University of New Jersey, 197 University Avenue, Newark, New Jersey 07102, USA. James Chrobak

is at the Department of Psychology, University

of Connecticut, 406 Babbidge Road, Storrs, Connecticut 06269, USA.


and relegate the old instrument to the recy-cling bin. Likewise, sometime after an event, memories that initially depend on the activity of hippocampal neuronal assemblies are trans-ferred to and consolidated in the neocortex and no longer depend on the hippocampus. How does this change take place, and how do patterns of activity within hippocampal cell assemblies transfer information to the neocor-tex and consolidate it there? And how can cell assemblies produce the patterns of neuronal

discharge required to induce synaptic change? Self-organized population discharges in the hippocampus such as hippocampal sharp wave-ripple (SPW-R) complexes, mainly seen in vivo, are thought to represent stored infor-mation that is then transferred to the neocor-tex. Nonetheless, the mechanisms responsible for the induction of these SPW-R complexes are unclear.

Now Behrens and colleagues1 report that they can induce in vitro SPW-R complexes,

20

05 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce



have found. During repeated stimulus trains (400 ms at 100 Hz), repeated every 40 s, SPW-R complexes gradually increased in inci-dence and amplitude after the third to fifth stimulus. Trains that failed to induce LTP failed to induce SPW-R. After approximately the fifth LTP train, the incidence of SPW-Rs reached a plateau, perhaps because the synapses involved in the generation of the spontaneous patterns were saturatedthat is, they reached their maximum possible strength.

Both MK-801 (a noncompetitive antagonist of the NMDA subtype of glutamate receptor) and D-AP5 (a competitive NMDA receptor antagonist) could prevent the induction of SPW-R complexes. However, once these com-plexes were established, these drugs did not prevent SPW-Rs. Indeed, the incidence of estab-lished SPW-R incidence increased in the pres-ence of these drugs, probably via mechanisms that involve decreased calcium influx through NMDA receptors and a subsequent reduction in the activation of SK2 calcium-activated potassium channels11. Importantly, the estab-lished SPW-R events were reduced or abolished by low-frequency stimulation, a protocol that induces long-term depression of synapses. The blockade of gap junctions with carbenoxolone also attenuated ripple occurrence, in accordance with earlier observations of SPW-R complexes in vitro12. In short, the spontaneously emerging SPW-R complexes show a strong parallel with synaptic changes observed in vivo13.

Using combined intracellular and extracel-lular recordings, Behrens and colleagues also examined intracellular responses in pyrami-dal cells during the development of SPW-R complexes. They observed extracellular post-

synaptic potentialintracellular postsynaptic potential (EPSP-IPSP) sequences, IPSP-EPSP sequences and prominent IPSPs, but they never found isolated EPSPs. Thus, inhibitory inputs are prominent in the development of ripple complexes and may contribute to the temporal precision of EPSP-spike cou-pling14,15. Importantly, the pattern of inhi-bition-excitation in any particular neuron remained stable during the development of SPW-Rs in individual neurons; this indicated that the discharge sequence of the participat-ing neurons during the SPW-R complexes is determined largely by a unique distribution of synaptic strengths at both excitatory and inhibitory connections2,13.

Although these findings show substantial homology between a hippocampus-gener-ated event in vivo and population events in vitro, some differences are worth pointing out. First, SPW-Rs are associated with large population events and ripple oscillations in both the CA1 and CA3 pyramidal layer in vitro. In contrast, the in-vivo ripple event in CA1 reflects a convergence of small-ampli-tude excitatory inputs from a large area of the CA3 region, without synchronous ripple oscillations. This difference may be inter-preted as an artificial augmentation of excita-tion of the truncated CA3 collateral circuitry, reminiscent of epileptiform activity. Second, in contrast to the regularly occurring and uniformly sized SPW-Rs in the slice, their in-vivo counterparts are very irregular in both their temporal distribution and magnitude.

Nevertheless, the replication of an endog-enous brain pattern in vitro allows for the investigation of a number of important

CA3

CA1

Sub

EC

Para

Figure 1 Self-organized burst of activity in the hippocampal CA3 region produces a field potential in the dendritic layer of CA1 and a short-lived fast-frequency field oscillation (200-Hz ripple) within stratum pyramidale, as well as a phase-related discharge of the neurons. Hippocampal output, in turn, produces similar sharp wave-ripple complexes in the subiculum (Sub), parasubiculum (Para) and deep layers of the entorhinal cortex (EC). Behrens and colleagues1 show that the rules of synaptic plasticity govern the emergence and the recruitment of particular cell groups in these hippocampal output events.

Ann

Tho

mso

n

similar to the fast-frequency ensemble pat-terns commonly seen in vivo. Moreover, SPW-R can be induced with stimulation protocols known to induce LTP, a popular neurophysiological model of learning and memory. The authors also demonstrate that these events induce synaptic change among CA3 neurons. Thus, they have established a link between the induction of LTP and the emergence of a physiological network pattern believed to be involved in shaping memories. Their report takes one big step in bridging the chasm separating synaptic mechanisms studied in vitro and the consolidation of memory traces in the intact brain.

The hippocampal SPW-R complex has features that make it a candidate pattern for the consolidation of synaptic plasticity and the transfer of neuronal patterns2. Importantly, it also has a widespread effect. In the approximately 100-ms time window of a hippocampal SPW, between 50,000 and 100,000 neurons1020% of the total neuronal population of the rat hip-pocampusdischarge simultaneously in the CA3-CA1subicular complexentorhinal axis, qualifying it as the most synchronous network pattern in the brain3 (Fig. 1). The SPW-R com-plex arises in the recurrent collateral system of the CA3 region and spreads downstream. Although its recruitment dynamics are delicately controlled by various classes of interneurons4, a three- to fivefold gain of network excitability is achieved transiently5. The observation that the SPW-R is shaped by previous experience6 gives further support for the fundamental role of these population patterns. However, there are several missing links in the story, including how SPW-R complexes emerge and whether this pattern is actually accompanied by changes in synaptic connectivity. Behrens and colleagues1 demonstrate that stimuli that induce LTP lead to the generation of SPW-R complexes in slices of the dorsal hippocampus of the rat. Further, the induction, but not the expression, of SPW-R complexes is NMDA receptor dependent. They also provide evidence that the induction of SPW-R complexes is paralleled by changes in both excitation and inhibition in the CA3 region.

SPW-R complexes in vitro have been observed only in the mouse hippocampus and from the ventral hippocampus of the rat7,8 and the mouse9,10. One possible explanation for the absence of spontaneous SPW-R in slices of the rat dorsal hippocampus is that the density of recurrent axon collaterals and the mutual excitation are simply not sufficient to bring about a self-organized population burst. If so, strengthening the surviving synapses could rescue the compromised circuit. This is precisely what Behrens and colleagues1

20

05 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce



mechanisms that are difficult to explore in vivo. For example, using simultaneous intra-cellular recordings from two or more neurons, future experiments will be able to reveal if the induced but otherwise self-generated patterns involve stable recruitment mechanisms. If so, such findings would provide evidence for the hypothesis that endogenous patterns preserve the information about the perturbations that gave rise to the patterns. Monitoring large numbers of neurons and modifying targeted parts of the circuit may identify the elementary mechanisms involved in the consolidation of network patterns. The finding that such modi-

fications can be brought about by stimulation protocols that induce LTP and can be altered by those that produce long-term depression is even more exciting. Memories may really be made in the hippocampus, and SPW-R com-plexes may contribute to the process after all.

1. Behrens, J., van den Boom, L P., de Hoz, L., Friedman, A. & Heinemann, U. Nat. Neurosci. 8, 15601567 (2005).

2. Buzsaki, G. Neuroscience 31, 551570 (1989).3. Chrobak, J.J. & Buzsaki, G. J. Neurosci. 14, 6160

6170 (1994).4. Klausberger, T. et al. Nature 421, 844848 (2003).5. Csicsvari, J., Hirase, H., Czurko, A., Mamiya, A. &

Buzsaki, G. J. Neurosci. 19, 274287 (1999).6. Wilson, M.A. & McNaughton, B.L. Science 265, 676

679 (1994).7. Papatheodoropoulos, C. & Kostopoulos, G. Brain Res.

Bull. 57, 187193 (2002).8. Kubota, D., Colgin, L.L., Casale, M., Brucher, F.A. &

Lynch, G. J. Neurophysiol. 89, 8189 (2003).9. Yanovsky, Y., Brankack, J. & Haas, H.L. Neuroscience

64, 319325 (1995).10. Maier, N. Nimmrich, & Draguhn, A. J. Physiol. (Lond.)

550, 873887 (2003).11. Colgin, L.L., Jia, Y., Sabatier, J.M. & Lynch, G.

Neurosci. Lett. 385, 4651 (2005).12. LeBeau, F.E., Traub, R.D., Monyer, H., Whittington,

M.A. & Buhl, E.H. Brain Res. Bull. 62, 313 (2003).13. King, C., Henze, D.A., Leinekugel, X. & Buzsaki, G.

J. Physiol. (Lond.) 521, 159167 (1999).14. Axmacher, N. & Miles, R. J. Physiol. (Lond.) 555,

713725 (2004).15. Pouille, F. & Scanziani, M. Science 293, 11591163

(2001).

Glial cells under remote controlKlaus-Armin Nave & Markus H Schwab

Not all axons in a peripheral nerve are myelinated. A recent study shows that the expression of neuregulin-1 on an axon membrane determines whether immature Schwann cells will differentiate into myelinating Schwann cells.

Klaus-Armin Nave and Markus Schwab are at the

Department of Neurogenetics, Max Planck Institute

of Experimental Medicine, Hermann-Rein-Str. 3,

37075 Gttingen, Germany.


Like the insulation on electrical wires in your house, myelin sheaths are essential for rapid impulse propagation throughout the verte-brate nervous system. The multilayered myelin membranes are synthesized by highly special-ized glial cells, termed oligodendrocytes, in the CNS and Schwann cells in the PNS. In the PNS, the decision of Schwann cells to myelin-ate resembles a cell lineage decision that is triggered by axonal signals, but the nature of these signals has remained unclear. Recently in Neuron1, Taveggia et al. demonstrated that axonal neuregulin-1 type III, a regulator of myelin growth, is required to induce the dif-ferentiation of myelinating Schwann cells in dorsal root ganglion (DRG) and superior cer-vical ganglion (SCG) explant cultures.

Ever since electron microscopists demonstrated the complexity of compact myelin sheaths, cellular neurobiologists have been hooked on the myelin-forming glia, their complex interaction with axons and their spectacular membrane growth. Numerous proteins of myelin and the axon-glia junc-tion have been identified; nevertheless, major questions remain unanswered. We do not know the driving force of glial ensheathment. Cross-sections suggest that myelin assembly is a spiral membrane growth process, but this has not been verified. It is unclear why axons larger than

1 m are myelinated, but small-caliber axons are only ensheathed, and dendrites seem not to interact at all with oligodendroglia. We do not know what signals provide specificity of axon-glia interaction and instruct glia to myelinate, or whether these mechanisms are the same for oligodendrocytes and Schwann cells. One would expect neurons and myelinating glia to communicate through a complex assembly of developmentally regulated signaling pro-teins. It is therefore surprising that, at least in the PNS, one signaling system, comprised of axonal neuregulin-1 and glial ErbB receptors, seems to operate at all stages of Schwann cell development and myelination.

Neuregulin-1 (Nrg1) comprises a family of more than 15 membrane-associated and secreted growth factors that are derived from a single gene by alternative splicing and pro-motor use. Major subgroups are Nrg1 type I (including proteins named NDF, heregulin, and ARIA) and type II (glial growth factor or GGF) isoforms, both of which are potentially secreted or shed upon proteolytic cleavage. In contrast, Nrg1 type III has a second transmem-brane domain and remains a membrane-asso-ciated ligand2. Common to all isoforms is an EGF-like domain that activates ErbB receptor tyrosine kinases. In the developing nerve, ErbB2 and ErbB3 are expressed at the cell surface of Schwann cell progenitors and are essential for their survival and subsequent differentiation3. Nrg1 signaling also contributes to synaptogen-esis (at least in vitro), the migration of cortical interneurons and cardiac development in the embryo. The latter has greatly hampered the

conventional genetic analysis of Nrg1 function in the postnatal nervous system.

Taveggia et al.1 now provide experimental evi-dence that neuregulin-1 type III is necessary for myelination in the PNS and can instruct imma-ture ensheathing cells to become true myelin-forming Schwann cells. These in vitro findings close a gap between related in vivo findings by other groups, including the requirement of neuregulin-1 for the survival of precursor cells and immature Schwann cells4, the requirement of glial ErbB2 receptors for normal myelina-tion in conditional mouse mutants5, and the identification of neuregulin-1 type III as an axo-nal signal that regulates myelin sheath thickness in mice with altered neuregulin gene dosage6.

Mice that selectively lack Nrg1 type III die perinatally and have a marked decrease in Schwann cells and degenerating motor and sensory nerves7. To study the competence of these mutant axons to be myelinated, Taveggia et al. used a coculture system in which in vitro myelination is initiated by the addition of ascorbate8. Sensory DRG neurons from wild-type and Nrg1 type IIIdeficient mice were cocultured with Schwann cells from normal rats. The Nrg1 type III mutant axons never became myelinated, even in the presence of a fivefold excess of Schwann cells. Lentiviral expression of Nrg1 type III was sufficient to restore myelination competence. In agreement with a report of hypermyelin-ation in mice that overexpress Nrg1 type III, but not type I (ref. 6), multiple examples of unusually thick myelin profiles were seen in rescued cocultures.

20

05 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce



Because Nrg1/ErbB2 signaling is required for myelination in vivo5, the authors then asked whether normally unmyelinated axons would become myelinated if they ectopically expressed Nrg1 type III. By transducing cultured SCG neu-rons (with axons that are normally ensheathed by Schwann cells, but not myelinated), they could show that indeed low levels of Nrg1 type III are sufficient to induce ectopic myelination. This suggests that increasing Nrg1 expression in neurons above a certain level may confer myelin-ation competence. An important next question is whether insufficient Nrg1 expression can also explain why axons smaller than 1 m in diameter (such as C-fibers devoted to pain perception) are typically unmyelinated, or whether axon size and membrane curvature pose neuregulin-indepen-dent thresholds for myelination.

What are the effector molecules in neu-regulin-induced Schwann cells? The inhibi-tion of PI3 kinase blocks myelin formation in cocultures9. PI3 kinase and MAP kinase in Schwann cells are activated by GGF, a soluble type II neuregulin, or by contact with neurite membranes. Now Taveggia et al. extend these observations by showing that neurite mem-branes from wild-type (but not from Nrg1 type IIIdeficient) neurons activate the PI3 kinase pathway. In contrast, MAPK activation was not impaired in mutant cultures. Thus, Nrg1 type III emerges as a critical activator of glial PI3 kinase, whereas other signaling mol-ecules underlie the activation of MAP kinases. These other molecules may or may not include the secreted neuregulin isoforms.

Assuming that the paracrine neuregulins (type I and II) are functionally not equiva-lent to the juxtacrine type III (localized to the axon surface), the authors asked whether a type III protein promotes myelination when offered as a paracrine signal. To address that issue, they added a recombinant type III ect-odomain to the coculture system. The soluble protein retained mitogenic activity, but did not support myelination. The authors also tested an ectopic juxtacrine source of Nrg1 type III. Schwann cells were cultured under myelinating conditions on a monolayer of CHO cells that stably expressed Nrg1 type III on the surface. Again, the factor retained some mitogenic potential, but Schwann cell differentiation (as determined by myelin gene expression) could not be induced. These observations suggest that neurons produce multiple neuregulin-1 isoforms, but only type IIIin concert with other axonal sig-nalspromotes ensheathment and myelina-tion as a contact-dependent signal.

Taveggia et al. also analyzed heterozygous Nrg1 type III mutant mice, which are phenotypically normal7. In agreement with previous work6, they

found that the sciatic nerves of adult mice are hypomyelinated and have reduced nerve conduc-tion velocity. Additionally, the authors observed a higher number of small caliber axons (C-fibers) grouped into so-called Remak bundles. The latter suggests that Nrg1 type III also signals between C-fibers and non-myelinating Schwann cells. Indeed, the importance of this interaction has been demonstrated before by dominant-negative perturbation of ErbB receptor function, which causes disorganization of Remak bundles and peripheral neuropathy in transgenic mice10.

Most likely, the conclusions presented here will be rederived, and possibly modified, in vivo using conditional mutants and transgenic mice. But when the new data1 are combined with those from other studies3,5,6,10 an unexpected picture emerges already: evolution of the nervous system has recruited a neuregulin-based signaling system for virtually all steps of Schwann cell differentia-tion (Fig. 1), including the fine-tuning of myelin

growth. It will be most interesting to extend this line of research into the CNS. Like Schwann cells, oligodendrocytes respond to Nrg1 (refs. 11,12). Myelination control in the brain and spinal cord differs with respect to other growth fac-tors13, and preliminary evidence suggests a more complex regulation by different Nrg1 isoforms (B. Brinkmann, M.H.S. and K.A.N., unpublished data). This research may also have clinical impli-cations, if axonal Nrg1 contributes to remyelin-ation (or the lack thereof) in multiple sclerosis.

From a neuronal point of view, glial cell differ-entiation emerges as largely remote controlled. If quantitative differences of neuronal Nrg1 gene expression determine Schwann cell fate, even myelin thickness, then how does a neuron know its own size, or how much neuregulin to make? It should matter whether an axon measures 1 or 10 m in diameter, and whether it is 1 mm or 1 meter long. This issue of whether axon length and caliber control neuronal gene expression

Neuronalcell bodies

Mitotic Schwann cellprogenitors

Immature Schwann cell

Remak bundle

MyelinatingSchwann cell

Neuregulin

Axon surface

Axon

1

23

4

5

Figure 1 Neuregulin-1 directs Schwann cell differentiation. In peripheral nerves, neural crestderived Schwann cell progenitors (1) proliferate and populate axon bundles. Later, as immature Schwann cells (2), they face a binary choice: they either stay tightly associated with several axons to form a Remak bundle (3), or alternatively they single out larger axons and differentiate into myelinating Schwann cells (4). Work in several laboratories indicates that the entire path of Schwann cell development and myelination is remote controlled by the neurons through expression of the neuregulin-1 (5) on the axonal surface.

Deb

bie

Mar

zels

20

05 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce



Senseless makes sense for spinocerebellar ataxia-1Vikram Khurana, Tudor A Fulga & Mel B Feany

Why are some neurons selectively targeted for death in neurodegenerative diseases? A recent paper combines genetics in the fruit fly and mouse to uncover mechanisms underlying the vulnerability of Purkinje cells in spinocerebellar ataxia-1.

The authors are at the Department of Pathology at Brigham and Womens Hospital and Harvard Medical School, 77 Louis Pasteur Ave., Boston, Massachusetts 02115, USA.e-mail: [email protected]

Neurodegenerative diseases share many fea-tures, including a progressive loss of neurons and the formation of proteinaceous aggregates. These similarities have motivated research into common underlying pathogenic processes, including dysfunction of the ubiquitin-pro-teasome system, impaired axonal transport and oxidative stress. A recent paper in Cell by Hiroshi Tsuda and colleagues1 reminds us that neurodegenerative diseases also have impor-tant features that distinguish them from one another, including the selective vulnerability of particular groups of neurons.

The authors focus on a type of spinocer-ebellar ataxia (SCA). SCAs are debilitating neurodegenerative diseases characterized by progressive gait incoordination and cerebel-lar atrophy. Tsuda et al. delineate a physical and functional interaction between the AXH domain of ataxin-1, a protein of unknown function, and the transcription factor known as Senseless in Drosophila melanogaster and Gfi-1 in vertebrates. The authors provide compelling evidence in animal models that this interac-tion contributes to the progressive demise of Purkinje neurons in SCA-1.

Autosomal-dominant polyglutamine (polyQ) expansion disorders, including Huntington disease and a number of SCAs, are all caused by the expansion of unstable CAG repeat sequences within the coding region of the causative gene2. Neurodegeneration accompanies the intraneuronal aggrega-tion of the polyQ-expanded proteins in each disease. The dominant mode of inheritance, together with the recapitulation of disease phenotypes in overexpression but not knock-

out animal models, suggests a toxic gain-of-function mechanism whereby the expanded polyQ tract confers new molecular functions upon the causative protein. Notably, however, despite ubiquitous expression in the nervous system, only certain neuronal groups are tar-geted for death in these diseases. Furthermore, differing polyQ repeat lengths are required to initiate neurodegeneration in the different diseases. These differences strongly implicate sequences outside the CAG repeat region in disease pathogenesis.

Tsuda et al.1 have now taken us a signifi-cant step closer to understanding the unique features of ataxin-1 that mediate degenera-tion of Purkinje cells in SCA-1. This disease is caused by a polyglutamine expansion in ataxin-1 and accompanied by nuclear aggre-gation of this protein in neurons. The authors previously showed that overexpressing human ataxin-1 (hAtx-1) with an expanded polyQ tract in Drosophila resulted in neurodegenera-tion3; further, they showed that phosphoryla-tion of Ser776 by the kinase Akt is critical to toxicity by enabling an interaction with 14-3-3 and increasing hAtx-1 stability4. Tsuda et al.1 now report that expressing the fly homolog of ataxin-1 (dAtx-1), but not a polyQ repeat alone, recapitulates hAtx-1induced phe-notypes in different fly tissues, albeit with reduced severity. Intriguingly, dAtx-1 does not contain a polyQ domain but shares an AXH (ataxin-1/HBP1) domain with hAtx-1, a domain recently implicated in RNA bind-ing and self-association5. The authors further demonstrate that dAtx-1 physically interacts with the transcription factor Senseless (Sens) by means of this domain. An in vitro tran-scriptional assay and a functional analysis in the fly reveal that both Sens activity and protein abundance are downregulated by dAtx-1. Furthermore, expressing hAtx-1 with an expanded polyQ tract reduces Sens levels more potently than dAtx-1, whereas over-

expressing the polyQ tract alone, or polyQ-expanded hAtx-1 with the AXH domain deleted, has no effect on Sens levels.

The study proceeds with a logical series of experiments relating the findings in Drosophila to a mouse model system, thus strengthening the relevance of AXH domain interactions to the human disease. The authors show that hAtx-1 binds to the ver-tebrate homolog of Sens, Gfi-1, also through its AXH domain. Importantly, Gfi-1 is maxi-mally expressed in the nervous system within the Purkinje neurons of the cerebellum, one group of neurons that selectively degenerate in SCA-1. In mammalian cells, as in flies, Gfi-1 levels are downregulated by hAtx-1, an effect that is post-translational and depends on the ubiquitin-proteasome system. Significantly, these findings are recapitulated in a mouse model of SCA- 1, where expression of polyQ-expanded hAtx-1 in Purkinje cells leads to an early decrease in the abundance of Gfi-1, preceding Purkinje cell loss and ataxia. Furthermore, removing a single copy of Gfi1 enhances the neurodegeneration phenotype in this model. In a final elegant proof-of-prin-ciple experiment, the authors show that a pro-gressive loss of Purkinje neurons accompanies ataxia in Gfi1 knockout mice, thus demon-strating that decreased Gfi-1 is sufficient to cause Purkinje cell loss.

These findings are important at multiple levels. Most directly, they suggest that the interaction between the AXH domain of hAtx-1 and Gfi-1 is important for mediating neurodegeneration and that this is poten-tially a therapeutic target. However, although the authors show that reducing Gfi-1 levels results in an enhancement of hAtx-1induced neurodegeneration, establishing whether ecto-pic expression of Sens or Gfi-1 rescues neu-rodegeneration in flies or mice, respectively, would further strengthen the case for Gfi-1 stabilization as a potential treatment. Beyond

may be relevant for many more proteins, but it has never been recognized as such. Once again, glial cells have evoked important questions.

1. Taveggia, C. et al. Neuron 47, 681694 (2005).2. Falls, D.L. Exp. Cell Res. 284, 1430 (2003).3. Jessen, K.R. & Mirsky, R. Nat. Rev. Neurosci. 6, 671

682 (2005).4. Meyer, D. & Birchmeier, C. Nature 378, 386390

(1995).5. Garratt, A.N., Voiculescu, O., Topilko, P., Charnay, P. &

Birchmeier, C. J. Cell Biol. 148, 10351046 (2000).6. Michailov, G.V. et al. Science 304, 700703 (2004).7. Wolpowitz, D. et al. Neuron 25, 7991 (2000).8. Eldridge, C.F., Bunge, M.B., Bunge, R.P. & Wood, P.M.

J. Cell Biol. 105, 10231034 (1987).9. Maurel, P. & Salzer, J.L. J. Neurosci. 20, 46354645

(2000).10. Chen, S. et al. Nat. Neurosci. 6, 11861193 (2003).11. Park, S.K., Miller, R., Krane, I. & Vartanian, T. J. Cell

Biol. 154, 12451258 (2001).12. Fernandez, P.A. et al. Neuron 28, 8190 (2000).13. Chan, J.R. et al. Neuron 43, 183191 (2004).

20

05 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce



the direct therapeutic implications, however, this study provides critical evidence implicat-ing sequences outside the polyQ region in the selectivity of neurodegeneration in polyQ dis-orders. Future studies might explore whether the hAtx-1Gfi1-1 interaction mediates neu-rodegeneration in other cell populations vul-nerable in SCA1, including the inferior olivary nucleus or spinocerebellar tracts, or whether other interactions are involved. In this regard it would be interesting to determine if, in addi-tion to Purkinje neurons, these populations are also vulnerable in Gfi-1null mice.

The methodology used by Tsuda et al.1 also deserves attention. Whereas several models of autosomal-dominant neurodegenera-tive diseases have been made by transgenic overexpression of causative human genes in Drosophila6, the present study adopts the nor-mal fly protein as a starting point. The obser-vation that expression of an expanded polyQ tract alone does not phenocopy certain phe-notypes shared by dAtx-1 and hAtx-1polyQ leads the authors to infer the existence of functionally important sequences outside the polyQ region. Flies certainly provide an ideal system to make such comparisons. Further, by recapitulating the biochemical and functional interactions in the mouse model system, the study supports the utility of Drosophila in modeling human diseases. Indeed, there are fly homologs for proteins, such as tau, that are involved in other neurodegenerative diseases, raising the possibility that such an approach might be fruitful for these diseases also.

How do we place the present findings in the context of what is known about the patho-genesis of SCA-1 and related disorders? PolyQ expansion clearly initiates the disease process. The model proposed in this study would implicate this expansion in the stabilization of hAtx-1, abnormally potentiating the AXH domainGfi-1 interaction. Neurotoxicity would follow from proteasomal degradation of Gfi-1 and transcriptional dysregulation (Fig. 1). Here, toxic gain-of-function is caused, not by the protein attaining an entirely aber-rant function, but rather from the abnormal activation of a physiological pathway. In con-trast, many previous studies have concentrated on abnormal interactions mediated by the polyQ tracts themselves. Because the nuclear localization of causative proteins is essential for neurotoxicity7, these studies have focused on abnormal effects on transcription, either directly or through sequestration of transcrip-tion factors8. For example, the polyQ tract of hAtx-1 binds to the nuclear protein PQBP-1 in a manner dependent on polyQ tract length9. A resultant complex forming between hAtx-1, PQBP-1 and RNA polymerase II led to

a decrease in basal transcription. Intriguingly, PQBP-1 is enriched in the cerebellum, and a polyQ-dependent interaction could there-fore contribute to selective neuronal vulner-ability in SCA-1 (ref. 10). Other groups have provided data supporting different toxic gain-of-function mechanisms in polyQ-asso-ciated diseases, including disruption of axo-nal transport and downregulation of survival pathways11. Wild-type ataxin-3 suppresses degeneration in multiple polyQ models in flies by proteasomal activation, implying not only that common mechanisms might oper-ate in different polyQ-associated diseases, but that loss-of-function mechanisms might also be involved12.

Taking these studies together, a picture emerges of both the common mechanisms in polyQ expansion disorders that could be mediated by the polyQ tracts themselves and the distinct effects that could be attributable to nonpolyQ-encoding sequences. In keep-ing with this idea, transcriptional profiling and microarray studies reveal both common and distinct changes in different polyQ animal models13. A clear challenge for future studies is to try and integrate the multiple pathways downstream of polyQ expansion into a cohe-sive picture. For example, in SCA-1, do the implicated transcription factors function as

a network influencing common downstream processes? Gfi-1, for example, downregu-lates proapoptotic genes14. Is it possible that downregulation of Gfi-1 by hAtx-1 leads to neuronal apoptosis? Could dysregulation of other transcription factors converge on apop-tosis also? To guide investigations into events downstream of transcriptional dysregulation, it would clearly be useful to define the molecu-lar pathways mediating cell death in these dis-eases, whether apoptotic or non-apoptotic. At present, the mechanisms of cell death in SCA-1 remain unclear, with neurodegeneration in the SCA-1 mouse seeming to be p53 dependent but not classically apoptotic15.

In summary, Tsuda et al. present us with an important and thought-provoking study that provides hope for targeted SCA-1 therapies in the future. Establishing the direct relevance of a physiological protein interaction to dis-ease pathogenesis has implications extending beyond SCA-1 and polyQ disorders to other neurodegenerative diseases for which toxic gain-of-function mechanisms have been proposed, including familial Alzheimer and Parkinson diseases. For mice and fruit flies, the message is loud and clear: resolving the similarities and differences among related neurodegenerative diseases should guarantee employment for many generations to come.

Figure 1 Tsuda et al. demonstrate homologous pathways that mediate ataxin-1induced neurodegeneration in mouse and Drosophila models of SCA-1. In Drosophila, dAtx-1 (which lacks polyQ repeats) binds, through its AXH domain, to the transcription factor Senseless and targets it for proteasomal degradation. A homologous interaction, potentiated by abnormal polyQ expansion and aggregation, occurs between hAtx-1 and Gfi-1 in mouse Purkinje neurons. The resultant transcriptional dysregulation mediates neurotoxicity in both model systems. Direct dysregulation of transcription by polyQ-expanded repeats may also make an important contribution to neurotoxicity9.

Nucleus

Nucleus

PolyQaggregationPolyQ

AXH

AXH

AXH

Gfi-1

Gfi-1

Proteasome

MousePurkinje neurons

Drosophila

hAtx1

dAtx1

Sens

Neurotoxicity

Ann

Tho

mps

on

20

05 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce



How the brain recovers following damageYalin Abdullaev & Michael I Posner

Individuals with neglect fail to process stimuli on the left. A new paper uses functional imaging to show that a restricted lesion, usually caused by a stroke, may influence the network of areas associated with attention shifts.

Yalin Abdullaev and Michael I. Posner are in the

Department of Psychology, University of Oregon,

Eugene, Oregon 97403, USA.


After right-hemisphere stroke, some people see objects in their world as having no left side (Fig. 1). In the acute stage immediately follow-ing the stroke, such individuals with spatial neglect may fail to orient to people approach-ing from their left, to recognize their left arm as their own and to eat from the left side of their plate. How can a stroke to a local area of the brain be associated with such a mysterious array of symptoms?

In this issue, Corbetta et al.1 report that the activity of an interconnected network of dorsal, ventral parietal and frontal areas may be influ-enced by comparatively restricted lesions of the ventral right hemisphere. The dorsal network includes the superior parietal lobe and the fron-tal eye fields, and the ventral network involves the temporal parietal junction and the ventral lateral prefrontal cortex1. In normal people, imaging studies show that this interconnected network is important for shifts of spatial attention2.

The authors tracked the recovery of this network in individuals with neglect through successive functional magnetic resonance imaging (fMRI) scans taken while the patients performed a task. In an fMRI scanner, partici-pants were instructed to direct their gaze to the center of the screen in front of them. An arrow appeared at this central point, directing the subjects to attend to either the left or the right side of the screen, without moving their eyes. Participants were then asked to press a button when they detected a target. The target appeared on the side indicated by the arrow most of the time, but occasionally it appeared on the other, unattended side. Normal par-ticipants have longer reaction times when the

activity in similar areas in the left hemisphere. Therefore, the dorsal parietal area, which is crit-ical for voluntary shifts of attention in normal people2, was the only brain area that showed increased activity on the lesioned side from the acute to the chronic stage. This was accompa-nied by reduced activity in the non-lesioned hemisphere. This effect is likely to be respon-sible for the observed reduction in rightward bias from the acute to the chronic stage.

How do these parietal findings relate to what is found in visually specific areas of the cor-tex? Several studies5 indicate that the source of attention effects (manifested as increased activ-ity in visual cortex during stimulus detection) lies in parietal areas. The results of Corbetta et al.1 show that activity in the left visual cortex is reduced in the chronic phase as compared to the acute phase, whereas right hemisphere visual cortex activation is increased. These findings mirror those in the parietal lobe.

Individuals recovering from neglect also

Figure 1 Sample drawings by individuals with spatial neglect. (a,b) The pictures demonstrate how they ignore their left visual field in trying to make simple line drawings of a clock (a) or a house (b).

target appears on the side opposite to where they are directing their attention3. People with neglect show particularly long reac-tion times in response to targets on the left side when they are first cued to attend to the right4. In the acute stage immediately after the stroke, they may miss such targets com-pletely, and even after many years they have a large deficit in reaction time3.

The Corbetta et al. study1 found a dramatic alteration in the pattern of activation in the pari-etal-frontal network four weeks after the stroke (the acute stage), even though the individual nodes showed no evidence of structural dam-age. Seven months later (in the chronic stage), the participants had considerably improved in their ability to orient and detect stimuli on the left. Functionally, the most striking change was that the dorsal right parietal lobe, which was not activated at all during the acute phase, was now strongly activated in the chronic phase. This was in contrast with an actual reduction of

1. Tsuda, H. et al. Cell 122, 633644 (2005).2. Taroni, F. & DiDonato, S. Nat. Rev. Neurosci. 5, 641

655 (2004).3. Fernandez-Funez, P. et al. Nature 408, 101106

(2000).4. Chen, H.K. et al. Cell 113, 457468 (2003).5. de Chiara, C. et al. FEBS Lett. 551, 107112 (2003).6. Muqit, M.M. & Feany, M.B. Nat. Rev. Neurosci. 3,

237243 (2002).7. Klement, I.A. et al. Cell 95, 4153 (1998).8. Margolis, R.L. & Ross, C.A. Trends Mol. Med. 7, 479

482 (2001).9. Okazawa, H. et al. Neuron 34, 701713 (2002).10. Humbert, S. & Saudou, F. Neuron 34, 669670

(2002).11. Lipinski, M.M. & Yuan, J. Curr. Opin. Pharmacol. 4,

8590 (2004).12. Warrick, J.M. et al. Mol. Cell 18, 3748 (2005).13. Sugars, K.L. & Rubinsztein, D.C. Trends Genet. 19,

233238 (2003).14. Jafar-Nejad, H. & Bellen, H.J. Mol. Cell. Biol. 24,

88038812 (2004).15. Shahbazian, M.D., Orr, H.T. & Zoghbi, H.Y. Neurobiol.

Dis. 8, 974981 (2001).

20

05 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce



improve their ability to spot targets on the unattended side while they focus their atten-tion on the side indicated by the arrow. This recovery seems to depend mostly on the right temporal parietal junction, a brain area thought to be responsible for interrupting attention and allowing a shift of attention toward the target location. This area is active only in the right hemisphere during shifts of attention to tar-gets in either direction, whereas most other areas are symmetric6. The right lateralization of this area seems to be a major reason for the strong neglect of the left side apparent in the participants performance (Fig. 1).

The Corbetta et al. study1 reveals that damage to the ventral areas of the spatial attention system produces neglect because it causes dysfunction of the dorsal system as well. Although this was not a specific focus of the current study, the results raise the ques-tion of why more dorsal lesions do not cause malfunctions of the ventral parietal area and thus also produce neglect. Apparently, there is an asymmetry between the ventral and dorsal parietal areas: lesions of the dorsal

parietal area alone do not cause neglect7. The mechanisms underlying this strong asymme-try in the remote effects of these two critical areas of the parietal lobe remain unclear.

Evidence for greater than normal left hemi-sphere activity during the acute stage of neglect provides some basis for rehabilitation methods as well. The authors suggest that competition between hemispheres may provide the basis for helping recovery, either by increasing activity in the ipsilesional cortex or by reducing it in the contralesional cortex. Increased right hemi-sphere activity through warning signals, already known to reduce neglect, could serve as one method for improving hemispheric balance8. The need for hemispheric balance could be an important reason why forcing the increased use of a paralyzed limb can foster recovery9. Among patients with neglect, those with right hemisphere lesions, who show a chronic right-ward bias, may also be helped by inhibiting left superior parietal lobe activity1.

There are more general lessons that might be gleaned from this work. Remote effects of lesions provide some explanation of how local-

ized computations, as revealed in many func-tional imaging studies, may be consistent with the occurrence of a syndrome with many com-plex sensory and motor features. This finding may be important for all of neuropsychology. The authors also argue that the general prin-ciples of recovery that they have found might apply to aphasia or sensory motor deficits. In any case, the paper provides a strong argument for using imaging during recovery to determine the effects of various forms of therapy.

1. Corbetta, M., Kincade, M.J., Lewis, C. & Sapir, A. Nat. Neurosci. 8, 16031610 (2005).

2. Corbetta, M. & Shulman, G.L. Nat. Rev. Neurosci. 3, 201215 (2002).

3. Posner, M.I. Q. J. Exp. Psychol. 32, 325 (1980).4. Posner, M.I., Walker, J.A., Friedrich, F.J. & Rafal, R.D.

J. Neurosci. 4, 18631874 (1984).5. Hillyard, S.A., DiRusso, F. & Martinez, A. in Functional

Neuroimaging of Visual Cognition (eds. Kanwisher, N. & Duncan, J.) 381388 (Oxford Univ. Press, Oxford, 2004).

6. Perry, R.J. & Zeki, S. Brain 123, 22732288 (2000).7. Friedrich, F.J., Egly, R., Rafal, R.D. & Beck, D.

Neuropsychology 3, 193207 (1998).8. Robertson, I.H., Mattingley, J.B., Rorden, C. & Driver,

J. Nature 395, 169172 (1998).9. Taub, E., Uswatte, G. & Elbert, T. Nat. Rev. Neurosci.

3, 228236 (2002).

Time to smell the rosesTiming is a thorny issue for the chemical senses. Principal neurons in the vertebrate olfactory bulb and insect antennal lobe have dynamic odor-evoked responses that can long outlast odorant exposure. The temporal pattern of these responses is thought to be important for distinguishing different odorants, but in a natural environment, odor stimuli have their own temporal structure. With an animals movements and the intermittent arrival of odors in the air, new odors are likely to interrupt responses to previous ones. Experience tells us that a second sniff of a rose still smells like a rose, but how does the olfactory system sort out these potentially conflicting time courses?

On page 1568 of this issue, Mark Stopfer and colleagues address this question in the olfactory system of the locust. They exposed adult locusts to trains of brief odorant pulses with natural interpulse intervals, and made intracellular and extracellular recordings from projection neurons (PNs) in the antennal lobe. In most cases, responses varied over successive pulses, showing that the temporal structure of odorant presentation did influence the temporal pattern of PN responses.

To determine the downstream consequences of this interference, the authors considered known features of the locust olfactory system. Each antennal lobe contains 830 PNs, and more than 100 PNs converge onto each of about 50,000 Kenyon cells in the mushroom body, a brain area important for olfactory memory. Odorant stimulation evokes oscillations in the antennal lobe, and Kenyon cells seem to integrate convergent PN input over 50 ms, approximately one oscillation cycle. The authors pooled 117 PN recordings from multiple experiments and constructed activity vectors over 50-ms windows as well as ensemble activity trajectories over the entire PN responses to different patterns of odor pulses.

In contrast to the varied responses in individual PNs, ensemble PN activity showed highly similar responses to multiple pulses. Response trajectories for repeated odor pulses overlapped, with successive pulses appearing to partially reset the circuitry such that each response began similarly and followed a similar trajectory, irrespective of the pulse timing. A template chosen from any 50-ms window of a response could accurately classify odorant identity for responses in different presentation patterns.

These results suggest that the problem of temporal interference may be adequately solved by converging PN input onto Kenyon cells, which also responded reliably to repeated odorant pulses. However, this solution depends on the ability of newly arriving odors to reset ensemble activity in the antennal lobe. What restricts PNs to repeatedly return to the same response trajectory despite ongoing dynamic activity? Can an ongoing response to one odorant also be reset by the arrival of a different smell? Researchers undoubtedly will continue to sniff around for these answers.

Cara Allen

Stac

ey B

row

n

20

05 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce


I N T RO D U C T I O NN E U R O B I O LO G Y O F A D D I C T I O N

Neurobiology of addiction

The pleasant sensation of sipping a drink after a hard days work is familiar to many people, but for some, recreational use easily slips into dependence and tolerance. Some users then progress to addiction. Even in the face of harmful consequences to self and others, addicts cannot resist the urge to engage in the addictive activity. Moreover, if they do stop taking drugs, even after years of abstinence, addicts may relapse into drug use under stress or when faced with otherwise benign cues that remind them of the addicting drug.

Drug use and addiction are pervasive. The 2005 World Drug Report from the United Nations estimates that 200 million people, or 5% of the global population, consumed illicit drugs at least once in the last 12 months. The US Department of Health estimates that in 2004, 22.5 million Americans aged 12 or older (9.4% of the population) experienced substance dependence or abuse. During this period, about 21.1 million people needed but did not get treatment for their addiction in the US alone.

Although drug abuse cuts across all societal strata and age groups, the young and poor are affected most. Addiction has very high overall health costs, once related factors such as heart disease, cancer and accidents are considered. The National Institute on Drug Abuse esti-mated the cost of drug and alcohol abuse at about $246 billion in 1992 (without consider-ing nicotine addiction). This figure includes health consequences from drug abuse and their effects on the health care system, crimi-nal behavior, negligent driving, job loss and the effects of impaired productivity on these individuals and their employers.

The progression from initial drug use to addiction is influenced by the drug, the users personality, peer influences and environ-mental stressors. These complex interactions determine why some individuals are more easily addicted than others. In this focus, we highlight the biology of the most commonly abused substances, explore the genetics of predisposition to addiction, and examine the components of addictive behavior itself.

Drug addiction can clearly vary with the drug. Cocaine, marijuana, LSD or amphet-

amine can create psychological dependence, in which the individual feels satisfaction and euphoria and is driven by a need to repeat the experience. Heroin or alcohol can produce physical dependence. Drugs also act on specific receptors and brain areas. Given this complexity, can addiction be treated as a unitary disorder? Are there common brain targets for all addictive substances that could be exploited to provide a magic bullet for addiction treatment? A perspective by Eric Nestler addresses this issue.

Exposure to drugs causes plasticity in neural circuits related to reward and motivation, sup-porting the idea that addiction is a biological disorder. Plasticity (of synapses and circuits) results from drug use and drug abuse. How do we make sense of the multitude of observations in so many different areas under different cir-cumstances? What animal models are likely to have the most validity for studying addiction, and what specific changes should we exam-ine? In three separate commentaries, George Koob and Michel LeMoal, Peter Kalivas, and Yavin Shaham and Bruce Hope discuss which changes are likely to be critical to addiction.

Taking drugs may begin as a voluntary choice to seek a pleasant stimulus, but for addicts, that choice is no longer volitional, even in the face of terrible personal consequences. Barry Everitt and Trevor Robbins review the cortical and subcortical circuits that mediate reinforcing effects of drugs, presenting a framework for how occasional behav-iors become habits and then compulsions through pavlovian and instrumental learning. Antoine Bechara proposes that volitional decisions involve a balance between neural systems signaling the immediate and delayed consequences of actions. He discusses how drugs may tip this balance, lead-ing to an inability to weigh future consequences and the urge to make impulsive decisions.

What makes certain individuals more vul-nerable to drug use and abuse? Mary Jeanne Kreek and colleagues discuss the genetic influences on complex personality traits such as impulsivity, risk taking and stress respon-siveness, and their relationship to addiction vulnerability. They also discuss the difficulty in teasing out genetic vulnerability factors, in

light of the strong comorbidity between addic-tion and other mental disorders.

Alcohol and nicotine are legal drugs that are prone to abuse. Nicotine is one of the most widely abused substances, and tobacco addiction kills more than 430,000 Americans each year. John Dani and Adron Harris review progress in understanding nicotine addiction and its comor-bidity with alcoholism. John Crabbe and David Lovinger discuss the neurobiology of alcohol abuse and genetic influences that may predispose animals (and humans) to alcoholism.

Despite the enormous social and economic cost of addiction, long-term treatments are few and far between. Only a handful of phar-maceutical therapies exist. In a commentary, Charles Dackis and Charles OBrien discuss social issues that may be hampering develop-ment and access to treatment, pointing out that loss of control, the hallmark of addic-tion, is the source of its societal stigma. A naive public is likely to conceptualize addiction as a character flaw rather than a bona fide brain disorder. Dackis and OBrien argue that to effectively develop treatments for addiction, we must change this perception.

We are grateful to the National Institute of Drug Abuse and the National Institute of Alcoholism and Alcohol Abuse for their gen-erous financial support for this focus issue. With their help, we are making the content of this focus freely available on the web for three months at http://www.nature.com/neuro/focus/addiction/index.html. Apart from the sponsors foreword, the editorial team of Nature Neuroscience is entirely responsible for the con-tent of the focus issue. We hope that our readers will find this collection of articles useful and enlightening and that it may contribute toward understanding and eventually solutions to this critical medical and societal problem.

I-han ChouAssociate Editor

Kalyani NarasimhanSenior EditorView background material on Connotea at http://www.connotea.org/user/NatNeurosci/tag/addictionfocus.

20

05 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce


S P O N S O R S F O R E WO R D

The neuroscience of addictionThe burden of substance abuse and addiction to society is enor-mous, with an estimated annual economic impact in the United States of approximately half a tril-lion dollars arising from medical consequences, loss of productivity, accidents and crime1. The impact of drugs and alcohol on children is particularly problematic, as adolescents are significantly more vulnerable than adults to sub-stance abuse and to addiction2. Also, because many of the molec-ular targets affected by drugs are

involved with brain development, substance abuse during childhood and adolescence has the potential to be particularly deleterious. Indeed, it has been shown that children who begin using alcohol early in child-hood (ages 14 or younger) are four times more vulnerable to becoming addicted to alcohol later in life than are those who begin drinking at 20 years of age or older3.

Scientists are n

nature neuroscience november 2005

Documents