research highlights - wofford college...nal female partner after sex and to spurn other females, in...

RESEARCH

HIGHLIGHTS

HIGHLIGHT ADVISORS

NANCY ANDREASEN

UNIVERSITY OF IOWA, IA, USA

ALLAN BASBAUM

UNIVERSITY OF CALIFORNIASAN FRANCISCO, CA, USA

RANDY BUCKNER

WASHINGTON UNIVERSITY,MO, USA

DAVID CLAPHAM

HARVARD MEDICAL SCHOOL,MA, USA

PIETRO DE CAMILLI

YALE UNIVERSITY SCHOOL OFMEDICINE, CT, USA

BARRY EVERITT

UNIVERSITY OF CAMBRIDGE,UK

GORDON FISHELL

SKIRBALL INSTITUTE, NY, USA

MARY KENNEDY

CALIFORNIA INSTITUTE OFTECHNOLOGY, CA, USA

LYNN NADEL

UNIVERSITY OF ARIZONA,

AZ, USA

DENNIS O’LEARY

THE SALK INSTITUTE, CA, USA

TERRY SEJNOWSKI

THE SALK INSTITUTE, CA, USA

WOLF SINGER

MAX-PLANCK-INSTITUT FÜRHIRNFORSCHUNG, GERMANY

CLAUDIO STERN

UNIVERSITY COLLEGE LONDON,UK

PATRICK TAM

CHILDREN'S MEDICALRESEARCH INSTITUTE, SYDNEY,AUSTRALIA

RICHARD W. TSIEN

STANFORD UNIVERSITYSCHOOL OF MEDICINE, CA, USA

RAFAEL YUSTE

COLUMBIA UNIVERSITY, NY, USA

Whether you are a home-loving,faithful type or more of a Casanovacould be further from your controlthan you might think. Reporting in Nature, Lim and colleagues ele-gantly show that the expression of asingle gene, V1aR, is responsible forpromiscuous versus monogamousbehaviour in voles.

Voles are ideal for studying socialbehaviour as different species of voleshow different levels of social attach-ment. For example, male prairievoles are faithful creatures that seekclose physical contact with theirmates and care for their young, incontrast to their more promiscuousrelatives, meadow voles, who livemore solitary lives and give minimalcare to their young.

Vasopressin and dopamine arethought to be key signals that regulate the extent of attachmentbetween animals. In fact, vasopressin1a receptors (V1aRs) are more abun-dant in the ventral forebrain — partof the brain’s reward system — inmale prairie voles than in malemeadow voles. Using viral vector-mediated gene transfer, Lim et al.showed that increasing the expres-sion of V1aR in the ventral forebrainof male meadow voles caused themto spend more time with their origi-nal female partner after sex and tospurn other females, in contrast to control animals. Moreover, theauthors show that dopamine has animportant role in this behaviour, asblocking dopamine receptors inmeadow voles, as previously shown

in prairie voles, cancels out the effectsof V1aR overexpression.

Lim et al. argue that these resultscan be attributed to the simultaneousactivation of reward and recognitionpathways causing parallel V1aR anddopamine-receptor activation in the ventral forebrain. This, they say,triggers an association between the rewarding nature of sex and theunique physical characteristics ofthe female vole, leading to partnerpreference.

The similarity in related molecu-lar and neuronal circuitry betweenthese two species of vole, and the dis-tinct effect of altering the expressionof one receptor in this circuitry, indi-cates the remarkable capacity of evo-lution to act selectively on a singlegene to induce profound changes insocial behaviour.

It is likely that other genes in thesepathways and their interactions withsocial and environmental factors alsocontribute to the variability in thisbehaviour. That said, key regulators,such as V1aR, will provide importantinsights into the molecular mecha-nisms in the brain that contribute tocomplex social behaviour, and mightin turn shed light on some disordersof social behaviour in humans.

Alison Rowan, Copy Editor,Nature Reviews Genetics

References and linksORIGINAL RESEARCH PAPER Lim, M. M. et al.Enhanced partner preference in a promiscuousspecies by manipulating the expression of a singlegene. Nature 429, 754–757 (2004)FURTHER READING Insel, T. R. & Young, L. J.The neurobiology of attachment. Nature Rev.Neurosci. 2, 129–136 (2001) | Balaban, E. Whyvoles stick together. Nature 429, 711–712 (2004)WEB SITELarry Young’s laboratory:http://www.emory.edu/YERKES/YOUNG/

Taming of the vole

B E H AV I O U R A L G E N E T I C S

596 | AUGUST 2004 | VOLUME 5 www.nature.com/reviews/neuro

Meadow voles. Image courtesy of M. Lim, Emory University.

© 2004 Nature Publishing Group

Throughout embryonic neurogenesis, neuronal precursorsgenerate progenitor cells and migratory neurons; the ratio ofthese two cell types is determined by whether the precursor celldivides symmetrically or asymmetrically. Zhang and colleagueshave investigated whether the same processes might operate inthe adult brain, and they report in The Journal of Neurosciencethat stroke not only increases neurogenesis in adult rats but alsoshifts the balance of dividing cells towards symmetric divisionand production of progenitor cells.

In early development, symmetric divisions predominate inthe ventricular zone, producing progenitor cells that areproliferative throughout life. Symmetric division, therefore, hasthe potential to contribute exponentially to neuronalproduction. By contrast, half the daughter cells that areproduced by asymmetric division remain progenitors, and theother half become young migratory neurons.

In the adult mammalian brain, the subventricular zone (SVZ)harbours neural stem cells that divide asymmetrically. Agreeingwith previous reports that neurogenesis in the SVZ is induced bystroke, Zhang et al. report that the number of dividing cells in therat SVZ increased after experimental stroke.

In addition, the authors assessed the orientation of mitoticcells as an indicator of whether the cells were dividingsymmetrically or asymmetrically. Horizontal and vertical cleavageorientations produce asymmetric and symmetric divisions,respectively. Four and seven days after stroke, Zhang et al. foundan increased percentage of vertically orientated dividing cells infixed sections of brain; the same increase was not found incontrol animals or more than seven days after stroke.

A shift in neurogenesis was also evident from the frequency of cells that assumed neuronal identity, as shown byimmunoreactivity to β-tubulin III. Four days after stroke, thefrequency had increased to 47% in cells that dividedsymmetrically, compared with 33% in cells that dividedasymmetrically.

The findings indicate that stroke reverses the usual preferenceof neuronal precursors in the SVZ to divide asymmetrically.As well as increasing total neurogenesis, stroke seems to shift theweighting towards the production of progenitor cells. This mightreflect a strategy of coping with neuronal loss by expanding thepool of progenitor cells.

Finally, the authors looked at the possible involvement of themembrane-associated signalling protein Numb, which directsasymmetric cell division during development. However, theyreport that distribution of the protein was unchanged afterstroke. So, identifying the signals that mediate these changes inneurogenesis remains a target for future studies.

Patrick Stevenson, Web Editorial Intern,Nature Publishing Group

References and linksORIGINAL RESEARCH PAPER Zhang, R. et al. Stroke transiently increases subventricularzone cell division from asymmetric to symmetric and increases neuronal differentiation in theadult rat. J. Neurosci. 24, 5810–5815 (2004)FURTHER READINGLo, E. H. et al. Neurological diseases: mechanisms, challenges and opportunities in stroke.Nature Rev. Neurosci. 4, 399–414 (2003)

PARP1 allows lasting memories

N E U R O N A L P L A S T I C I T Y

Activation of polyADP-ribose-polymerase-1 (PARP1) is normally aresponse to DNA damage, but newfindings indicate that it also con-tributes to the formation of long-termmemories — perhaps by facilitatinggene transcription.

PARP1 is responsible for poly-ADP-ribosylation — a temporarymodification that influences thebinding of nuclear proteins to DNA.Although PARP1 is usually acti-vated by DNA damage, it can alsobe activated in neurons by depolar-ization, so Cohen-Armon et al.investigated whether it is involvedin the cellular processes that under-lie memory formation in the seaslug Aplysia californica.

A. californica responds to anunpleasant stimulus by sensitizing itswithdrawal reflexes. A single noxiousstimulus produces short-term sensiti-zation, but repeated stimulation pro-duces long-term sensitization that canlast weeks. This sensitization is due tofacilitation of synapses between sensory neurons and motor neuronsby the neuromodulator serotonin (5-HT, 5-hydroxytryptamine), andlike other forms of long-term memoryit requires gene expression.

The authors showed that repeatedapplication of 5-HT to isolatedpleural-pedal ganglia (which containthe facilitated neurons) caused acti-vation of PARP1, although a single 5-HT application (which causes tem-porary facilitation) did not. In the

intact animal, repeated noxious stim-ulation also caused PARP1 activation,specifically in the pleural-pedal gan-glion on the same side as the stimulus.

Another form of training can also produce long-term memory inA. californica. Feeding responses arereduced for several days after a train-ing session in which the animals aregiven inedible food that cannot beswallowed. This training also acti-vated PARP1, this time in the centraland buccal ganglia that control feed-ing. Furthermore, blocking PARP1activation by applying 3-amino-benzamide (3-AB) just before train-ing prevented the development oflong-term memory, although it didnot affect short-term memory.

It is unclear how these types ofstimulation cause PARP1 to be acti-vated. Another important question iswhat function polyADP-ribosylationhas in the development of long-termmemory. Cohen-Armon and col-leagues also show that PARP1 acts onthe linker histone H1 in response totreatment with 5-HT. As polyADP-ribosylation of H1 causes relaxationof chromatin structure, this couldallow rapid access to DNA for genetranscription, which is essential forlong-term memory.

Rachel Jones

References and linksORIGINAL RESEARCH PAPER Cohen-Armon, M.et al. Long-term memory requires polyADP-ribosylation. Science 304, 1820–1822 (2004)FURTHER READING Kraus, W. L. & Lis, J. T. PARPgoes transcription. Cell 113, 677–683 (2003)

Stroke symmetry

N E U R O G E N E S I S

NATURE REVIEWS | NEUROSCIENCE VOLUME 5 | AUGUST 2004 | 597



R E S E A R C H H I G H L I G H T S

IN BRIEF

Local-feature assembling in visual pattern recognitionand generalization in honeybees. Stach, S. et al. Nature 429, 758–761 (2004)

Continuing on a theme of recent years in demonstratingincreasingly complex sensory and cognitive abilities in the tinybrains of bees, the authors show that honeybees can linkdifferent features in learning a visual pattern, and can generalizetheir responses to novel stimuli that share the layout of thetrained stimuli. Generalization depended on the number ofedges that were the same between two layouts. The patternrecognition required stimulation of the achromatic L-photoreceptor system.

In vivo convergence of BMP and MAPK signalingpathways: impact of differential Smad1phosphorylation on development and homeostasis. Aubin, J. et al. Genes Dev. 18, 1482–1494 (2004)

Effector molecules called Smads integrate signalling pathwaysthat are important during development and in homeostasis.Aubin et al. show that a carboxy-terminal mutant Smad thatcannot undergo transcriptional activation by BMP (bonemorphogenetic protein) recapitulates many of the features of aSmad-null mutant, whereas a mutant in which Smad cannot bephosphorylated by MAPK (mitogen-activated protein kinase)has various other phenotypes. It seems that signalling by BMPand MAPK through Smads needs to be precisely balanced.

Localized chemical release from an artificial synapse chip.Peterman, M. C. et al. Proc. Natl Acad. Sci. USA 101, 9951–9954 (2004)

The authors have developed a chip that uses electro-osmosis todeliver tiny quantities of chemical compounds through 5-µmapertures. They cultured PC12 cells on this ‘artificial synapse chip’and can stimulate individual cells by controlled release ofbradykinin. As a prototype neural interface, the chip could havemany applications in basic research and clinical treatments.

Matching behavior and the representation of value inthe parietal cortex.Sugrue, L. P. et al. Science 304, 1782–1787 (2004)

Both reward history and reward probability influence decisionmaking. To investigate how these factors are represented byneurons, Sugrue et al. used an eye-movement task in whichmonkeys were rewarded with different probability for choosingdifferent targets. The behaviour of the monkeys tracked changingreward probabilities and could be predicted by a simple model.The authors found that neurons in the parietal cortex representedthe relative value of available choices as predicted by the model.

C O G N I T I O N

N E U R OT E C H N I Q U E S

D E V E LO P M E N T

V I S U A L P R O C E S S I N G

‘Bad mothers’ have long been held responsible for behavioural problems intheir children; remember, for example, the now disgraced idea that‘refrigerator mothers’ bear the blame for children’s autism. In rats, we knowthat the amount of nursing care a dam lavishes on her pups influences theiradult response to stress. New work from Michael Meaney’s group, publishedin Nature Neuroscience, indicates a plausible mechanism to explain how earlymaternal behaviour might wield such long-term influence over the offspring.

Stress activates the hypothalamic–pituitary–adrenal (HPA) axis, whichincreases plasma glucocorticoid levels. By means of a negative-feedbackloop, glucocorticoids dampen the HPA response. In adult rats that enjoyedhigh levels of maternal care in their first week of life, glucocorticoid receptor(GR) expression in the hippocampus is increased, making the whole HPAaxis more sensitive to glucocorticoid feedback. Possible mechanisms havebeen suggested for how maternal care might acutely increase GR expression,but how this effect is maintained throughout life remains a mystery.

The regulation of gene transcription is highly complex, and onemechanism that is involved is DNA methylation. Weaver et al. comparedmethylation of the GR promoter in the adult offspring of high-caring andlow-caring dams. In the ‘high-care’ group, methylation of a particularcytosine in the binding site for the transcription factor NGFI-A was 90%less than in the ‘low-care’ group. This cytosine is highly methylated innewborn pups of either high- or low-caring dams, but in pups that werenursed (not necessarily born) by a high-caring dam, methylation droppedwithin six days, and remained low thereafter. Reduced methylationcorrelated with increased NGFI-A binding to the promoter, which isexpected to increase GR transcription.

To perturb the system, the authors infused trichostatin A (TSA), aninhibitor that should indirectly reduce DNA methylation, into the brains of adult rats. As hypothesized, TSA treatment of ‘low-care’ rats decreasedmethylation of the vital cytosine, increased NGFI-A binding, and increasedGR expression in the hippocampus. Finally, TSA treatment of ‘low-care’ ratsreduced their plasma levels of the glucocorticoid corticosterone under bothbasal and stressed conditions, implying normalized HPA axis function.

DNA methylation patterns tend to be stable over time in postmitoticcells. Indeed, one of the questions arising from the Weaver et al. study ishow exactly the de-methylation of the GR promotor cytosine isaccomplished in ‘high-care’ pups. But the very stability of methylation alsooffers an intriguing explanation for how the effects of maternal care mightlast throughout a rat’s lifespan.

Annette Markus, Associate Editor, Nature Neuroscience

References and linksORIGINAL RESEARCH PAPER Weaver, I. C. G. et al. Epigenetic programming by maternal behavior.Nature Neuroscience 27 June 2004 (doi:10.1038/nn1276)FURTHER READING Meaney, M. J. Maternal care, gene expression, and the transmission of individualdifferences in stress reactivity across generations. Annu. Rev. Neurosci. 24, 1161–1192 (2001) | Gross, C.& Hen, R. The developmental origins of anxiety. Nature Rev. Neurosci. 5, 545–552 (2004)

Imag

e by

S. F

enw

ick.

Loving mothers ease stressthrough epigenetics

B E H AV I O U R




Rico the wonder dogMove over, Lassie — there’sa new canine superhero intown. Rico the border colliemight not have rescued hisowner from a well, but asJulia Fischer’s team at theMax-Planck Institute forEvolutionary Anthropology inLeipzig report in Science(11 June 2004), he has aremarkable way with words.

Not only can nine-year-oldRico identify more than 200objects by name, he is alsocapable of a form ofreasoning called fast-mapping, which was thoughtto be exclusive to humans.This skill enables him to inferthat an unfamiliar wordcorresponds to a new object,and he can still remember theobject after being deniedaccess to it for a month.

According to The Telegraph(News online 11 June),“[Fischer and colleagues]believe they have shownsomething more than theintelligence of the breed,saying that seeminglycomplex language skills onlyseen before in children appearto be found in other species.”

However, some expertswarn against crediting Ricowith ‘language skills’. Talkingto BBC News (11 June),behavioural ecologist AlexKacelnik said, “We don’tknow that Rico has aparticular ability to linksounds to objects, ratherthan link any arbitrarystimulus to objects. Ricoprobably has the generalability to connect things —not a language ability.”

Also, some dog ownersremain unconvinced thatobedience equalsintelligence. Science writerJames Gorman cites theexample of a former familypet, a Pomeranian who“Would do anything, as longas there was food involved.When I thought I wasteaching him to come when Icalled, by giving him a treat,what he learned was that ifhe ran away from me as oftenas possible, I would have tocall him more often. Thatwould mean more treats”(New York Times, 21 June).

Heather Wood

IN THE NEWS

Axon guidance is not just a matter ofsending axons in the right direction —they must also terminate in the rightplace. The vertebrate retinotectal sys-tem is an excellent model for studyingthis mechanism, and John Flanaganand colleagues have developed aningenious new in vitro assay to investi-gate how ephrin signals regulate thetermination of retinal ganglion cell(RGC) axon growth in the tectum.

Axons from RGCs in the temporalhemisphere of the retina terminate inthe anterior tectum, whereas axonsfrom the nasal region terminate inthe posterior tectum. Ephrin-A2 and-A5 are expressed in a low-anterior to high-posterior gradient across the tectum, and their receptors, theEphAs, are expressed in a high-temporal to low-nasal gradient inRGCs. Previous in vitro studiesfocused predominantly on how RGCaxons make binary decisions, such aswhether or not to grow on a stripe ofephrin-expressing tissue. However,Hansen et al. introduced a quantita-tive dimension by exposing RGCsfrom different nasal–temporal levelsto varying ephrin concentrations.

The authors cultured explants ofmouse retina (numbered 1–8 todenote their position along thenasal–temporal axis) on membrane‘carpets’ from various regions of thechick tectum. They found that bothnasal and temporal RGC axons grewbetter on anterior than posterior tectal

tissue. As might be expected, nasalaxons grew better on posterior tectaltissue than did the temporal axons.

Next, the authors cultured retinalexplants on carpets of membrane vesi-cles from cell lines that had been tran-siently transfected with an ephrin-A2or ephrin-A5-expressing construct.They observed a smoothly gradedresponse to ephrins across the retina.On ephrin-A5 carpets, explants from the nasal region (position 1)showed the most prolific axon out-growth, whereas on ephrin-A2 car-pets, outgrowth peaked at position 3.Interestingly, this pattern mirrors theexpression gradients of these mole-cules across the mouse superior col-liculus (the mammalian equivalent ofthe tectum) — ephrin-A5 expressionpeaks at the posterior end, whereasephrin-A2 is expressed in a biphasicgradient, peaking around the middleof the posterior half.

To quantify the response toephrin-A2 further, the authors pre-pared vesicle carpets consisting ofvarying proportions of membranesfrom ephrin-A2-transfected cells andnon-transfected anterior tectal cells.At the highest ephrin concentration(100% ephrin-A2 membranes), therewas no axon outgrowth from any ofthe retinal explants, whereas with 50%ephrin-A2 membranes there was anabrupt switch from limited outgrowthto no outgrowth between the nasaland temporal halves. With 15%

ephrin-A2 membranes there was agraded biphasic response, with maxi-mal outgrowth occurring at retinalposition 3.

Intriguingly, there was little or noaxon outgrowth from any of theexplants with 0% ephrin-A2 mem-branes, indicating that low levels ofephrin-A2 have a growth-promotingeffect on RGC axons. The positioningof RGC axons in the tectum is thoughtto require a counterbalance betweenattraction and repulsion, with thegrowth cones stopping where the twoforces balance out. Hansen and col-leagues’ findings indicate that ephrinscould provide both of these forces, andthe data point to a model to explainhow attraction turns to repulsion. Therepellent actions of ephrins seem todepend on intracellular signalling,whereas ephrin-mediated attractionseems to rely on adhesion. Therefore,they propose that adhesion predomi-nates at low ephrin concentrations,but as the levels increase, intracellularsignalling pathways are activated,thereby generating a repulsive forcethat overrides the attraction.

Heather Wood

References and linksORIGINAL RESEARCH PAPER Hansen, M. J. et al. Retinal axon response to ephrin-As shows agraded, concentration-dependent transition fromgrowth promotion to inhibition. Neuron 42,717–730 (2004)FURTHER READING Wilkinson, D. G. Multipleroles of Eph receptors and ephrins in neuraldevelopment. Nature Rev. Neurosci. 2, 155–164(2001)

Knowing when to stop

A X O N G U I D A N C E




Retinoic acid (RA) is known to havemany roles in neural

development, but itseffects at the developingsynapse are relatively

unexplored.In a study

reported inthe Journal of

Cell Science, Liao andcolleagues investigated

how RA influences neuro-transmitter release at embryonic

neuromuscular synapses, and theyfound that it enhances acetylcholine(ACh) release by means of a novelmechanism that does not require

protein synthesis or transcriptionalchanges.

The authors applied RA to nerve–muscle cultures from Xenopus laevisembryos, and measured spontaneoussynaptic currents (SSCs) by takingwhole-cell recordings from the post-synaptic muscle cells. They showedthat RA increased the frequency ofthese currents, indicating an increasein either neurotransmitter releasefrom the presynaptic terminal, or thesensitivity of the postsynaptic cell toACh. They ruled out the possibility ofan increase in postsynaptic sensitivityby showing that there was no changein the SSC amplitude distribution,and they concluded that RA was act-ing primarily on the presynaptic cell.

So, how does RA enhance AChrelease at the neuromuscular synapse?

As RA is lipophilic, it might simplydisrupt the presynaptic membrane.However, Liao et al. showed that all-trans retinol, which is more disruptive to the membrane thanRA, had no effect on SSC frequency.The rapid onset and decay of theRA-induced enhancement wasincompatible with an event thatrequired protein synthesis or transcriptional changes, and it wasunaffected by inhibitors of proteinsynthesis, such as anisomycin orcycloheximide. The authors specu-late that RA modulates the activityof components of the secretorymachinery through a more directmechanism, possibly involvingpost-translational modification.

Interestingly, an agonist of theRA receptor RARβ — one of thenuclear receptors that usually medi-ates RA’s effects on transcription —also enhanced ACh release at theneuromuscular synapse. Therefore,

Polyglutamine (polyQ) repeat expansions areinvolved in several human disorders, perhapsthe most famous among them beingHuntington’s disease (HD). The HD gene wascloned in 1993, and subsequently, much efforthas gone into the molecular characterizationof the disease. Two significant new strides inthis direction have just been made — Schaffaret al. show that transcription factors drawninto the bad company of mutant huntingtin(HTT) cannot bind their DNA targets,whereas Choo et al. find that HTT associateswith the mitochondrial outer membrane andthat mutant HTT alters mitochondrialpermeability.

Expanded tracts of polyQ might causedisease through negative interactions withthe ubiquitin–proteasome system or withimportant proteins, such as TATA-box-binding proteins (TBP) and transcriptionalco-activator CREB-binding proteins (CBP),which also contain polyQ repeats. Schaffar et al. used an in vitro system, mouseneuroblastoma cells and a yeast model tofocus on the second hypothesis.

By using a combination of mutant andwild-type HTT and TBP with or withoutpolyQ tracts, they showed that TBP binds to

HTT only when the polyQ tracts are present.Not surprisingly, aggregated TBPs cannotbind their targets. Mutant HTT is able toexert this toxic effect on transcription factorseven before the formation of themultiprotein aggregates, which are alsoknown as inclusion bodies and which are ahallmark of the disease at the cellular level.The authors also show in vitro and in mouseneuroblastoma cells that insoluble pre-existing aggregates of HTT cannot interactwith TBP; only the newly synthesized,soluble HTT can.

Toxic forms of polyQ only form after aconformational change in mutant HTT —Schaffar et al. show that the non-toxicglutathione S-transferase (GST)-taggedHTT becomes toxic when the GST iscleaved. Drawing a parallel with the in vivosituation, they suggest that similar to theGST tag in the experimental setting, the restof the HTT protein shields the polyQrepeats from adopting the toxicconformation. They propose that thisshielding process might account for the lateonset of the disease.

The late onset of the disease might alsoinvolve Hsp70/Hsp40 chaperones, which also

prevent HTT polyQ tracts from adopting thetoxic conformation. These chaperones mightalso protect polyQ tracts on other proteins,such as TBPs, and so prevent the formation ofcomplexes between these proteins and HTT.The authors suggest that as age takes its tollon the efficiency of the chaperone system, thesymptoms of disease begin to appear.

HD has also been associated withmitochondrial dysfunction and this aspect ofthe disease was the focus of the study byChoo et al. Not only is the wild-type andmutant HTT associated with the outermitochondrial membrane, but, as the authorsshow in a mouse knock-in HD model, theexpanded polyQ tract causes mitochondrialmembranes to be leaky, leading to what isknown as mitochondrial permeabilitytransition — an important contributor toapoptosis and necrosis.

The gene that is responsible for HD wascloned more than 10 years ago and althoughwe still do not know what HTT, we arecoming ever closer to understanding howexpanded polyQ repeats cause this and otherpolyQ-associated diseases.

Magdalena Skipper, Editor,Nature Reviews Genetics

References and linksORIGINAL RESEARCH PAPERS Schaffar, G. et al. Cellulartoxicity of polyglutamine expansion proteins: mechanism oftranscription factor deactivation. Mol. Cell 15, 95–105 (2004) |Choo, Y. S. et al. Mutant huntingtin directly increasessusceptibility of mitochondria to the calcium-inducedpermeability transition and cytochrome c release. Hum. Mol.Genet. 13, 1407–1420 (2004)

Drawn into bad company

H U M A N G E N E T I C S

A rapid response to RA

S Y N A P T I C P H Y S I O LO G Y




The hindbrain is organized in segmental unitscalled rhombomeres that give rise to cranialsensory ganglia and motor nuclei. Loss-of-function mutants show that Hoxb1 is required forthe development of both central and peripheralcomponents of the developing nervous system atthe level of rhombomere 4, but its precise role isstill unclear. Reporting in Genes andDevelopment, Arenkiel and colleagues show thatHoxb1 is important in establishing andmaintaining the VII cranial nerve circuitrybetween the rhombomere-born motor nuclei andthe neural crest-derived branchial arch tissues.

They first conducted a genetic chimaeraanalysis by forming blastomere aggregatescomprised of cells from both wild-type andHoxb1–/– embryos. The authors found thatanimals with high numbers of Hoxb1–/– cells hadvery few of the neuronal progenitors that wouldnormally give rise to facial branchiomotorneurons (FBMs). Interestingly, the few remainingwild-type cells that express Hoxb1 maintain anormal developmental program. These findingsindicate that Hoxb1 is required cell autonomouslyfor early neuronal specification.

Arenkiel et al. then studied the role of Hoxb1 inperipheral tissues by using Cre/loxP-mediatedconditional mutagenesis to remove the conditionalallele of Hoxb1 (Hoxb1C) from specific tissues.They used two constructs, Wnt1-Cre and AP2-Cre,which targeted pre- and post-migratory neuralcrest cells, respectively, for Cre expression andconcomitant deletion of Hoxb1. By removingHoxb1 in the pre-migratory neural crest cells, theauthors reproduced null mutant phenotypes suchas the failure to blink or to move whiskers, whichwere consistent with defects in FBM development.By contrast, there was no behavioural deficit inconditional mutants in which expression of Hox1bin post-migratory neural crest cells was removed.

By comparing the number of motor neurons inHoxb1C/–/Wnt1-Cre+/– mutants at different stagesof embryogenesis, the authors were able topinpoint a time when the loss of Hoxb1 in theneural crest affected FBM development. Theyfound that the neuron numbers in the mutantsbegan to decrease at embyonic day (E) 14.5, whennormal neuronal pruning, through apoptosis, isinitiated during development. By E16.5, when cellloss is almost complete, the mutants had only athird as many motor neurons as wild-typeanimals. Therefore, expression of Hoxb1 in thepre-migratory neural crest cells is important forthe survival of motor neurons at a stage whenneuronal elimination occurs.

The authors next investigated thedevelopmental fates of Hoxb1-expressing neuralcrest cells. Genetic lineage analysis shows that, inadult animals, peripheral Hoxb1 derivativesinvolved in FBM circuits give rise primarily toglia and that a large percentage of these glial cellsare associated with the VII nerve. Interestingly,more than 95% of the FBM axons are myelinatedby glia that express the same set of Hox genes,which is the molecular signature of cells derivedfrom a particular rhombomere segment. Thisindicates that there is a correlation betweenmotor neurons born in rhombomere 4 and thesubsequent ensheathment by glia derived fromthe same segment.

This elegant study reveals novel functions forHoxb1 in FBM development and delineates thetemporal and physical element of its action.The study also sheds light on how complexdevelopmental programmes are coordinatedthrough dynamic regulation of gene expression inmultiple cell and tissue types at a given axial level.It opens doors for future research to identifydownstream targets involved in this process.

Jane QiuReferences and links

ORIGINAL RESEARCH PAPER Arenkiel, B. R. et al. Hoxb1 functionsin both motoneurons and in tissues of the periphery to establish andmaintain the proper neuronal circuitry. Genes Dev. 18, 1539–1552(2004)FURTHER READING Trainor, P. A. & Krumlauf, R. Patterning thecranial neural crest: hindbrain segmentation and Hox gene plasticity.Nature Rev. Neurosci. 1, 116–124 (2000)

Hoxb1’s double act

D E V E LO P M E N TRA might act through a classicalreceptor to facilitate spontaneousneurotransmitter release by meansof a non-classical mechanism.

RA has recently been implicatedin long-term synaptic plasticity inthe brain, and these new findingsindicate that it also regulates synapticefficacy in the developing peripheralnervous system. Neuronal activity isimportant for synaptic competitionand maturation at the neuromuscu-lar junction, and for the assembly ofpostsynaptic structures, so RA mighthave an important role in regulatingthese processes.

Heather Wood

References and linksORIGINAL RESEARCH PAPER Liao, Y.-P. et al.Non-genomic regulation of transmitter release byretinoic acid at developing motoneurons inXenopus cell culture. J. Cell Sci. 117, 2917–2924(2004)FURTHER READING Maden, M. Retinoidsignalling in the development of the centralnervous system. Nature Rev. Neurosci. 3,843–853 (2002)




Educational efforts

• IBRO-edu:http://www.ibroedu.org

IBRO — the InternationalBrain Research Organization— is undertaking anambitious project on itsrecently relaunched website.‘IBRO-edu’ is a web resourcethat, in the short term, aimsto provide a registry ofeducational resourcesrelated to neuroscience thatare already available on theinternet. In the longer term,IBRO intends to add to theeducational tools bysoliciting and creating newcontent — presumably to fillin any gaps that are leftuncovered.

The creators of the site are inviting authors ofeducational material that isposted on the Web tosubmit it for inclusion inIBRO-edu. If you have aneducational website thatcould be included, you canregister it online. The qualityof the resources that areincluded will be ensured bythe editorial committee,headed by the Editor-in-Chief of IBRO-edu, AntePadjen (McGill University,Montréal, Canada).

The site is now up andrunning, with some materialalready included. You cansearch for material related toa particular topic, or you canbrowse through subjectareas. Although coverage israther patchy at present,there is some useful materialon the site. However, it willonly fulfil its potential if thecreators of the vast amountsof educational materialalready present on the webjoin in by submitting detailsof their sites to theorganizers.

IBRO’s shiny new site alsohas plenty of other essentialreading, particularly foryounger neuroscientists. The‘Map and Compass’ section,which bills itself as a ‘survivalguide for youngneuroscientists’, containstips on writing a grant,preparing a poster,submitting work to journals,and more.

Rachel Jones

WEB WATCH

Successful synaptic transmission involves intricateinteractions of an array of proteins at the active zone.How these biochemical interactions are translated intophysiological output is still largely unknown. Reportingin Neuron, Calakos and colleagues provide a detailedanalysis of the molecular mechanism whereby an activezone protein, RIM1α, regulates neurotransmitterrelease.

RIM1α is localized at the presynaptic active zone andhas been implicated in short- and long-term plasticity. Itcontains several protein-binding domains, and isregarded as a scaffold that interacts with many proteinsthat are involved in the late stages of neurotransmitterrelease. Previous studies in mice and worms indicatethat RIM1α functions at a stage after the docking ofsynaptic vesicles at the active zone. However, severalregulated steps lead up to transmitter release, includingvesicle priming, binding of calcium to presynapticsensors and fusion of vesicles with the plasmamembrane, so which steps require RIM1α? Calakos et al.addressed this issue by carrying out a detailed analysis ofpresynaptic funtion in RIM1α-deficient neurons .

The authors applied the whole-cell recordingtechnique to hippocampal autaptic cultures preparedfrom RIM1α–/– mice. They noted that RIM1α–/– synapseshad a 50% reduction in excitatory postsynaptic charge,which could result from a decrease in synapse numbers,postsynaptic receptor response or the synapticprobability of neurotransmitter release (P

r). RIM1α–/–

and wild-type mice had the same number ofsynapses, and there was no difference in the amplitudeor frequency of miniature excitatory postsynapticcurrents (mEPSCs), so the authors conjectured thatcompromised synaptic responses in RIM1α–/– neuronswere probably a result of decreased P

r.

RIM1α binds directly to Munc13-1, an active zoneprotein that is essential for synaptic vesicle priming, soCalakos et al. asked whether priming was affected in theabsence of RIM1α. The priming process influences P

r

by determining the number of vesicles in the ‘readily

releasable pool’ (RRP). The authors found that the RRPwas reduced by 50% in RIM1α–/– synapses, and thatthere was no evidence for a deficiency in subsequentvesicle exocytosis.

Having established RIM1α as a priming factor,Calokos et al. turned to examine the effect of RIM1αdeficiency on short-term plasticity. They found thatwhen RIM1α–/– synapses were challenged with high-frequency stimulus trains, they could sustain responsesthroughout, whereas there was a 50% reduction in EPSCamplitude in wild-type cultures. Intriguingly, althoughthere was a reduction in initial P

rin RIM1α–/– synapses,

Prat steady state during high-frequency activity was not

altered. The authors discovered that this was due not to adifference in activity-dependent refilling of RRP, but toan increase in the vesicle release probability.

Previous studies have indicated that calcium has animportant role in regulating interactions of RIM1α withother synaptic proteins. Therefore, Calakos et al.suspected that calcium-dependent neurotransmitterrelease might be abnormal in RIM1α–/– synapses. Theyfound that although the overall calcium responsivenesswas unchanged, the ‘asynchronous’ (slow) component ofcalcium-dependent release was markedly reduced inRIM1α-deficient synapses.

This detailed analysis shows that RIM1α is a keyregulator of vesicle maturation at the active zone, frompriming to calcium-dependent triggering of synapticvesicle fusion. These results support the idea thatRIM1α acts as a scaffold to localize various active-zonecomponents and to integrate their actions. The studyalso provides the first direct link between a synapticprotein and the poorly understood processes thatcontrol asynchronous neurotransmitter release.

Jane Qiu

References and linksORIGINAL RESEARCH PAPER Calakos, N. et al. Multiple roles for the activezone protein RIM1α in late stages of neurotransmitter release. Neuron 42,889–896 (2004)FURTHER READING Rizo, J. & Sudhof, T. C. Snares and Munc18 in synapticvesicle fusion. Nature Rev. Neurosci. 3, 641–653 (2002)

A scaffolder at the active zone

S Y N A P T I C P H Y S I O LO G Y


research highlights - wofford college...nal female partner after sex and to spurn other females, in...

Documents