nature reviews - molecular cell biology -

November 2000 Vol 1 No 2 CONTENTS

81 | In this issue doi:10.1038/35040108

Highlights PDF [1847K]

83 | CELL DIVISION Foreman in the histone factory doi:10.1038/35040000

84 | WEB WATCH A forum for all doi:10.1038/35040018

84 | CYTOSKELETON Subunits 1: Polymers 0 doi:10.1038/35040020

84 | DNA REPAIR Recognition is crystal clear doi:10.1038/35040003

85 | RNA PROCESSING Control freak doi:10.1038/35040110

85 | IN BRIEF TRANSCRIPTION | CELL POLARITY | CELL SIGNALLING doi:10.1038/35040023

86 | WEB WATCH From signal to sequence doi:10.1038/35040025

86 | CELL MOTILITY The sting in WASP's tail doi:10.1038/35040027

87 | CHROMOSOME BIOLOGY Sisters find a slick way to stick doi:10.1038/35040030

87 | TRANSCRIPTION A molecular Swiss-army knife doi:10.1038/35040032

88 | DEVELOPMENT AND CANCER Tampering with the cell cycle's brakes doi:10.1038/35040034

91 | CADHERINS IN EMBRYONIC AND NEURAL MORPHOGENESIS Ulrich Tepass, Kevin Truong, Dorothea Godt, Mitsuhiko Ikura & Mark Peifer doi:10.1038/35040042

[1620K]

101 | THE EXPANDING POLYMERASE UNIVERSE Myron F. Goodman & Brigette Tippin doi:10.1038/35040051

[714K]

110 | SECRETS OF ACTIN-BASED MOTILITY REVEALED BY A BACTERIAL PATHOGEN Lisa A. Cameron, Paula A. Giardini, Frederick S. Soo & Julie A. Theriot doi:10.1038/35040061

[678K]

120 | APOPTOSIS IN NEURODEGENERATIVE DISORDERS Mark P. Mattson doi:10.1038/35040009

[1647K]

130 | GRABBING THE CAT BY THE TAIL: MANIPULATING MOLECULES ONE BY ONE Carlos Bustamante, Jed C. Macosko & Gijs J. L. Wuite doi:10.1038/35040072

[1148K]

137 | NUCLEAR COMPARTMENTALIZATION AND GENE ACTIVITY Claire Francastel, Dirk Schübeler, David I. K. Martin & Mark Groudine doi:10.1038/35040083

[851K]

145 | TIMELINE THE METEORIC RISE OF REGULATED INTRACELLULAR

of 2toc_r

12/3/2006http://www.nature.com.ezproxy.library.wisc.edu/nrm/journal/v1/n2/toc_r.html

88 | APOPTOSIS Viral pirates hijack Bcl-2 doi:10.1038/35040037

89 | MEMBRANE TRAFFIC FYVE fingers grab endosomes doi:10.1038/35040006

89 | IN BRIEF APOPTOSIS | TRANSLOCATION | STEM CELLS | TECHNOLOGY doi:10.1038/35040040

PROTEOLYSIS R. John Mayer doi:10.1038/35040090

[523K]

149 | TIMELINE BIOLOGICAL MACHINES: FROM MILLS TO MOLECULES Marco Piccolino doi:10.1038/35040097

[815K]

153 | OPINION SLOW AXONAL TRANSPORT: STOP AND GO TRAFFIC IN THE AXON Anthony Brown doi:10.1038/35040102

[1117K]

157 | NatureView doi:10.1038/35040112

back to top NATURE REVIEWS | MOLECULAR CELL BIOLOGY

© 2000 Nature Publishing Group

of 2toc_r

12/3/2006http://www.nature.com.ezproxy.library.wisc.edu/nrm/journal/v1/n2/toc_r.html

HIGHLIGHTS

NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 1 | NOVEMBER 2000 | 83

The cell, like any factory, must oftenstep up supply to meet demand. Forexample, during S phase of the cellcycle, the supply of histones has to beincreased to decorate the newly syn-thesized DNA. Two papers in Genesand Development explain how thekinase CDK2 and its regulatory part-ner cyclin E coordinate the synthesisof histones and DNA through aCDK2 substrate called NPAT.

Humans have two clusters of his-tone genes, on chomosomes 1 and 6,but the transcription factors drivinghistone expression vary. Zhao andcolleagues reasoned that there mustbe a ‘master regulator’ of histoneexpression and set out to find it.Having previously identified NPAT ina screen for cyclin E–CDK2 sub-strates, they used immunofluores-cence to study its cellular localization.This revealed two tiny dots of NPATin non-S-phase cells, but four in Sphase. This localization overlappedwith that of coilin, a component of anuclear organelle called the Cajalbody or coiled body (see picture) — afinding corroborated by Ma and col-leagues. Cajal bodies often associatewith histone gene clusters, so thisfinding provided an intriguing linkbetween cyclin E–CDK2 and histonegenes. Furthermore, fluorescence insitu hybridization showed thatNPAT’s association with the histonegene cluster on chromosome 1 wascell cycle dependent, explaining whythe number of NPAT dots increasesduring S phase.

Next, Zhao et al. found a large

increase in gene expression driven bythe histone H4 promoter when theNPAT gene was cotransfected into thecells. NPAT also enhanced expressionfrom the H2B and H3 promoters. Forthe H4 promoter, the authors nar-rowed down the NPAT-responsiveregion to a sequence that binds aputative transcription factor calledH4TF-2. Mutations in this sequencethat abolish H4TF-2 binding blockedthe effect of NPAT, whereas cotrans-fection of cyclin E- and CDK2-expressing plasmids enhanced NPAT-mediated transcriptional activationof histone genes.

Ma and colleagues determined theCDK2 phosphorylation sites onNPAT, then used phospho-NPAT-specific antibodies to show that phos-pho-NPAT colocalizes with bothcyclin E and coilin in Cajal bodies,and that the combination of phos-pho-NPAT and cyclin E–CDK2 ispresent in Cajal bodies only during Sphase. Furthermore, mutation ofNPAT’s CDK2 phosphorylation sitesto alanine reduced NPAT’s ability toactivate transcription from the his-tone 2B promoter.

So cyclin E–CDK2 gives theorders, and NPAT ensures that they’recarried out. Appreciating NPAT’sskills will be our next lesson in thistour of the histone factory: how doesNPAT manage its staff — presumablythe histone-gene-specific transcrip-tion factors — and does it have otherteams with responsibilities beyondhistone production?

Cath Brooksbank

Foreman in the histone factory

C E L L D I V I S I O N

HIGHLIGHTS ADVISORS

JOAN S. BRUGGE

HARVARD MEDICAL SCHOOL,BOSTON, MA, USA

PASCALE COSSART

INSTITUT PASTEUR, PARIS,FRANCE

GIDEON DREYFUSS

UNIVERSITY OF PENNSYLVANIA,PHILADELPHIA, PA, USA

PAMELA GANNON

CELL AND MOLECULARBIOLOGY ONLINE

JEAN GRUENBERG

UNIVERSITY OF GENEVA,SWITZERLAND

ULRICH HARTL

MAX-PLANCK-INSTITUTE,MARTINSRIED, GERMANY

NOBUTAKA HIROKAWA

UNIVERSITY OF TOKYO, JAPAN

STEPHEN P. JACKSON

WELLCOME/CRC INSTITUTE,CAMBRIDGE, UK

ROBERT JENSEN

JOHNS HOPKINS UNIVERSITY,BALTIMORE, MD, USA

VICKI LUNDBLAD

BAYLOR COLLEGE OFMEDICINE, HOUSTON, TX, USA

TONY PAWSON

SAMUEL LUNENFELD RESEARCHINSTITUTE, TORONTO, CANADA

NORBERT PERRIMON

HARVARD MEDICAL SCHOOL,BOSTON, MA, USA

THOMAS D. POLLARD

THE SALK INSTITUTE, LA JOLLA, CA, USA

JOHN C. REED

THE BURNHAM INSTITUTE, LA JOLLA, CA, USA

KAREN VOUSDEN

NATIONAL CANCER INSTITUTE,FREDERICK, MD, USA

JOHN WALKER

MRC DUNN HUMAN NUTRITIONUNIT, CAMBRIDGE, UK

One of Cajal’s drawings of his ‘accessorybodies’, from a paper published in 1910.This organelle was then forgotten for 60years until it was rediscovered and namedthe coiled body by W. Bernard. Imagecourtesy of Joseph Gall, CarnegieInstitution, Baltimore, Maryland, USA.

References and linksORIGINAL RESEARCH PAPERS Zhao, J. et al.NPAT links cyclin E–Cdk2 to the regulation ofreplication-dependent histone gene transcription.Genes Dev. 14, 2283–2297 (2000) | Ma, T. et al.Cell cycle-regulated phosphorylation of p220NPAT

by cyclinE/Cdk2 in Cajal bodies promotes histonegene transcription. Genes Dev. 14, 2298–2313(2000) FURTHER READING Ewan, M. E. Where the cellcycle and histones meet. Genes Dev. 14,2265–2270 (2000)

© 2000 Macmillan Magazines Ltd

84 | NOVEMBER 2000 | VOLUME 1 www.nature.com/reviews/molcellbio

H I G H L I G H T S

The field of slow axonal transport isdivided between those who believethat cytoskeletal subunits are trans-ported along the axon (the subunitmodel) and those who believe thatentire filaments move (the polymermodel). What both camps have incommon is ignorance of the precisetransport mechanism, whatever thecargo might be. Reporting in Cell,Hirokawa and colleagues now pro-vide evidence that slow axonal move-ment requires kinesin motors andmicrotubule tracks.

The visualization of slow axonaltransport is not trivial (see AnthonyBrown’s article on page 153 of thisissue), and finding an appropriatemodel system is half the job.Hirokawa and colleagues used thesquid giant axon in their studiesbecause it has two advantages: it istranslucent and it is big. They injectedfluorescent tubulin into the axon andmeasured the speed at which it

moved away from the cell body. Thevalues obtained with this experimen-tal set-up were similar to thosereported for mammalian axons.

In a series of pharmacologicalstudies, the speed of movement wasconsiderably reduced when micro-tubules were depolymerized, butremained constant in the absence ofpolymerized actin microfilaments.Similarly, transport was slowed downwhen kinesin’s motor activity wasinhibited, but myosin seemed to bedispensable for this process.

Hirokawa and colleagues alsoobserved that the diffusion rate ofthe transported tubulin is lower thanthe diffusion rate of creatine kinase(another cargo for slow axonal trans-port), but higher than the diffusionrate of taxol-stabilized microtubules.The authors interpret this finding asan indication that tubulin is trans-ported in a complex that is large butdifferent from a fully polymerized

microtubule. This would tilt the bal-ance in favour of the subunit modelagain, at least for the transport oftubulin. Figuring out the polymer-ization state of tubulin transportedin this complex will hopefully clarifythis issue.

Raluca Gagescu

A forum for allSorting the wheat from thechaff when it comes to websites is always tricky,especially in fields of publicinterest. But for anybodyseeking to learn more aboutAlzheimer’s disease, theAlzheimer Research Forum isa good place to start.Although the site has beenaround for some years(indeed, the archive goesback to February 1996), it isregularly updated andcontains an impressivevariety of sections and links.

First, though, the visitormust identify themself —layperson, physician orresearcher? After logging inthe layperson is directed toan ‘Alzheimer generalinformation directory’containing basic informationabout the latest research,and links to Alzheimerassociations and supportgroups around the world. Asa physician or researcher,however, you are directed toa different home page.

Here you’ll find theexpected lists of relevantpapers, some of which havelinks through to PubMed,and ‘Abstracts in Advance’from The Journal ofAlzheimer’s Disease. There’salso a directory listing “genesthat have been studied inrelation to their role inAlzheimer’s disease” — againwith useful links to publicdatabases. Other featuresthat catch the eye are the‘Forum Interviews’ with well-known researchers such asDennis Selkoe and BruceYankner, the variousmutations directories (APP,presenilins and tau), and the‘Virtual Conferences’, whereyou can listen to recordingsof the speakers.

There are a few glitches inthis otherwise excellent site.For example, many of thenewest ‘Papers of the Week’do not yet contain PubMedlinks, and the ‘MilestonePapers’ section needsupdating (for instance, thereis no mention of the recent γ-secretase studies). Butoverall this is an easilynavigable, useful site.

Alison Mitchell

WEB WATCH

DNA polymerases are not infallible— they make mistakes whilereplicating DNA. Sometimes theseerrors are deliberate (see, forexample, the review by Goodmanand Tippin on page 101 of thisissue). But often they are not, andthat’s when mismatch repair kicksin to protect against mutation. Thissystem identifies DNA bases thathave been incorrectly paired up,and allows the correct base to bereinserted. But how does itrecognize the mismatches? Twogroups, reporting in the 12 Octoberissue of Nature, have studied thebacterial mismatch-repair proteinMutS to address this question.

In the first paper, Obmolova et al.present the crystal structures of aThermus aquaticus MutShomodimer, both alone and as acomplex with DNA containing asingle unpaired thymidine. The

authors showthat the DNA adoptsan unusual kinkedshape owing tointeractions withtwo domains fromeach MutSmonomer(domainsI and IV).However,theseinteractions are asymmetric,with domain I from monomerA in the diagram donating thephenylalanine residue (yellow ring)that interacts specifically with theunpaired base.

This message — that the MutShomodimer is actually aheterodimer at the structural level— also emerges from the secondpaper by Lamers and colleagues.These authors report the crystal

structure of Escherichia coli MutSbinding to a G•T mismatch.Mismatch binding is known toinduce the uptake of ATP, and bothgroups show that the ATPasedomains also differ between thetwo MutS monomers. Thisasymmetry between the monomersin DNA and ATP binding couldexplain the specificity of MutS for

Subunits 1: Polymers 0

C Y TO S K E L E TO N

C-l C-l'

N-l

N-l'

Domain I (B)

N-IVDomain IV (B) N-IV'

Domain IV (A)

Recognition is crystal clear

D N A R E PA I R

The axonal cytoskeleton. Courtesy of N. Hirokawa, University of Tokyo, Japan.


H I G H L I G H T S


References and linksORIGINAL RESEARCH PAPER Terada, S., Kinjo, M. & Hirokawa, N. Oligomeric tubulin in large transporting complex is transported viakinesin in squid giant axons. Cell 103, 141–155(2000) FURTHER READING Hirokawa, N. Themechanisms of fast and slow transport in neurons:identification and characterization of the newkinesin superfamily motors. Curr. Opin. Neurobiol.7, 605–614 (1997)

The yeast Saccharomyces cerevisiaelikes to be in control, especially whenit comes to gene expression. Wealready knew that transcription,splicing, messenger RNA export,mRNA stability, mRNA translationand post-translational modificationare regulated processes. Tollervey andcolleagues now report in Cell thatdegradation of unspliced pre-mRNAs occurs in the nucleus in aregulated manner, providing yetanother regulatory mechanism forgene expression.

Using several mutants defective inmRNA processing, Tollervey and co-workers found that unspliced pre-mRNAs are degraded in the nucleusin a 3′ to 5′ direction by a large pro-tein complex containing severalexoribonucleases (the exosome), or ina 5′ to 3′ direction by the exonucleaseRat1p. This is, in fact, similar to whatis known to happen in the cytosol,where mRNAs are degraded either 3′to 5′ by the exosome or 5′ to 3′ by thecytosolic exonuclease Xrn1p. Butcontrary to what happens in thecytosol, exosome-mediated degrada-tion is predominant in the nucleus.

Nuclear degradation of pre-mRNA seems to compete with splic-ing. It is increased in the presence ofglucose — yeast’s favourite food —indicating that it is a physiologicalregulatory pathway for gene expres-sion. The 3′ to 5′ Rat1p-dependentpathway is probably inhibited by thecap structure of the pre-mRNA, pro-viding another level of control.

Degradation of inaccuratelyspliced pre-mRNAs has also beenobserved in mammalian cells, andthere again, the activity seems to benuclear. An obvious experiment willbe to test whether homologues of thegenes identified in yeast have thesame function in mammalian cells.

Raluca Gagescu

References and linksORIGINAL RESEARCH PAPER

Bousquet–Antonelli C., Presutti, C. & Tollervey, D.Identification of a regulated pathway for nuclearpre-mRNA turnover. Cell 102, 765–775 (2000)FURTHER READING Mitchell, P. & Tollervey, D.Musing on the structural organization of theexosome complex. Nature Struct. Biol. 7,843–846 (2000)

mispaired DNA bases;presumably the observedconformational changes occuronly in response to binding amismatch.

Why are these papers sosignificant? The human cousinsof bacterial MutS — MSH2 andMSH6 — are mutated inhereditary nonpolyposiscolorectal cancer (HNPCC). Notonly can these mutations bemapped to the new MutSstructures but, according toObmolova et al.,“the crystalstructures … provide amolecular framework forunderstanding HNPCCmutations”.

Alison Mitchell

References and linksORIGINAL RESEARCH PAPERS Obmolova,G. et al. Crystal structures of mismatch repairprotein MutS and its complex with a substrateDNA. Nature 407, 703–710 (2000) | Lamers,M. H. et al. The crystal structure of DNAmismatch repair protein MutS binding to a G•Tmismatch. Nature 407, 711–717 (2000) FURTHER READING Kolodner, R. D.Guarding against mutation. Nature 407,687–689 (2000)

Control freak

R N A P R O C E S S I N G IN BRIEF

Dynamic association of capping enzymes withtranscribing RNA polymerase II.Schroeder, C. et al. Genes Dev. 14, 2435–2440 (2000)

Different phosphorylated forms of RNA polymerase IIand associated mRNA processing factors duringtranscription.Komarnitsky, P., Cho, E.-J. & Buratowski, S. Genes Dev. 14, 2452–2460 (2000)

RNA polymerase II is a molecular platform to which manymessenger RNA-processing factors bind during transcription.Are such factors associated simultaneously with RNA pol II, ordo they interact in a transient and sequential manner? UsingmRNA-capping enzymes, both papers indicate that RNA-processing enzymes associate dynamically with differentlymodified forms of the polymerase at different stages of thetranscriptional cycle.

Plasma membrane compartmentalization in yeast bymessenger RNA transport and a septin diffusionbarrier. Takizawa, P. A. et al. Science 290, 341–344 (2000)

Saccharomyces cerevisiae restricts the cellular distribution of thetranscription factor Ash1p by transporting its messenger RNAinto the forming bud. Takizawa et al. now present a list of othermRNAs that also localize to the bud, and the transportmechanism for at least one of them — which encodes atransmembrane protein — is the same as for ASH1 mRNA. Theprotein is then retained in the bud by a diffusion barrierinvolving septins that is functionally similar to tight junctions inepithelial cells.

RasGRP is essential for mouse thymocytedifferentiation and TCR signalling.Dower, N. A. et al. Nature Immunol. 1, 317–321 (2000)

Control of pre-T cell proliferation and differentiation bythe GTPase Rac-1.Gomez, M. et al. Nature Immunol. 1, 317–321 (2000)

These papers describe the actions of two small GTPases, Ras andRac, in T-cell responses. The first sheds light on how RasGRP, aRas activator that’s directly sensitive to diacylglycerol, mediatesresponses that had previously been assumed to be a result of thearchetypal diacylglycerol-sensitive enzyme — protein kinase C.The second uses Rac mutants that can control actin dynamics,but not other downstream targets of Rac such as mitogen-activated protein kinases, to show that changes in actindynamics are sufficient to drive some stages of T-celldifferentiation.

C E L L S I G N A L L I N G

C E L L P O L A R I T Y

T R A N S C R I P T I O N



From signal to sequenceNo scientist is an island and,as the Human GenomeProject testifies, we canachieve remarkable thingswhen we work as a team.Biocarta has adopted acommunity approach tomapping cellular pathways: itprovides the tools; thescientific community provides— and constantly updates —the data.

The Biocarta websitecontains several types ofinformation, but its pathwaydiagrams, which cover allareas of cellular regulationfrom cell division toapoptosis, immunology toneuroscience, are the mainattraction. Clicking on apathway category gives you alist of pathways: for example,clicking on cell-cycleregulation links to a menucontaining the ATM pathway,regulation by cyclins, the p53pathway and more. Clickingon any single pathwayprovides a stylized,information-packed diagramthat evolves as users providenew information. In the future,pathways will also have‘gurus’ responsible forassessing submittedinformation on their pathway.If you don’t agree with whatyou see, you can sendcomments to a discussiongroup and help the pathwayto evolve.

Each component of thepathway is a gateway to awealth of information: clickingon any protein takes you to atable containing links to justabout any public-domaindatabase you can think of.Whether you wantsequences, structures,information on geneticdiseases, or just somerelevant abstracts, you canlink to them from here.

If your favourite pathwayisn’t in the list, you can submitit. Biocarta even provides atemplate with which to drawyour pathway diagram. If thetemplate doesn’t have theright components, you cansend diagrams in any format,“even a paper napkin”. Sodon’t sit back and watch thissite evolve: join in!

Cath Brooksbank

WEB WATCH

H I G H L I G H T S

How are the many stimuli that tell cells to movetranslated into changes in actinpolymerization? Evidence points tomembers of the Wiskott–Aldrichsyndrome protein (WASP) family asthe interpreters but, for scientists, the language ofcell movement has proved difficult to learn. Two papersin The Journal of Cell Biology provide someclues as to how a lipid and a proteincollaborate to activate two WASP-familymembers. The details seem protein specificbut the general message is the same— activation of WASPsinvolves stopping them frombiting their own tails.

Actin nucleation is stimulated by theArp2/3 complex, which is activated by WASPs. Asmall GTPase, Cdc42, and a phospholipid,phosphatidylinositol-4,5-bisphosphate(PtdIns(4,5)P

2), activate WASPs, but how they do so

is controversial: recombinant WASPs often havesome constitutive activity, and inactive, GDP-boundCdc42 sometimes seems capable of activating them.

Henry Higgs and Tom Pollard sought to end thiscontroversy — at least for the haematopoietic-cell-specific member of the family, WASP — bypurifying native WASP from bovine thymus. They findthat purified WASP alone has no effect on actinpolymerization rates, but that micelles containingPtdIns(4,5)P

2activate polymerization through WASP. In

the presence of Cdc42, PtdIns(4,5)P2

micelles, or vesiclescontaining either PtdIns(4,5)P

2or another acidic

phospholipid, phosphatidylserine, produce halos ofpolymerized actin surrounding the phospholipid (seepicture). This effect requires PtdIns(4,5)P

2or

phosphatidylserine, and Cdc42 must be both GTP-bound and prenylated, indicating that it needs to bemembrane associated to do its job.

Rohatgi and colleagues, working with a recombinantform of the widely expressed N-WASP, have a differentstory: they find that Cdc42, but not PtdIns(4,5)P

2, can

partly activate N-WASP, and that the two moleculessynergize to activate N-WASP fully. They identify a basicregion, close to the Cdc42-binding domain, that seems tobind PtdIns(4,5)P

2. In actin nucleation assays, a mutant

N-WASP lacking this domain remains sensitive to Cdc42but is insensitive to the additive effects of Cdc42 andPtdIns(4,5)P

2. These researchers previously showed that

PtdIns(4,5)P2

stimulates actin polymerization inXenopus egg extracts. They now show that this effectdepends on N-WASP but, curiously, the deletion mutantcan also translate a PtdIns(4,5)P

2signal into limited

actin polymerization, albeit more slowly.WASP’s carboxyl terminus is constitutively active,

suggesting an autoinhibitory mechanism. Both groupsshow that a separate Cdc42-binding domain can curb the

activity , including the C motif and an acidic region (Cand A in the figure), of the carboxyl terminus in trans.Full-length N-WASP, however, can’t inhibit N-WASP’scarboxyl terminus, presumably because the full-lengthprotein is folded into its autoinhibited conformation.Higgs and Pollard find that inhibition of WASP’scarboxyl terminus is relieved by GTP–Cdc42 but not byPtdIns(4,5)P

2, whereas Rohatgi and co-workers find that

their intermolecular inhibitory complex is regulated inexactly the same way as wild-type N-WASP.

We’re still straining to understand what WASPs, Cdc42and PtdIns(4,5)P

2are saying to each other. Perhaps

WASP and N-WASP respond to their two activatorsslightly differently, or maybe the discrepancies are due tovariations between purified and recombinant proteins.Is WASP’s PtdIns(4,5)P

2-binding domain equivalent to

N-WASP’s? And can N-WASP be activated byphosphatidylserine? Further studies should clarify howCdc42 and acidic phospholipids unleash WASP’s sting.

Cath Brooksbank

References and linksORIGINAL RESEARCH PAPERS Higgs, H. N. & Pollard, T. D. Activation byCdc42 and PIP2 of Wiskott–Aldrich syndrome protein (WASP) stimulates actinnucleation by Arp2/3 complex. J. Cell Biol. 150, 1311–1320 (2000) | Rohatgi,R. et al. Mechanism of N-WASP activation by CDC42 and Phosphatidylinositol4,5-bisphosphate. J. Cell Biol. 150, 1299–1309 (2000)FURTHER READING Cameron, L. A. et al. Secrets of actin-based motilityrevealed by a bacterial pathogen. Nature Rev. Mol. Cell Biol. 1, 110–119 (2000)

How Cdc42 and acidic phospholipids might activate WASP.Inset shows an actin-filament halo (red) surrounding a vesicle(green). Photo courtesy of Henry Higgs and Tom Pollard, SalkInstitute, La Jolla, California, USA.

Actin polymerization

C

C

N

N

CA

A

WH1 PolyprolineCDC42-binding

WH1 PolyprolineCDC42-binding

GTP-boundCDC42

GDP-boundCDC42

Inactive WASP(Autoinhibited)

Active WASP

Acidicphospholipid

Prenyl group

Pi

C

The sting in WASP’s tail

C E L L M OT I L I T Y



A molecularSwiss-army knife

T R A N S C R I P T I O N

Like a bottle of wine, heterochromatin is nouse unless you can open it. TAF250 is oneprotein that helps to uncork the DNA byloosening histone H1’s grip on it. A paper byPham and Sauer in Science suggests thatTAF250 is more a Swiss-army knife than acorkscrew, though: it was already known tophosphorylate and acetylate histone H1, butnow it adds ubiquitylation to its list oftalents list.

Polyubiquitylation — adding a chain ofubiquitin molecules to proteins — is a signalfor destruction, but monoubiquitylation isless well understood. Knowing that histoneH1 can be monoubiquitylated, Pham andSauer set out to discover what was doing it,and found TAF250. Ubiquitylation requires aminimum of two enzyme activities, anactivator and a conjugator, which are usuallyon separate proteins. In vitro assays revealedthat TAF250 has both. But what about invivo? Having identified the region of TAF250that’s necessary for histonemonoubiquitylation, they created mutantsthat lack the activity and are heterozygous fordorsal — a maternally expressed geneimportant for early Drosophila development.In these mutants, expression of the dorsal-response genes twist and snail wassignificantly reduced, resulting in a twistedphenotype. Levels of monoubiquitylatedhistone were also reduced. More substrates ofTAF250’s ubiquitylating activity, as well asfactors that ubiquitylate the other histones,may await discovery.

Cath BrooksbankReferences and links

ORIGINAL RESEARCH PAPER Pham, A.-D. & Sauer, F.Ubiquitin-activating/conjugating activity of TAF250, a mediator ofactivation of gene expression in Drosophila. Science 289,2357–2360 (2000) FURTHER READING Mizzen, C. A. & Allis, C. D. New insights into an old modification. Science 289, 2290–2291(2000)

In some ways, meiotic division is like countrydancing. During meiosis I, homologous chro-mosome pairs line up opposite one another andthen, at anaphase I, whirl apart to opposite polesof the cell. But in the next reel, meiosis II, thesister chromatids themselves separate. So whatis the molecular ‘caller’ that tells the partnerswhen to stick together and when to separate?Dean Dawson and colleagues, reporting inOctober’s Current Biology, have identified a keyplayer in the process.

Clues to what’s going on in meiosis can begleaned from events during mitosis. Sister chro-matids are initially bound along their length by aphysical tie — the so-called cohesin complex.Consisting of four subunits, Smc1p, Smc3p,Scc1p and Scc3p, cohesin establishes this connec-tion and maintains it until anaphase. At thispoint, sister-chromatid cohesion is lost — firstalong the arms and then at the centromeres —by proteolytic cleavage of the Scc1p subunit. Thiscohesion must be maintained during meiosis Iand, in addition, the kinetochores (multiproteincomplexes at the centromeres) must be regulatedsuch that both sisters are pulled to one polerather than being dragged away from each other.

Dawson and colleagues realized that yeastwith a mutation in the SLK19 (synthetic lethalkar3) gene show a defect in meiosis consistentwith a failure in the control of sister-chromatidbehaviour: rather than forming tetrads contain-ing four haploid spores, these mutants formdyads with two diploid spores. The authorsfound that although slk19 mutants enter meio-sis I efficiently, the sister chromatids then sepa-rate to opposite poles of the cell. Moreover,most of the mutants do not go through a sec-ond meiotic division — that’s why they formdyads. This could be due to a breakdown of sis-ter-chromatid cohesion, to impaired kineto-chore function, or to both.

Dawson and co-workers next used indirectimmunofluorescence to monitor the localiza-tion of Slk19p during meiosis. The pictureshows synaptonemal complexes (red; these runalong paired meiotic chromosomes) andSlk19p–GFP (green). The authors found thatSlk19p localizes to centromeric regions duringprophase I, and that it remains there untilanaphase I. After this point, however, Slk19ploses its association with the centromeres, andlocalizes instead along the spindles, as it is leftbehind when the centromeres migrate pole-wards at meiosis I.

These localization patterns indicate that

Slk19p might keep the sister chromatids togeth-er during the first meiotic division. But how?The authors speculated that it could affectRec8p, which is the yeast meiosis-specifichomologue of Scc1p. Rec8p must be localized atthe centromeres to ensure sister-chromatidcohesion through meiosis I, so perhaps Slk19pprevents the degradation of Rec8p? To test this,Dawson and colleagues looked at the effect ofSlk19p on localization of Rec8p. They foundthat, during the early stages of meiosis I, thelocalization of Rec8p is indistinguishablebetween slk19 and wild-type strains. Inanaphase I, however, Rec8p is maintained at thecentromeres in 80% of the wild-type cells, butin most of the slk19 cells there is almost noRec8p staining.

How can a protein found in mitotic cellsspecifically regulate the behaviour of sister chro-matids during meiosis? The answer may be thatSlk19p acts downstream of another meiosis-spe-cific factor, and the authors’ top candidate is aputative meiotic regulatory protein,Spo13p.Bothspo13 and slk19 mutants have defects in thebehaviour of sister chromatids during meiosis.But the defect in spo13slk19 double mutants is nomore serious than that in the spo13 mutant alone,indicating that the two proteins may be involvedin a common pathway. The next step in studyingthe molecular dance of meiosis is to work outhow Slk19p and Spo13p might prevent the part-ners from separating too soon.

Alison Mitchell

References and linksORIGINAL RESEARCH PAPER Kamieniecki, R. J., Shanks, R.M. Q. & Dawson, D. S. Slk19p is necessary to preventseparation of sister chromatids. Curr. Biol. 10, 1182–1190 (2000)FURTHER READING Nasmyth, K., Peters, J.-M. & Uhlmann, F.Splitting the chromosome: cutting the ties that bind sisterchromatids. Science 288, 1379–1384 (2000)

Sisters find a slick way to stick

C H R O M O S O M E B I O LO G Y



Viruses are modern-day buccaneers. Theyride the cellular seas, hijacking proteins andusing them to promote their own survival.But how? Ojala et al., reporting in theNovember issue of Nature Cell Biology,describe how the Kaposi’s sarcomaherpesvirus (aka human herpesvirus 8;HHV8) interferes with apoptotic signallingpathways in its host.

The virus encodes a pirated cyclin (v-cyclin), which forms a complex with acellular cyclin-dependent kinase, CDK6.This complex can induce apoptosis, andOjala et al. now show that it probably doesso by phosphorylating — and inactivating— the cellular anti-apoptotic molecule Bcl-2. The authors find that Bcl-2 andCDK6 associate in cell extracts, and theyshow that the targets for phosphorylationare two serine residues in an unstructured‘loop’ region of cellular Bcl-2.

But herein lies a paradox. Why shouldthe virus inactivate a molecule that protectsthe host cell from apoptotic death? Onemight think it in the best interests of thevirus to keep its host alive. It turns out,however, that cellular Bcl-2 has otherfunctions that the virus finds less thansavoury; for instance, it has been reportedto impair cell-cycle progression whenoverexpressed. Moreover, host cell death isadvantageous to the virus in that it allowsthe spread of viral particles.

There is a problem, though — the riskthat, when the virus blocks cellular Bcl-2,the host cell will die before the viralreplication cycle is complete. Here, in anultimate act of viral skulduggery, HHV8produces its own Bcl-2 homologue. Thisvirus-encoded protein lacks the crucialunstructured loop, so it cannot bephosphorylated or inactivated by v-cyclin–CDK6.

Infection of host cells with HHV8 haspreviously been shown to be linked toapoptosis, and the lesions associated withKaposi’s sarcoma contain some apoptoticcells. The next step, then, will be to work outhow the acts of viral piracy uncovered byOjala et al. link HHV8 to tumour formationand to the development of Kaposi’ssarcoma.

Alison Mitchell

References and linksORIGINAL RESEARCH PAPER Ojala, P. M. et al. Theapoptotic v-cyclin-CDK6 complex phosphorylates andinacticates Bcl-2. Nature Cell Biol.2, 819–825 (2000)FURTHER READING Hardwick, J. M. Cyclin’ on the viral pathto destruction. Nature Cell Biol. 2, E203–E204 (2000) |Desagher, S. & Martinou, J. C. Mitochondria as the centralcontrol point of apoptosis. Trends Cell Biol. 10, 369–377 (2000).

To stop the cell cycle, tumour suppressorssuch as the retinoblastoma protein (Rb) applythe brakes. A Nature paper by Anna Lasorellaand colleagues describes how Id2 — a domi-nant-negative inhibitor of helix–loop–helixDNA-binding proteins — gets the wheelsturning again.

Retinoblastoma protein is essential formammalian development: knockouts dieduring embryogenesis. Lasorella et al. reportthat knocking out Id2 rescues Rb–/– embryos.Defective myogenesis kills Id2–/–Rb–/– miceshortly after birth, but they show none of thehallmarks of Rb–/– mice — too much prolifer-ation and apoptosis in the haematopoieticand nervous systems. Id2 therefore perpe-trates the Rb–/–phenotype; but how?Immunoprecipitates revealed that active,hypophosphorylated Rb binds Id2 and, to tipthe balance against tumour suppression, Id2has to be in molar excess of Rb.

An intact Rb pathway is needed to preventtumorigenesis, so Lasorella and colleagues

asked whether Id2 is overexpressed in neu-roblastoma cell lines, in which N-myc ampli-fication typically bypasses Rb. Remarkably,Id2 expression correlated with N-myc ampli-fication. What’s more, Myc’s effect on Id2expression (which also extends to c-Myc) isdirect, owing to two high-affinity Myc-bind-ing sites in the Id2 promoter. Deletion ofthese sites abolished Id2 expression inresponse to Myc.

By producing a surfeit of Id, then, Myc canoverride Rb’s attempts to halt the cell cycle.But what targets of Id are responsible for theRb–/– phenotype? And what about the otherId family members, Id1 and Id3? These ques-tions, and others, await the next cycle tour.

Cath Brooksbank

References and linksORIGINAL RESEARCH PAPER Lasorella, A. et al. Id2 is aretinoblastoma protein target and mediates signalling by Myconcoproteins. Nature 407, 592–598 (2000)FURTHER READING Yokota, Y. et al. Development ofperipheral lymphoid organs and natural killer cells depends onthe helix-loop-helix inhibitor Id2.Nature 397, 702--706 (1999)

Tampering with the cell cycle’s brakes

D E V E LO P M E N T A N D C A N C E R

Viral pirates hijack Bcl-2

A P O P TO S I S

H I G H L I G H T S


H I G H L I G H T S


Over the years, many investigatorshave set out to look for a protein reg-ulator of their favourite protein andinstead found a phosphoinositide.Most labs have therefore developed atleast a peripheral interest in this fami-ly of lipids. So they will be happy toread in the EMBO Journal that thelaboratories of Harald Stenmark andRob Parton have devised a new probeto study the cellular localization ofphosphatidylinositol-3-phosphate(PtdIns(3)P).

The products of phosphatidyli-nositol-3-OH kinases function inprocesses as diverse as signal trans-duction, cytoskeletal organizationand apoptosis. PtdIns(3)P is particu-larly interesting for membrane trafficaficionados, as it regulates transportalong the endocytic pathway in allspecies where this has been studied.One of its activities is to recruit pro-teins that contain PtdIns(3)P-bindingFYVE finger domains to membranes.But to which membranes?

Gillooly et al. reasoned that ifPtdIns(3)P binds FYVE domains,then FYVE domains should bindPtdIns(3)P. They built a probe(2XFYVE) consisting of two FYVEdomains from Hrs, a protein that actsin the endocytic pathway. A series ofcontrol experiments showed that2XFYVE binds selectively toPtdIns(3)P, both in vivo and in vitro.The probe effectively competes withendogenous proteins for PtdIns(3)Pbinding when transfected into cells,and can be used for immunofluores-cence as well as for immunoelectronmicroscopy studies.

The next step was to use 2XFYVEto localize PtdIns(3)P in the cell. TheFYVE finger protein EEA1 is known to

localize exclusively to early endosomes,where it is involved in membranefusion. So it was clear from the startthat there must be substantial amountsof PtdIns(3)P in the membrane ofearly endosomes. This was confirmedin this study — 2XFYVE, shown in redin the picture, colocalized extensivelywith EEA1, shown in green.

More surprisingly, immunoelec-tron microscopy using 2XFYVErevealed that PtdIns(3)P is also pre-sent on internal membranes of multi-vesicular late endosomes. This obser-vation led the authors to speculatethat the PtdIns(3)P-containingintralumenal vesicles arise frominvagination of the endosomal mem-brane. This could sequesterPtdIns(3)P away from the surface,stopping it from recruiting cytoplas-mic proteins such as EEA1 to lateendosomes. The origin of the convo-luted morphology of multivesicularendosomes is still mysterious, and2XFYVE might prove a useful tool tostudy this process.

This probe is not the first on themarket. The pleckstrin homology(PH) domain of phospholipase Cδ1and the PH domains of ARNO andBruton’s tyrosine kinase have beenused to detect PtdIns(4,5)P

2and

PtdIns(3,4,5)P3, respectively. But

these probes cannot be used for elec-tron microscopy, and neither of themshows the exquisite specificity for itstarget that 2XFYVE seems to have.

Raluca Gagescu

References and linksORIGINAL RESEARCH PAPER Gillooly, D. J. etal. Localization of phosphatidylinositol 3-phosphate in yeast and mammalian cells. EMBOJ. 17, 4577–4588 (2000) FURTHER READING Corvera, S., D’arrigo, A. &Stenmark, H. Phosphoinositides in membranetraffic. Curr. Opin. Cell Biol. 11, 460–465 (1999)

FYVE fingers grab endosomes

M E M B R A N E T R A F F I C IN BRIEF

P53AIP1, a potent mediator of p53-dependent apopto-sis, and its regulation by Ser-46-phosphorylated p53.Oda, K. et al. Cell 102, 849–862 (2000)

The star of this story — p53AIP1 — is a newly cloned gene, theprotein product of which localizes to the mitochondria. It causesapoptotic cell death by dissipating the mitochondrialtransmembrane potential, and it could help to mediate p53-dependent apoptosis. After severe DNA damage, phosphorylationof a specific residue (serine 46) on p53 leads to apoptosis. Theauthors show that substitution of Ser 46 not only inhibits p53-dependent apoptosis, but also blocks expression of p53AIP1.

Two intermembrane space TIM complexes interactwith different domains of Tim23p during its import intomitochondria.Davis, A. J. et al. J. Cell Biol. 150, 1271–1282 (2000)

How are mitochondrial proteins targeted for either insertion intothe mitochondrial inner membrane or translocation into thematrix? Davis et al. show that the inner membrane proteinTim23p interacts with both known intermembrane space TIMcomplexes before reaching the Tim22p inner membranetranslocon. But only its interaction with one of these complexes— Tim9p–Tim10p — is essential for correct targeting, leaving themystery of what the Tim8p–Tim13p complex does intact.

Effects of eight growth factors on the differentiation ofcells derived from human embryonic stem cells.Schuldiner, M. et al. Proc. Natl Acad. Sci. USA 97, 11307–11312 (2000)

Until now there has been no systematic attempt to determine howgrowth factors affect the lineage choice of embryonic stem cells.This broad study correlates different growth-factor treatmentswith cell morphology and expression of markers for 11 tissues,derived from all three germ layers. It shows that what you put inbiases, but doesn’t absolutely determine, what you get out, and hasobvious implications for stem-cell therapy.

Harnessing the ubiquitination machinery to target thedegradation of specific cellular proteins.Zhou, P. et al. Mol. Cell 6, 751–756 (2000)

This paper uses a neat trick to functionally ‘knock out’ stableproteins — by targeting them for proteasomal degradation. Byengineering specific protein–protein interaction domains intoone component of the SCF complex, a multimeric ubiquitin-conjugating machine, Zhou and colleagues can send proteins totheir death in both yeast and mammalian cells, and can measurethe phenotypic consequences.

T E C H N O LO G Y

S T E M C E L L S

T R A N S LO C AT I O N

A P O P TO S I S


REVIEWS

In 1963, two publications appeared that focused atten-tion on the adhesive mechanisms that contribute toembryonic morphogenesis and govern structural differ-entiation of the nervous system. Malcolm Steinberg laidthe groundwork for the ‘differential adhesion hypothe-sis’, suggesting that the segregation or ‘sorting out’ ofdifferent embryonic cell types into separate tissuesinvolves qualitative or quantitative differences in celladhesion1. Roger Sperry proposed the ‘chemoaffinityhypothesis’, which holds that specific synaptic contactsform according to differences in the adhesive propertiesof individual GROWTH CONES and synapses2. Bothhypotheses were the result of decades of experimentalwork, but were formulated before any adhesion mole-cules had been identified. It is now apparent that a largenumber of adhesion molecules exist that can begrouped into several superfamilies. The cadherin andIMMUNOGLOBULIN-TYPE ADHESION MOLECULES are the maingroups of cell–cell adhesion receptors, whereas the inte-grins are the predominant contributors to cell–sub-strate adhesion3,4. Adhesive mechanisms that contributeto embryonic or neural morphogenesis share manysimilarities, revealing that Steinberg’s and Sperry’shypotheses are essentially similar proposals applied todifferent cell populations.

Morphogenesis involves two interrelated themes:structure and movement. For example, the differentadhesive properties of two mixed cell populationsinduce cell movement, leading to the sorting out of thetwo groups of cells. After the sorting process is complet-ed, adhesive differences maintain the segregation andrelative position of the two cell groups, therefore pre-

serving a specific tissue architecture5–8. Neuronal mor-phogenesis follows a similar pattern, in which the neu-ronal growth cone has to move through a complexenvironment using differential adhesive cues. On reach-ing the target, synaptic contacts are formed and main-tained by specific adhesive interactions9.

The first cadherins to draw scientists’ attention werevertebrate classic cadherins, which were independentlyidentified for their ability to mediate calcium-depen-dent adhesion among cultured cells and for their role inthe epithelialization of the early mouse embryo10. So far,the sequences of over 300 vertebrate cadherins havebeen reported, and the virtually complete sets of cad-herins encoded by the genomes of Caenorhabditis ele-gans and Drosophila melanogaster are now known.

Structural diversity of the cadherin superfamilyThe recent explosion in genomic sequencing of variousanimals has shed new light on the diversity of the cad-herin superfamily. In humans, more than 80 membersof the cadherin superfamily have been sequenced.Current annotation of the C. elegans and the Drosophilagenomes reveals 14 and 16 cadherin genes, respectively.Cadherins are defined by the presence of the cadherindomain (CD), a roughly 110 amino-acid peptide thatmediates calcium-dependent homophilic interactionsbetween cadherin molecules (FIG. 1). The CD is typicallyorganized in tandem repeats. Calcium ions associatewith the linker region that connects two CDs, andrequire interaction with amino acids from both CDs(FIG. 1; see below).

Here we present a classification of cadherins into

CADHERINS IN EMBRYONIC ANDNEURAL MORPHOGENESISUlrich Tepass*, Kevin Truong‡, Dorothea Godt*, Mitsuhiko Ikura‡ and Mark Peifer§

Cadherins not only maintain the structural integrity of cells and tissues but also control a widearray of cellular behaviours. They are instrumental for cell and tissue polarization, and theyregulate cell movements such as cell sorting, cell migration and cell rearrangements.Cadherins may also contribute to neurite outgrowth and pathfinding, and to synapticspecificity and modulation in the central nervous system.


*Department of Zoology,University of Toronto, 25Harbord Street,Toronto,Ontario M5S 3G5, Canada.‡Division of Molecular andStructural Biology, OntarioCancer Institute andDepartment of MedicalBiophysics, University ofToronto, Toronto M5G2M9, Ontario, Canada.§Department of Biology,University of NorthCarolina, Chapel Hill,North Carolina 27599, USA.Correspondence to U.T.e-mail:[email protected]

GROWTH CONE

Exploratory tip of an extendingneuronal process such as anaxon.

IMMUNOGLOBULIN-TYPE

ADHESION MOLECULES

Family of adhesion moleculescharacterized by the presence ofimmunoglobulin-like domains,which are also found inantibody molecules.



R E V I E W S

tions that are required for the multicellular organizationof METAZOANS. The function of classic cadherins duringthe formation and maintenance of epithelial tissues andcell–cell ADHERENS JUNCTIONS — two other metazoaninventions — is of particular importance. Within theclassic cadherin subfamily, an interesting shift in proteinorganization has taken place during evolution — chor-date classic cadherins lack non-chordate classic cadherindomains (NCCDs), laminin G (LG) and epidermalgrowth factor (EGF) domains and consistently containfive CDs, in contrast to the domain structure of classiccadherins in three other phyla: ECHINODERMS, ARTHROPODS

and NEMATODES (FIG. 2, TABLE 1). This finding indicates that,during the early evolution of chordates, the structure ofclassic cadherins was modified and that a single progen-itor might have given rise to the numerous classic cad-herins found in vertebrates today, a conclusion support-ed by the phylogenetic analysis of chordate cadherins12.

The cadherin families of C. elegans and Drosophilaare small and very similar in size, and these cadheringenes are scattered throughout the genome without anyobvious clustering, indicating that gene duplicationevents might not have occurred recently. So most ofthese genes were presumably established early duringmetazoan evolution. In particular, the progenitors of thefour conserved subfamilies of cadherins are the result ofa GENE RADIATION event that occurred before nematodes,arthropods and chordates diverged. In contrast, classiccadherins and the closely related desmosomal cad-herins, as well as the more divergent protocadherins,were all amplified within the chordate lineage, resulting

subfamilies on the basis of the domain layout of individ-ual cadherins, which includes the number and sequenceof CDs, and the presence of other conserved domainsand sequence motifs (FIG. 2, TABLE 1). This analysis revealsthat four cadherin subfamilies are conserved betweenC. elegans, Drosophila and humans: classic cadherins,Fat-like cadherins, seven-transmembrane cadherins anda new subfamily of cadherins that is related toDrosophila Cad102F. Classic cadherins break up in foursubgroups, as listed in TABLE 1. Fat-like cadherins containa subgroup of highly related molecules that we callFatoid cadherins. These include all known vertebrateFat-like cadherins, Drosophila Cad76E and C. elegansCdh-4. Cadherins containing protein kinase domainsare found in vertebrates (RET cadherins) and inDrosophila. Desmosomal cadherins are presumablyderived from type I classic cadherins within the CHORDATE

lineage, as neither desmosomal cadherins nor DESMO-

SOMES are found in Drosophila or C. elegans. Finally, pro-tocadherins also seem to be limited to chordates. Thegrouping of cadherins into seven subfamilies, which islargely on the basis of the overall protein domain archi-tecture, is corroborated by sequence comparison of CDsonly (see online supplementary materials). Note thatonly about half of the cadherins found in C. elegans andDrosophila are part of identified subfamilies.

Cadherins are not found in yeast, and only a single,poorly conserved CD has been reported in a secretedprotein from Dictyostelium11, indicating that transmem-brane proteins of the cadherin superfamily might haveevolved to meet the need for the complex cell interac-

E_CAD DWVIPPISC––PENEK–GEFPKNLVQIKSNRDKET–––KVFYSITGQGADKPPVGVFIIERETGWLKVTQ–PLDREAIAKYILYSHAVSSNGEAVEDPMEIVITVTDQNDNRPEFN_CAD DWVIPPINL––PENSR–GPFPQELVRIRSDRDKNL–––SLRYSVTGPGADQPPTGIFIINPISGQLSVTK–PLDRELIARFHLRAHAVDINGNQVENPIDIVINVIDMNDNRPEF

E_CAD ––TQEVFEGSVAEGAVPGTSVMKVSATDADDDVNTYNAAIAYTIVSQDPELPHKNMFTVNRDTGVISVLTSGLDRESYPTYTLVVQAADLQG–––EGLSTTAKAVITVKDINDNAPVFN_CAD ––LHQVWNGSVPEGSKPGTYVMTVTAIDADDPN–ALNGMLRYRILSQAPSTPSPNMFTINNETGDIITVAAGLDREKVQQYTLIIQATDMEGNPTYGLSNTATAVITVTDVNDNPPEF

1

111 121 131 141 151 161 171 181 191 201 211

11 21 31 41 51 61 71 81 91 101

βA-A'

βA'

βB

βB

βA

βC

βC

βD

βD

βE

βE

βF

βF

βG

βG

αA

αA

αA

αB

αB

αB

αB

R

R

N

E

E31 K D D67

A70

I4 P5

K25

G58

G58

V34

S82

E86

P6 I S C9

A V E D P M E I V I T V98

V A H S Y L I Y K A72

F Y S I T G Q41

R E I I F V G49

W L K V T63

I Q V L N K P18

D103NQD100

E11Calcium-bindingpocket

N

C

a

c

b

Putative homophilicbinding surface

Linker

CHORDATES

Phylum that comprises animalswith a notochord and includesall vertebrates.

DESMOSOME

A patch-like adhesiveintercellular junction found invertebrate tissues that is linkedto intermediate filaments.

METAZOANS

Refers to the kingdom Animalia(animals) that comprisesroughly 35 phyla ofmulticellular organisms.

ADHERENS JUNCTIONS

Cell–cell or cell–matrixadhesive junctions that arelinked to microfilaments.

ECHINODERMS

Animal phylum of marineinvertebrates including seaurchins and starfish.

ARTHROPODS

Largest animal phylumcomposed of invertebrates thathave a segmented body,segmented appendages and anexternal skeleton. This includesinsects, spiders and crustaceans.

NEMATODES

Animal phylum ofunsegmented roundworms.

GENE RADIATION

Process that leads to theformation of gene families inwhich gene amplificationthrough gene duplicationevents is followed by thediversification of gene structureand function.

Figure 1 | Structure of the cadherin domain. The structure of the first cadherin domain (CD) of mouse E-cadherin is shown in a |and b |. It was solved by NMR spectroscopy18 and, subsequently, the crystal structure of the first CD of N-cadherin revealed asimilar topology19. The CD consists of a seven-strand β-sheet with the amino and carboxyl termini located at opposite ends of themolecule. The segment connecting strands B and C adopts an apparently helical structure consisting of a succession of β-turn andβ-like hydrogen bonds. This unique quasi-β-helix structure is characteristic of the CD. a | Schematic of topology of the amino-terminal CD of mouse E-cadherin. βA, βA ′, βB, βE, and βD (green) and βC, βF, and βG (yellow) form β-sheets. The α-helices areshown in magenta. The putative homophilic binding surface including the amino acids HAV (red), and the Ca2+-binding pocket withthe amino acids that interact with Ca2+ (blue), are indicated by the dotted lines. b | Ribbon structure of the CD. The colour coding isthe same as in (a). c | Alignment of the first two N-terminal CDs of mouse E- and N-cadherin. The colour coding is the same as in(a). The CDs are connected by a roughly 10 amino-acid linker region. Note that the amino acids that form a single Ca2+ bindingpocket (indicated in blue) are found in the first and second CDs and include animo acids in the linker region.



R E V I E W S

region 5q31–32, and many classic cadherins are orga-nized in gene clusters13–16. A similar gene amplificationis seen in other unrelated gene clusters, such as in theC. elegans collagen genes. It has been proposed that the

in numerous genes with pronounced clustering13,14. Inhumans, all six desmosomal cadherin genes are foundin chromosomal region 18q12.1; three genes thatencode more than 50 protocadherins are found in

Table 1 | Cadherin subfamilies

Cadherin subfamily Type Cadherin Species No. of CDs

(I) Classic cadherins Vertebrate Type I classic cadherins Examples include:• Highly conserved cytoplasmic • HAV motif in first CD E-cadherin (CDH1) Hs 5domain that binds to catenins P-cadherin (CDH3) Hs 5

• Often found at adherens junctions N-cadherin (CDH2) Hs 5R-cadherin (CDH4) Hs 5

Vertebrate Type II classic cadherins Examples include:• No HAV motif in first CD VE-cadherin (CDH5) Hs 5

Cadherin-7 (CDH7) Hs 5Cadherin-8 (CDH8) Hs 5Br-cadherin (CDH12) Hs 5

Ascidian classic cadherins CI-cadherin Ci 5BS-cadherin Bs 5

Non-chordate classic cadherins LvG-cadherin Lv 17• Variable number of CDs DE-cadherin Dm 8• LG and EGF domains DN-cadherin Dm 18• NCCD domain, only found in these cadherins Hmr-1 (splice products 1a/1b) Ce 3/19*

(II) Fat-like cadherins Fatoid cadherins Examples include:• Very large extracellular domain • More than 30 CDs, closely related in sequence hFat1 Hs 34

with up to 34 CDs • Flamingo box in some members hFat2 Hs 34• Heterogeneous subfamily • LG and EGF domains DCad76E (CG7749) Dm 34

• Conserved region in the cytoplasmic domain Cdh4 (F25F2.2) Ce 32(between vertebrates and fly)

Other Fat-like cadherins Fat Dm 34• Variable number of CDs Dachsous Dm 27• Flamingo box in Fat Cdh-3 (ZK112.7) Ce 19• LG and EGF domains in Fat and Cdh-3 Cdh-1 (R10F2.2) Ce 25

(III) Seven-pass • Seven-pass transmembrane domain similar to hFlamingo1 Hs 9transmembrane cadherins G-protein linked receptors (Secretins) hFlamingo2 Hs 9

• Flamingo box Flamingo/Starry night Dm 9• LG and EGF domains Cdh-6 (F15B9.7) Ce 9

(IV) DCad102F-like cadherins • Sequence conservation throughout much of KIAA0911 Hs 2the protein KIAA0726 Hs 2

• LG domain DCad102F (CG11059) Dm 2• Glu/Ser-rich cytoplasmic domain Cdh-11 (B0034.3) Ce 2

(V) Protein kinase cadherins • Tyrosine kinase domain RET Hs 2(RET-Cadherins) DRet (CG1061+CG14396) Dm 1‡

• Putative Ser/Thr kinase DCad96Ca (CG10244) Dm 1

(VI) Desmosomal cadherins Desmocollins Desmocollin-1 Hs 5• Only found in vertebrates • Conserved cytoplasmic domain Desmocollin-2 Hs 5• Localize at desmosomes Desmocollin-3 Hs 5• Interact with plakoglobin, Desmogleins Desmoglein-1 Hs 5desmoplakin and plakophillins • Conserved cytoplasmic domain Desmoglein-2 Hs 5

Desmoglein-3 Hs 5

(VII) Protocadherins Protocadherins (Pcdh) α, β and γ Examples include:• Only found in vertebrates • 52 protocadherins encoded by 3 genes Pcdh-α3 Hs 6

• all Pcdh-α/CNR proteins have a constant Pcdh-β1 Hs 6C-terminal cytoplasmic domain that interacts Pcdh-γA9 Hs 6with Fyn tyrosine kinase

Other protocadherins Examples include:Pcdh-1 Hs 6Pcdh-8 (Arcadlin) Hs 6

*J. Petite, personal communication; ‡ R. Cagan, personal communication (Species: Hs, Homo sapiens; Dm, Drosophila melanogaster; Ce, Caenorhabditis elegans, Ci,Ciona intestinalis; Bs, Botryllus schlosseri; Lv, Lytechinus variegatus. Domains: CD, cadherin domain; CNR, cadherin-related neuronal receptor; EGF, epidermal growthfactor; LG, laminin G; NCCD, non-chordate classic cadherin domain.) Uncharacterized cadherins of Drosophila are named according to their cytological map position (forexample, DCad 102F).



R E V I E W S

carboxyl termini located at opposite ends of the mole-cule (FIG. 1b). The crystal structures of peptides contain-ing the first and second CDs (CD1 and CD2) of E-cad-herin20,21, or of N-cadherin22, indicate that calcium iscentral in cis-dimer formation. Each dimer associateswith six calcium ions through residues that are locatedin the linker region between CD1 and CD2 (FIG. 3a).Single amino-acid substitutions in the calcium bindingsites can disrupt cell aggregation in vivo23. Calcium bind-ing makes CDs arrange in a rigid structure 20,21,24,25 that isresistant to proteolysis26.

Crystallographic analysis on the first and secondCDs of E-cadherin and N-cadherin have provided cluesas to how cadherin molecules induce lateral clusteringessential for the formation of a stable adhesive interfacebetween adjacent cells. Although different mechanismsunderlying cis dimerization have been observed in thecrystal structures of different cadherin molecules19,20,22,an emerging theme is that two cadherin molecules forma cis dimer that functions as a building block for lateralclustering. These and other studies indicate that cisdimerization or more extensive lateral clustering is aprerequisite for stable cell adhesion27–29. Although cisdimers might primarily form as homodimers, the for-mation of functional cis heterodimers between N- andR-cadherin has been reported30.

Adhesion between opposing cell membranesrequires the formation of trans dimers (FIG. 3b). Themechanism of trans-dimer formation is, at present, con-troversial. Several studies indicate that trans dimers formby interactions between the amino-terminal CDs ofopposing cadherin molecules19,21,28,31. These data are cor-roborated by early findings that located the homophilicbinding specificity of classic cadherins within theamino-terminal CD32,33. On the basis of the crystalstructure of the first CD from N-cadherin, a zippermodel for trans dimerization was proposed thatinvolved only the tip of the amino-terminal CD19. Thismodel provided an early foundation for understandingthe mechanics of the cell adhesion interface. However,the subsequent crystal structures of CD1 and CD2 fromN-cadherin22 and E-cadherin20,21 did not show the adhe-sion interface seen in the first CD of N-cadherin. Inaddition, a recent biophysical study indicates a differenttype of trans-dimer association, in which the five CDsshow variable degrees of lateral overlap, including thecomplete anti-parallel overlap of all five CDs (FIG. 3b)34.In the presence of calcium, the extracellular part of ver-tebrate classic cadherins forms a rod-like structure ofabout 20 nm in length with each individual CD span-ning about 4.5 nm (REFS 19,20,25). Full lateral overlap oftrans dimers would imply a distance between adjacentplasma membranes of 20–25 nm, a value consistent withthe distance between plasma membranes at adherensjunctions that is found in ultrastructural studies.

Adhesive contacts and adherens junctionsTwo types of adhesive contacts are mediated by classiccadherins: diffuse adhesive contacts all along a cell–cellcontact surface, and more discrete contacts by ultra-structurally defined adherens junctions, such as the

formation of collagen gene clusters was caused by theevolution of a complex extracellular matrix, the nema-tode cuticle17. The amplification of cadherins in verte-brates might be explained by the more complex tissueinteractions found in humans and other vertebratescompared with invertebrates, particularly the largeincrease in size and complexity in the central nervoussystem.

Structural basis of cell adhesionAlthough it is broadly accepted that the predominantrole of cadherins is to mediate adhesive interactionsbetween cells, the mechanism of adhesive contact for-mation is still a matter of intense research. Structuralstudies have focused on vertebrate classic cadherins.These molecules are believed to form two types ofdimers. Cadherins associate laterally within the sameplasma membrane to form parallel cis dimers, and cad-herins protruding from adjacent plasma membranesassociate in an anti-parallel fashion to form transdimers. The structure of the first CD of E-cadherin18

and of N-cadherin19 revealed that the cadherin foldconsists of a seven-strand β-sheet with its amino and

Figure 2 | Structural diversity of the cadherin superfamily. Representatives of each of theseven subfamilies of cadherins are shown. Subfamilies I to VII are conserved betweennematodes (Caernorhabditis elegans), arthropods (Drosophila melanogaster) and chordates(humans). Members of subfamily V are found in chordates and Drosophila, whereas cadherinsof subfamilies VI and VII are at present only known in vertebrates. Binding partners for thecytoplasmic tail of cadherins have been characterized for classic cadherins (subfamily I), fordesmosomal cadherins (subfamily VI), and for Pcdhα/CNR protocadherins (subfamily VII).These interacting factors are listed at the right. It was recently shown that DE-cadherin isproteolytically cleaved within the NCCD domain during maturation (arrowhead)95. (NCCD, non-chordate classic cadherin domain; EGF, epidermal growth factor; LG, Laminin G)

G

G G

G

G

G

E-cadherin

DE-cadherin

Fat-like cadherins

III.

Classic cadherins

Flamingo

RET

Desmocollin

Pcdhα/CNR

DCad102F-like cadherins

Catenins

PlakophillinsPlakoglobinDesmoplakin

Catenins

Fyn

?

?

?

?

Plasma membrane

Cadherin domain LG domain NCCD domain

Flamingo box Tyrosine-kinase domain

EGF domain/C-rich

DCad76E

I.

II.

IV.

V.

VI.

VII.

Seven-passtransmembranecadherins

Protein kinase cadherins

Desmosomal cadherins

Protocadherins

TK

TK



R E V I E W S

that Afadin is important for junctional organizationand epithelial integrity44. So adherens junctions containtwo complexes that interact with each other and withthe actin cytoskeleton.

Cell and tissue polarityEpithelial cells provide a clear example of cell polarity,with various molecules, including proteins, sorting todistinct apical and basolateral membrane domains. Thecrucial role of classic cadherins and their associatedcatenins in epithelial differentiation has been well docu-mented, and these protein complexes seem to be broad-ly important for forming and maintaining epithelial tis-sues45,46. Conversely, downregulation of classiccadherins, such as E-cadherin or DE-cadherin, is oftenassociated with a loss of epithelial morphology duringnormal development and in many carcinomas. Thezinc-finger transcription factor Snail is important forrepressing the expression of DE-cadherin and E-cad-herin in non-epithelial cells47–49.

Epithelial cells usually form a continuous tissuestructure. However, at certain times during normaldevelopment, or in experimental cell-culture models,epithelial cells have free edges that approach each otherto establish new lateral contacts36,50,51. The initial contactbetween cells is made by FILOPODIA or LAMELLIPODIA, andsuch contacts are stabilized by classic cadherins. Whenthese contacts broaden, cadherins concentrate in dis-crete puncta. The adhesive interactions are further sta-bilized through linkage of cadherins to the cytoskeletonand, eventually, by the formation of mature adherensjunctions. Cadherin-mediated adhesion leads to

ZONULA ADHERENS. Diffuse adhesive contacts probablyinvolve the oligomerization of cadherin trans dimers, asindividual trans dimers provide little adhesivestrength25,29. Adherens junctions could simply representvery large arrays of trans dimers. However, the situationseems to be more elaborate, as cadherins might not bethe principal components of adherens junctions, at leastin some cases. Indeed, adherens junctions can form inthe absence of Hmr-1 cadherin, the only classic cad-herin in C. elegans, or in the absence of its associatedcatenins35,36. In mouse and Drosophila embryos, whereE-cadherin or DE-cadherin, respectively, are essentialfor adherens junction assembly and epithelial integrity,markedly reduced levels of these cadherins can still sup-port the formation of normally sized adherens junc-tions37–39. These observations are inconsistent with amodel in which adherens junctions simply represent alarge array of cadherins and associated cytoplasmicproteins. Instead, cadherin trans dimers probably formsmall clusters separated by other proteins, and the den-sity of these clusters in the adherens junctions may varyconsiderably without affecting the size of the adherensjunction. A novel protein complex has recently beencharacterized that is concentrated at adherensjunctions40–42. This complex is composed of Nectin, atransmembrane protein of the immunoglobulin super-family that interacts with the PDZ-DOMAIN protein Afadin,which in turn can bind to Ponsin, a protein containingthree SH3 DOMAINS (FIG. 4a). This complex can interact withthe actin cytoskeleton. Nectin and cadherin complexesinteract with each other and are recruited together toadherens junctions43. Initial functional studies indicate

Figure 3 | Ca2+-mediated cis- and trans-dimer formation of vertebrate classic cadherins. a | Dimer interface betweentwo N-terminal repeats of E-cadherin domains 1 (Ecad1) and 2 (Ecad2). Each cadherin molecule binds three calcium ions thatare important in the rigidification and cis-dimer association of cadherins20,21. b | A cis dimer consists of two cadherin moleculeswithin the same plasma membrane that are associated laterally. The pairs of cadherin molecules from opposing cells thatassociate with one another are referred to as trans dimers. Different models for trans-dimer formation have been proposed thatsuggest different extents of lateral overlap between the extracellular regions. Red dots indicate the location of Ca2+ ions betweenadjacent CDs.

Plasma membrane

Cis dimer Trans dimer

Ecad1

Ca2+

Ecad2

NN

a b

20–25 nm

ZONULA ADHERENS

A cell–cell adherens junction thatforms a circumferential beltaround the apical pole ofepithelial cells.

PDZ DOMAINS

Protein–protein interactiondomain, first found in PSD-95,DLG and ZO-1.

SH3 DOMAINS

Src homology region 3 domains.Protein sequences of about 50amino acids that recognize andbind sequences rich in proline.

FILOPODIUM

Finger-like exploratory cellextension found in crawling cellsand growth cones.

LAMELLIPODIUM

Thin sheet-like cell extensionfound at the leading edge ofcrawling cells or growth cones



R E V I E W S

In addition to their role in apical–basal polarity, cad-herin superfamily members were recently implicated ina second form of cell polarity called planar epithelialpolarity (FIG. 5). This property is found in many epithe-lia. One example is the fly wing epithelium, where eachcell polarizes its actin cytoskeleton along theproximal–distal axis, such that a bundle of actin fila-ments polymerizes and projects from the surface at thedistal-most vertex of each cell, ultimately forming awing hair (FIG. 5a). Genes involved in establishing the pla-nar polarity of the wing epithelium include componentsof the Wnt/Frizzled signalling pathway, Frizzled andDishevelled53, and three different cadherins, Fat,Dachsous54 and Flamingo/Starry Night55,56. Althoughthe mechanism by which cadherins affect planar polari-ty is unknown, it was found that Flamingo/Starry Nightadopts an asymmetric distribution during polarityestablishment, becoming enriched at the proximal anddistal cell surfaces (FIG. 5b)55. Planar polarity also influ-ences polymerization of the microtubule cytoskeleton,manifesting itself through orientation of the mitoticspindle and thereby the axis of cell division along thebody axes. Studies in both C. elegans and Drosophilaimplicate the Wnt/Frizzled pathway in this process53.Many planar polarity genes have not been examined inthis context, but it is at least clear that Flamingo/StarryNight is essential for spindle orientation57.

Two other examples that highlight important roles ofcadherins in generating asymmetric tissue organizationare the contribution of DE-cadherin to the formation ofthe anterior–posterior axis in Drosophila, and the func-tion of N-cadherin in setting up left–right asymmetry inthe chick. A cell sorting process that is driven by differ-ent levels of DE-cadherin directs the oocyte to the poste-rior pole of the egg chamber during Drosophila oogene-sis8,58. This highly reproducible positioning event allowsthe oocyte to interact with a specific group of FOLLICLE

CELLS, thereby initiating a cascade of cell interactions thatare crucial for the formation of the embryonic anteri-or–posterior axis. Disruption of the function of N-cad-herin during chick GASTRULATION leads to a random ori-entation of the heart along the left–right axis59.Asymmetric N-cadherin expression and cell movementsthat prefigure the position of the heart and other organsalong the left–right axis are seen during gastrulation.How N-cadherin contributes to these asymmetric cellmovements remains a mystery.

Cell movementMany of the changes in cell shape or movementobserved during development occur while cells are indirect contact and require, therefore, dynamic changesin adhesive interactions. These changes may play a per-missive role, as the release of adhesion is important forthe relative movement of cells that are in contact.However, adhesive interactions also directly promotemovement, as traction must be generated between cellsfor cell rearrangements to occur in solid tissues. Todetermine whether changes in cadherin activity play apermissive or a more active role can be difficult, as illus-trated by the analysis of C-cadherin function during

recruitment of specific cytoskeletal factors, such as theactin-associated factor Mena51 and other transmem-brane proteins, to cell–cell contact sites46.

Cadherins seem to be directly involved in maintain-ing cell polarity by directing the localization of thesec6/8 complex, which specifies vesicle targeting to thelateral membrane52. This recruitment, and the continu-ous polarized delivery of specific molecular compo-nents to the lateral membrane, establishes and main-tains the lateral membrane domain of epithelial cellsand contributes to epithelial apical–basal polarity46,52.Interestingly, in fully polarized epithelial cells, the sec6/8complex is not found along the entire lateral membranebut is concentrated in close association with apicaladherens junctions, indicating a potentially direct mole-cular link between the cadherin–catenin complex andthe vesicle targeting machinery (FIG. 4a)52.

Figure 4 | Comparison between cadherin-mediated adhesive interactions at epithelialand synaptic adherens junctions. a | Schematic of the zonula adherens, a circumferentialadherens junction found in epithelial cells. This junction contains the classic cadherin–catenincomplex and the recently identified nectin/afadin/ponsin complex40–42. Both complexes interactwith the actin cytoskeleton and with each other10,43. The zonula adherens is closely associatedwith a vesicle docking site that contains the Sec6/8 complex52. b | At the interneuronal synapse,we also find a close association between adherens junctions and vesicle docking zones. Thesec6/8 complex was found to associate with the postsynaptic membrane only duringsynaptogenesis96. The classic cadherin–catenin complex is a principal component of synapticadherens junctions similar to the zonula adherens. Protocadherins that localize to synapsesinclude Arcadlin84 and the Pcdhα/CNR protocadherins75. Whether protocadherins contribute tosynaptic adhesion remains to be established. The Pcdhα/CNR protocadherins interact with thecytoplasmic protein kinase Fyn, and seem to reside within the active zone, indicating that theymight have a primary role in signalling rather than adhesion13,75.

Actin cytoskeleton

Actin cytoskeleton

Actin cytoskeleton

Actin cytoskeleton

Actin cytoskeleton

Actin cytoskeleton

Sec6/8 complex

Sec6/8 complex

PonsinAfadin

PonsinAfadin

AfadinPonsin

AfadinPonsin

Catenins Catenins

Catenins

Catenins CateninsFyn

Catenins CateninsFyn

CateninsNectin Classic

cadherin

? ?

? ?

Classiccadherin

Protocadherin

Synaptic vesicle docking

Active zone(Sec6/8 complex)

a

b

FOLLICLE CELLS

In this review, the term folliclecells refers to cells that surroundthe developing insect egg andsecrete the egg membranes, thechorion and vitelline envelope.

GASTRULATION

Series of morphogeneticmovements observed duringthe early development of mostanimals that leads to theformation of a multilayeredembryo with an outer cell layer(ectoderm), an inner cell layer(endoderm), and anintermediate cell layer(mesoderm).



R E V I E W S

beginning at NEURULATION. This expression pattern led tothe speculation that N-cadherin might be critical for thesegregation of neural and epidermal tissues duringneural tube formation. However, an essential role inneurulation was disproved by the knockout of mouseN-cadherin70. After neurulation, but before neuronaldifferentiation, many classic cadherins, including N-cadherin, are expressed in the developing central ner-vous system (CNS) in a region-specific manner thatoften coincides with morphological boundaries71. Themost direct evidence that these cadherins contribute tothe subdivision of the neuroepithelum has come fromanalysis of the Xenopus type II classic cadherin F-cad-herin72. The expression of F-cadherin confines neuroep-ithelial cells to the sulcus limitans, a region separatingthe dorsal and ventral halves of the caudal neural tube.One apparent consequence of F-cadherin expression in

convergent extension movements in Xenopus gastrula-tion. In this process, cells move towards the dorsal mid-line of the embryo, thereby rearranging by cell interca-lation, leading to an extension of the embryo along theanterior–posterior axis (FIG. 6a)60. Adhesion mediated bythe classic cadherin C-cadherin must be reduced to per-mit these movements to occur61. However, the disrup-tion of C-cadherin activity causes defects not only dur-ing gastrulation movements, but also in tissuestructure62, raising the possibility that the disruptions incell movement might be a secondary consequence of acompromised cell architecture. Similar difficulties haveemerged from the analysis of DE-cadherin in embryon-ic morphogenesis where its role in epithelial mainte-nance might mask a function in promoting cellrearrangements37,38.

Paraxial protocadherin (PAPC) seems to be directlyinvolved in convergent extension in Xenopus andzebrafish embryos, where it is expressed in the mesodermduring gastrulation63,64. PAPC, which can promotehomotypic cell adhesion, is required for convergentextension of the mesoderm. Notably, overexpression ofPAPC can promote convergent extension under certainexperimental conditions63. These findings argue thatPAPC acts as an adhesion receptor that directly pro-motes cell movement, perhaps providing traction neededfor cell motility. Alternative and non-exclusive possibili-ties are that PAPC is primarily a signalling receptor, aswas suggested for other protocadherins13. PAPC activitymight also generate the tissue polarization that isobserved during convergent extension (FIG. 6a)60,63, func-tioning similarly to the activity of other cadherins inplanar epithelial polarization, outlined above.Intriguingly, cell polarization during convergent exten-sion resembles planar polarity in that it also requiresWnt/Frizzled signalling65–67.

The requirement for DE-cadherin in cell migrationduring Drosophila oogenesis is a convincing examplefor a direct role of classic cadherins in cell migration ona cellular substrate. DE-cadherin is involved in themigration of a small group of somatic cells, the ‘border’cells, on the surface of the much larger germline cells68

(FIG. 6b). It is required in both the somatic and germlinecells for this movement to occur. DE-cadherin is notrequired for the formation of the border cell clusterand, more importantly, is not required for maintainingintegrity of the border cell cluster during migration. Inthe case of integrin-based cell migration, it was shownthat intermediate levels of adhesion to the substratepromotes maximal migration speed, with both positiveand negative deviations slowing or halting motility69.Similarly, reduction in the level of DE-cadherin reducesthe speed of border cell migration68, indicating that DE-cadherin might not have just a permissive role, butmight be the key adhesion molecule that provides trac-tion for border cells to travel over germline cells.

Organization of the nervous systemVarious cadherins are expressed in the nervous systemin complex patterns. The first example was N-cadherin,which is broadly expressed in the NEUROEPITHELIUM,

Figure 5 | Cadherins in apicobasal and planar epithelialpolarity. a | The example depicted here is the wingepithelium of Drosophila melanogaster, shown in crosssection. Classic cadherins mediate lateral cell contacts(yellow) between epithelial cells that can take the form ofeither a diffuse adhesive contact, or an adherens junction,such as the zonula adherens, that can be seen in electronmicrographs as an electron-dense specialization of theplasma membrane. Epithelial sheets are obviously differentacross their apical–basal axis, but many epithelial cells canalso discern directions in the plane of the epithelium withrespect to the organ or body axis of which they are a part.They use this information to polarize their actin andmicrotubule cytoskeletons along this axis. The mostobvious indication of planar polarity in the wing epithelium isthe hair that is formed by each cell. Hairs emerge from adistal region of the apical cell surface and all point distally. b | The array of cells in a Drosophila wing epithelium,viewed from above. A wild-type array is shown on the left,illustrating the uniform planar polarity (wing hair orientation)and the distribution of Flamingo/Starry Night (green), whichaccumulates at the proximal and distal surfaces of everycell. The right array shows a group of cells that containFlamingo/Starry Night mutant cells (absence of green), inwhich the orientation of wing hairs and therefore the axis ofplanar polarity has shifted55,56.

Planar polarity

Zonulaadherens

Diffuseadhesivecontact

Apical

Basal

Distal Distal

a

b

NEUROEPITHELIUM

Epithelial layer of cells that givesrise to the nervous system.

NEURULATION

Morphogenetic process duringwhich the progenitors of thenervous system segregate fromthe ectoderm.



R E V I E W S

expressed in the developing embryonic CNS78. Nullmutations in DN-cadherin and the Drosophila cateningene armadillo affect axon outgrowth, although in a mildfashion, with many axons finding their targets appropri-ately78,79. In this respect, cadherins resemble various otheraxon guidance cues that direct axon outgrowth in a com-binatorial fashion, with individual cues having subtlefunctions9. In addition, it was recently found that the pat-terning of dendrites that extend from Drosophila sensoryneurons requires Flamingo/Starry Night80.

Both classic and protocadherins localize to synapses,indicating that they may contribute to the generation ofadhesive specificity needed to build complex neural net-works75,81–84. The synapse is an adhesive contact betweentwo neurons, with the transmitter release zone framed byadherens junctions that are ultrastructurally similar toepithelial adherens junctions, and that share with themthe cadherin–catenin complex as a principal molecularcomponent (FIG. 4b)82,83. In synapses, as in epithelial cells,adherens junctions are closely associated with vesicle-release zones (FIG. 4). It is believed that classic cadherinsare important during synaptic adhesion, whereas theadhesive role of protocadherins at synapses remains to beclarified. Recent intriguing evidence indicates that synap-tic activity can change the distribution and adhesive stateof N-cadherin85. Cadherins, in turn, can influence synap-tic activity84,86,87. These findings indicate an intimate rela-tionship between synaptic adhesion and activity, raisingthe possibility that cadherins are important regulators ofsynaptic plasticity and activity modulation13,89.

Some classic cadherins, such as cadherin-6, areexpressed in groups of neurons that form neural cir-cuits, indicating that cadherins might functionally inte-grate such neural circuits71,90. Protocadherins are alsoexpressed in divergent and restricted patterns in theCNS, indicating that they might have a function in inte-grating neural circuits91,92. Moreover, the protocadheringenes Pcdhα, Pcdhβ and Pcdhγcan give rise to over 50protocadherins18 and the Pcdhα/CNR proteins have adifferential expression pattern within individual brainareas75. These findings raise the possibility that a ‘cad-herin code’ exists, which could identify individual neu-rons and their synaptic contacts75,89, although neurexinsand immunoglobulin-type adhesion receptors have alsobeen proposed to contribute to synaptic specificity93,94.The structures of the Pcdhα, Pcdhβ and Pcdhγ genesshow provocative similarities to the gene organization ofimmunoglobulins or T-cell receptors, which has led tothe proposal that gene rearrangement might function indetermining protocadherin expression patterns13,15,16, aspeculation that remains to be proved.

The futureThe analysis of cadherins emphasizes the similaritiesbetween embryonic and neural morphogenesis.Cadherins have emerged as the predominant group ofcell–cell adhesion molecules involved in embryonicmorphogenesis, determining cell and tissue architecture,and controlling dynamic changes in cell shape and posi-tion. The role of individual cadherins in several specificmorphogenetic processes has been determined, which

the sulcus limitans is that these cells remain a coherentgroup and do not participate in the extensive cell rear-rangments that take place during neurulation.

Protocadherins also contribute to CNS regionaliza-tion by controlling the migration of neurons that willorganize into different cortical layers during brain mor-phogenesis. Although cadherins, including protocad-herins, are generally viewed as homophilic adhesionmolecules, recent work indicates that the Pcdhα/CNRprotocadherins might also function as receptors or co-receptors for extracellular ligands in the brain13,73. Thiswork began with studies of the secreted molecule Reelin,identified because mutant mice have a marked behav-ioural disorder. Two protein families have been shown tofunction as reelin receptors, perhaps as a heteromericcomplex: members of the LDL-receptor-related proteinfamily, which couple to the cytoplasmic adaptor proteinmDab1; and members of the Pcdhα/CNR family, whichbind the non-receptor tyrosine kinase Fyn73,74. AsPcdhα/CNR protocadherins show considerable molecu-lar diversity and differential expression patterns withinlocal brain areas, it is possible that Reelin receptor com-plexes that contain different Pcdhα/CNR protocad-herins are instructive in positioning and differentiatingneuronal sub-populations within the cortex73,75.

Classic cadherins are also important during the out-growth of NEURITES and during axonal patterning andFASCICULATION (FIG. 6c). Early studies that indicated thatN-cadherin can function as a substrate for neuriteextension in cultured cells were reinforced by the find-ing that N-cadherin is required for the normal out-growth and guidance of retinal axons76,77. InDrosophila, DN-cadherin is the only classic cadherin

Figure 6 | Cell movements that involve cadherins. a |Convergent extension is a cell rearrangement that involvesthe transient polarization of cells, which then move,converging towards the centre of the tissue. Cell intercalationleads to an extension of the tissue perpendicular to the axisof convergence. This movement is seen, for example, duringfrog gastrulation, where it may involve C-cadherin andparaxial protocadherin61,63. b | The border cell cluster (green)is a small group of Drosophila follicle cells that migrate on thesurface of much larger germline cells (green/yellow). Thismovement is driven by DE-cadherin, which is required in bothborder cells and germline cells68. c | N-cadherin and DN-cadherin mediate the movement of neuronal growth cones(red) on cellular substrates such as axon bundles (green)76–78.

a

b c

NEURITE

Process extended by a nerve cellthat can give rise to an axon or adendrite.

FASCICULATION

Bundling of axonal processes ofneurons.



R E V I E W S

tant during neural morphogenesis, although the func-tional significance of cadherins in neural developmentremains less well understood. One important challengewill be to determine the exact mechanism underlyingthe specificity and stability of cadherin-mediatedcell–cell adhesion, and to explore the variation in adhe-sive mechanisms and cellular responses of differenttypes of cadherins. A second important challenge willbe to substantiate the conjecture that cadherins providean adhesive code that controls synaptic specificity.Elucidating whether and how different families of adhe-sion receptors cooperate in this process will represent anenormous advance in our understanding of complexneural network formation.

will now allow the study of how the adhesive activity ofcadherins is modulated by cell signalling to facilitatecoordinated cell behaviour. Cadherins are also impor-

Links

DATABASE INFORMATION | Cadherin domain | Cadherins | RET | Protocadherins | EGF-like domain | LG domain | E-cadherin | N-cadherin | R-cadherin | Catenins | DE-cadherin | Nectin | Afadin | Ponsin | Snail | Mena | Frizzled | Disheveled | Fat | Dachsous |Flamingo/ Starry night | Reelin | LDL-receptor | mDab1 | Fyn | DN-Cadherin | Armadillo| Cadherin-6 | Pcdhα | Pcdhβ | Pcdhγ | NeurexinFURTHER INFORMATION The cadherin resource | Godt and Tepass labs Drosophilacadherin resource | Cadherin web site at the LMB, Cambridge | Pfeifer lab page | Ikuralab page | Godt lab page | Tepass lab pageENCYCLOPEDIA OF LIFE SCIENCES Adhesive specificity and the evolution ofmulticellularity

1. Steinberg, M. S. Reconstruction of tissue by dissociatedcells. Science 141, 401–408. (1963).

2. Sperry, R. W. Chemoaffinity in the orderly growth of nervefiber patterns and connections. Proc. Natl Acad. Sci. USA50, 703–710 (1963).

3. Gumbiner, B. M. Cell adhesion: The molecular basis oftissue architecture and morphogenesis. Cell 84, 345–579(1996).

4. Hynes, R. O. Cell adhesion: Old and new questions.Trends Cell Biol. 9, M33–M37 (1999).

5. Townes, P. L. & Holtfreter, J. Directed movements andselective adhesion of embryonic amphibian cells. J. Exp.Zool. 128, 53–120 (1955).

6. Nose, A., Nagafuchi, A, & Takeichi, M. Expressedrecombinant cadherins mediate cell sorting in modelsystems. Cell 54, 993–1001 (1988).This paper shows for the first time cadherin-mediated cell sorting in cell-culture cells.

7. Steinberg, M. S. & Takeichi, M. Experimental specificationof cell sorting, tissue spreading, and specific spatialpatterning by quantitative differences in cadherinexpression. Proc. Natl Acad. Sci. USA 91, 206–209(1994).

8. Godt, D. & Tepass, U. Drosophila oocyte localization ismediated by differential cadherin-based adhesion. Nature395, 387–391 (1998).This paper documents for the first time an in vivo cellsorting process that is driven by differentialexpression of a cadherin.

9. Tessier-Lavigne, M. & Goodman, C. S. The molecularbiology of axon guidance. Science 274, 1123–1133(1996).

10. Takeichi, M. Cadherin cell adhesion receptors as amorphogenetic regulator. Science 251, 1451–1455 (1991).

11. Wong, E. F. S., Brar, S. K., Sesaki, H., Yang, C. & Siu, C. H. Molecular cloning and characterization of DdCAD-1,a Ca2+-dependent cell–cell adhesion molecule, inDictyostelium discoideum. J. Biol. Chem. 271,16399–16408 (1996).

12. Gallin, W. J. Evolution of the ‘classical’ cadherin family ofcell adhesion molecules in vertebrates. Mol. Biol. Evol. 15,1099–1107 (1998).

13. Yagi, T. & Takeichi, M. Cadherin superfamily genes:Functions, genomic organization, and neurologic diversity.Genes Dev. 14, 1169–1180 (2000).

14. Nollet, F., Kools, P. & van Roy, F. Phylogenetic analysis ofthe cadherin superfamily allows identification of six majorsubfamilies besides several solitary members. J. Mol. Biol.299, 551–72 (2000).

15. Wu, Q. & Maniatis, T. A. striking organization of a largefamily of human neural cadherin-like cell adhesion genes.Cell 97, 779–790 (1999).

16. Sugino, H. et al. Genomic organization of the family ofCNR cadherin genes in mice and humans. Genomics 63,75–87 (2000).

17. Hutter, H. et al. Conservation and novelty in the evolutionof cell adhesion and extracellular matrix genes. Science287, 989–994 (2000).

18. Overduin, M. et al. Solution structure of the epithelialcadherin domain responsible for selective cell adhesion.Science 267, 386–389 (1995).

19. Shapiro, L. et al. Structural basis of cell–cell adhesion bycadherins. Nature 374, 327–337 (1995).

20. Nagar, B., Overduin, M., Ikura, M. & Rini, J. M. Structuralbasis of calcium-induced E–cadherin rigidification anddimerization. Nature 380, 360–364 (1996).

This is one of several papers (see also refs 18–26,31)that describe the three-dimensional structure of thecadherin domain, and analyse the role of calciumions in the formation of cadherin dimers.

21. Pertz, O. et al. A new crystal structure, Ca2+ dependenceand mutational analysis reveal molecular details ofE–cadherin homoassociation. EMBO J. 18, 1738–1747(1999).

22. Tamura, K., Shan, W. S., Hendrickson, W. A., Colman, D. R. & Shapiro, L. Structure–function analysis of celladhesion by neural (N-) cadherin. Neuron 20, 1153–1163(1998).

23. Ozawa, M., Engel, J. & Kemler, R. Single amino acidsubstitutions in one Ca2+ binding site of uvomorulin abolishthe adhesive function. Cell 63,1033–1038 (1990).

24. Pokutta, S., Herrenknecht, K., Kemler, R. & Engel, J.Conformational changes of the recombinant extracellulardomain of E–cadherin upon calcium binding. Eur. J.Biochem. 223, 1019–1026 (1994).

25. Baumgartner, W. et al. Cadherin interaction probed byatomic force microscopy. Proc. Natl Acad. Sci. USA 97,4005–4010 (2000).

26. Hyafil, F., Babinet, C. & Jacob, F. Cell–cell interactions inearly embryogenesis: a molecular approach to the role ofcalcium. Cell 26, 447–454 (1981)

27. Brieher, W. M., Yap, A. S., & Gumbiner, B. M. Lateraldimerization is required for the homophilic binding activityof C-cadherin. J. Cell Biol. 135, 487–496 (1996).

28. Alattia, J. R. et al. Lateral self-assembly of E-cadherindirected by cooperative calcium binding. FEBS Lett. 417,405–408 (1997).

29. Yap, A. S., Brieher, W. M., Pruschy, M. & Gumbiner, B. M.Lateral clustering of the adhesive ectodomain: afundamental determinant of cadherin function. Curr. Biol.7, 308–315 (1997).

30. Shan, W. S. et al. Functional cis-heterodimers of N- and R-cadherins. J. Cell Biol. 148, 579–590 (2000).

31. Tomschy, A., Fauser, C., Landwehr, R. & Engel, J.Homophilic adhesion of E-cadherin occurs by a co-operative two-step interaction of N–terminal domains.EMBO J. 15, 3507–3514 (1996).

32. Nose, A., Tsuji, K. & Takeichi, M. Localization of specificitydetermining sites in cadherin cell adhesion molecules. Cell61,147–155 (1990).

33. Blaschuk, O. W., Sullivan, R., David, S. & Pouliot, Y.Identification of a cadherin cell adhesion recognitionsequence. Dev. Biol. 139, 227–229 (1990).

34. Sivasankar, S., Brieher, W., Lavrik, N., Gumbiner, B. &Leckband, D. Direct molecular force measurements ofmultiple adhesive interactions between cadherinectodomains. Proc. Natl Acad. Sci. USA 96,11820–11824 (1999).A recent paper that suggests a new model ofcadherin trans-dimer formation.

35. Costa, M. et al. A putative catenin–cadherin systemmediates morphogenesis of the Caenorhabditis elegansembryo. J. Cell Biol. 141, 297–308 (1998).

36. Raich, W. B., Agbunag, C. & Hardin, J. Rapid epithelial-sheet sealing in the Caenorhabditis elegans embryorequires cadherin-dependent filopodial priming. Curr. Biol.9, 1139–1146 (1999).This paper analyses the dynamics of cadherin-dependent contact formation between epithelialcells during C. elegans embryogenesis.

37. Uemura, T. et al. Zygotic Drosophila E-cadherin expressionis required for processes of dynamic epithelial cell

rearrangement in the Drosophila embryo. Genes Dev. 10,659–671 (1996).

38. Tepass, U. et al. shotgun encodes Drosophila E–cadherinand is preferentially required during cell rearrangement inthe neuroectoderm and other morphogenetically activeepithelia. Genes Dev. 10, 672–685 (1996).

39. Ohsugi, M., Larue, L., Schwarz, H. & Kemler, R. Cell-junctional and cytoskeletal organization in mouseblastocysts lacking E-cadherin. Dev. Biol. 185, 261–271(1997).

40. Mandai, K. et al. Afadin: A novel actin filament-bindingprotein with one PDZ domain localized at cadherin-basedcell-to-cell adherens junction. J. Cell Biol. 139, 517–528(1997).

41. Mandai, K. et al. Ponsin/SH3P12: An l-afadin- andvinculin-binding protein localized at cell-cell and cell-matrixadherens junctions. J. Cell Biol. 144, 1001–1017 (1999).

42. Takahashi, K. et al. Nectin/PRR: An immunoglobulin-likecell adhesion molecule recruited to cadherin-basedadherens junctions through interaction with Afadin, a PDZdomain-containing protein. J. Cell Biol. 145, 539–549(1999).One of a series of papers (refs 40–44) that describe anew protein complex that localizes to adherensjunctions, and that might interact with thecadherin–catenin complex functionally.

43. Tachibana, K. et al. Two cell adhesion molecules, nectinand cadherin, interact through their cytoplasmic domain-associated proteins. J. Cell Biol. 150, 1161–1175 (2000).

44. Ikeda, W. et al. Afadin: A key molecule essential forstructural organization of cell–cell junctions of polarizedepithelia during embryogenesis. J. Cell Biol. 146,1117–1132 (1999).

45. Tepass, U. Genetic analysis of cadherin function in animalmorphogenesis. Curr. Opin. Cell Biol. 11, 540–548 (1999).

46. Yeaman, C., Grindstaff, K. K. & Nelson, W. J. Newperspectives on mechanisms involved in generatingepithelial cell. Phys. Rev. 79, 73–98 (1999).

47. Oda, H., Tsukita, S. & Takeichi, M. Dynamic behavior of thecadherin-based cell–cell adhesion system duringDrosophila gastrulation. Dev. Biol. 203, 435–450 (1998).

48. Cano, A. et al. The transcription factor snail controlsepithelial–mesenchymal transitions by repressing E-cadherin expression. Nature Cell Biol. 2, 76–83 (2000).

49. Batlle, E. et al. The transcription factor snail is a repressorof E–cadherin gene expression in epithelial tumour cells.Nature Cell Biol. 2, 84–89 (2000).

50. Adams, C. L. et al. Mechanisms of epithelial cell–celladhesion and cell compaction revealed by high-resolutiontracking of E-cadherin–green fluorescent protein. J. CellBiol. 142, 1105–1119 (1998).

51. Vasioukhin, V., Bauer, C., Yin, M. & Fuchs, E. Directedactin polymerization is the driving force for epithelialcell–cell adhesion. Cell 100, 209–219 (2000).

52. Grindstaff, K. K. et al. Sec6/8 complex is recruited tocell–cell contacts and specifies transport vesicle delivery tothe basal–lateral membrane in epithelial cells. Cell 93,731–740 (1998).This paper shows a close association between thesec6/8 complex, which is involved in lateral vesicletargeting, and cadherin-based adherens junctions.

53. Peifer, M. & Polakis, P. Wnt signaling in oncogenesis andembryogenesis — a look outside the nucleus. Science287, 1606–1609 (2000).

54. Adler, P. N., Charlton, J. & Liu, J. Mutations in the cadherinsuperfamily member gene dachsous cause a tissue



R E V I E W S

polarity phenotype by altering Frizzled signaling.Development 125, 959–968 (1998).

55. Usui, T. et al. Flamingo, a seven-pass transmembranecadherin, regulates planar cell polarity under the control ofFrizzled. Cell 98, 585–595 (1999).

56. Chae, J. et al. The Drosophila tissue polarity gene starrynight encodes a member of the protocadherin family.Development 126, 5421–5429 (1999).References 55 and 56 show a role for theFlamingo/Starry Night cadherin in planar epithelialpolarity. Reference 55 also shows that thesubcellular distribution of Flamingo/Starry Nightdepends on the direction of Wnt/Frizzled signalling.

57. Lu, B., Usui, T., Uemura, T., Jan, L. & Jan, Y. N. Flamingocontrols the planar polarity of sensory bristles andasymmetric division of sensory organ precursors inDrosophila. Curr. Biol. 9, 1247–1250 (1999).

58. Gonzalez-Reyes, A. & St Johnston, D. The Drosophila APaxis is polarised by the cadherin-mediated positioning ofthe oocyte. Development 125, 3635–3644 (1998).

59. Garcia-Castro, M. I., Vielmetter, E. & Bronner-Fraser, M. N-Cadherin, a cell adhesion molecule involved inestablishment of embryonic left–right asymmetry. Science288, 1047–1051 (2000).

60. Shih, J. & Keller, R. Cell motility driving mediolateralintercalation in explants of Xenopus laevis. Development116, 901–914 (1992).

61. Zhong, Y., Brieher, W. M. & Gumbiner, B. M. Analysis of C-cadherin regulation during tissue morphogenesis with anactivating antibody. J. Cell Biol. 144, 351–359 (1999).

62. Lee, C. H. & Gumbiner, B. M. Disruption of gastrulationmovements in Xenopus by a dominant–negative mutantfor C-cadherin. Dev. Biol. 171, 363–373 (1995).

63. Kim, S. H., Yamamoto, A., Bouwmeester, T., Agius, E. &Robertis, E. M. The role of paraxial protocadherin inselective adhesion and cell movements of the mesodermduring Xenopus gastrulation. Development 125,4681–4690 (1998).Experiments in this paper indicate an important rolefor a protocadherin in the convergent extensionmovements during frog gastrulation.

64. Yamamoto, A. et al. Zebrafish paraxial protocadherin is adownstream target of spadetail involved in morphogenesisof gastrula mesoderm. Development 125, 3389–3397(1998).

65. Heisenberg, C. P. et al. Silberblick/Wnt11 mediatesconvergent extension movements during zebrafishgastrulation. Nature 405, 76–81 (2000).

66. Wallingford, J. B. et al. Dishevelled controls cell polarityduring Xenopus gastrulation. Nature 405, 81–85 (2000).

67. Tada, M. & Smith, J. C. Xwnt11 is a target of Xenopusbrachyury: regulation of gastrulation movements viadishevelled, but not through the canonical wnt pathway.Development 127, 2227–2238 (2000).

68. Niewiadomska, P., Godt, D. & Tepass, U. DE–cadherin isrequired for intercellular motility during Drosophila

oogenesis. J. Cell Biol. 144, 533–547 (1999).This paper documents a cadherin-dependent cellmigration process.

69. Palecek, S. P., Loftus, J. C., Ginsberg, M. H.,Lauffenburger, D. A. & Horwitz, A. F. Integrin-ligand bindingproperties govern cell migration speed throughcell–substratum adhesiveness. Nature 385, 537–540(1997).

70. Radice, G. L. et al. Developmental defects in mouseembryos lacking N-cadherin. Dev. Biol. 181, 64–78 (1997).

71. Redies, C. Cadherins in the central nervous system. Prog.Neurobiol. 61, 611–648 (2000).

72. Espeseth, A., Marnellos, G. & Kintner, C. The role ofF–cadherin in localizing cells during neural tube formationin Xenopus embryos. Development 125, 301–312 (1998).

73. Senzaki, K., Ogawa, M. & Yagi, T. Proteins of the CNRfamily are multiple receptors for Reelin. Cell 99, 635–647(1999).Protocadherins of the CNR family are identified asreceptors of the extracellular matrix protein Reelin,an interaction that might contribute to the migrationand differentiation of neurons within the braincortex.

74. Gilmore, E. C. & Herrup, K. Cortical development:receiving reelin. Curr. Biol. 10, R162–R166 (2000).

75. Kohmura, N. et al. Diversity revealed by a novel family ofcadherins expressed in neurons at a synaptic complex.Neuron 20, 1137–1151 (1998).This paper identifies a closely related group ofprotocadherins that are differentially expressed inthe brain and localize to synapses.

76. Riehl, R. et al. Cadherin function is required for axonoutgrowth in retinal ganglion cells in vivo. Neuron 17,837–848 (1996).

77. Inoue, A. & Sanes, J. R. Lamina-specific connectivity in thebrain: Regulation by N-cadherin, neurotrophins, andglycoconjugates. Science 276, 1428–1431 (1997).

78. Iwai, Y. et al. Axon patterning requires DN-cadherin, anovel neuronal adhesion receptor, in the Drosophilaembryonic CNS. Neuron 19, 77–89 (1997).

79. Loureiro, M. et al. Anomalous origin of the left pulmonaryartery (Sling): A case report and review of the literature.Rev. Port. Cardiol. 17, 811–815 (1998).

80. Gao, F. B., Brenman, J. E., Jan, L. Y. & Jan, Y. N. Genesregulating dendritic outgrowth, branching, and routing inDrosophila. Genes Dev. 13, 2549–2561 (1999).

81. Yamagata, M., Herman, J. P. & Sanes, J. R. Lamina-specific expression of adhesion molecules in developingchick optic tectum. J. Neurosci. 15, 4556–4571 (1995).

82. Fannon, A. M. & Colman, D. R. A model for centralsynaptic junctional complex formation based on thedifferential adhesive specificities of the cadherins. Neuron17, 423–434 (1996).

83. Uchida, N., Honjo, Y., Johnson, K. R., Wheelock, M. J. &Takeichi, M. The catenin/cadherin adhesion system islocalized in synaptic junctions bordering transmitter release

zones. J. Cell Biol. 135, 767–779 (1996).References 82 and 83 show that classic cadherinsare components of synaptic adherens junctions.

84. Yamagata, K. et al. Arcadlin is a neural activity-regulatedcadherin involved in long term potentiation. J. Biol. Chem.274, 19473–19479 (1999).

85. Tanaka, H. et al. Molecular modification of N–cadherin inresponse to synaptic activity. Neuron 25, 93–107 (2000).

86. Tang, L., Hung, C. P. & Schuman, E. M. A role for thecadherin family of cell adhesion molecules in hippocampallong-term potentiation. Neuron 20, 1165–1175 (1998).

87. Manabe, T. et al. Loss of Cadherin–11 adhesion receptorenhances plastic changes in hippocampal synapses andmodifies behavioral responses. Mol. Cell. Neurosci. 15,534–546 (2000).

88. Uemura, T. The cadherin superfamily at the synapse: Moremembers, more missions. Cell 93, 1095–1098 (1998).

89. Shapiro, L. & Colman, D. R. The diversity of cadherins andimplications for a synaptic adhesive code in the CNS.Neuron 23, 427–430 (1999).

90. Suzuki, S. C., Inoue, T., Kimura, Y., Tanaka, T. & Takeichi,M. Neuronal circuits are subdivided by differentialexpression of type-II classic cadherins in postnatal mousebrains. Mol. Cell. Neurosci. 9, 433–447 (1997).

91. Obata, S. et al. A common protocadherin tail: Multipleprotocadherins share the same sequence in theircytoplasmic domains and are expressed in differentregions of brain. Cell. Adhes. Commun. 6, 323–333(1998).

92. Hirano, S., Yan, Q. & Suzuki, S. T. Expression of a novelprotocadherin, OL–protocadherin, in a subset of functionalsystems of the developing mouse brain. J. Neurosci. 19,995–1005 (1999).

93. Missler, M. & Südhof, T. C. Neurexins: Three genes and1001 products. Trends Genet. 14, 20–26 (1998).

94. Schmücker, D. et al. Drosophila Dscam is an axonguidance receptor exhibiting extraordinary moleculardiversity. Cell 101, 671–684 (2000).

95. Oda, H., & Tsukita, S. Nonchordate classic cadherins havea structurally and functionally unique domain that is absentfrom chordate classic cadherins. Dev. Biol. 216, 406–422(1999).

96. Hazuka, C. D. et al. The sec6/8 complex is located atneurite outgrowth and axonal synapse-assembly domains.J. Neurosci. 19, 1324–1334 (1999).

AcknowledgementsWe would like to thank Y. Takai, J. Petite, R. Cagan and T. Uemurafor communicating unpublished results. The work on cadherins inthe authors’ laboratories is funded by grants from the NationalCancer Institute of Canada with funds from the Canadian CancerSociety (to U.T. and M.I.), the Canadian Institute for HealthResearch (to U.T. and D.G.), University of Toronto ConnaughtCommittee (to D.G.), the National Institutes of Health (to M.P.), theHuman Frontier Science Program (to M.P.) and the US Army BreastCancer Research Program (to M.P.).


In the beginning there was just one DNA polymerase —Escherichia coli DNA polymerase I (pol I), discovered byArthur Kornberg and colleagues1,2 in 1956. Thirteenyears later, Paula de Lucia and John Cairns, at StonyBrook, New York, isolated an E. coli mutant, polA (itsdesignation being a play on de Lucia’s first name, as pro-posed to Cairns by Julian Gross) that seemed to have lessthan 1% of the normal pol I activity3. From this strain, anew DNA-polymerizing enzyme, pol II, was isolated4.

The polA strain was much more sensitive to ultravi-olet (UV) radiation than wild-type cells, suggesting thatpol I might be involved in DNA repair in addition tochromosomal replication. Shortly after, using this samepolA strain, Thomas Kornberg and Malcolm Gefter5,and Friedrich Bonhoeffer, Heinz Schaller andcolleagues6, independently discovered DNA polymeraseIII (pol III). Isolation of a conditionally lethal tempera-ture-sensitive pol III mutant7 showed that this enzymeis required for replicating the E. coli chromosome6. Incontrast, pol II remained an enigma until last year,when it was shown8 to be pivotal in restarting replica-tion in UV-irradiated cells.

Last year also saw the identification of a new class ofDNA polymerases — the UmuC/DinB/Rev1p/Rad30superfamily (TABLE 1) — on the basis of five conservedsequence motifs present in all of these proteins (FIG. 1).The yeast Rev1 protein had been shown to contain DNA-template-dependent DCMP TRANSFERASE activity nearly threeyears earlier9, but it was not until 1999 that the other fam-ily members were isolated and shown to be capable ofreplicating DNA using all four bases. Biological functions

have been established for some members, including theE. coli UmuD′

2C complex (now known as pol V), the

yeast Rev1 protein and human DNA polymerase eta (polη/Rad30). However, the functions of the remainingmembers of the UmuC/DinB/Rev1p/Rad30 polymerasesuperfamily are less certain.

A feature common to many of these polymerases istheir tendency to copy undamaged DNA with remark-ably poor fidelity, whether or not they are involved intranslesion synthesis. As its name suggests, translesionsynthesis is the unimpaired copying of aberrant bases(see below) at which other cellular polymerases stall.With undamaged DNA, these low-fidelity polymerasesincorporate an incorrect nucleotide once every100–1,000 bases on average10–12 (TABLE 1). For compari-son, normal polymerases that do not PROOFREAD misin-corporate nucleotides in the range of once every 104–106

bases13. Examples of low-fidelity polymerases includeE. coli pol V, which preferentially misincorporates Gopposite a 3′ T of a T–T 6–4 PHOTOPRODUCT; E. coli DNApolymerase IV (pol IV/DinB), which adds a nucleotideonto the end of a misaligned primer; Rev1p, whichincorporates C opposite a non-coding ABASIC LESION; andhuman DNA polymerase iota (pol ι/Rad30B), whichfavours misincorporation of G opposite T on undam-aged DNA. All of these events lead to mutation. There isalso the remarkable case of pol η, which copies pyrimi-dine T–T DIMERS accurately, resulting in mutation avoid-ance at this type of DNA damage (FIG. 2).

The number of DNA polymerases has now grownfrom 3 to 5 in E. coli, and from 5 to at least 14 and count-

THE EXPANDING POLYMERASEUNIVERSEMyron F. Goodman and Brigette Tippin

Over the past year, the number of known prokaryotic and eukaryotic DNA polymerases hasexploded. Many of these newly discovered enzymes copy aberrant bases in the DNA templateover which ‘respectable’ polymerases fear to tread. The next step is to unravel their functions,which are thought to range from error-prone copying of DNA lesions, somatic hypermutationand avoidance of skin cancer, to restarting stalled replication forks and repairing double-stranded DNA breaks.


University of SouthernCalifornia, Department ofBiological Sciences andChemistry, Stauffer Hall ofScience 172, Los Angeles,California 90089-1340, USA.e-mail:[email protected]

DCMP TRANSFERASE

The DNA template-directedreaction catalysed by the yeastRev1 protein, where C isfavoured for incorporationopposite an abasic templatesite, and to a much lesser extentopposite a normal G site.

PROOFREADING

Excision of a misincorporatednucleotide at a growing 3′-primer end by a 3′ exonucleaseassociated with the polymerase.

T–T 6–4 PHOTOPRODUCT

A form of damage occurringwhen DNA is exposed to UVradiation, in which a covalentbond is formed between the 6and 4 pyrimidine ringpositions, coupling adjacentthymines on the same DNAstrand.

R E V I E W S



R E V I E W S

that there is a special class of polymerases to copy dam-aged DNA templates. Instead, proteins such as UmuCand UmuD′ (the mutagenically active form of UmuD;BOX 1) were thought to reduce the fidelity of E. coli pol III,enabling a blocked REPLICATION FORK to carry out error-prone translesion synthesis21. We now know that this isnot the case — indeed, the cellular function of severalnew, errant DNA polymerases is translesion DNA syn-thesis22–25 (TABLE 1). Further studies using these ‘sloppiercopier’ DNA polymerases are now revealing a rich bio-chemical tapestry. For example, one member of this fam-ily — the E. coli UmuD′

2C polymerase (pol V)26–28 —

does not act alone, but requires three further proteins tocatalyse translesion synthesis12,28. Moreover, yeast Rev1prequires DNA polymerase zeta (pol ζ) to copy past abasicsites9 and T–T 6–4 photoproducts29. Other family mem-bers probably also use accessory proteins.

The E. coli pol V mutasomeA good place to start any tour of the new DNA poly-merases is with the E. coli pol V mutasome. The DNA-damage-inducible SOS response in E. coli was discov-ered more than 25 years ago (BOX 1). Many of the 30 ormore SOS-regulated genes are involved in repairingDNA damage30,31, but two genes, umuC and umuD, areinstead required for SOS-induced mutagenesis18,32–34.Although SOS mutation rates are typically 100-foldhigher than spontaneous rates31, increased mutagenesiscannot occur unless UmuD is first converted (by cleav-age) to the mutagenically active UmuD′ protein in areaction that depends on another SOS-induced protein,RecA (REF. 35). The UmuC and UmuD′ proteins theninteract to form a tight complex36,37, UmuD′

2C (pol

V)27,28, with intrinsic DNA polymerase activity.Working alone on an undamaged primed DNA tem-

plate, pol V is a poor DISTRIBUTIVE POLYMERASE26,27. However,pol V cannot copy damaged DNA by itself — it requiresRecA, single-stranded DNA binding protein (SSB) andβ/γcomplex12,27 (where β is the PROCESSIVITY CLAMP and γthe CLAMP-LOADER component of the replicative pol IIIholoenzyme). This multiprotein system, consisting ofpol V, RecA, SSB and β/γ,is called the pol V “mutasome”(FIG. 3), a term coined by Harrison Echols38. The specificactivity of pol V is amplified by an extraordinary15,000-fold in the presence of RecA-coated template,allowing it to copy past damaged DNA bases12.

Although the SOS system typically introduces muta-tions at sites of DNA damage, there is also an increase inuntargeted mutations in the absence of damage39. Allthree common forms of DNA damage (FIG. 2) are copiedefficiently by the pol V mutasome, but synthesis by eitherthe pol III holoenzyme or pol IV (DinB) is blocked12. Thespecificity of incorporation by the pol V mutasome oppo-site the three forms of lesion mimics the in vivo mutation-al data12. For example, the 3′ T of a T–T 6–4 photoprod-uct is a T→C mutational ‘hotspot’ caused by themisincorporation of G opposite T (FIG. 2b) — preciselythe reaction favoured by the pol V mutasome12. In con-trast, pols III and IV preferentially incorporate A, whichagrees with the ‘A-rule’40, but not with the in vivo data.

What is the mechanism of translesion synthesis by

ing in eukaryotes (TABLE 1). Indeed, in a ‘back to thefuture’ moment during a recent conversation with BobLehman, Arthur Kornberg remarked, “In 1955, whowould have imagined that there could be five DNA poly-merases in E. coli?”. So what were the events that led tothe discovery of these polymerases, and what do we nowknow of their biochemical functions and cellular proper-ties? And why are there so many of them in eukaryoticcells? Whereas prokaryotic cells have just one choice —replicate damaged DNA or die — eukaryotic cells can, inprinciple, use programmed cell death (apoptosis) as an‘escape hatch’ to avoid a potential catastrophe.

A growing familyGenetic studies in Saccharomyces cerevisiae and E. colihave been instrumental in defining groups of proteinsrequired for mutagenesis. For example, yeast lacking theREV3 (REF. 14), REV7 (REFS 15,16) or REV1 (REF. 17) genesshow significantly decreased spontaneous and UV-induced mutation rates. In E. coli, SOS mutagenesis(BOX 1) requires the umuC and umuD genes18.

In 1968, Dean Rupp and Paul Howard-Flanders19

observed discontinuities (daughter-strand gaps) inDNA synthesized in an excision-defective strain of E.coli after UV irradiation. Because these strains cannotcarry out NUCLEOTIDE-EXCISION REPAIR20, Rupp and Howard-Flanders suggested that a single pyrimidine dimer isenough to kill the cell, presumably by blocking DNAreplication. But although it may be advantageous tocopy a variety of template lesions as an alternative to celldeath, there is no such thing as a free lunch. The associ-ated cost of survival is an increased number of muta-tions, targeted at the lesion sites. In E. coli, this is referredto as UV-induced SOS error-prone repair (BOX 1).

Because both E. coli and yeast were known to havethree DNA polymerases, there was no reason to suspect

BRCTH2N

H2N

H2N

H2N

H2N COOH

COOH

COOH

COOH

COOH

I II III

IV V

E. coli UmuC

H. sapiens pol ι

H. sapiens pol η

E. coli pol IV/DinB

S. cerevisae Rev1p

e

f

c

a

d

H2N COOHH. sapiens pol κ/ HDINB1b

ABASIC LESION

A common form of DNAdamage in which a base is lostfrom a strand of DNA,spontaneously or by the actionof DNA repair enzymes such asapurinic endonucleases oruracil glycosylase, while leavingthe phosphodiester bond intact.

T–T DIMER

A form of damage occurringwhen DNA is exposed to UVradiation, in which two covalentbonds are formed between boththe 5 and 6 positions of thepyrimidine ring on adjacentthymines located on the sameDNA strand.

NUCLEOTIDE-EXCISION REPAIR

The main pathway for removalof UV-damaged bases.

REPLICATION FORK

Site in double-stranded DNA atwhich the template strands areseparated, allowing a newlyformed copy of the DNA to besynthesized, with the forkmoving in the direction ofleading strand synthesis.

DISTRIBUTIVE POLYMERASE

A polymerase that dissociatesfrom the primer–template DNAafter incorporating one (or atmost a few) nucleotides.

Figure 1 | Representative members of the UmuC/DinB/Rev1p/Rad30 superfamily. Fivehighly conserved domains (indicated by roman numerals I–V) are believed to contain thenucleotide binding and catalytic residues. The subgroups within the family can be easilydistinguished by the presence or absence of unique domains. a, b | The DinB subgroupcontains a further three small domains near the carboxyl terminus of the protein (red boxes),whereas zinc finger motifs are uniquely found in b | pol κ/HDINB1 (C2HC type, yellow diamonds)and c | pol η (C2H2 type, dark grey diamond) that may be involved in DNA binding and selectivetargeting. d | Rev1p is the longest member of the family and contains two regions that are onlyconserved within the Rev1p subgroup (light grey boxes) as well as a BRCT (BRCA-1 carboxy-terminal) domain believed to mediate protein–protein interactions for cell cycle checkpoints andDNA repair. e | UmuC and f | pol ι are both characterized by unique carboxy-terminal ends inwhich no known functional domains have been identified. These unique regions could possiblymediate protein interactions that stimulate and target UmuC or pol ι to their cellular destinations.



R E V I E W S

filament, driven in the 3′→5′ direction by pol V–SSBand in the opposite direction by ATP hydrolysis, con-fines SOS mutations to the sites of DNA damage41,although SOS untargeted mutations do occur, albeit ata much lower frequency. After disassembly of the RecAfilament and dissociation of pol V, the pol III holoen-zyme presumably resumes replication on undamagedDNA downstream from the lesion43.

The DinB subfamilyEscherichia coli pol IV (DinB) is believed to copy undam-aged DNA at stalled replication forks44, which arise in vivofrom mismatched or misaligned primer ends that are notproofread.A function for pol IV in alleviating stalling ofthe pol III holoenzyme is potentially significant, given theestimate that E. coli replication forks probably stall at leastonce during each replication cycle45. Overexpression ofpol IV results in increased frameshift mutagenesis44, con-sistent with the ability of the enzyme to extend mis-aligned primer termini46 (FIG. 2). Whereas DinB homo-logues are among the most conserved members of theUmuC/DinB/Rev1p/Rad30 superfamily, almost nothingis known about what they do in other organisms.

A second function for pol IV has been found, howev-er, in adaptive mutation, a process in which non-prolif-erating microbial populations accumulate mutationswhen placed under non-lethal selective pressure47. InE. coli, pol IV is responsible for roughly half the lacZadaptive frameshift mutations occurring on a plasmid ina wild-type background, and essentially all of theincreased frameshifts in the absence of pol II48. So muta-

pol V? The key to arriving at quantitative, kinetic-basedconclusions for the effects of RecA, SSB and β/γon pol V-catalysed lesion bypass is to have the pol V mutasomebound in a confined region just before the lesion41. Twointeractions occur: the first is between pol V and RecA;the second is between pol V and SSB (FIG. 3).Assembly ofa RecA filament requires ATP binding (it proceeds 5′→3′along a single-stranded DNA template in the presence ofATP or a poorly hydrolysable analogue,ATPγS). But dis-assembly of the filament in the same direction requiresATP hydrolysis42. So in the presence of ATPγS, RecA isbound stably to DNA as a helical filament. The otherpolymerases (pols II, III or IV) cannot copy DNA in theform of a RecA filament, even if the template is undam-aged. Remarkably, however, pol V, along with SSB andβ/γ,copies damaged and undamaged stabilized filamentswith high processivity12,41, perhaps providing the key tounlock the lesion-copying mechanism.

The RecA filament is 100 Å in diameter, whereas theβ-clamp has an inner diameter of only 35 Å. But pro-cessive synthesis takes place on the filament. The obvi-ous explanation is that pol V, acting in conjunctionwith SSB, strips RecA off the DNA in a 3′→5′ direction— a 100 Å RecA molecule cannot be threaded throughthe eye of a 35 Å β-dimer ‘needle’41. The strippingprocess is akin to the action of a locomotive ‘cowcatch-er’ (a pointed device attached to the front of trains topush obstacles off the track). In this case, the RecA‘cow’ is pushed off the DNA template ahead of theadvancing pol V–SSB ‘locomotive’41. We have recentlyproposed that bidirectional disassembly of the RecA

Table 1 | The expanding polymerase universe

DNA polymerase Error rate Properties Possible function

UmuC/DinB/Rev1p/Rad30 superfamily

E. coli pol V (umuDC) 12,27,28,99 10–2 –10–3 Translesion synthesis SOS lesion-targeted and untargeted Low-fidelity synthesis mutagenesis

E. coli pol IV (dinB) 12,46 10–3–10–4 Mismatch extension Untargeted and lesion-targeted mutagenesisH. sapiens pol κ/θ (HDINBI) 49,50,100 Rescues stalled replication forks

H. sapiens pol η (XPV) 11,61,62 10–2 –10–3 Error-free synthesis of Prevents sunlight-induced skin cancerT–T UV photodimers xeroderma pigmentosum

S. cerevisiae pol η (RAD30) 10

H. sapiens pol ι (HRAD30B) 59,79 101–10–4 Low-fidelity synthesis Somatic hypermutation

H. sapiens Rev1 (HREV1) 56,57 Incorporation of C opposite abasic sites UV mutagenesisS. cerevisiae Rev1 (REV1) 9,29

Family B

S. cerevisiae pol ζ (REV3/REV7) 58,59 10–4–10–5 Mismatch extension at lesions UV mutagenesisH. sapiens pol ζ (HREV3/HREV7) 55,66,67,101

Family X

H. sapiens pol λ/pol β2 (POLL/POLβ2) 82,83 Meiosis-associated DNA repair

H.sapiens pol µ (POLM) 81 Somatic hypermutation

S. cerevisiae pol κ (TRF4) 84* Sister-chromatid cohesion

Family A

H. sapiens pol θ (POLQ) 85* Repair of DNA crosslinksD. melanogaster MUS308 86

* The HDINB1 polymerase has also been designated pol κ/pol θ.

PROCESSIVITY CLAMP

A doughnut-shaped proteincomplex that threads the DNAthrough its hole while tetheringthe polymerase to DNA,typically increasing theprocessivity of the polymerase(the number of nucleotidesincorporated into DNA perpolymerase–template bindingevent).

CLAMP LOADER

A protein complex that bindsand then assembles theprocessivity clamp onto theDNA at a 3′-OH primer end, ina reaction requiring ATP.



R E V I E W S

interactions that target or regulate the enzyme50. In vitro,polκ∆C can bypass an abasic site by preferential insertionof A opposite the lesion, creating a –1 frameshift muta-tion by a template loop-out mechanism. This occurswhen the abasic site is followed by a T in the template; ifthe abasic site is followed by a template A, only a simplebase substitution is observed (FIG. 2). Polκ∆C can also syn-thesize past an N-2-acetylaminofluorine (AAF)-modifiedG in an error-prone manner by preferential incorpora-tion of T, without generating a frameshift mutation50.

Rev1p C transferaseWhen confronted with a missing DNA template base —that is, an abasic site (FIG. 2) — most DNA polymerasesfavour an ‘A-rule’ default mechanism in which A isstrongly preferred (about tenfold) for incorporationopposite the non-templating lesion40,51–53. However, whenChristopher Lawrence and colleagues54 used plasmidDNA containing a site-directed abasic moiety to infectyeast cells, they observed preferential incorporation of C,not A (FIG. 2). This unexpected effect depends on the Rev1protein, which is required for UV mutagenesis17.Mutations in REV1, REV3 or REV7 eliminate more than95% of base-substitution mutations in yeast55,56. Thehuman homologue of yeast Rev1p has since been foundto be required for UV-induced mutagenesis56, and it alsobehaves as a template-dependent dCMP transferase57.

There is very little bypass of an abasic site in vivo inthe absence of Rev1p, and what little bypass does occurobeys the A-rule29. Rev1p also shows weak incorporationof C opposite G, at about a tenfold lower rate comparedwith incorporation opposite an abasic site. Because itincorporates only C, Rev1p is perhaps better character-ized as a template-based dCMP transferase rather than abona fide DNA polymerase9.

But Rev1p does not act alone in catalysing translesionreplication — for this it requires pol ζ (the Rev3 and Rev7proteins)58 (FIG. 4). Pol ζ has the remarkable property ofadding correct nucleotides onto mismatched 3′-primerends with exceptionally high efficiencies, only 10–100-fold less than observed for correct primer extension59. Soit is likely that pol ζ takes over from Rev1p, which incor-porates C opposite an abasic lesion but cannot go fur-ther58. Lawrence and co-workers have also reported29 thatRev1p is needed to copy past pyrimidine 6–4 photoprod-ucts but, in contrast to bypass of abasic sites, C is notincorporated. Rev1p therefore seems to have two distinctfunctions in copying DNA damage. One requires its Ctransferase activity (FIG. 4a), whereas the other facilitatestranslesion synthesis by another polymerase, most proba-bly pol ζ (FIG. 4b). However, a direct interaction betweenRev1p and pol ζ has not been reported.

DNA polymerase ηDNA pol η, a human homologue of the yeast Rad30protein60, was identified as the product of the XPVgene61,62 last year. Xeroderma pigmentosum (XP) ischaracterized by mutations in eight GENETIC COMPLEMENTA-

TION GROUPS, seven of which code for enzymes involved innucleotide-excision repair31. The eighth is the XPV gene.Although XPV cells can carry out nucleotide-excision

tor polymerases provide flexibility in dealing with envi-ronmental stress, particularly in prokaryotic organisms.By investigating competition for survival using E. colistrains containing combinations of single, double andtriple pol II, pol IV and pol V mutants, it should be pos-sible to determine the contribution of each polymeraseto the relative fitness of the organism.

Little is known about the in vivo function of humanDINB (DNA polymerase kappa, pol κ). Purified pol κ,with a carboxy-terminal truncation (polκ∆C) thatdeletes two zinc clusters (FIG. 1) found only in the highereukaryote homologues49, retains its polymerase activity.This implies that the carboxy-terminal region is dispens-able for binding and catalysis, but that it mediates protein

Figure 3 | The pol V mutasome. The pol V mutasome consists of pol V (UmuD′2C), activatedRecA (RecA*), β sliding clamp, γ clamp loading complex and the single-stranded DNA bindingprotein (SSB). Pol V associates at a 3′-primer end (vacated by the pol III core), while establishingdirect contact with SSB, and the 3′ tip of a RecA filament.

X

RecA*

pol V SSB

5'3'

β

Figure 2 | Biochemical properties of new polymerases. a | Error-prone translesionsynthesis (TLS) by E. coli pol V results in misincorporation of G opposite the 3′ T of a T–T6–4 photoproduct. b | DNA polymerase η incorporates two A bases as it replicates acrossa T–T cis–syn photodimer, thereby avoiding mutation. c | The DNA-dependent dCMPtransferase activity of Rev1p incorporates C opposite an abasic site. d | Misincorporationof G opposite T on an undamaged template is carried out by pol ι , in preference to correctincorporation of A, resulting in a high incidence of A to G transitions. e | Misalignedprimer–template ends are extended efficiently by E. coli pol IV, leading to frameshift errors.Extension of a mismatched primer end (not shown) would lead to a base-substitutionmutation. f | Bypass of an abasic site by pol κ (HDINB1) results in a –1 frameshift mutationwhen the lesion is followed by T.

T G A5'

5'

pol V

Error-prone TLS

T–T (6-4) photoproduct

HN

N

O

OH

O

N

NO

123

456

12

345

6

XC

Rev1p5'

5'Error-prone dCMP incorporation

Abasic site

O

O

O-CH2

TG

pol ι5'

5'Error-pronesynthesis

T T

A A5'5'

pol η

Error-free TLS

T–T cis-syn photodimer

NO

O

NO

O

HNHN

12

34

5

6

T

pol IV 5'

5'

T

A A A CMisalignedprimer–templateextension

G

A

pol κ (HDINB1) 5'

5'

X

T G A CError-prone TLS by primer–templatemisalignment

G

Figure 1

a

b

c

d

e

f

T

T T

TC

GENETIC COMPLEMENTATION

GROUPS

A distinct group of genescoding for separatepolypeptides (proteins)required in the same biologicalpathway.


when copying T–T photodimers does not translate intoerror-free synthesis on undamaged DNA. Indeed, errorrates for pol η on natural DNA templates can be as highas about 5% for T→C mutations (T•dGMP mispairs)11,with most base-substitution errors in a range of around0.5–1% (REFS 10,11). In comparison, most non-proof-reading cellular polymerases (which do not have anassociated 3′→5′ exonuclease activity for editing outmisinserted nucleotides13) have error rates of about10–3–10–5. So relaxed active-site specificity, enablingpol η to copy ‘blocking’ T–T dimers accurately, is proba-bly responsible for its low fidelity on undamaged DNA.

In XPV cells, error-prone replication of T–T pho-todimers by some other polymerase could potentiallycause an increase in mutations. One candidate is ahuman homologue of the yeast pol ζ. The yeast poly-merase is composed of a complex of the Rev3 and Rev7proteins, and, as already discussed, is required in an error-prone translesion-synthesis pathway58. Indeed, humanREV3 (REF. 66) and REV7 (Ref. 67) homologues havebeen identified, but even if human pol ζ is not responsi-ble for error-prone replication in XPV cells, there areplenty of other candidates to choose from (TABLE 1).

Candidates for somatic hypermutationThe kind of mutagenesis discussed so far is not the onlyprocess in which errors can be introduced into DNA.Take somatic hypermutation, for example, which is oneof the processes responsible for generating the roughlyone billion antibody variants in humans68. An initialrepertoire of antibodies results from non-random V(D)J

RECOMBINATION. After exposure to an antigen, activationof B cells expressing the correct antibody starts a secondphase of diversity, termed ‘affinity maturation’, causedby somatic hypermutation in rapidly dividing GERMINAL

CENTRE cells69. These mutations occur exclusively in thevariable region of the immunoglobulin gene; they beginproximal to the promoter and diminish about 1–2 kilo-bases downstream70. The base-substitution error rate ofaround 3×10–4 per base pair per generation is about106-fold above spontaneous background levels68. Inother words, somatic hypermutation is exquisitely tar-geted, and is unaccompanied by a global alteration inthe fidelity of B-cell replication.

Two cis-acting transcriptional enhancers locateddownstream of the variable region in light and heavychains regulate somatic hypermutation71 (FIG. 5). Theintronic enhancer (E

i) and flanking matrix attachment

region (MAR) of the κ light chain are both essential,eliminating somatic hypermutation completely whendeleted72. Another κ light chain 3′ enhancer (E

3′) affects

mutations to a lesser extent73. A promoter sequenceupstream of the immunoglobulin gene variable regionis also essential, but any promoter can be used, and anyDNA inserted into the variable region can act as themutational target.

Before the discovery of error-prone polymerases,several models for somatic hypermutation were pro-posed. One suggests that hypermutation rates couldarise from repetitive application of transcription-cou-pled repair, by which stalled replication forks in the

repair, they are deficient in copying UV-damagedDNA63. People who carry defects in XP genes showincreased susceptibility to sunlight-induced skin cancer.

The prevalent form of UV damage to DNA is theT–T cis–syn photodimer (FIG. 2). T–T photodimers blockreplication by various polymerases in vitro, but they donot significantly impede human or yeast pol η, both ofwhich copy these photodimers by correctly incorporat-ing two A bases opposite each T site64,65 (FIG. 2). This isconsistent with a cellular role for pol η in the error-freereplication of UV photodimers: error-free synthesis


R E V I E W S

Box 1 | SOS mutagenesis

Escherichia coli keeps anarsenal of regulatedpathways that help it tosurvive when under stress31.One of these is the ‘SOS’regulon, which is thought tobe induced in response toregions of single-strandedDNA — presumably ahallmark of large-scale DNAdamage. Normally, the LexArepressor binds to theoperators of more than 30SOS genes and keeps themrepressed. But in thepresence of single-strandedDNA, the RecA proteinforms a close-packed‘activated’ RecA filament,RecA*, which acts as a co-protease to cleave any LexAreleased from low-affinityoperators. Further cleavageof LexA frees up the moreweakly bound operators,and the SOS genes arerelieved from repression(see figure).

The SOS proteins aremainly involved innucleotide-excision andrecombination-repairpathways to remove theDNA damage. However, thetwo ‘UV mutagenesis’(umu) genes, umuC andumuD, are instead required for replication past unrepaired lesions in the DNAtemplate. They leave behind mutations targeted to sites of DNA damage. To be active,UmuD must be post-translationally cleaved to UmuD′ on the RecA* filament31,35,97,98

(see figure). UmuC and UmuD′ then form a tight complex, UmuD′2C, which has an

intrinsic, low-fidelity DNA polymerase activity43.A replication fork blocked by DNA damage is dealt with by two SOS-induced DNA

polymerases — pol II and pol V (UmuD′2C). About two minutes after SOS induction,

pol II reinitiates replication downstream from the lesion, leaving a gapped structurethat is resolved by homologous recombination43. Replication restart is an error-freerepair process. Pol V appears 30–45 minutes later. It binds at the 3′-OH adjacent to thelesion, then copies past the lesion, often inserting the wrong base opposite it. Thisprocess also requires RecA*, single-stranded DNA binding protein (SSB), and the β/γprocessivity proteins (FIG. 3).

umuD

UmuDprotein

UmuD'(mutationally active)

umuC

UmuD'2C/ pol V

RecA*

SOS on

LexA(inactivated)

RecA*umuD

x

umuC x

umuD x x

LexA(bound)

LexA(bound)

LexA(free)

LexA(free)

SOS off

umuC

UmuC protein

V(D)J RECOMBINATION

The site-specific recombinationof immunoglobulin codingregions from multiple copies inthe germ line to just onevariable (V), one diversity (D)and one joining (J) region inthe process of forming afunctional immunoglobulingene in B cells.



R E V I E W S

another experiment59, incorporation of both G and Topposite a template T is favoured by about tenfold andfivefold, respectively, relative to incorporation of Aopposite T. Most of the other mispairs occur in therange of 10–2–10–3. The pol ι misincorporation prefer-ences seem consistent with immunoglobulin mutationalspectra in which TRANSITIONS are favoured 2:1 over TRANS-

VERSIONS, and A mutates more often than T (REF. 80).Human pol µ is most closely related (41% amino-

acid identity) to terminal deoxynucleotidyltransferase(TdT)81, a template-independent DNA-synthesizingenzyme. Some similiarity (23% identity) is alsoobserved between pol µ and polymerase beta (pol β).Pol µ has weak intrinsic TERMINAL TRANSFERASE activity, andcan also act as a DNA-dependent polymerase that showspoor base selection when manganese replaces magne-sium as a cofactor in replication reactions in vitro81. Thispolymerase is expressed preferentially in peripheral lym-phoid tissues and, based on analysis of the expressedsequence tag database, could be overrepresented inhuman B-cell germinal centres, which are critical formaturation of the immune response81.

Should experiments using knockout mouse strainsreveal a requirement for one (or perhaps several) of theerrant polymerases in somatic hypermutation, it will bejust the beginning of the story. Biochemical reconstitu-tion of somatic hypermutation in vitro is likely to be achallenge. Any model for somatic hypermutation willhave to account for the localization, polarity, magnitudeand specificity of the point mutations — a tall order.

Although it is premature to speculate on specificmechanisms of somatic hypermutation in vitro, at leasttwo models can be envisaged. In the first, interactionsbetween transcription factors, somatic hypermutation-specific enhancer elements and co-activator proteinsresult in formation of a DNA secondary structure thatrecruits a mutator polymerase (FIG. 6a). In the secondmodel, the errant polymerase is recruited to the site of aDNA nick, short gap, or perhaps even to a double-strandbreak (FIG. 6b). In both models, DNA synthesis by themutator enzyme across a short gap — analogous to base-excision repair by pol β — generates mutations targetedto the variable-region gene. Moreover, nucleotide misin-corporation and mismatch extension might require theaction of separate polymerases. For example, pol ι oftenmakes misincorporation errors, but seems to have diffi-culty in extending a mismatched primer end59. On theother hand, pol ζ synthesizes DNA with essentially nor-

variable region recruit transcription-coupled repairproteins74. Repeated replication of the ‘repair’ regioncould increase the chance of mutation. But transcrip-tion-coupled repair shows a large bias for the activelytranscribed strand, a feature that is uncharacteristic ofsomatic hypermutation75. Moreover, a huge amount ofrepetitive replication would be necessary to reach muta-tion rates of about 3×10–4 per base pair. In anothermodel, reverse transcriptase is thought to synthesize acomplementary DNA copy from an elongating messen-ger RNA that could replace the gene by homologousrecombination76. However, this would be a convolutedway to obtain chromosomal mutations.

Two other models invoke a collision between stalledtranscription forks and moving replication forks, caus-ing either a reduction in fidelity of a normalpolymerase77, or a signal from a stalled fork for an error-prone polymerase to take over78. The idea that somatichypermutation is caused by error-prone DNA poly-merases has been given fresh impetus by the discoveryof the errant polymerases (although the models invok-ing transcription-coupled repair, reverse transcripts,and collisions between transcription and replicationmachinery remain alive, albeit tenuously). At a meetingof The Royal Society on ‘Hypermutation in antibodygenes’ (5–6 July 2000), which devoted one of its foursessions to the new polymerases, two favoured candi-date polymerases emerged — pol ι (Rad30B) and DNApolymerase mu (pol µ) (TABLE 1).

Interest in human pol ι stems from its preference forincorporating G opposite a template T, making a G•Twobble base pair with a 3:1 preference over aWatson–Crick A•T pair79 (FIG. 2). A T•T mispair is alsoeasily formed, about 70% as efficiently as A•T (REF. 79). In

Figure 5 | Genetic elements required for somatichypermutation in the kappa light-chain immunoglobulingene. Both a promoter (P) and leader (L) sequence arerequired, but may be replaced with non-immunoglobulincounterparts from other genes. The intronic enhancer element(Ei) and associated nuclear matrix attachment region (MAR),as well as the 3′ enhancer (E3′), must be present forhypermutation in the variable region (VJ). This region is flankedby the upstream promoter and downstream MAR/Eisequences. The constant domain of the kappa light chain (Cκ)is not a target for somatic hypermutation.

L VJ E3'CκP MAR EiEi

GERMINAL CENTRE

A highly organized structurethat develops around follicles inperipheral lymphoid organs,such as the spleen and lymphnodes, in which B cells undergorapid proliferation andselection on formation ofantigen–antibody complexesduring the immune response.

TRANSITION

A point mutation in which apurine base (A or G) issubstituted for a differentpurine base, and a pyrimidinebase (C or T) is substituted for adifferent pyrimidine base, forexample, an A•T→G•Ctransition.

TRANSVERSION

A point mutation in which apurine base is substituted for apyrimidine base and vice versa,for example, an A•T→C•Gtransversion.

TERMINAL TRANSFERASE

An enzyme found primarily inthe thymus gland thatincorporates nucleotidesrandomly onto the 3′ end ofsingle-stranded DNA (a non-templated reaction), in contrastto a polymerase, whichincorporates nucleotides onto a3′-primer-end in a double-stranded, template-directedreaction.

Figure 4 | Rev1p/pol ζ lesion bypass. a | Rev1p dCMPtransferase activity incorporates C opposite a non-instructional abasic site in a DNA template in the absence ofpol ζ, but it cannot extend the primer beyond the mismatch.Pol ζ can then take over for Rev1p and efficiently extend themismatched primer terminus. However, it is not knownwhether Rev1p stays associated with the DNA or directlyinteracts with pol ζ during this process. b | Pol ζ can efficientlyincorporate two A bases opposite a T–T 6–4 photoproduct invitro, resulting in error-free bypass of the template lesion, but itcan only do so in the presence of Rev1p, although the Ctransferase activity of Rev1p is not involved.

X

C

Rev1p

5'

5'

X T

CA 5'

5'

pol ζ

pol ζ

5'

5'T T

A A

Rev1p

Rev1p

a

b



R E V I E W S

that cells lacking TRF function do not completely repli-cate the genome. In vitro, Trf4 has a DNA polymeraseactivity with an elevated K

m(the substrate concentration

that allows the reaction to proceed at half its maximumrate) for nucleotides, designated pol κ (this name hasbeen applied to two other DNA polymerases; TABLE 1).One suggested function for Trf4 DNA polymerase is tofacilitate progression of the replication fork throughchromatid cohesion sites that might inhibit other DNApolymerases, such as pol δor pol ε, causing the replica-tion fork to collapse84.

Why so many polymerases?There are well established biological roles for pol V(UmuD′

2C), pol η (Rad30) and Rev1p. Although less

certain, pol IV (DinB) is probably required to rescuestalled replication complexes. What about the othernew polymerases? The DNA polymerase theta (polθ/POLQ)85 and MUS308 (REF. 86) proteins, both familyA polymerases (TABLE 1), may be involved in repairingDNA crosslinks, but functions for pols ι (Rad 30B), µ, λand κ (HDINB1) remain speculative (TABLE 1). It is rea-sonable to conclude that pol V and pol η usually copydamaged DNA templates, with pol V required for error-prone and pol η for error-free repair. Because pol Vcauses SOS untargeted mutations in the absence ofDNA damage in E. coli 31, it is probably also importantin natural selection and evolution87.

In eukaryotes, the new polymerases might fill inshort gaps emanating from non-homologous end-join-ing and from homologous recombination. Indeed, amarked increase in sister-chromatid exchange in trans-formed XPV cells led to the discovery88, earlier this year,of a relationship between the S-phase checkpoint of thecell cycle and an X-ray-induced recombination pathwayfor repairing double-strand breaks88. Even E. coli seemsto have a DNA-damage checkpoint in which UmuCand uncleaved UmuD coordinate progression throughthe cell cycle, signalling when it is safe to switch fromstationary phase to exponential growth89,90.

Recently, two groups have independently demon-strated the embryonic lethality of disrupting themouse homologue of the REV3 gene91,92, the presumedcatalytic subunit of mouse pol ζ . These studiesemphasize the potential importance of specializedDNA polymerases in development and raise evenmore interesting questions regarding the extent oflesions that may occur during rapid cell proliferation,or rather if pol ζ might be closely linked to the mitoticcheckpoints in the absence of DNA damage91.

Macromolecular traffic controlThe fact that DNA polymerases have become a ‘growthindustry’ in the cell raises concerns about traffic control.To copy damaged DNA, rescue blocked replicationforks, catalyse somatic hypermutation or fill in gapsduring homologous and non-homologous recombi-nation, the enzymes have to show up where and whenthey are needed and then depart when finished —and not a moment later. The basic idea, and it is not anew one, is that DNA repair proteins might always be

mal fidelity (10–4–10–5) but seems very efficient at mis-match extension (10–1–10–2). So the sequential action ofboth polymerases may be necessary for translesion syn-thesis in eukaryotic cells59.

Other new polymerasesPol λ/β2. Pol µ is not the only new polymerase relatedto pol β. Another DNA polymerase (pol λ) was identi-fied in mouse82, and later in humans (POL β2)83, thatshares a 32% amino-acid identity with pol β and con-tains the conserved family X (TABLE 1) residues criticalfor DNA and nucleotide binding, as well as catalysis82.Pol λ/β2 is expressed to significant levels only in thetestes and ovaries, indicating that this enzyme may beinvolved in meiotic cell division82, but this remains tobe shown. Purified pol λ/β2 has polymerase activity,although nothing is known about its fidelity or pre-ferred DNA substrates82,83. A BRCA1-containing car-boxy-terminal (BRCT) domain located in the amino-terminal region can also be deleted from human polλ/β2 without significant reduction in polymeraseactivity83. The in vivo function of BRCT domains inboth pol λ/β2 and Rev1p pathways is an area thatawaits further investigation.

Pol κ (Trf4). In S. cerevisiae, another β-like polymerasehas been identified as the product of the TRF4 gene.With its close homologue Trf5, theTrf4 protein isinvolved in maintaining sister-chromatid cohesion dur-ing S-phase replication. Fluorescence in situ hybridiza-tion (FISH) in trf4 mutant cells84 has revealed a markedincrease in nuclei that fail to maintain cohesion of sisterchromatids near centromeres and on chromosome arms.Moreover, trf4ts/trf5 double mutant cells show delays inthe G1 to S-phase transition and contain levels of DNAbetween those found in G1 and G2. These results imply

Figure 6 | Models for somatic hypermutation by an error-prone polymerase.a | Enhancer/primer-mediated targeting of somatic hypermutation. Strong interactionsbetween enhancer-binding proteins (shown in white) and transcription-associated factors(TAFs) at the promoter (P) mediate the formation of a unique open DNA complex. Error-pronepolymerases (pol), possibly pol ι or pol µ, may preferentially substitute for the normal replicativepolymerases in this region of the DNA, resulting in mutations. Mutation is shown on bothstrands as there is no evidence for a strand bias. b | Nick- or gap-dependent targeting ofsomatic hypermutation. The variable region of the immunoglobulin genes may contain smallnicks or gaps as a by-product of V(D)J recombination or very active transcription. The DNAbreaks may be substrates for an error-prone polymerase to bind and generate mutations.(MAR, matrix attachment region; Ei, enhancer element.)

MAR E iLPVJ

PL

MARE i

TAF

pol

a

b

pol

pol

pol

pol


macromolecular complex from individually purifiedprotein components. The technology is available — insitu immunofluorescence, immunoprecipitation andmultiple-hybrid screening can all be used to identifyinteractions between the new polymerases and otherproteins. Even if such a fishing expedition bears fruit,to mix a metaphor, it still will not be easy to reconsti-tute a replisome–mutasome macromolecular com-plex in vitro. But there are successful precedents. Forexample, prokaryotic and eukaryotic replication,repair and recombination complexes have been builtfrom the ground up with purified polymerases, pri-mases, processivity factors, DNA-binding proteins,mismatch-binding proteins, recombinases, helicasesand ligases. Indeed, this tried and tested approach isthe first of Arthur Kornberg’s96 Ten commandments:Lessons from the enzymology of DNA replication —“rely on enzymology to clarify biological questions”.

Update — added in proofM. Goldsmith et al.102 have shown that the MucB protein,a plasmid-encoded homologue of the E. coli. UmuC pro-tein, is a DNA polymerase capable of translesion synthe-sis past an abasic site in the presence of MucA′ (theUmuD′ homologue), RecA and SSB protein.


R E V I E W S

present at the replication fork, perhaps bound to thereplication complex. A multiprotein complex, com-posed of two interconnected polymerase holoenzymesfor coordinated leading- and lagging-strand synthesis,a lagging-strand PRIMOSOME, DNA helicase and SSBprotein, would have the added baggage of other spe-cialized polymerases, to be used sparingly when calledfor. This picture is easier to imagine if the replicationcomplex is stationary, with the DNA moving throughthe ‘replication factory’93. In contrast, however, thetextbook version of events is that polymerase boundto accessory proteins traverses a DNA track.Nevertheless, an in situ assay using a Bacillus subtilispolymerase (PolC) tagged with green fluorescent pro-tein identified the enzyme at discrete intracellularloci94, indicating that the DNA may be movingthrough an anchored DNA polymerase.

Interactions between proteins of the replication com-plex and a superfamily polymerase have indeed beenfound in E. coli. Last year, Graham Walker and co-work-ers95 reported differential binding between componentsof pol V (UmuD/UmuD′) and the α-, β- and ε-subunitsof pol III. Further evidence27 comes from the stabiliza-tion of a thermolabile pol III α-subunit, at non-permis-sive temperature, in the presence of pol V. These datahint that there could be a coordinated exchange betweenhigh- and low-fidelity polymerases, acting as partners ina macromolecular complex at sites of DNA damage.

However, the challenge, as Arthur Kornberg hasoften cautioned, is the need “to capture it alive” —that is, to reassemble an intact ‘replisome–mutasome’

1. Kornberg, A., Lehman, I. R., Bessman, M. J. & Simms, E. S.Enzymatic synthesis of deoxyribonucleic acid. Biochim.Biophys. Acta 21, 197–198 (1956).

2. Lehman, I. R., Bessman, M. J., Simms, E. S. & Kornberg, A.Enzymatic synthesis of deoxyribonucleic acid I. Preparationof substrates and partial purification of an enzyme fromEscherichia coli. J. Biol. Chem. 233, 163–170 (1958).

3. de Lucia, P. & Cairns, J. Isolation of an E. coli strain with amutation affecting DNA polymerase. Nature 224,1164–1166 (1969).

4. Knippers, R. DNA polymerase II. Nature 228, 1050–1053(1970).

5. Kornberg, T. & Gefter, M. L. Purification and DNA synthesisin cell-free extracts: properties of DNA polymerase II. Proc.Natl Acad. Sci. USA 68, 761–764 (1971).

6. Nusslein, V., Otto, B., Bonhoeffer, F. & Schaller, H. Functionof DNA polymerase 3 in DNA replication. Nature New Biol.234, 285–286 (1971).

7. Gefter, M. L., Hirota, Y., Kornberg, T., Wechsler, J. A. &Barnoux, C. Analysis of DNA polymerases II and III inmutants of Escherichia coli thermosensitive for DNAsynthesis. Proc. Natl Acad. Sci. USA 68, 3150–3153(1971).

8. Rangarajan, S., Woodgate, R. & Goodman, M. F. A phenotype for enigmatic DNA polymerase II: a pivotalrole for pol II in replication restart in UV-irradiatedEscherichia coli. Proc. Natl Acad. Sci. USA 96, 9224–9229(1999).

9. Nelson, J. R., Lawrence, C. W. & Hinkle, D. C. Deoxycytidyltransferase activity of yeast REV1 protein. Nature 382,729–731 (1996). Shows that yeast Rev1 protein has template-dependent dCMP transferase activity, and is the firstmember of the UmuC/DinB /Rev1p/Rad30superfamily found to be involved in DNA synthesis.

10. Washington, M. T., Johnson, R. E., Prakash, S. &Prakash, L. Fidelity and processivity of Saccharomycescerevisiae DNA polymerase η. J. Biol. Chem. 274,36835–36838 (1999).Yeast Rad30 protein is shown to have low-fidelityDNA polymerase activity and is designated as pol η.

11. Matsuda, T., Bebenek, K., Masutani, C., Hanaoka, F. &

Kunkel, T. A. Low fidelity DNA synthesis by human DNApolymerase η. Nature 404, 1011–1013 (2000).

12. Tang, M. et al. Roles of E. coli DNA polymerases IV and V inlesion-targeted and untargeted mutagenesis. Nature 404,1014–1018 (2000).The E. coli DNA pol V mutasome is shown to beresponsible for SOS-induced mutagenesis througherror-prone translesion synthesis.

13. Echols, H. & Goodman, M. F. Fidelity mechanisms in DNAreplication. Annu. Rev. Biochem. 60, 477–511 (1991).

14. Lawrence, C. W. & Christensen, R. B. Ultraviolet-inducedreversion of CYC1 alleles in radiation-sensitive strains ofyeast. III. REV3 mutant strains. Genetics 92, 397–408(1979).

15. Lawrence, C. W., Das, G. & Christensen, R. B. REV7, anew gene concerned with UV mutagenesis in yeast. Mol.Gen. Genet. 200, 80–85 (1985).

16. Lawrence, C. W., Nisson, P. E. & Christensen, R. B. UVand chemical mutagenesis in REV7 mutants of yeast. Mol.Gen. Genet. 200, 86–91 (1985).

17. Lawrence, C. W. & Christensen, R. Ultraviolet-inducedreversion of cyc1 alleles in radiation sensitive strains ofyeast. I. rev1 mutant strains. J. Mol. Biol. 122, 1–22 (1978).

18. Sommer, S., Knezevic, J., Bailone, A. & Devoret, R.Induction of only one SOS operon, umuDC, is required forSOS mutagenesis in E. coli. Mol. Gen. Genet. 239,137–144 (1993).

19. Rupp, W. D. & Howard-Flanders, P. Discontinuities in theDNA synthesized in an excision-defective strain ofEscherichia coli following ultraviolet radiation. J. Mol. Biol.31, 291–304 (1968).

20. Sancar, A. & Sancar, G. B. DNA repair enzymes. Annu.Rev. Biochem. 57, 29–67 (1988).

21. Bridges, B. A. & Woodgate, R. The two-step model ofbacterial UV mutagenesis. Mutat. Res. 150, 133–139(1985).

22. Walker, G. Skiing the black diamond slope: Progress onthe biochemistry of translesion DNA synthesis. Proc. NatlAcad. Sci. USA 95, 10348–10350 (1998).

23. Woodgate, R. A plethora of lesion-replicating DNApolymerases. Genes Dev. 13, 2191–2195 (1999).

24. Goodman, M. F. & Tippin, B. Sloppier copier DNA

polymerases involved in genome repair. Curr. Opin. Genet.Dev. 10, 162–168 (2000).

25. Friedberg, E., Feaver, W. J. & Gerlach, V. L. The manyfaces of DNA polymerases: Strategies for mutagenesis andfor mutational avoidance. Proc. Natl Acad. Sci. USA 97,5681–5683 (2000).

26. Tang, M. et al. Biochemical basis of SOS mutagenesis inEscherichia coli: Reconstitution of in vitro lesion bypassdependent on the UmuD′2C mutagenic complex and RecAprotein. Proc. Natl Acad. Sci. USA 95, 9755–9760 (1998).The first evidence for the intrinsic polymeraseactivity in UmuD′2C (now pol V).

27. Tang, M. et al. UmuD′2C is an error-prone DNApolymerase, Escherichia coli pol V. Proc. Natl Acad. Sci.USA 96, 8919–8924 (1999).Purified UmuD′2C from a temperature-sensitive DNApolymerase III mutant possesses intrinsicpolymerase activity and can do translesionsynthesis, demonstrating that UmuD′2C is a bonafide DNA polymerase.

28. Reuven, N. B., Arad, G., Maor-Shoshani, A. & Livneh, Z.The mutagenic protein UmuC is a DNA polymeraseactivated by UmuD′, RecA, and SSB and is specialized fortranslesion replication. J. Biol. Chem. 274, 31763–31766(1999).The E. coli UmuC protein was shown to have DNApolymerase activity.

29. Nelson, J. R., Gibbs, P. E. M., Nowicka, A. M., Hinkle, D. C. & Lawrence, C. W. Evidence for a second function forSaccharomyces cerevisiae Rev1p. Mol. Microbiol. 37,549–553 (2000).

30. Walker, G. C. Inducible DNA repair systems. Annu. Rev.Biochem. 54, 425–457 (1985).

31. Friedberg, E. C., Walker, G. C. & Siede, W. in DNA Repairand Mutagenesis Vol. 1, 407–522 (ASM Press, WashingtonDC, 1995).

32. Kato, T. & Shinoura, Y. Isolation and characterization ofmutants of Escherichia coli deficient in induction ofmutagenesis by ultraviolet light. Mol. Gen. Genet. 156,121–131 (1977).

33. Steinborn, G. Uvm mutants of Escherichia coli K12deficient in UV mutagenesis. I. Isolation of uvm mutants

Links

DATABASE LINKS REV3 | REV7 | REV1 | umuC | umuD |pol κ | pol η | Rad30 | pol ι | pol µ | pol λ | TRF4 | pol θENCYCLOPEDIA OF LIFE SCIENCES DNA polymerasefidelity mechanisms | Eukaryotic replication fork

PRIMOSOME

A complex of proteins whoserole is to initiate DNA synthesisby the de novo synthesis of anoligonucleotide RNA primer ona DNA template strand; aprimosome is typically used toinitiate synthesis at a replicationorigin or to re-initiate synthesisdownstream of a stalledreplication fork.



R E V I E W S

and their phenotypical characterization in DNA repair andmutagenesis. Mol. Gen. Genet. 165, 87–93 (1978).

34. Bagg, A., Kenyon, C. J. & Walker, G. C. Inducibility of agene product required for UV and chemical mutagenesis inEscherichia coli. Proc. Natl Acad. Sci. USA 78, 5749–5753(1981).

35. Burckhardt, S. E., Woodgate, R., Scheuremann, R. H. &Echols, H. UmuD mutagenesis protein of Escherichia coli:Overproduction, purification, and cleavage by RecA. Proc.Natl Acad. Sci. USA 85, 1811–1815 (1988).

36. Woodgate, R., Rajagopalan, M., Lu, C. & Echols, H. UmuCmutagenesis protein of Escherichia coli: Purification andinteraction with UmuD and UmuD′. Proc. Natl Acad. Sci.USA 86, 7301–7305 (1989).

37. Bruck, I., Woodgate, R., McEntee, K. & Goodman, M. F.Purification of a soluble UmuD’C complex from Escherichiacoli: Cooperative binding of UmuD’C to single-strandedDNA. J. Biol. Chem. 271, 10767–10774 (1996).

38. Echols, H. & Goodman, M. F. Mutation induced by DNAdamage: A many protein affair. Mutat. Res. 236, 301–311(1990).

39. Witkin, E. M. Ultraviolet mutagenesis and inducible DNArepair in Escherichia coli. Bacteriol. Rev. 40, 869–907(1976).

40. Strauss, B. S. The ‘A rule’ of mutagen specificity: aconsequence of DNA polymerase bypass of non-instructional lesions? Bioessays 13, 79–84 (1991).

41. Pham, P., O’Donnell, M., Woodgate, R. & Goodman, M. F.A ‘cowcatcher’ model for SOS–lesion targeted mutationsin E. coli involving pol V, RecA, SSB and β sliding clamp.(submitted).

42. Kuzminov, A. Recombinational repair of DNA damage inEscherichia coli and bacteriophage λ. Microbiol. Mol. Biol.Rev. 63, 751–813 (1999).

43. Goodman, M. F. Coping with replication ‘train wrecks’ inEscherichia coli using pol V, pol II, and RecA proteins.Trends Biochem. Sci. 25, 189–195 (2000).

44. Kim, S.-R. et al. Multiple pathways for SOS-inducedmutagenesis in Escherichia coli: An overexpression ofdinB/dinP results in strongly enhancing mutagenesis in theabsence of any exogenous treatment to damage DNA.Proc. Natl Acad. Sci. USA 94, 13792–13797 (1997).

45. Cox, M. M. et al. The importance of repairing stalledreplication forks. Nature 404, 37–41 (2000).

46. Wagner, J. et al. The dinB gene encodes a novelEscherichia coli DNA polymerase (DNA pol IV). Mol. Cell 4,281–286 (1999).E. coli DinB/pol IV protein is shown to havedistributive polymerase activity that can also extendmisaligned 3′-primer ends.

47. Foster, P. L. Adaptive mutation: The uses of adversity.Annu. Rev. Microbiol. 47, 467–504 (1993).

48. Foster, P. L. Adaptive mutation in Escherichia coli. ColdSpring Harb. Symp. Quant. Biol. (in the press).

49. Gerlach, V. L. et al. Human and mouse homologs ofEscherichia coli DinB (DNA polymerase IV), members of theUmuC/DinB superfamily. Proc. Natl Acad. Sci. USA 96,11922–11927 (1999).

50. Ohashi, E. et al. Error-prone bypass of certain DNA lesionsby the human DNA polymerase κ. Genes Dev. 14,1589–1594 (2000).

51. Boiteux, S. & Laval, J. Coding properties ofpoly(deoxycycytidylic acid) templates containing uracil orapyrimidinic sites: In vitro modulation of mutagenesis byDNA repair enzymes. Biochemistry 21, 6746–6751 (1982).

52. Schaaper, R. M., Kunkel, T. A. & Loeb, L. A. Infidelity ofDNA synthesis associated with bypass of apurinic sites.Proc. Natl Acad. Sci. USA 80, 487–491 (1983).

53. Sagher, D. & Strauss, B. S. Insertion of nucleotidesopposite apurinic/apyrimidinic sites in deoxyribonucleicacid during in vitro synthesis: uniqueness of adeninenucleotides. Biochemistry 22, 4518–4526 (1983).

54. Lawrence, C. W., Borden, A., Banerjee, S. K. & LeClerc, J. E. Mutation frequency and spectrum resulting from asingle abasic site in a single-stranded vector. Nucleic AcidsRes. 18, 2153–2157 (1990).

55. Gibbs, P. E., McGregor, W. G., Maher, V. M., Nisson, P. &Lawrence, C. W. A human homolog of the Saccharomycescerevisiae REV3 gene, which encodes the catalytic subunitof DNA polymerase ζ. Proc. Natl Acad. Sci. USA 95,6876–6880 (1998).

56. Gibbs, P. E. M. et al. The function of the human homolog ofSaccharomyces cerevisiae REV1 is required formutagenesis induced by UV light. Proc. Natl Acad. Sci.USA 97, 4186–4191 (2000).

57. Lin, W. et al. The human REV1 gene codes for a DNAtemplate-dependent dCMP transferase. Nucleic AcidsRes. 27, 4468–4475 (1999).

58. Nelson, J. R., Lawrence, C. W. & Hinkle, D. C.Thymine–thymine dimer bypass by yeast DNA polymeraseζ. Science 272, 1646–1649 (1996).

59. Johnson, R. E., Washington, M. T., Haracska, L., Prakash,S. & Prakash, L. Mutagenic bypass of DNA lesions bysequential action of eukaryotic DNA polymerases ι and ζ.Nature 406, 1015–1019 (2000).

60. Johnson, R. E., Prakash, S. & Prakash, L. Requirement ofDNA polymerase activity of yeast Rad30 protein for itsbiological function. J. Biol. Chem. 274, 15975–15977(1999).

61. Masutani, C. et al. The XPV (xeroderma pigmentosumvariant) gene encodes human DNA polymerase η. Nature399, 700–704 (1999).Purification of human xeroderma pigmentosumvariant (XPV) and demonstration that it is the humanhomologue of yeast DNA pol η.

62. Johnson, R. E., Kondratick, C. M., Prakash, S. & Prakash,L. hRAD30 mutations in the variant form of Xerodermapigmentosum. Science 285, 263–265 (1999).

63. Cordeiro-Stone, M., Zaritskaya, L. S., Price, L. K. &Kaufmann, W. K. Replication fork bypass of a pyrimidinedimer blocking leading strand DNA synthesis. J. Biol.Chem. 272, 13945–13954 (1997).

64. Masutani, C. et al. Xeroderma pigmentosum variant (XP-V)correcting protein from HeLa cells has a thymine dimerbypass DNA polymerase activity. EMBO J. 18, 3491–3501(1999).

65. Johnson, R. E., Prakash, S. & Praskah, L. Efficient bypassof a thymine–thymine dimer by yeast DNA polymerase, Polη. Science 283, 1001–1004 (1999).

66. Lin, W., Wu, X. & Wang, Z. A full-length cDNA of hREV3 ispredicted to encode DNA polymerase ζ for damage-induced mutagenesis in humans. Mutat. Res. 433, 89–98(1999).

67. Murakumo, Y. et al. A human REV7 homolog that interactswith the polymerase ζ catalytic subunit hREV3 and thespindle assembly checkpoint protein hMAD2. J. Biol.Chem. 275, 4391–4397 (2000).

68. Berek, C. & Milstein, C. The dynamic nature of the antibodyrepertoire. Immunol. Rev. 105, 5–26 (1988).

69. Zhang, J., MacLennan, I. C., Liu, Y. J. & Lane, P. J. Is rapidproliferation in B centroblasts linked to somatic mutationand memory B cell clones? Immunol. Lett. 18, 2393–2400(1988).

70. Kim, S., Davis, M., Sinn, E., Patten, P. & Hood, L. Antibodydiversity: somatic hypermutation of rearranged VH genes.Cell 27, 573–581 (1981).

71. Neuberger, M. S. et al. Monitoring and interpreting theintrinsic features of somatic hypermutation. Immunol. Rev.162, 107–116 (1998).

72. Betz, A. G. et al. Elements regulating somatichypermutation of an immunoglobulin kappa gene: criticalrole for the intron enhancer/matrix attachment region. Cell77, 239–248 (1994).

73. Winter, D. B. & Gearhart, P. J. Another piece in thehypermutation puzzle. Curr. Biol. 5, 1345–1346 (1995).

74. Storb, U. The molecular basis of somatic hypermutation ofimmunoglobulin genes. Curr. Opin. Immunol. 8, 206–214(1996).

75. Milstein, C., Neuberger, M. S. & Staden, R. Both DNAstrands of antibody genes are hypermutation targets. Proc.Natl Acad. Sci. USA 95, 8791–8794 (1998).

76. Steel, E. J., Rothenfluh, H. S. & Blanden, R. V. Mechanismof antigen-driven somatic hypermutation of rearrangedimmunoglobulin V(D)J genes in the mouse. Immunol. Cell.Biol. 75, 82–95 (1997).

77. Winter, D. B., Sattar, N. & Gearhart, P. J. in Current topicsin Microbiology and Immunology (eds Kelsoe, G. & Flajnits,M. F.) 1–10 (Springer, Berlin, 1998).

78. Diaz, M. & Flajnik, M. F. Evolution of somatic hypermutationand gene conversion in adaptive immunity. Immunol. Rev.162, 13–24 (1998).An excellent summary of the key features of somatichypermutation spectra.

79. Tissier, A., McDonald, J. P., Frank, E. G. & Woodgate, R.pol ι , a remarkably error-prone human DNA polymerase.Genes Dev. 14, 1642–1650 (2000).Demonstration that human Rad30B/pol ι prefers tomake the wobble G–T base pair over the correct A–Tbase pair, and overall has extremely poor fidelity.

80. Foster, S. J., Dorner, T. & Lipsky, P. E. Somatichypermutation of VκJκ rearrangements: targeting ofRGYW motifs on both DNA strands and preferentialselection of mutated codons within RGYW motifs. Eur. J.Immunol. 29, 4011–4021 (1999).

81. Dominguez, O. et al. DNA polymerase mu (Pol µ),homologous to TdT, could act as a DNA mutator ineukaryotic cells. EMBO J. 19, 1731–1742 (2000).

82. Garcia-Diaz, M. et al. DNA polymerase lambda (Pollambda), a novel eukaryotic DNA polymerase with apotential role in meiosis. J. Mol. Biol. 301, 851–867 (2000).

83. Nagasawa, K.-I. et al. Identification and characterization of

human DNA polymerase β2, a DNA polymerase β–relatedenzyme. J. Biol. Chem. 275, 31233–31238 (2000).

84. Wang, Z., Castano, I. B., de Las Penas, A., Adams, C. &Christman, M. F. Pol κ: A DNA polymerase required forsister chromatid cohesion. Science 289, 774–779 (2000).Discovery that the yeast Trf4 protein is a polymeraseessential for sister-chromatid cohesion, indicatingthat possible polymerase exchange mechanismsmay be needed to replicate chromosomes fully.

85. Sharief, F. S., Vojta, P. J., Ropp, P. A. & Copeland, W. C.Cloning and chromosomal mapping of the human DNApolymerase θ (POLQ), the eighth human DNA polymerase.Genomics 59, 90–96 (1999).

86. Oshige, M., Aoyage, N., Harris, P. V., Burtis, K. C. &Sakaguchi, K. A new DNA polymerase species fromDrosophila melanogaster: a probable mus308 geneproduct. Mutat. Res. 433, 183–192 (1999).

87. Radman, M. Enzymes of evolutionary change. Nature 401,866–869 (1999).

88. Limoli, C. L., Giedzinski, E., Morgan, W. F. & Cleaver, J. E.Polymerase η deficiency in the Xeroderma pigmentosumvariant uncovers an overlap between the S phasecheckpoint and double-strand break repair. Proc. NatlAcad. Sci. USA 97, 7939–7946 (2000).

89. Opperman, T., Murli, S., Smith, B. T. & Walker, G. C. A model for umuDC-dependent prokaryotic DNA damagecheckpoint. Proc. Natl Acad. Sci. USA 96, 9218–9223(1999).

90. Murli, S., Opperman, T., Smith, B. T. & Walker, G. C. A rolefor the umuDC gene products of Escherichia coli inincreasing resistance to DNA damage in stationary phaseby inhibiting the transition to exponential growth. J.Bacteriol. 182, 1127–1135 (2000).

91. Bemark, M., Khamlichi, A. A., Davies, S. L. & Neuberger,M. S. Disruption of mouse polymerase ζ (Rev3) leads toembryonic lethality and impairs blastocyst development invitro. Curr. Biol. 10, 1213–1216 (2000).

92. Wittschieben, J. et al. Disruption of the developmentallyregulated Rev3I gene causes embryonic lethality. Curr.Biol. 10, 1217–1220 (2000).

93. Shapiro, L. & Losick, R. Dynamic spatial regulation in thebacterial cell. Cell 100, 89–98 (2000).

94. Lemon, K. P. & Grossman, A. D. Localization of bacterialDNA polymerase: evidence for a factory model ofreplication. Science 282, 1516–1519 (1998).

95. Sutton, M. D., Opperman, T. & Walker, G. C. TheEscherichia coli SOS mutagenesis proteins UmuD andUmuD′ interact physically with the replicative DNApolymerase. Proc. Natl Acad. Sci. USA 96, 12373–12378(1999).

96. Kornberg, A. Ten commandments: Lessons from theenzymology of DNA replication. J. Bacteriol. 182,3613–3618 (2000).

97. Perry, K. L., Elledge, S. J., Mitchell, B. B., Marsh, L. &Walker, G. C. umuDC and mucAB operons whoseproducts are required for UV light- and chemical-inducedmutagenesis: UmuD, MucA, and LexA proteins sharehomology. Proc. Natl Acad. Sci. USA 82, 4331–4335(1985).

98. Shinagawa, H., Iwasaki, T., Kato, T. & Nakata, A. RecAprotein-dependent cleavage of UmuD protein and SOSmutagenesis. Proc. Natl Acad. Sci. USA 85, 1806–1810(1988).

99. Maor-Shoshani, A., Reuven, N. B., Tomer, G. & Livneh, Z.Highly mutagenic replication by DNA polymerase V (UmuC)provides a mechanistic basis for SOS untargetedmutagenesis. Proc. Natl Acad. Sci. USA 97, 565–570(2000).

100. Johnson, R. E., Prakash, S. & Prakash, L. The humanDINB1 gene encodes the DNA polymerase pol θ. Proc.Natl Acad. Sci. USA 97, 3838–3843 (2000).

101. Morelli, C., Mungall, A. J., Negrini, M., Barbanti–Brodano,G. & Croce, C. M. Alternative splicing, genomic structure,and fine chromosome localization of REV3L. Cytogenet.Cell Genet. 83, 18–20 (1998).

102. Goldsmith, M., Sarov-Blat, L. & Livneh, Z. Plasmid-encoded MucB protein is a DNA polymerase (pol RI)specialized for lesion bypass in the presence of MucA′,RecA, and SSB. Proc. Natl Acad. Sci. USA 97,11227–11231 (2000).

AcknowledgementsThis work was supported by grants from the National Institutesof Health. I want to express heartfelt gratitude to my collabora-tors, Roger Woodgate, Mike O’Donnell, John-Stephen Taylor,Kevin McEntee and especially to Hatch Echols. I also want toexpress my sincere appreciation to the students in my laborato-ry, Phuong Pham, Mengjia Tang, Xuan Shen, Irina Bruck andJeffrey Bertram.


110 | OCTOBER 2000 | VOLUME 1 www.nature.com/reviews/molcellbio

R E V I E W S

The ability to move in a directed, purposeful manner isone of the properties we most closely associate with liv-ing cells. Many forms of cell motility, such as the intra-cellular movement of organelles by molecular motors,rely on discrete and stable protein machines. By cou-pling energy release to a protein conformationalchange, myosin and kinesin carry their cargo a singlestep along their substrate, and larger-scale movement issimply the linear addition of many such discrete steps.These types of motility have been extensively studied inpurified or semi-purified systems, and a great deal isknown about the molecular and biophysical require-ments for movement1.

In contrast, AMOEBOID MOTILITY is not driven by dis-crete machines acting additively. Instead, it is a complexprocess involving an interconnecting network of non-equilibrium, dynamic, whole-cell events. Althoughdetailed descriptive studies of amoeboid motility havegraced the cell biological literature for over fifty years, itcould not be easily investigated at the molecular levelusing classic biochemical or genetic techniques.However, the past ten years have seen remarkableadvancements in our understanding of the molecularbasis of amoeboid motility. The breakthrough came,oddly, from a bacterial pathogen called Listeria monocy-togenes. Like many cell biologists, this pathogen choseactin-based motility as its field of study, but it has hadthe advantage of millions of years of evolutionaryexperimentation. This review tells the story of how the

secrets of amoeboid motility known to this tiny bacteri-um have been revealed.

Actin dynamics in locomoting cellsCrawling cells, such as epithelial cells, FIBROBLASTS or neu-rons, have at their front a broad, flat region, usually less

SECRETS OF ACTIN-BASEDMOTILITY REVEALED BY ABACTERIAL PATHOGENLisa A. Cameron, Paula A. Giardini, Frederick S. Soo and Julie A. Theriot

Actin-based cell motility is a complex process involving a dynamic, self-organizing cellularsystem. Experimental problems initially limited our understanding of this type of motility, but theuse of a model system derived from a bacterial pathogen has led to a breakthrough. Now, allthe molecular components necessary for dynamic actin self-organization and motility havebeen identified, setting the stage for future mechanistic studies.

Department ofBiochemistry, StanfordUniversity School ofMedicine, 279 CampusDrive West, Stanford,California 94305-5307,USA.e-mail:[email protected] to: J.A.T.

Figure 1 | A rapidly moving cell; a keratocyte from the skinof a fish. This is a phase-contrast micrograph, a single framefrom a video sequence. The lamellipodium and cell body arelabelled. This cell is moving in the direction of the large arrow.

Movie online

Cell body

Lamellipodium

AMOEBOID MOTILITY

A distinctive form of cellcrawling typified by Amoebaproteus, which involvesextension of pseudopodia andcytoplasmic streaming.

FIBROBLAST

Common cell type found inconnective tissue in many partsof the body, which secretes anextracellular matrix rich incollagen and othermacromolecules and connectscell layers.



R E V I E W S

forward7. In slowly moving fibroblasts, the actin mesh-work moves rearward with respect to the substrate, evenas the leading edge continues to move forward8. To usethe treadmilling activity of actin assembly in the lamel-lipodium for efficient forward extension of its leadingedge, the cell must anchor the actin meshwork throughthe plasma membrane to the underlying substrate. Akeratocyte has an efficient ‘clutch’ mechanism thatallows rapid forward protrusion and little or no rear-ward flux, whereas a fibroblast has a ‘slippery clutch’ thatresults in significant rearward flux and slow forwardprotrusion. Within a given cell type, there is no correla-tion between the rate of centripetal movement of actinand the rate of lamellipodial protrusion, so the compo-nents that control the rate of protrusion by regulatingactin dynamics must be localized at the leading edge8.

The depolymerization of actin from the meshworkseems to be tightly controlled. Depending on the celltype, the average lifetime of actin filaments in the lamel-lipodium is very short — around 20 seconds to 2 min-utes7,8. The rate of filament loss is correlated with cellspeed: rapidly moving cells have more labile actin fila-ments in their lamellipodia, whereas filaments in slowlymoving cells are more stable9. But most importantly, inall cells, the turnover of actin filaments is at least twoorders of magnitude faster than the turnover of pureactin filaments in solution, indicating that other pro-teins inside the cell must be actively disassembling thefilaments in the lamellipodia.

Such cell-based experiments established organiza-tional principles for making a functional lamellipodium

than one micrometre thick, filled with a dense mesh-work of actin filaments. Referred to as a lamellipodium,this region of the cell contains all of the machinery nec-essary for amoeboid motility. Rapidly moving cells,such as the fish epidermal KERATOCYTE, consist basically ofa large lamellipodium that carries the cell body on itsdorsal surface (FIG. 1) (Movie 1). Small lamellipodialfragments sliced off from a cell body can crawl on theirown, essentially forming a tiny nucleus-free cell2,3. Thecrawling process can be broken down into three sub-processes: the assembly of actin into a coherent mesh-work at the LEADING EDGE of the lamellipodium, thecoupling of this meshwork to the external substrate,and the controlled depolymerization of the meshworkfor recycling and reuse of the actin monomer.Understanding how each of these subprocesses is regu-lated and how they interconnect and work together iscritical to the study of how cells crawl.

Several experiments established the principle thatactin assembly (BOX 1) occurs primarily at the front oflamellipodia. In fibroblasts and neuronal GROWTH CONES,a spot photobleached in the actin meshwork of thelamellipodium translocates backwards slowly andmoves rearward relative to the leading edge in a coher-ent fashion4,5. When a stationary neuronal growth coneis allowed to recover after the actin meshwork is com-pletely depolymerized by treatment with the toxincytochalasin, the actin network reforms exclusively atthe leading edge and then moves rearward6.

Across cell types, the rate of rearward flux of theactin meshwork is negatively correlated with the speedof forward protrusion. In rapidly moving fish kerato-cytes, photoactivated spots of fluorescent actin that haveincorporated into the lamellipodium remain stationarywith respect to the substrate as the leading edge moves

Figure 2 | Movement of Listeria monocytogenes in aninfected host cell. This is a phase-contrast micrograph, asingle frame from a video sequence. The kidney epithelial cellwas infected about five hours before the acquisition of thisvideo sequence. All of the bacteria in this cell are clonaldescendants of a single individual. A bacterium and itsassociated comet tail are labelled. Bacteria are moving in thedirection of the blue arrows.

Movie online

Bacterium

Comet tail

Box 1 | Actin filament dynamics

Actin is one of the most abundant proteins in eukaryotic cells, and is a primarydeterminant of cell shape and cytoplasmic structure. It exists in two forms, G-actin (forglobular), the soluble 43 kDa protein subunit, and F-actin (for filamentous), a helicalpolymer of arbitrary length where individual subunits self-associate in a head-to-tailfashion. About half the actin in a typical cell (up to 50 µM) is in the form of G-actin,and the other half is in the form of F-actin. Actin is an ATPase, and ATP hydrolysisaffects the kinetics of polymerization.

The rate-limiting step in the formation of F-actin from a solution of pure G-actin isthe formation of a stable ‘nucleus’. When two molecules of G-actin collide in solution,they will form a dimer, but the dimer comes apart rapidly and no filament can grow.When three or four molecules collide simultaneously, they form a more stable trimeror tetramer, which can be rapidly elongated by further collisions of individual subunitswith either end of the growing filament. In cells, spontaneous nucleation is rare. Cellsregulate the location of new F-actin formation by regulating nucleation.

Within the cell, the dynamic behaviour of F-actin and G-actin is modified andregulated by a group of over 100 actin-binding proteins. These include proteins thatbind to G-actin and prevent it from polymerizing, proteins that bind to F-actin andprevent it from depolymerizing, accessory proteins that affect the rate of nucleotidehydrolysis, proteins that sever long filaments into smaller bits, proteins that bypass theslow steps of nucleation, myosin motors that carry cargo along filaments … in short,proteins to speed up or slow down every dynamic behaviour of this remarkablepolymer. In addition, F-actin crosslinking proteins can assemble multiple filamentsinto larger-scale structures, including bundles where all the filaments align in paralleland meshworks where the filaments cross orthogonally.

KERATOCYTE

A small, motile cell type foundin the epidermis of fish andamphibians.

LEADING EDGE

The thin margin of alamellipodium spanning thearea of the cell from the plasmamembrane to about 1 µm backinto the lamellipodium.

GROWTH CONE

Motile tip of the axon ordendrite of a growing nerve cell,which spreads out into a largecone-shaped appendage.



R E V I E W S

Actin-based motility is essential in the L. monocyto-genes life cycle. In a natural food-borne infection, thebacteria induce phagocytosis by the epithelial cells lin-ing the small intestine, an event that can be replicated inthe laboratory in a wide variety of tissue-culture lines.The bacteria then secrete a pore-forming toxin (listeri-olysin O) that degrades the enclosing membrane, andescape into the cytoplasm of the host cell. After a fewhours of growth and division, host-cell actin filamentsbegin to form a dense cloud on the surface of the bacte-ria. Subsequently, the actin cloud becomes polarizedinto a comet tail made up of an oriented, crosslinkednetwork of actin filaments with their barbed endspointing towards the bacterium12. Bacteria associatedwith comet tails move very rapidly within the host cyto-plasm, at rates of up to 1 µm per second16 (FIG. 2) (Movie2). Finally, the infection spreads as the bacteria pushtheir way into neighbouring cells through plasmamembrane projections17,18.

The L. monocytogenes comet tail resembles a simpli-fied lamellipodium, and the bacterial surface imitatesthe plasma membrane at the leading edge. Labelledexogenous actin monomers preferentially incorporateinto the actin tail near the bacterial surface in living andpermeabilized cells19,20, recapitulating the behaviour inlamellipodia, where new incorporation is primarily atthe leading edge. Fluorescence photoactivation experi-ments reveal that filaments in the tail of a moving bac-terium remain stationary and only the bacteriummoves forward21, as in rapidly moving lamellipodia.The rate of filament depolymerization in the comet tailis independent of either position in the tail or bacterialspeed, and the filaments have very short half-lives, ofthe order of 30 seconds21.

In contrast to amoeboid motility, bacterial actin-based motility does not require the host cell plasmamembrane and could therefore be reconstituted in acell-free cytoplasmic extract, which has facilitated bio-chemical approaches to study the regulation of actin

that could not have been predicted from in vitro prop-erties of actin polymerization. First, actin filaments arenucleated and grow primarily at the leading edge,immediately adjacent to the plasma membrane.Second, filaments are crosslinked into a coherent mesh-work that either remains stationary with respect to thesubstrate as the cell moves forward (in rapidly movingcells) or moves rearward towards the cell body (in sta-tionary or slowly moving cells). And last, actin filamentsin the bulk of the meshwork, away from the leadingedge, depolymerize rapidly so that steady-state, self-organized movement can be maintained. Cell motilityrequires that these three processes be properly coordi-nated in space and time.

In a lamellipodium, there must be modulatory fac-tors that govern these phenomena. This raises severalspecific molecular questions. What factors catalysenucleation and elongation of actin filaments at the lead-ing edge? How are filament nucleation and elongationsuppressed elsewhere? What is causing the older fila-ments to depolymerize so rapidly? What holds themeshwork together, and is it important for motility thatthis meshwork be coherent?

Technical problems have impeded attempts toanswer these questions. Genetic systems have been oflimited use in identifying the full complement of com-ponents that make up the machinery necessary foractin-based motility.Yeast, unfortunately, do not crawl.Genetic and reverse genetic approaches in model meta-zoans and in Dictyostelium successfully defined the cel-lular functions of some of the individual componentsof the motility apparatus. However, because the actinpolymerization machinery necessary for amoeboidmotility is so critical for other aspects of cellular behav-iour, many cells with lesions in important cytoskeletalloci are inviable and therefore difficult to evaluate formotility phenotypes. In addition, functional redun-dancy is rampant in the actin cytoskeleton, so manynull mutants in genes that encode interesting proteinshave no detectable phenotype. Biochemical reconstitu-tion of amoeboid motility has been hindered by theneed for an intact cell plasma membrane, which mustserve to localize filament nucleation10 and might con-tribute to force generation11. A decade ago, the identifi-cation of a genetically manipulable model system thatcould mimic the actin filament dynamics of lamellipo-dial protrusion without the requirement for a plasmamembrane was desperately needed to understandactin-based motility at the molecular level.

Actin-based motility of bacterial pathogensIn the late 1980s, several research groups found that F-actin is responsible for the intracellular movement oftwo unrelated bacterial pathogens, Listeria monocyto-genes12,13 and Shigella flexneri14,15, which live within thecytoplasm of the host cell. Because L. monocytogenes isless virulent and easier to handle experimentally than S.flexneri, most laboratories investigating this form ofactin-based motility have chosen to focus on L. monocy-togenes, and we will concentrate on the L. monocyto-genes model system in this review.

Figure 3 | Reconstitution of Listeria monocytogenesmotility in a cytoplasmic extract. Top row | Phase-contrast images showing the position of the bacterium.Adjacent frames are separated by ten-second intervals.Bottom row | Fluorescent signal arising from rhodamine–actinadded as a tracer. The fluorescent images were captured lessthan one second after the corresponding phase image. Theonline movie shows a polystyrene bead coated with ActAmoving in a similar extract.

Movie online



R E V I E W S

a comet tail, indicating that it probably interacts withother host-cell factors. As ActA is a surface protein, ini-tial efforts focused on host-cell factors that localize tothe bacterial surface and not throughout the comet tail.Two proteins that fulfil this criterion were initially iden-tified by immunofluorescence surveys: the G-actin-binding protein profilin22 and an F-actin associated pro-tein called vasodilator-stimulated phosphoprotein(VASP)28. Both proteins require ActA to associate withthe bacterial surface.VASP binds directly to ActA28, andprofilin binds to VASP29. It seemed unlikely that thesefactors were responsible for localized actin nucleation,however, as cytoplasmic extracts depleted of profilincould still support nucleation of actin clouds by L.monocytogenes22,30. Furthermore,VASP binds to F-actin29

but does not show nucleating activity, and profilin sig-nificantly inhibits nucleation31. Profilin can serve as anucleotide exchange factor for actin32, and it can alsolower the effective critical concentration for actin poly-merization in a cellular environment33. So, it was con-cluded that the combination of VASP and profilin mayaccelerate filament elongation at the bacterial surface,but neither one is the nucleator.

Systematic ActA deletion studies carried out in bac-teria indicated that VASP and profilin are localized by acentral polyproline-rich region in the middle of the pro-tein, and that four consensus FPPPP repeats act in anadditive fashion to bind multiple molecules of VASP34.Bacteria containing an ActA construct that lacks theVASP-binding domain still mediate actin nucleation,although they move more slowly than wild-type bacte-ria in both cytoplasmic extracts35 and in infected cells34.Interestingly, the rate of motility is linearly related to thenumber of proline-rich repeats present34.Immunolocalization shows that the central proline-richdomain recruits VASP36, and biochemical experimentsshow that the consensus motif is sufficient for VASPbinding37. Subsequently, the VASP-related proteinsmammalian Enabled (Mena) and Ena/VASP-like pro-tein (Evl) were found to act interchangeably with VASPin bringing profilin to the bacterial surface and in

dynamics22 (FIG. 3). The exclusive localization of actin fil-ament growth close to the bacterial surface indicatedthat factors either secreted by the bacterium orexpressed on the bacterial surface must trigger actinpolymerization. Because L. monocytogenes continue tomove in cells in the presence of drugs that inhibit bacte-rial protein synthesis20, the key factor was probably astable protein on the surface of the bacterium ratherthan a factor secreted continually by the bacterium.

The bacterial factors involved in actin polymerizationare required for the spread of bacteria from cell to cell. Toidentify the gene(s) required for actin assembly, screenswere designed to identify mutant bacteria deficient inthe ability to spread from cell to cell, but capable of nor-mal initial cell invasion, membrane lysis and bacterialdivision. The only gene ever isolated in such L. monocy-togenes screens is actA23,24. Furthermore, ActA confersactin-based motility on normally immotile bacteria, forexample if actA is expressed in the non-pathogenicstrain Listeria innocua25, or if purified ActA protein isattached asymmetrically to Streptococcus pneumoniae26.Polystyrene beads coated with purified ActA proteinhave been shown to form comet tails and move in cyto-plasmic extracts, proving that no other bacterial surfacecomponents are required for motility27 (Movie 3).

Dissection of the comet tail and lamellipodiumA flurry of experimental work followed the identificationof ActA and the reconstitution of motility in cytoplasmicextracts. Today, researchers in the field largely agree onthe identities and the functions of all the main moleculesrequired for regulating actin filament dynamics in theself-organized motile system of the comet tail. The simi-larities, and differences, between the simplified bacterialcomet tail and the far more complicated, dynamic lamel-lopodial structure have given great insight into the orga-nization and regulation of these systems at many levels.

Factors catalysing nucleation and elongation. AlthoughActA is sufficient to cause polymerization at the bacteri-al surface, it does not interact directly with actin to form

Box 2 | Special features of the lamellipodium

In general, it is thought that the dynamic behaviour of the actin-binding proteins described here is comparablebetween comet tails and lamellipodia. However, there are important differences. For example, ActA is not found ineukaryotic systems. The search for the eukaryotic equivalent of ActA led to the characterization of a new proteinfamily called WASP/Scar. Wiskott–Aldrich syndrome protein (WASP) is expressed only in human haematopoietic cellsand contains a GTPase binding domain69 that binds the small GTPases, Cdc42 and Rac, known to be involved inregulating the triggering of actin polymerization in fibroblasts70. Its close relative N-WASP is expressed widely invertebrate cells71, and causes filopodial formation when co-expressed with Cdc42 in cultured cells72. The moredistantly related protein Scar (for suppressor of cyclic AMP receptor mutation) was discovered in Dictyostelium, whereits deletion causes cytoskeletal defects73. WASP and Scar interact with the p21 subunit of Arp2/374 and, like ActA, Scaractivates Arp2/3 to nucleate actin filaments75. Finally, polystyrene beads coated with WASP are capable of formingactin comet tails and moving in cytoplasmic extracts, in a manner apparently identical to the movement of L.monocytogenes or ActA-coated beads76.

Much of what is known about cell motility and lamellipodial protrusion has come from descriptive observations.Two of the most visually striking behaviours of lamellipodia include ruffling and rearward flux. Ruffling is aphenomenon where the protruding leading edge detaches from the substrate and folds back on the dorsal surface ofthe lamellipodium. Rearward flux, described in the section ‘Actin dynamics in locomoting cells’, requires myosin77.Neither of these characteristic behaviours can be investigated using L. monocytogenes as a model system.



R E V I E W S

gation by VASP may not directly translate into anincrease in crawling speed in mammalian cells, possiblyhighlighting an important difference between lamel-lipodia and L. monocytogenes motility (for other differ-ences, see BOX 2).

But regardless of the nature of the accelerator, whatturns the ignition key by catalysing F-actin nucleation inthe first place? The answer to this question came from abiochemical study. Platelet extracts, rich in actin-associ-ated cytoskeletal proteins and easy to obtain in largequantities, were fractionated, and the fractions examinedusing a visual assay to determine which could supportthe formation of F-actin clouds around L. monocyto-genes. The most purified active fraction contained atightly associated protein complex consisting of sevenpolypeptide chains42. This complex, named Arp2/3 aftertwo of its members (actin related proteins 2 and 3), hadinitially been isolated as a profilin-binding complex byaffinity chromatography of Acanthamoeba castellaniicytosol43, but its function in the regulation of actindynamics had previously been unclear. The nucleatingactivity of Arp2/3 can be measured in vitro44, but it is

enhancing the speed of actin-based motility38,39.Profilin tagged with green fluorescent protein(GFP–profilin) associates with moving bacteria ininfected cells and, strikingly, the concentration ofGFP–profilin at the bacterial surface is closely correlatedwith bacterial speed40. This indicates that the proline-rich repeats, VASP and profilin may act together as anaccelerator for bacterial movement.

VASP and profilin may also have independent effectson filament elongation. Experiments using humanplatelet extracts show that movement of L. monocyto-genes is still enhanced in the presence of a mutant pro-filin that does not bind proline-rich sequences andtherefore cannot associate with VASP. Conversely,VASPstill accelerates L. monocytogenes motility in profilin-depleted extracts39. Surprisingly, overexpression ofmembers of the Ena/VASP family in mammalian cellscauses them to move at less than half the speed of wild-type cells, and removal of Ena/VASP proteins by seques-tration to the mitochondrial surface causes cells tomove faster than wild-type cells41. This observationindicates that enhancement of the rate of filament elon-

Figure 4 | Diagram of the molecular components required for actin-based motility of Listeria monocytogenes.a | Interactions between host-cell proteins and ActA at the bacterial surface. Two domains of ActA are required for normalmotility. The amino-terminal domain activates actin filament nucleation through Arp2/3. The central proline-rich domain bindsVASP and profilin interacts with VASP, enhancing filament elongation. b | Host protein functions throughout the comet tail. Inaddition to the factors that act at the bacterial surface, capping protein binds to the barbed end of actin filaments to preventelongation of older filaments, α-actinin crosslinks filaments to stabilize the tail structure, and ADF/cofilin disassembles oldfilaments. (VASP, vasodilator-stimulated phosphoprotein.)

C

C

CC

C

CC

P P P

P

P P P

CC

C CC

C

C

C

C

C

CC

CActin/Arp2/3interaction

Proline-richrepeats

VASP

VASP

P

PP

a b

P Capping proteinArp2/3 complex Cofilin α-ActininProfilin Actin

Bacterium

Bacterium

ActA

ActA

Table 1 | Actin-based motility in other systems

System Structure formed Host factors Special features References

Shigella flexneri Comet tail N-WASP, Arp2/3, profilin • Bacterial factor IcsA (VirG) has no homology with ActA 14, 15, 25,similar to Listeria • Actin dynamics are identical to L. monocytogenes 50, 80, 81

Rickettsia spp. Comet tail, made of • Bacterial factor not identified 82–85long twisted bundles • Tails are straight

• Dynamics are distinct from L. monocytogenes and S. flexneri

Enteropathogenic Pedestal WASP, Arp2/3 • Signalling from bacterium occurs across host cell membrane 86, 87Escherichia coli • Bacterial factors intimin and Tir required

Vaccinia virus Comet tail N-WASP, Nck, WIP, • Intracellular enveloped form moves 88–90similar to Listeria Src-family tyrosine kinase • Actin-based movement may contribute to viral budding

Vesicles Transient comet tail PtdIns(4,5)P2, Cdc42, • Endosomal rocketing induced by phorbol esters, metal ions 91–95similar to Listeria N-WASP, Arp2/3 • May occur normally with nascent endosomes

Nck, non-catalytic region of tyrosine kinase; PtdIns(4,5)P2, phosphatidylinositol-4,5-bisphosphate; N-WASP, neural Wiskott–Aldrich syndrome protein; WASP,Wiskott–Aldrich syndrome protein; WIP, Wiskott–Aldrich syndrome protein interacting protein.



R E V I E W S

The simplest mechanism to achieve this would involvecapping the growing barbed ends of the older filaments.Several proteins with barbed-end F-actin capping activi-ty are known, including capping protein (also known asCapZ) and gelsolin. Biochemical studies indicate thatActA may suppress capping close to the bacterial sur-face30, although ActA probably exerts this effect indirect-ly. Capping protein is strongly associated with comettails51. Gelsolin is localized throughout the tail and, para-doxically, is enriched at the bacterial surface52, but it maybe inactive there. The combination of barbed-end cap-ping suppression at the bacterial surface, exclusive local-ization of the elongation enhancers VASP and profilin atthe bacterial surface, and potent activation of Arp2/3 byActA, seems to be sufficient to enable nucleation andelongation only at the front of the comet tail.

Factors causing filament depolymerization. An impor-tant feature of both lamellipodial and bacterial motilitydynamics is the rapid depolymerization of the actinmeshwork far from the leading edge and the bacterialsurface, suggesting the existence of depolymerizing fac-tors. Another functional study using L. monocytogeneselucidated the specific role of the protein that controls fil-ament depolymerization. This protein, called ADF (actindepolymerizing factor) or cofilin, was first identified inbiochemical assays as a factor that accelerates actindepolymerization.ADF/cofilin binds cooperatively to thesides of actin filaments and increases the twist of F-actin53, destabilizing the filament structure and causingan increase in the rate of spontaneous filament breakage54

and a significant acceleration of subunit dissociationfrom the filament pointed end55. It has a higher affinityfor ADP-containing filaments, and so preferentially accel-erates the turnover of old filaments after nucleotidehydrolysis has occurred, rather than the newest filaments,which still contain ATP. This combination of activitiesmakes cofilin the leading candidate as the factor responsi-ble for the 10–100-fold higher actin turnover rate in cellscompared with the turnover rate of pure actin.

Immunodepletion of ADF/cofilin from cytoplas-mic extracts supporting L. monocytogenes motilityalters the morphology of the comet tails, making themfive times longer than normal56. Conversely, additionof excess exogenous ADF/cofilin to extracts causesshortening of the actin tail and increases the rate ofbacterial motility55. ADF/cofilin is localized through-out the L. monocytogenes comet tail56, consistent withthe previous finding that depolymerization occurs uni-formly everywhere in the tail21.

Experiments done in intact cells also show thatADF/cofilin is important for acceleration of actinturnover57. ADF/cofilin is localized to the leading edgeand ruffling membrane of motile cells58. Rapidly mov-ing keratocytes show an ADF/cofilin-free zone at thevery leading edge of the lamellipodium47. This narrowADF/cofilin-free margin may be the geometrical corre-late of the hydrolysis kinetics of ATP; the newest ATP-containing filaments at the front are briefly protectedfrom rapid disassembly by ADF/cofilin. The spatial sep-aration between nucleation and elongation at the front,

markedly activated by the presence of ActA (REF. 45). Theamino-terminal domain of ActA, which was implicatedby the deletion studies as being sufficient to cause actinnucleation inside of cells, is also sufficient for full activa-tion of Arp2/3 (REF. 46). Arp2/3 can also bind to the sideof a pre-existing actin filament and initiate nucleation ofa new filament at that location, creating a branch at a 70°angle from the original filament44. Such branches arefound throughout the lamellipodium47 with Arp2/3 atthe branch points.Arp2/3 is localized to the leading edgeof several cell types42,43,48, and it is found throughout theactin comet tail associated with L. monocytogenes49. So,activated Arp2/3 is responsible for the nucleation ofactin polymerization at the bacterial surface (FIG. 4a).

Activation of Arp2/3 by ActA is essential for L. mono-cytogenes motility. No full-length homologues of ActAexist in mammalian cells, so how is Arp2/3 activated inlamellipodia? Similarly, IcsA (VirG), the bacterial surfaceprotein required for S. flexneri actin-based motility, doesnot interact directly with Arp2/3, indicating that someother factor mediates the activation of Arp2/3 in this sys-tem. Recently, it has been found that Wiscott–Aldrichsyndrome protein (WASP) and its relatives, N-WASP andScar (for suppressor of cyclic AMP receptor mutation),have a function similar to ActA, activating Arp2/3 down-stream of signalling through small GTPases (BOX 2). In thecase of S. flexneri motility, N-WASP binds IcsA (VirG)and activates Arp2/3 at the bacterial surface50 (TABLE 1).

Factors suppressing nucleation and elongation. Newactin filaments are continuously nucleated and elongat-ed exclusively at the bacterial surface or at the leadingedge of the cell, suggesting that some mechanism existsto prevent the continuing elongation of old filaments.

Figure 5 | Functions of similar proteins in the lamellipodium. N-WASP activates Arp2/3 tonucleate actin filaments. VASP and profilin, which are localized to the leading edge, facilitateelongation. Capping protein caps the barbed ends of older filaments. Filamin crosslinksfilaments into an actin network. Finally, ADF/cofilin accelerates depolymerization throughout thelamellipodium, except for a cofilin-free zone at the immediate leading edge (reviewed in REF. 78).The localization of N-WASP is not well known. Here we show it at the leading edge, binding andactivating Arp2/3 in the cytoplasm, and associated with Arp2/3 at filament branches. (VASP,vasodilator-stimulated phosphoprotein; N-WASP; neural Wiskott–Aldrich syndrome protein.)

P

C

C C

C

C

C

C

C

C

C

P

P

P

PP

P P

P

PP

P

PP

P

P

P

P

N-WASPN-WASP

N-WASP N-WASP

N-WASP

N-WASP

N-WASP N-WASP

N-WASP

N-WASP

N-WASP

N-W

ASP

N-WAS

P

Capping proteinArp2/3 complex Cofilin FilaminProfilin Actin

Plasma membrane

VASP

CC

C

C

C

C

C

C

CC



R E V I E W S

plished62 (FIG. 4b). Of these proteins, only actin, Arp2/3,ADF/cofilin and capping protein are absolutely requiredfor motility.VASP and profilin increase the rate of move-ment, and α-actinin stabilizes the tail. The success of thispurified system definitively establishes the minimalcomponents necessary for this type of self-organizedactin-based motility. In addition, it clearly shows thatthis form of force generation does not require a myosinmotor62, and that actin polymerization alone can act as abona fide molecular motor63. Moreover, the functionalconservation between the lamellipodium and the comettail is striking (TABLE 2, FIG. 5), proving that L. monocyto-genes is an excellent tiny cell biologist.

Open questionsThe establishment of a purified protein motility systemthat includes all of the bacterial and host componentsnecessary and sufficient for actin-based movement62 rep-resents a satisfying culmination of the past decade ofmolecular research in this field. Work from numerouslaboratories has brought us to the point where we nowknow most of the critical molecules involved in actin reg-ulation at the leading edge of cells. However, despite thislist of molecules, we still cannot formulate the set of rulesneeded to generate a motile cell, or a lamellipodium, oreven a comet tail. The purified system has confirmed thesupposition that actin polymerization alone must pro-duce the force necessary for motility, but provides no fur-ther information about the actual microscopic mecha-nism of force generation. It will, however, provide auseful experimental system for investigating some of theopen mechanistic questions about cell movement.

Translating the minimal requirements for bacterialactin-based motility to a mechanism of cell motility oreven the protrusion of a lamellipodium requires a signif-icant increase in complexity, the most important differ-ence being the presence of a plasma membrane. Furtherunderstanding of cell motility will also require knowl-edge of how the regulation of cell adhesion and nucleartranslocation are integrated with lamellipodial protru-

and disassembly further back, helps to maintain thesteady-state organization of the motile F-actin mesh-work in both lamellipodia and comet tails.

Factors that crosslink filaments. Because Arp2/3 fre-quently binds to the side of a pre-existing filament as itnucleates the growth of a new filament, the meshworkforming at the front of the comet tail or the leading edgeof the lamellipodium is effectively crosslinked at birth, ina dendritic web44,47. In addition, numerous F-actincrosslinking proteins are found throughout the comettail, including fimbrin59 and α-actinin16. Microinjectionof a dominant-negative fragment of α-actinin, whichinhibits crosslinking by the endogenous protein, causesL. monocytogenes in infected cells to stop moving60. Thisobservation indicates that strong crosslinking is impor-tant for movement through the highly viscous cytoplasmof a living cell, although its mechanical contribution maybe less important in cytoplasmic extracts (see below).

Fimbrin and α-actinin, which crosslink F-actin toform tight parallel bundles, are not generally found inlamellipodia. Instead, lamellipodia are enriched in a dif-ferent type of crosslinking protein, filamin (also calledABP-280), which tends to crosslink filaments at rightangles to form a web. Mutant melanoma cells that fail toexpress filamin show very poor motility61. These com-plementary results in comet tails and in lamellipodiaseem to indicate that crosslinking is indeed importantfor mechanical stability of a protrusive self-organizedactin structure, but that the nature of the crosslinkerrequired is probably different depending on the detailsof filament organization in each case.

Establishment of a purified system. Using the molecularinformation provided by the studies detailed above, wemight hypothesize that L. monocytogenes motility couldbe reconstituted in vitro with a mixture of the followinghost proteins: actin,Arp2/3,VASP, profilin, capping pro-tein,ADF/cofilin and α-actinin, along with a steady sup-ply of ATP. This impressive feat has recently been accom-

Table 2 | Similarities between lamellipodia and comet tails

Protein Function Localization in lamellipodia Localization in comet tails

F-actin* Cell shape and Enriched in stress fibres and focal Throughout tail16

cytoplasmic structure contacts

Arp2/3* Nucleation Leading edge and throughout Bacterial surface and Filament crosslinking lamellipodium 47,49 throughout tail42

Pointed-end capping Filament branches47

Capping protein* Barbed-end capping Cytoplasm and cell–cell contacts96 Throughout tail51

Gelsolin‡ Barbed-end capping, Cytoplasm and focal contacts97 Bacterial surface andsevering throughout tail52,56

VASP Binds F-actin and Leading edge98 and Bacterial surface34

profilin, binds proline- focal contacts29

rich region of ActA

Profilin Actin monomer binding Leading edge, focal contacts40 Bacterial surface22

ADF/cofilin* Depolymerizes ADP filaments Lamellipodium58 (1 µm from edge47) Throughout tail56

α-actinin Crosslinking Focal contacts and stress fibres99 Throughout tail16

* Absolutely required for motility in the reconstituted system. ‡ Can be substituted for capping protein.

Lamellipodium

Comet tail


Links

DATABASE LINKS actin | actA | profilin | VASP | Mena | Evl | Ena | Arp2/3 | VirG | WASP |N-WASP | Scar | CapZ | gelsolin | | cofilin | fimbrin | α-actinin | filaminFURTHER INFORMATION L. monocytogenes | Theriot lab homepageENCYCLOPEDIA OF LIFE SCIENCES Actin and actin filaments


R E V I E W S

1. Vale, R. D. & Milligan, R. A. The way things move: lookingunder the hood of molecular motor proteins. Science 288,88–95 (2000).

2. Euteneuer, U. & Schliwa, M. Persistent, directional motilityof cells and cytoplasmic fragments in the absence ofmicrotubules. Nature 310, 58–61 (1984).

3. Verkhovsky, A. B., Svitkina, T. M. & Borisy, G. G. Self-polarization and directional motility of cytoplasm. Curr. Biol.9, 11–20 (1999).

4. Wang, Y. L. Exchange of actin subunits at the leading edgeof living fibroblasts: possible role of treadmilling. J. CellBiol. 101, 597–602 (1985).These photobleaching studies show that actinmonomers are added at the leading edge and thatthe actin meshwork translocates backward towardsthe centre of the cell in a stationary lamellipodium.

5. Okabe, S. & Hirokawa, N. Actin dynamics in growth cones.J. Neurosci. 11, 1918–1929 (1991).

6. Forscher, P. & Smith, S. J. Actions of cytochalasins on theorganization of actin filaments and microtubules in aneuronal growth cone. J. Cell Biol. 107, 1505–1516(1988).

7. Theriot, J. A. & Mitchison, T. J. Actin microfilamentdynamics in locomoting cells. Nature 352, 126–131(1991).Shows that the rate of cell motility is directly relatedto the rate of actin filament assembly at the leading

sion to move the entire cell. But even within the compar-atively simple system of the actin comet tail, many basicmechanistic questions remain to be addressed. We sug-gest that the answers to the following questions willbegin to reveal the rules needed to organize an actincomet tail, and may lead to insights into how to addressthe more difficult questions of whole-cell locomotion.

How is force generated? In theory, motile force for actin-based motility can come solely from the chemicalpotential of actin polymerization63,64. Several models forthe mechanism by which this conversion takes placehave been proposed, including BROWNIAN RATCHET-TYPE

MODELS65 and BULK ELASTIC MODELS66. Current data areinconclusive about which of these models, if any, arecorrect, and have been limited in large part by the natur-al variations in the extract and cultured-cell prepara-tions used to assay motility. With the purified proteinsystem, force generation by the self-organizing actinpolymerization machinery should now be amenable todetailed biophysical experimentation analogous to thestudies that defined the protein conformational changesresponsible for force generation by the discrete motorproteins myosin and kinesin1.

How is movement initiated? Actin-associated bacteria incells seem to exist in two states: moving with a comettail, or stationary with a uniform cloud12. The same twostates are seen in cytoplasmic extracts, both for bacteriaand for ActA-coated polystyrene beads27. These twostates can readily interconvert in a classic bistable sys-tem. What makes a bacterium or a bead start moving?Stochastic modelling based on the elastic Brownianratchet mechanism for force generation has suggestedthat this symmetry breaking event might be caused by aform of dynamic positive feedback. Small variations inthe polymerization rate on one side of the bacteriumversus the other side can be amplified to cause large-scale symmetry breaking67. The quantitative theoreticalpredictions of this model could be confirmed or refutedusing the purified system.

What causes variations in speed? Within a populationof genetically identical bacteria moving in a singlehost-cell type, there are wide variations in averagespeed from one bacterium to another17,21,34. This widevariation is not due to sampling error, but rather to

the fact that some individual bacteria are intrinsicallyfaster than other individual bacteria (P.A.G and J.A.T.,unpublished observations). What is responsible forthese intrinsic differences? Changes in ActA surfacedensity do not affect speed27,34, so the source of varia-tion must lie elsewhere. Even within the trajectory ofan individual bacterium, there are significant varia-tions in speed over time, often over an order of mag-nitude within a period of a few minutes. Are varia-tions in the subcellular environment responsible forthis? If so, can this property be used to map out thepositions of biochemically distinct microenviron-ments within a living cell?

What causes curvature in trajectories? It is extremely rareto find a bacterium that moves in a straight line; mostbacterial tails are gently curved (FIG. 2, FIG. 3).What causesthese curves? Is there a correlation between variations incurvature and variations in speed? Some bacterial strainscarrying point mutations in ActA show curvature behav-iours that are very different from wild-type bacteria,including one mutant that makes tighter smooth curvesand one mutant that can ‘skid’, making occasional sharpturns at apparently random intervals (P. Lauer, S.Rafelski, D. Portnoy and J.A.T, unpublished observa-tions). How do these mutations affect interactions withthe host-cell proteins that govern actin self-organizationand movement?

These mechanistic questions, and many others, cannow be definitively addressed in the purified proteinsystem62, exploiting the ability of ActA to confer motilityon artificial particles whose geometry can be con-trolled27. So, with a set of proteins in hand, and a simplesystem in which to study these interactions, the time isripe for a powerful convergence of molecular, biochemi-cal and physical techniques on a single area: the organi-zation and control of the actin cytoskeleton.

This review has detailed an experimental success story,in which an unusual bacterial system has been successful-ly exploited as a robust, reproducible proxy for eukaryoticactin-based amoeboid motility. Using immunofluores-cence, biochemical and genetic techniques, proteins puta-tively involved in motility were identified and found to benecessary and sufficient for motility, as shown by a mini-mal reconstituted system. As a whole, this experimentaljourney can serve as a lesson on how to approach molec-ular mastery of a dynamic, self-organizing cellular sys-tem. Now that the molecular basis of this type of motilityis fairly well understood, attention can be focused on thebiophysical mechanisms68, bringing us one step closer to adetailed understanding of the beautiful and complexprocess of amoeboid motility.

BROWNIAN RATCHET MODEL

(ELASTIC)

A proposed model for actin-based motility in which actinfilaments are thought to flexaway from the bacterial surfaceto allow addition of monomerat the end of the filament.When the filament flexes back,it is one subunit longer andpushes the bacterium forwardthat distance.

BULK ELASTIC MODEL

A proposed model for actin-based motility, which treats theactin comet tail as a cohesiveelastic gel that respondselastically to deformation. Thisindicates that the energy fromactin polymerization may bestored as elastic energy in theactin gel to produce force thatpropels the bacterium forward.



R E V I E W S

edge and that filaments further back in thelamellipodium remain stationary as the rapidlymoving cell translocates over them.

8. Theriot, J. A. & Mitchison, T. J. Comparison of actin andcell surface dynamics in motile fibroblasts. J. Cell Biol. 119,367–377 (1992).

9. Zigmond, S. H. Recent quantitative studies of actinfilament turnover during cell locomotion. Cell Motil.Cytoskeleton 25, 309–316 (1993).

10. Shariff, A. & Luna, E. J. Diacylglycerol-stimulated formationof actin nucleation sites at plasma membranes. Science256, 245–247 (1992).

11. Oster, G. F. & Perelson, A. S. The physics of cell motility. J.Cell Sci. 8, S35–S54 (1987).

12. Tilney, L. G. & Portnoy, D. A. Actin filaments and thegrowth, movement, and spread of the intracellular bacterialparasite, Listeria monocytogenes. J. Cell Biol. 109,1597–1608 (1989).This is the initial publication showing that thebacterial pathogen, Listeria monocytogenes,associates with host cytoplasmic actin, suggestingthat these bacteria move within and between cells insome manner involving the actin cytoskeleton.

13. Mounier, J., Ryter, A., Coquis-Rondon, M. & Sansonetti, P. J.Intracellular and cell-to-cell spread of Listeriamonocytogenes involves interaction with F-actin in theenterocytelike cell line Caco- 2. Infect. Immunol. 58,1048–1058 (1990).

14. Makino, S., Sasakawa, C., Kamata, K., Kurata, T. &Yoshikawa, M. A genetic determinant required forcontinuous reinfection of adjacent cells on large plasmid inS. flexneri 2a. Cell 46, 551–555 (1986).

15. Bernardini, M. L., Mounier, J., d’Hauteville, H., Coquis-Rondon, M. & Sansonetti, P. J. Identification of icsA, aplasmid locus of Shigella flexneri that governs bacterialintra- and intercellular spread through interaction with F-actin. Proc. Natl Acad. Sci. USA 86, 3867–3871 (1989).

16. Dabiri, G. A., Sanger, J. M., Portnoy, D. A. & Southwick, F. S.Listeria monocytogenes moves rapidly through the host-cell cytoplasm by inducing directional actin assembly.Proc. Natl Acad. Sci. USA 87, 6068–6072 (1990).

17. Robbins, J. R. et al. Listeria monocytogenes exploitsnormal host cell processes to spread from cell to cell. J.Cell Biol. 146, 1333–1350 (1999).

18. Theriot, J. A. The cell biology of infection by intracellularbacterial pathogens. Annu. Rev. Cell. Dev. Biol. 11,213–239 (1995).

19. Sanger, J. M., Sanger, J. W. & Southwick, F. S. Host cellactin assembly is necessary and likely to provide thepropulsive force for intracellular movement of Listeriamonocytogenes. Infect. Immunol. 60, 3609–3619 (1992).Shows that actin monomers add to the tail only atthe bacterial surface whereas α–actinin andtropomyosin are found throughout the tail, andsuggests that actin polymerization may provide theforce for motility.

20. Tilney, L. G., DeRosier, D. J., Weber, A. & Tilney, M. S.How Listeria exploits host cell actin to form its owncytoskeleton. II. Nucleation, actin filament polarity, filamentassembly, and evidence for a pointed end capper. J. CellBiol. 118, 83–93 (1992).

21. Theriot, J. A., Mitchison, T. J., Tilney, L. G. & Portnoy, D. A.The rate of actin-based motility of intracellular Listeriamonocytogenes equals the rate of actin polymerization.Nature 357, 257–260 (1992).Shows that the rate of bacterial motility is the sameas the rate of actin polymerization and that actindepolymerization is independent of position in thecomet tail.

22. Theriot, J. A., Rosenblatt, J., Portnoy, D. A., Goldschmidt-Clermont, P. J. & Mitchison, T. J. Involvement of profilin inthe actin-based motility of L. monocytogenes in cells andin cell-free extracts. Cell 76, 505–517 (1994).

23. Kocks, C. et al. L. monocytogenes-induced actinassembly requires the actA gene product, a surfaceprotein. Cell 68, 521–531 (1992).Shows that the bacterial surface protein ActA isrequired for actin-based motility of Listeriamonocytogenes.

24. Domann, E. et al. A novel bacterial virulence gene inListeria monocytogenes required for host cellmicrofilament interaction with homology to the proline-richregion of vinculin. EMBO J. 11, 1981–1990 (1992).

25. Kocks, C. et al. The unrelated surface proteins ActA ofListeria monocytogenes and IcsA of Shigella flexneri aresufficient to confer actin-based motility on Listeria innocuaand Escherichia coli respectively. Mol. Microbiol. 18,413–423 (1995).

26. Smith, G. A., Portnoy, D. A. & Theriot, J. A. Asymmetricdistribution of the Listeria monocytogenes ActA protein is

required and sufficient to direct actin-based motility. Mol.Microbiol. 17, 945–951 (1995).

27. Cameron, L. A., Footer, M. J., van Oudenaarden, A. &Theriot, J. A. Motility of ActA protein-coated microspheresdriven by actin polymerization. Proc. Natl Acad. Sci. USA96, 4908–4913 (1999).

28. Chakraborty, T. et al. A focal adhesion factor directly linkingintracellularly motile Listeria monocytogenes and Listeriaivanovii to the actin-based cytoskeleton of mammaliancells. EMBO J. 14, 1314–1321 (1995).This was the first identification of a host-cellcytoskeletal protein (VASP) able to interact directlywith a bacterial virulence factor (ActA).

29. Reinhard, M. et al. The proline-rich focal adhesion andmicrofilament protein VASP is a ligand for profilins. EMBOJ. 14, 1583–1589 (1995).

30. Marchand, J. B. et al. Actin-based movement of Listeriamonocytogenes: actin assembly results from the localmaintenance of uncapped filament barbed ends at thebacterium surface. J. Cell Biol. 130, 331–343 (1995).

31. Carlsson, L., Nystrom, L. E., Sundkvist, I., Markey, F. &Lindberg, U. Actin polymerizability is influenced by profilin,a low molecular weight protein in non-muscle cells. J. Mol.Biol. 115, 465–483 (1977).

32. Goldschmidt-Clermont, P. J. et al. The control of actinnucleotide exchange by thymosin β 4 and profilin. Apotential regulatory mechanism for actin polymerization incells. Mol. Biol. Cell 3, 1015–1024 (1992).

33. Pantaloni, D. & Carlier, M. F. How profilin promotes actinfilament assembly in the presence of thymosin β 4. Cell 75,1007–1014 (1993).

34. Smith, G. A., Theriot, J. A. & Portnoy, D. A. The tandemrepeat domain in the Listeria monocytogenes ActA proteincontrols the rate of actin-based motility, the percentage ofmoving bacteria, and the localization of vasodilator-stimulated phosphoprotein and profilin. J. Cell Biol. 135,647–660 (1996).

35. Lasa, I., David, V., Gouin, E., Marchand, J. B. & Cossart, P.The amino-terminal part of ActA is critical for the actin-based motility of Listeria monocytogenes; the centralproline-rich region acts as a stimulator. Mol. Microbiol. 18,425–436 (1995).

36. Pistor, S., Chakraborty, T., Walter, U. & Wehland, J. Thebacterial actin nucleator protein ActA of Listeriamonocytogenes contains multiple binding sites for hostmicrofilament proteins. Curr. Biol. 5, 517–525 (1995).

37. Niebuhr, K. et al. A novel proline-rich motif present in ActAof Listeria monocytogenes and cytoskeletal proteins is theligand for the EVH1 domain, a protein module present inthe Ena/VASP family. EMBO J. 16, 5433–5444 (1997).

38. Gertler, F. B., Niebuhr, K., Reinhard, M., Wehland, J. &Soriano, P. Mena, a relative of VASP and DrosophilaEnabled, is implicated in the control of microfilamentdynamics. Cell 87, 227–239 (1996).

39. Laurent, V. et al. Role of proteins of the Ena/VASP family inactin-based motility of Listeria monocytogenes. J. CellBiol. 144, 1245–1258 (1999).

40. Geese, M. et al. Accumulation of profilin II at the surface ofListeria is concomitant with the onset of motility andcorrelates with bacterial speed. J. Cell Sci. 113,1415–1426 (2000).

41. Bear, J. E. et al. Negative regulation of fibroblast motility byEna/VASP proteins. Cell 101, 717–728 (2000).

42. Welch, M. D., Iwamatsu, A. & Mitchison, T. J. Actinpolymerization is induced by Arp2/3 protein complex atthe surface of Listeria monocytogenes. Nature 385,265–269 (1997).Using a visual assay to detect nucleation of an actincloud at the surface of the bacterial pathogen, L.monocytogenes, a seven-member protein complexresponsible for actin filament nucleation wasisolated.

43. Machesky, L. M., Atkinson, S. J., Ampe, C.,Vandekerckhove, J. & Pollard, T. D. Purification of a corticalcomplex containing two unconventional actins fromAcanthamoeba by affinity chromatography on profilin-agarose. J. Cell Biol. 127, 107–115 (1994).

44. Mullins, R. D., Heuser, J. A. & Pollard, T. D. The interactionof Arp2/3 complex with actin: nucleation, high affinitypointed end capping, and formation of branchingnetworks of filaments. Proc. Natl Acad. Sci. USA 95,6181–6186 (1998).Shows that Arp2/3 has a high affinity for the pointedend of actin filaments in addition to side binding, andleads to a model for actin filament networkbranching that is the current model for actinassembly at the leading edge.

45. Welch, M. D., Rosenblatt, J., Skoble, J., Portnoy, D. A. &Mitchison, T. J. Interaction of human Arp2/3 complex andthe Listeria monocytogenes ActA protein in actin filament

nucleation. Science 281, 105–108 (1998).46. Skoble, J., Portnoy, D. A. & Welch, M. D. Three regions

within ActA promote Arp2/3 complex-mediated actinnucleation and Listeria monocytogenes motility. J. CellBiol. 150, 527–537 (2000).

47. Svitkina, T. M. & Borisy, G. G. Arp2/3 complex and actindepolymerizing factor/cofilin in dendritic organization andtreadmilling of actin filament array in lamellipodia. J. CellBiol. 145, 1009–1026 (1999).This is a beautiful study using correlative light andelectron microscopy to compare protein localizationand structure of the actin network in the lamellipodiaof fast moving keratocytes and slower movingfibroblasts.

48. Mullins, R. D., Stafford, W. F. & Pollard, T. D. Structure,subunit topology, and actin-binding activity of the Arp2/3complex from Acanthamoeba. J. Cell Biol. 136, 331–343(1997).

49. Welch, M. D., DePace, A. H., Verma, S., Iwamatsu, A. &Mitchison, T. J. The human Arp2/3 complex is composedof evolutionarily conserved subunits and is localized tocellular regions of dynamic actin filament assembly. J. CellBiol. 138, 375–384 (1997).

50. Suzuki, T., Miki, H., Takenawa, T. & Sasakawa, C. NeuralWiskott-Aldrich syndrome protein is implicated in theactin-based motility of Shigella flexneri. EMBO J. 17,2767–2776 (1998).Shows that N–WASP, a known activator of Arp2/3,mediates actin tail formation at the pole of S. flexnerithrough the bacterial surface protein VirG (IcsA),linking bacteria to the other host cell factors thathelp assemble the comet tail.

51. David, V. et al. Identification of cofilin, coronin, Rac andcapZ in actin tails using a Listeria affinity approach. J. CellSci. 111, 2877–2884 (1998).

52. Laine, R. O. et al. Gelsolin, a protein that caps the barbedends and severs actin filaments, enhances the actin-basedmotility of Listeria monocytogenes in host cells. Infect.Immunol. 66, 3775–3782 (1998).

53. McGough, A., Pope, B., Chiu, W. & Weeds, A. Cofilinchanges the twist of F-actin: implications for actin filamentdynamics and cellular function. J. Cell Biol. 138, 771–781(1997).

54. Maciver, S. K., Zot, H. G. & Pollard, T. D. Characterizationof actin filament severing by actophorin fromAcanthamoeba castellanii. J. Cell Biol. 115, 1611–1620(1991).

55. Carlier, M. F. et al. Actin depolymerizing factor (ADF/cofilin)enhances the rate of filament turnover: implication in actin-based motility. J. Cell Biol. 136, 1307–1322 (1997).Shows that the actin depolymerization rate in vitro isdirectly affected by ADF/cofilin in a concentration-dependent fashion and that addition of ADF to cellextracts increases turnover rate in L.monocytogenes actin tails.

56. Rosenblatt, J., Agnew, B. J., Abe, H., Bamburg, J. R. &Mitchison, T. J. Xenopus actin depolymerizing factor/cofilin(XAC) is responsible for the turnover of actin filaments inListeria monocytogenes tails. J. Cell Biol. 136, 1323–1332(1997).

57. Lappalainen, P. & Drubin, D. G. Cofilin promotes rapidactin filament turnover in vivo. Nature 388, 78–82 (1997).Using a yeast genetic study, provides in vivoevidence that cofilin, an actin depolymerizing factor,does in fact enhance filament depolymerization incells.

58. Bamburg, J. R. & Bray, D. Distribution and cellularlocalization of actin depolymerizing factor. J. Cell Biol. 105,2817–2825 (1987).

59. Prevost, M. C. et al. Unipolar reorganization of F-actin layerat bacterial division and bundling of actin filaments byplastin correlate with movement of Shigella flexneri withinHeLa cells. Infect. Immunol. 60, 4088–4099 (1992).

60. Dold, F. G., Sanger, J. M. & Sanger, J. W. Intact α-actininmolecules are needed for both the assembly of actin intothe tails and the locomotion of Listeria monocytogenesinside infected cells. Cell Motil. Cytoskeleton 28, 97–107(1994).

61. Cunningham, C. C. et al. Actin-binding protein requirementfor cortical stability and efficient locomotion. Science 255,325–327 (1992).

62. Loisel, T. P., Boujemaa, R., Pantaloni, D. & Carlier, M. F.Reconstitution of actin-based motility of Listeria andShigella using pure proteins. Nature 401, 613–616 (1999).Identifies the minimal set of proteins required for thereconstitution of actin-based motility with purifiedcomponents for both L. monocytogenes and S.flexneri motility, and demonstrates that no myosinmotor is required.

63. Theriot, J. A. The polymerization motor. Traffic 1, 19–28



R E V I E W S

(2000).64. Hill, T. L. & Kirschner, M. W. Bioenergetics and kinetics of

microtubule and actin filament assembly-disassembly. Int.Rev. Cytol. 78, 1–125 (1982).

65. Mogilner, A. & Oster, G. Cell motility driven by actinpolymerization. Biophys. J. 71, 3030–3045 (1996).

66. Gerbal, F. et al. Measurement of the elasticity of the actintail of Listeria monocytogenes. Eur. Biophys. J. 29,134–140 (2000).

67. van Oudenaarden, A. & Theriot, J. A. Cooperativesymmetry-breaking by actin polymerization in a model forcell motility. Nature Cell Biol. 1, 493–499 (1999).

68. Kuo, S. C. & McGrath, J. L. Steps and fluctuations ofListeria monocytogenes during actin–based motility.(submitted).

69. Symons, M. et al. Wiskott–Aldrich syndrome protein, anovel effector for the GTPase CDC42Hs, is implicated inactin polymerization. Cell 84, 723–734 (1996).

70. Hall, A. Rho GTPases and the actin cytoskeleton. Science279, 509–514 (1998).

71. Miki, H., Miura, K. & Takenawa, T. N-WASP, a novel actin-depolymerizing protein, regulates the cortical cytoskeletalrearrangement in a PIP2-dependent manner downstreamof tyrosine kinases. EMBO J. 15, 5326–5335 (1996).

72. Miki, H., Sasaki, T., Takai, Y. & Takenawa, T. Induction offilopodium formation by a WASP-related actin-depolymerizing protein N-WASP. Nature 391, 93–96(1998).

73. Bear, J. E., Rawls, J. F. & Saxe, C. L., SCAR, a WASP-related protein, isolated as a suppressor of receptordefects in late Dictyostelium development. J. Cell Biol.142, 1325–1335 (1998).

74. Machesky, L. M. & Insall, R. H. Scar1 and the relatedWiskott-Aldrich syndrome protein, WASP, regulate theactin cytoskeleton through the Arp2/3 complex. Curr. Biol.8, 1347–1356 (1998).

75. Machesky, L. M. et al. Scar, a WASp-related protein,activates nucleation of actin filaments by the Arp2/3complex. Proc. Natl Acad. Sci. USA 96, 3739–3744(1999).

76. Yarar, D., To, W., Abo, A. & Welch, M. D. The Wiskott-Aldrich syndrome protein directs actin-based motility bystimulating actin nucleation with the Arp2/3 complex. Curr.Biol. 9, 555–558 (1999).

77. Lin, C. H., Espreafico, E. M., Mooseker, M. S. & Forscher,

P. Myosin drives retrograde F-actin flow in neuronal growthcones. Neuron 16, 769–782 (1996).

78. Pollard, T. D., Blanchoin, L. & Mullins, R. D. Molecularmechanisms controlling actin filament dynamics innonmuscle cells. Annu. Rev. Biophys. Biomol. Struct. 29,545–576 (2000).

79. Pantaloni, D., Boujemaa, R., Didry, D., Gounon, P. &Carlier, M. F. The Arp2/3 complex branches filamentbarbed ends: functional antagonism with capping proteins.Nature Cell Biol. 2, 385–391 (2000).

80. Goldberg, M. B. & Theriot, J. A. Shigella flexneri surfaceprotein IcsA is sufficient to direct actin- based motility.Proc. Natl Acad. Sci. USA 92, 6572–6576 (1995).

81. Mimuro, H. et al. Profilin is required for sustaining efficientintra- and intercellular spreading of Shigella flexneri. J. Biol.Chem. 275, 28893–28901 (2000).

82. Heinzen, R. A., Hayes, S. F., Peacock, M. G. & Hackstadt,T. Directional actin polymerization associated with spottedfever group Rickettsia infection of Vero cells. Infect.Immunol. 61, 1926–1935 (1993).

83. Teysseire, N., Chiche-Portiche, C. & Raoult, D. Intracellularmovements of Rickettsia conorii and R. typhi based onactin polymerization. Res. Microbiol. 143, 821–829 (1992).

84. Heinzen, R. A., Grieshaber, S. S., Van Kirk, L. S. & Devin,C. J. Dynamics of actin-based movement by Rickettsiarickettsii in vero cells. Infect. Immunol. 67, 4201–4207(1999).

85. Gouin, E. et al. A comparative study of the actin-basedmotilities of the pathogenic bacteria Listeriamonocytogenes, Shigella flexneri and Rickettsia conorii. J.Cell Sci. 112, 1697–1708 (1999).

86. Kenny, B. et al. Enteropathogenic E. coli (EPEC) transfersits receptor for intimate adherence into mammalian cells.Cell 91, 511–520 (1997).

87. Kalman, D. et al. Enteropathogenic E. coli acts throughWASP and Arp2/3 complex to form actin pedestals.Nature Cell Biol. 1, 389–391 (1999).

88. Cudmore, S., Cossart, P., Griffiths, G. & Way, M. Actin-based motility of vaccinia virus. Nature 378, 636–638(1995).Demonstrates that a virus moves by actin-basedmotility in host cells, expanding the realm of thistype of motility to another system which has adifferent shape and possibly a new way to form apolarized actin tail.

89. Frischknecht, F. et al. Actin-based motility of vaccinia virusmimics receptor tyrosine kinase signalling. Nature 401,926–929 (1999).

90. Moreau, V. et al. A complex of N-WASP and WIPintegrates signalling cascades that lead to actinpolymerization. Nature Cell Biol. 2, 441–448 (2000).

91. Ma, L., Cantley, L. C., Janmey, P. A. & Kirschner, M. W.Corequirement of specific phosphoinositides and smallGTP-binding protein Cdc42 in inducing actin assembly inXenopus egg extracts. J. Cell Biol. 140, 1125–1136(1998).

92. Rohatgi, R. et al. The interaction between N-WASP andthe Arp2/3 complex links Cdc42-dependent signals toactin assembly. Cell 97, 221–231 (1999).

93. Taunton, J. et al. Actin-dependent propulsion ofendosomes and lysosomes by recruitment of N-WASP. J.Cell Biol. 148, 519–530 (2000).

94. Merrifield, C. J. et al. Endocytic vesicles move at the tips ofactin tails in cultured mast cells. Nature Cell Biol. 1, 72–74(1999).

95. Rozelle, A. L. et al. Phosphatidylinositol 4,5-bisphosphateinduces actin-based movement of raft-enriched vesiclesthrough WASP-Arp2/3. Curr. Biol. 10, 311–320 (2000).

96. Schafer, D. A., Mooseker, M. S. & Cooper, J. A.Localization of capping protein in chicken epithelial cells byimmunofluorescence and biochemical fractionation. J. CellBiol. 118, 335–346 (1992).

97. Cooper, J. A., Loftus, D. J., Frieden, C., Bryan, J. & Elson,E. L. Localization and mobility of gelsolin in cells. J. CellBiol. 106, 1229–1240 (1988).

98. Rottner, K., Behrendt, B., Small, J. V. & Wehland, J. VASPdynamics during lamellipodia protrusion. Nature Cell Biol.1, 321–322 (1999).

99. Lazarides, E. & Burridge, K. α–actinin: immunofluorescentlocalization of a muscle structural protein in nonmusclecells. Cell 6, 289–298 (1975).

AcknowledgementsWe thank Rachael Ream for digital time-lapse used to make Figure1 and Movie 1. We would like to apologize to those researcherswhose work could not be cited due to space limitations. J.A.T. issupported by grants from the National Institutes of Health and aFellowship in Science and Engineering from the David and LucilePackard Foundation.


120 | OCTOBER 2000 | VOLUME 1 www.nature.com/reviews/molcellbio

R E V I E W S

For proliferative tissues to maintain a constant size andfunction properly, the older cells must die to make wayfor new cells. Such ‘programmed’ death involves astereotyped sequence of biochemical and morphologicalchanges that allows the cell to die without adverselyaffecting its neighbours — a process called apoptosis(FIG. 1). In many diseases, aberrant regulation of apopto-sis is the central abnormality. For example, resistance ofcells to apoptosis is thought to be responsible for manytypes of cancer, and identification of the molecular alter-ations responsible for such cell immortalization is animportant area in cancer research1. Although details ofthe definition of apoptosis vary among investigators,there is general agreement that apoptosis is a cell deathprocess involving caspase activation and a lack of cellswelling with maintenance of organelle (mitochondriaand endoplasmic reticulum) integrity. Apoptosis occurswithin a tissue in a ‘spotty’ pattern such that healthy anddying cells are intermingled. This contrasts with anotherform of cell death, called necrosis, in which cellularorganelles swell and the plasma membrane lyses,resulting in massive death of groups of cells throughouta tissue.As described below, the criteria for apoptosis arefulfilled in many neurological disorders in whichneuronal death is a central feature.

In contrast to the rapid turnover of cells in prolifera-tive tissues, neurons commonly survive for the entire

lifetime of the organism — this enduring nature ofneurons is necessary for maintaining the function ofthose cells within neuronal circuits. For example, motorneurons must maintain connections to skeletal muscles,and long-term memories probably require the contin-ued survival of the neurons in the regions of the brainin which those memories are encoded. During develop-ment of the central and peripheral nervous systems,many neurons undergo apoptosis during a time win-dow that coincides with the process of SYNAPTOGENESIS2.Signals that determine whether or not developing neu-rons live or die may include competition for a limitedsupply of target-derived NEUROTROPHIC FACTORS and acti-vation of receptors for the excitatory neurotransmitterglutamate3. Initial overproduction of neurons, followedby death of some, is probably an adaptive process thatprovides enough neurons to form nerve cell circuitsthat are precisely matched to their functional specifica-tions4. Accordingly, the decision as to which neurons dieis made by cellular signal transduction pathways thatare ‘tuned’ to the functionality of neuronal circuits.

Unfortunately, many people experience excessivedeath of one or more populations of neurons as theresult of disease or injury. For example, death of hip-pocampal and cortical neurons is responsible for thesymptoms of Alzheimer’s disease; death of midbrainneurons that use the neurotransmitter dopamine under-

APOPTOSIS INNEURODEGENERATIVE DISORDERSMark P. Mattson

Neuronal death underlies the symptoms of many human neurological disorders, includingAlzheimer’s, Parkinson’s and Huntington’s diseases, stroke, and amyotrophic lateral sclerosis.The identification of specific genetic and environmental factors responsible for these diseaseshas bolstered evidence for a shared pathway of neuronal death — apoptosis — involvingoxidative stress, perturbed calcium homeostasis, mitochondrial dysfunction and activation ofcysteine proteases called caspases. These death cascades are counteracted by survivalsignals, which suppress oxyradicals and stabilize calcium homeostasis and mitochondrialfunction. With the identification of mechanisms that either promote or prevent neuronalapoptosis come new approaches for preventing and treating neurodegenerative disorders.

Laboratory ofNeurosciences, NationalInstitute on Aging,Gerontology ResearchCenter, 5,600 Nathan ShockDrive, Baltimore, Maryland21224, USA.e-mail:[email protected]

SYNAPTOGENESIS

The process of formation ofsynapses, the sites whereneurons communicate throughrelease of neurotransmittersfrom the presynaptic terminaland activation of receptors onthe postsynaptic neuron.

NEUROTROPHIC FACTORS

Proteins produced by neuronsand glial cells that promoteneuron survival and growth.



R E V I E W S

logical and pathological settings (TABLE 1), but many ofthe subsequent biochemical events that execute the celldeath process are highly conserved. One classification ofthis death programme is based on compelling evidencethat mitochondrial changes are pivotal in the cell deathdecision in many cases13. Mitochondria in cells under-going apoptosis show increased oxyradical production,opening of pores in their membranes and release ofcytochrome c (FIG. 1). These changes are central to thecell death process because agents such as manganesesuperoxide dismutase and cyclosporin A, which actdirectly on mitochondria to suppress oxidative stressand membrane pore formation, also prevent neuronaldeath in experimental models14.

The biochemical alterations that occur during theearly stages of apoptosis may induce mitochondrial dys-function either directly or indirectly (TABLE 2). The B-celllymphoma-2 (Bcl-2) family of proteins includes bothpro- and anti-apoptotic members15. The best-studiedanti-apoptotic members in neurons are Bcl-2 and Bcl-x

L; pro-apoptotic members include Bcl-2-associated X-

protein (Bax) and Bcl-associated death promoter (Bad).Overexpression of Bcl-2 in cell cultures and in trans-genic mice increases resistance of neurons to deathinduced by excitotoxic, metabolic and oxidative insultsrelevant to Alzheimer’s disease, stroke and other disor-ders16,17. Conversely, neurons lacking Bax are protectedagainst apoptosis18. The mechanism by which Bcl-2 pro-teins control cell death is not clear, but it may involveinteractions among family members and association ofthe proteins with mitochondria, resulting in altered ionmovements across mitochondrial membranes19.

Further mechanisms that can regulate the early stagesof apoptosis involve caspases, the prostate apoptosis

lies Parkinson’s disease; Huntington’s disease involves thedeath of neurons in the striatum, which control bodymovements; and death of lower motor neurons mani-fests as amyotrophic lateral sclerosis (FIG. 2). The numberof people with such neurodegenerative disorders israpidly increasing as the average lifespan gets longer.

Neuronal death cascadesMany signals can initiate or ‘trigger’ apoptosis in neu-rons (FIG. 3). The best-studied signal is lack of neu-rotrophic factor support, which may trigger apoptosisduring development of the nervous system and possiblyin neurodegenerative disorders2–4. Most neurons in themammalian central nervous system possess receptorsfor another trigger of apoptosis — the excitatory neuro-transmitter glutamate. Overactivation of glutamatereceptors can induce apoptosis by a mechanism involv-ing calcium influx5,6, and such ‘excitotoxicity’ may occurin acute neurodegenerative conditions such as stroke,trauma and severe epileptic seizures, as well as inAlzheimer’s disease and motor system disorders7,8. Athird trigger of neuronal death is increased oxidativestress, in which free radicals (such as the superoxideanion radical and the hydroxyl radical) damage cellularlipids, proteins and nucleic acids by attacking chemicalbonds in those molecules9,10. METABOLIC STRESS, as occursafter a stroke or during ageing, may also initiate neu-ronal apoptosis. Finally, environmental toxins caninduce neuronal apoptosis, and several such toxins caninduce patterns of brain damage and behaviouralphenotypes remarkably similar to Parkinson’s andHuntington’s diseases11,12.

The genetic and environmental factors that triggerneuronal apoptosis may be different in various physio-

Figure 1 | Morphological and biochemical features of apoptosis. During the initiation phase of apoptosis, the death signalactivates an intracellular cascade of events that may involve increases in levels of oxyradicals and Ca2+, production of Par-4 andtranslocation of pro-apoptotic Bcl-2 family members (Bax and Bad) to the mitochondrial membrane. Certain caspases (caspase-8, for example) can also act early in the cell death process before, or independently of, mitochondrial changes. The effector phaseof apoptosis involves increased mitochondrial Ca2+ and oxyradical levels, the formation of permeability transition pores (PTP) inthe mitochondrial membrane, and release of cytochrome c into the cytosol. Cytochrome c forms a complex with apoptoticprotease-activating factor 1 (Apaf-1) and caspase-9. Activated caspase-9, in turn, activates caspase-3, which begins thedegradation phase of apoptosis in which various caspase and other enzyme substrates are cleaved, resulting in characteristicchanges in the plasma membrane (blebbing and exposure of phosphatidylserine on the cell surface, which is a signal thatstimulates cell phagocytosis by macrophages/microglia). Finally, the nuclear chromatin becomes condensed and fragmented.

Death signals: • Oxidative stress• Glutamate• Decreased growth-factors• Genetic mutation

Oxygen radicals/Ca2+

Par-4

PTP

Phosphatidylserine

Deathsubstrates

Bax, Bad

Mitochondrion

Initiation phase Effector phase Degradation phase

Cytochrome cOxyradicals/Ca2+

Cytochrome cApaf-1

Caspase-9

Blebs

Caspase-3

ProteolysisEndonucleases

Nucleus

METABOLIC STRESS

Conditions in which levels ofglucose, oxygen and othermolecules required for ATP(energy) production aredecreased.



R E V I E W S

vous system during evolution to develop mechanismsthat guard against neuronal death. The marked symp-toms of neurodegenerative disorders emphasize theimportance of mechanisms that promote neuron sur-vival and plasticity (FIG. 3, BOX 1). There are severalprominent anti-apoptotic signalling pathways3.Neurotrophic factors have been identified that can pro-tect neurons against apoptosis by activating receptorslinked through kinase cascades to production of cell-survival-promoting proteins. For example, brain-derivedneurotrophic factor (BDNF), nerve growth factor(NGF) and basic fibroblast growth factor (bFGF) canprevent death of cultured neurons, in part by stimulat-ing production of antioxidant enzymes, Bcl-2 familymembers and proteins involved in regulation of calci-um homeostasis3,28. Cytokines such as tumour necrosisfactor-α (TNF-α), ciliary neurotrophic factor (CNTF)and leukaemia inhibitory factor (LIF) can prevent neu-ronal death in experimental models of natural neuronaldeath and neurodegenerative disorders29–31.

Several neurotrophic factors and cytokines use a

response-4 (Par-4) protein and telomerase. Caspases areevolutionarily conserved cysteine proteases central toapoptosis of many cell types20. Some caspases are activat-ed during the early phase of apoptosis. A prominentexample is the activation of caspase-8 in neurons inresponse to ligation of ‘death receptors’ such as Fas andthe p75 neurotrophin receptor21. These upstream caspas-es can then activate ‘effector’ caspases such as caspase-3either directly or indirectly, and may thereby elicit apop-tosis independently of mitochondrial alterations. Effectorcaspases can also be activated in response to mitochondr-ial changes and cytochrome c release; these caspases canthen activate a DNase that cleaves DNA into oligonucleo-some-sized fragments22. Caspases can also cleave varioussubstrate proteins that may coordinate the cell deathprocess, including enzymes such as poly-ADP-ribosepolymerase and ataxia telangiectasia mutated (ATM)kinase; ion channels including subunits of the α-amino-3-hydroxy-5-methyl-4-isoxazole propionate (AMPA)subtype of neuronal glutamate receptor; and cytoskeletalproteins such as actin and spectrin (for a review, see REF.

20). Par-4 was identified as being upregulated in prostatetumour cells undergoing apoptosis, but is now known tobe essential in developmental and pathological neuronaldeath23,24. Levels of Par-4 increase rapidly in response tovarious apoptotic stimuli through enhanced translationof Par-4 messenger RNA. A LEUCINE ZIPPER domain in thecarboxyl terminus of Par-4 is essential for its pro-apop-totic function, and interactions of Par-4 with other pro-teins, including protein kinase Cζ and Bcl-2, through thiszipper may be central to the mechanism by which Par-4induces mitochondrial dysfunction.

Telomerase adds a six-base DNA sequence(TTAGGG) to the ends of chromosomes, preventingtheir shortening and protecting them during chromo-some segregation in mitotic cells25. Telomerase consistsof a catalytic reverse-transcriptase subunit (TERT), anRNA template and several associated regulatory pro-teins. Telomerase activity is increased during cellimmortalization and transformation (and is thereforestrongly implicated in the pathogenesis of many can-cers), and in many tissues including the brain duringdevelopment, but is downregulated in all somatic tis-sues during late embryonic and early postnatal develop-ment. Telomerase activity and expression of TERT areassociated with increased resistance of neurons to apop-tosis in experimental models of developmental neu-ronal death and neurodegenerative disorders26,27. Theanti-apoptotic action of TERT in neurons is exerted atan early step in the cell death pathway before mitochon-drial alterations and caspase activation. These newfindings indicate that telomerase may be important inneural development and injury responses.

Cell survival mechanismsMuch of the structural and functional complexity of thenervous system arises because neurons do not divide.The persistence of neurons throughout life allows thenervous system to maintain continuous function overlong distances and to encode enduring memories.Therefore considerable pressure was placed on the ner-

Figure 2 | Brain regions in which neurodegenerativeconditions are typified by selective apoptosis ofneurons. a | In Alzheimer’s disease, neurons in thehippocampus and certain regions of the cerebral cortexdegenerate. b | In Huntington’s disease, neurons in thestriatum die. In Parkinson’s disease, dopamine neurons in thesubstantia nigra (not shown) undergo apoptosis, and instroke, the neurons that die are those supplied by anoccluded or ruptured blood vessel.

Cerebral cortex

Hippocampus

Striatum

a

b

LEUCINE ZIPPER

A leucine-rich domain within aprotein that binds to otherproteins with a similar domain.



R E V I E W S

In addition to extracellular signal-mediated neuropro-tection pathways, several intracellular signalling pathwayshave been identified that can protect neurons againstapoptosis. For example, stress can induce the expressionof neurotrophic factors and heat-shock proteins34,35. Theneurotrophic factors, in turn, act in an autocrine orparacrine manner to activate cell surface receptor-medi-ated kinase signalling pathways that ultimately induceexpression of genes encoding survival-promoting pro-teins such as antioxidant enzymes. Heat-shock proteinsact as ‘chaperones’ for many proteins, thereby maintain-ing protein stability. They may also interact directly withcaspases, inhibiting their activation.

Calcium is arguably the most versatile and importantintracellular messenger in neurons36. Interestingly,although calcium may often promote neuronal death, itcan also activate pathways that promote survival. Forexample, calcium can promote survival through a path-way involving activation of protein kinase B (PKB/Akt)by calcium/calmodulin-dependent protein kinase37.Calcium is a prominent regulator of cellular responses tostress, activating transcription through the cyclic-AMPresponse element-binding protein (CREB), which canpromote neuron survival in experimental models ofdevelopmental cell death38. Calcium can also activate arapid neuroprotective signalling pathway in which thecalcium-activated actin-severing protein gelsolin inducesactin depolymerization, resulting in suppression of calci-um influx through membrane NMDA (N-methyl-D-aspartate) receptors and voltage-dependent calciumchannels39. This may occur through intermediary actin-binding proteins that interact with NMDA receptor andcalcium channel proteins. Finally, signals such as calciumand secreted amyloid precursor protein-α (sAPP-α),which increase cyclic GMP production, can induce acti-vation of potassium channels and the transcription fac-tor NF-κB, and thereby increase resistance of neurons toexcitotoxic apoptosis40.

Apoptosis and neurodegenerative disordersFor each of the disorders described below, analyses ofpost-mortem tissue from patients and studies ofexperimental animal and cell-culture models haveimplicated neuronal apoptosis. Studies of the patho-genic mechanisms of genetic mutations that causeearly-onset autosomal dominant forms of Alzheimer’sand Huntington’s diseases and amyotrophic lateralsclerosis have been particularly valuable in implicatingapoptosis in age-related neurological disorders41–43.Nevertheless, it is very difficult to demonstrate apopto-sis convincingly in the brains of patients. This isbecause apoptosis usually occurs quite rapidly overseveral hours to a day, making it hard to ‘catch’ manycells showing classic features of apoptosis at any onetime. In addition, experiments cannot be done inhumans to establish whether blocking a step in theapoptotic biochemical cascade can prevent neuronaldeath. For these reasons, much of the evidence sup-porting an apoptotic mode of neuronal death comesfrom studies of animal and cell-culture models of neu-rodegenerative disorders.

Figure 3 | Roles for altered synaptic signalling in neurodegenerative disorders. Age-and disease-related stressors promote excessive activation of apoptotic (death) biochemicalcascades in synaptic terminals and neurites. For example, overactivation of glutamatereceptors under conditions of reduced energy availability or increased oxidative stress (fromreactive oxygen species, ROS) results in Ca2+ influx into postsynaptic regions of dendrites. Ca2+

entering the cytoplasm through plasma membrane channels and endoplasmic reticulum (ER)channels induces apoptotic cascades (lower left) that involve Par-4, pro-apoptotic Bcl-2 familymembers (Bax and Bad), and/or p53. These factors act on mitochondria to induce Ca2+ influx,oxidative stress, opening of permeability transition pores (PTP) and release of cytochrome c,which forms a complex with apoptotic protease-activating factor 1 (Apaf-1). This results incaspase activation and execution of the cell death process. Anti-apoptotic (life) signallingpathways are also concentrated in synaptic compartments (upper right). For example,activation of receptors (R) for neurotrophic factors (NTF) in axon terminals stimulates kinasecascades and transcription factors and increased production of survival-promoting proteinssuch as Bcl-2, Bcl-xL and manganese superoxide dismutase (Mn-SOD) (which act at the levelof mitochondria) and inhibitor of apoptosis proteins (IAPs) (which inhibit caspases).

BaxBad

p53Bcl-2

Par-4

Apaf-1Caspases

Mitochondrion

Death cascade

ER

ROSCa2+

Ca2+

PTP

Cytochrome c

Mitochondrion

Caspases

IAPs

ER

Life cascade

Ca2+

Ca2+

Ca2+

PTP

Bcl-2Bcl-xL

Mn-SOD

Dendrite

Axon

Glutamate

Kinases andtranscription factors

NTF

Death cascade

Life cascade

R

Ca2+

survival pathway involving the transcription factor NF-κB (REF. 31). Activation of NF-κB can protect culturedneurons against death induced by diverse stimuli,including trophic-factor withdrawal and exposure toexcitotoxic, oxidative and metabolic insults. Studies inmice that lack the p50 subunit of NF-κB show that NF-κB is also anti-apoptotic in the brain in vivo32. Gene tar-gets that mediate the survival-promoting action of NF-κB may include manganese superoxide dismutase, Bcl-2and inhibitor of apoptosis proteins (IAPs). However,NF-κB activation in MICROGLIA can promote neuronalapoptosis by inducing production of oxyradicals andEXCITOTOXINS33, so NF-κB may either prevent or promoteneuronal death depending on specific conditions.

MICROGLIA

Phagocytic immune cells in thebrain that engulf and removecells that have undergoneapoptosis.

EXCITOTOXINS

Compounds such as glutamate,kainic acid and N-methyl-D-aspartate that can kill neuronsby activating excitatory amino-acid (glutamate) receptors.



R E V I E W S

tion of amyloid plaques formed by aggregates of amy-loid-β peptide — a 40–42 amino-acid fragment gen-erated by proteolytic processing of the amyloid pre-cursor protein (APP)44,45. Increased DNA damage andcaspase activity, and alterations in expression of apop-tosis-related genes such as Bcl-2 family members, Par-4 and DNA damage response genes have been foundin neurons associated with amyloid deposits in thebrains of Alzheimer’s patients25,46,47. Expression-profileanalysis of thousands of genes in brain tissue samplesfrom Alzheimer’s patients and age-matched controlpatients revealed a marked decrease in expression ofan anti-apoptotic gene called NCKAP1 (for NCK-associated protein 1)48.

Exposure of cultured neurons to amyloid-β caninduce apoptosis directly49, and can greatly increasetheir vulnerability to death induced by conditions suchas increased oxidative stress and reduced energy avail-ability that are known to occur in the brain during age-ing10. The mechanism by which amyloid-β sensitizesneurons to death involves membrane lipid peroxida-tion. This impairs the function of membrane ion-motive ATPases and glucose and glutamate trans-porters, resulting in membrane depolarization, ATPdepletion, excessive calcium influx and mitochondrialdysfunction. Accordingly, antioxidants that suppresslipid peroxidation and drugs that stabilize cellular calci-umhomeostasis can protect neurons against amyloid-β-induced apoptosis10. Moreover, neurotrophic factorsand cytokines known to prevent neuronal apoptosis canprotect neurons against amyloid-β-induced death3.Further evidence for the involvement of apoptotic cas-cades in Alzheimer’s disease comes from studies show-ing that APP is a substrate for caspase-3 (REF. 50). Inaddition to promoting apoptosis by generating amy-loid-β, β- and/or γ-secretase cleavage of APP produces amembrane-associated carboxy-terminal fragment ofAPP that can induce apoptosis, possibly by a pathwayinvolving an APP-binding protein called APP-BP1,which may drive neurons into a mitotic cycle that endsin apoptosis51. Caspase-mediated cleavage of APP canrelease a carboxy-terminal peptide called C31 that is apotent inducer of apoptosis52.

Mutations in three genes, each inherited in an auto-somal dominant manner, can cause early-onset inherit-ed forms of Alzheimer’s disease — one gene encodesAPP, and the other two genes encode presenilins 1 and2. APP mutations seem to cause Alzheimer’s disease byaltering proteolytic processing of APP such that levels ofamyloid-β are increased and levels of sAPP-α are

In most of the neurodegenerative disorders consid-ered here, the disease process lasts for years or evendecades. It is therefore likely that, although individualneurons may be dysfunctional for extended periods inthese disorders, they may die rapidly once the apoptoticcascade is fully activated. In this view, the progressivedeficits that occur in chronic neurodegenerative disor-ders are the result of progressive attrition of individualneurons. Although the following section presents someof the salient evidence supporting the involvement ofapoptosis in a select group of neurodegenerative disor-ders, it should be noted that this is a controversial areaof research with many gaps to be filled.

Alzheimer’s disease. Alzheimer’s disease is characterizedby progressive impairment of cognition and emotion-al disturbances that are strongly correlated withsynaptic degeneration and death of neurons in LIMBIC

STRUCTURES, such as the hippocampus and the amyg-dala, and associated regions of the cerebral cortex (FIG. 2).Degenerating neurons show aggregates of hyperphos-phorylated tau protein, and evidence of excessive calci-um-mediated proteolysis and oxidative stress44 (FIG. 4). Adefining feature of Alzheimer’s disease is accumula-

Table 1 | Factors that may modulate apoptosis in neurodegenerative disorders

Disorder Genetic factors Environmental factors

Alzheimer’s47–51 APP, presenilin mutations, ApoE Head trauma, low education, calorie intake

Parkinson’s α-synuclein, parkin mutations Head trauma, toxins, calorie intake

Huntington’s61,62,66 Poly-CAG expansions in huntingtin

ALS77 Cu/Zn-SOD mutations Toxins, autoimmune response

Stroke83,84 Cadasil mutations Smoking, dietary calories and fat

Box 1 | Synaptic signalling in neuronal death and survival

Components of signalling pathways that initiate or prevent apoptosis are highlyconcentrated in synaptic terminals — the sites of intercellular communication betweenneurons. For example, receptors for glutamate are located in postsynaptic regions ofdendrites and receptors for neurotrophic factors are in both pre- and postsynapticterminals. Much of the biochemical machinery involved in apoptosis can be activatedin synaptic terminals, where it can alter synaptic function and promote localizeddegeneration of synapses and NEURITES105,106 (FIG. 3). For example, Par-4 production,mitochondrial alterations, caspase activation and release into the cytosol of factors thatmay cause nuclear apoptosis can be induced in SYNAPTOSOME preparations and neuritesof cultured brain neurons by insults that induce apoptosis in intact neurons105,106.Caspase-mediated cleavage of synaptic proteins may control the process of neuronalapoptosis. For example, AMPA receptor subunits are selectively degraded inhippocampal neurons after exposure to an apoptotic dose of glutamate, resulting indecreased calcium influx, thereby preventing excitotoxic necrosis6. The lattermechanism might also allow neurons to ‘withdraw’ from participation in neuronalcircuits, permitting them to recover from potentially lethal conditions. Apoptoticpathways may also function in synaptic plasticity, particularly under conditions ofstress and injury. Studies showing that TNF-α and NF-κB activation modify long-termdepression and potentiation of synaptic transmission in the hippocampus107 providefurther evidence that anti-apoptotic signalling can modulate synaptic plasticity.Finally, changes in mitochondrial membrane permeability in synaptic terminals havebeen associated with impaired synaptic plasticity in the hippocampus93, suggesting arole for apoptotic mitochondrial alterations in synaptic function.

LIMBIC STRUCTURES

Brain structures such as thehippocampus, amygdala andseptum that function inlearning and memory, and inemotions.

NEURITES

Generic name for processes(axons and dendrites)elaborated from neuronal cellbodies.

SYNAPTOSOME

A structure consisting of pre-and postsynaptic terminalsprepared from homogenizedbrain tissue with cellularfractionation techniques.



R E V I E W S

tetrahydropyridine (MPTP) show Parkinson’s-likesymptoms. Analyses of brain tissue from patients withParkinson’s disease implicate apoptosis-related DNAdamage and gene activation in the death of dopamineneurons56. Moreover, levels of Par-4 are selectivelyincreased in dopamine neurons of the substantia nigrabefore their death, and suppression of Par-4 expressionprotects dopamine neurons against death12. Caspase-1inhibition, drugs that suppress macromolecular synthe-sis, and neurotrophic factors, such as glial cell-derivedneurotrophic factor (GDNF), can protect dopamineneurons in Parkinson’s disease models58,59. Mutations inα-synuclein, a component of the brain lesions calledLEWY BODIES, are responsible for a small percentage ofParkinson’s disease cases57, and expression of mutant α-synuclein in cultured cells promotes apoptosis60.

Huntington’s disease is an inherited disorder inwhich neurons in the striatum degenerate, resulting inuncontrolled body movements. It is caused by expan-sions of a trinucleotide (CAG) sequence in the hunt-ingtin gene, producing a protein containing increasedpolyglutamine repeats61. Studies of patients withHuntington’s disease, and of rodents given the mito-chondrial toxin 3-nitropropionic acid, indicate thatimpaired mitochondrial function and excitotoxic deathmay be central to the disease11.

How does mutant huntingtin promote selectivedegeneration of striatal neurons? Although this is notknown, activation of an apoptotic programme is impli-cated. Studies of LYMPHOBLASTS from patients withHuntington’s disease have revealed increased stress-induced apoptosis associated with mitochondrial dys-function and increased caspase-3 activation62, suggestingan adverse effect of mutant huntingtin that is not limit-ed to neurons. Caspase-8 is redistributed to an insolublefraction in striatal tissue from patients, and expressionof mutant huntingtin in cultured cells induces caspase-

decreased (FIG. 5). Presenilin mutations may promoteneuronal degeneration by enhancing γ-secretase cleav-age of APP, thereby increasing production of neurotoxicamyloid-β (1–42)45. When mutant presenilin-1 proteinis expressed in cultured cells and in transgenic andknock-in mice, neurons become susceptible to deathinduced by various insults, including trophic-factorwithdrawal, exposure to amyloid-β or glutamate, andenergy deprivation53. Mutant presenilin-1 acts at an earlystep before Par-4 production, mitochondrial dysfunc-tion and caspase activation. Instead, calcium homeosta-sis in the endoplasmic reticulum (ER) is disturbed suchthat more calcium is released when neurons are exposedto potentially damaging oxidative and metabolicinsults54. Agents that suppress ER calcium release,including dantrolene and xestospongin, can counteractthe endangering effects of the mutations54, indicatingthat enhanced calcium release is central to the pathogen-ic action of mutant presenilin-1. Presenilin-2 may alsofacilitate apoptosis, although the underlying mechanismhas not been established55.

Motor system disorders. These disorders includeParkinson’s disease, Huntington’s disease and amy-otrophic lateral sclerosis. Patients with Parkinson’s dis-ease show profound motor dysfunction owing todegeneration of dopamine neurons in their SUBSTANTIA

NIGRA. Although the cause of Parkinson’s disease isunknown, increased oxidative stress and mitochondrialdysfunction in dopamine neurons are central to the dis-ease56. There is also a deficit in MITOCHONDRIAL COMPLEX I,

which may arise from, or contribute to, increased cellu-lar oxidative stress. Both environmental and genetic fac-tors may sensitize dopamine neurons to age-relatedincreases in oxidative stress and energy deficits56,57.Environmental toxins are implicated — monkeys andpeople exposed to the toxin 1-methyl-4-phenyl-1,2,3,6-

Table 2 | Examples of proteins that promote or suppress neuronal apoptosis

Pro-apoptotic

Caspases Cleavage of various enzyme, cytoskeletal and ion-channel substrates

Bax, Bad Pore formation in mitochondrial membrane; cytochrome c release

Glutamate receptor proteins Calcium influx; activation of kinase and proteases

Fas Initiates death cascade involving caspase-8

Par-4 Mitochondrial dysfunction; suppression of survival signals (NF-κB)

p53 Transcription of death genes; enhancement of Bax actions

Anti-apoptotic

Bcl-2, Bcl-xL Stabilize mitochondrial function; suppress oxidative stress

IAPs Caspase inhibition

Trophic factors/cytokines Induced expression of antioxidant enzymes, calcium regulating proteins, IAPs, Bcl-2;phosphorylation of Akt and other substrates

Telomerase Prevents telomere shortening; modulates DNA-damage responses

Anti-oxidant enzymes Suppress oxidative stress

Protein kinase Cζ Stimulates survival-gene expression (NF-κB)

Calcium-binding proteins Stabilize calcium homeostasis

SUBSTANTIA NIGRA

A part of the midbrain thatcontains dopamine-producingneurons whose axons innervatethe striatum and therebycontrol body movements.

MITOCHONDRIAL COMPLEX I

A group of proteins located atthe inner mitochondrialmembrane that function veryearly in the electron transportchain.

LEWY BODIES

Eosinophilic, cytoplasmicneuronal inclusions thatcontain aggregates of theproteins α-synuclein andubiquitin.

LYMPHOBLASTS

Bone-marrow-derived cells thatgive rise to lymphocytes.



R E V I E W S

tions with hydrogen peroxide or superoxide anion, themutant enzyme may induce oxidative damage to mem-branes and disturbances in mitochondrial function thatmake neurons vulnerable to excitotoxic apoptosis74.

In people with ALS, DNA starts to fragment betweenthe nucleosomes (a sign of nuclear apoptosis) in neu-rons within the spinal-cord anterior horn and motorcortex. The DNA damage is associated with increasedmitochondrial localization of Bax and decreased associa-tion of Bcl-2 (REF. 75). Levels of Bax, but not Bcl-2, areincreased in spinal-cord motor neurons of ALS patients,and a similar expression pattern of Bcl-2 family mem-bers is observed in Cu/Zn-SOD mutant mice. Theinvolvement of apoptosis in ALS is further suggested bythe fact that overexpression of Bcl-2 and administrationof caspase inhibitors delays degeneration and death ofmotor neurons in Cu/Zn-SOD mutant mice76,77.

Stroke. Focal ischaemic stroke, a leading cause of deathand disability worldwide, results in brain damage char-acterized by an INFARCT with a necrotic core, in which allcells die rapidly, and a surrounding ISCHAEMIC PENUMBRA,

in which neurons die over days to weeks78. Metaboliccompromise, overactivation of glutamate receptors, cal-cium overload and increased oxyradical productionoccur in neurons subjected to ischaemia. In addition,complex cytokine cascades involving microglial cellsand the cerebrovasculature may be important in pro-moting or preventing neuronal death after stroke79.Cells in the ischaemic penumbra show DNA damageand activation of the DNA damage-responsive proteinsPARP and Ku80. In rodent stroke models, neurons inthe ischaemic penumbra show morphological and mol-ecular changes consistent with apoptosis, including cas-pase activation, expression of pro-apoptotic genes andrelease of cytochrome c (REF. 78).

Signalling pathways involving hydrolysis of mem-brane phospholipids are implicated in neuronal apop-tosis in stroke. Cleavage of membrane sphingomyelinby acidic sphingomyelinase (ASMase) generates thelipid mediator ceramide. Focal cerebral ischaemia inmice induces large increases in ASMase activity andceramide levels, and production of inflammatorycytokines80. If mice lack ASMase or are given a drug thatinhibits production of ceramide, cytokine production issuppressed, brain damage is decreased, and symptomsare improved80. Mice lacking phospholipase-A

2show

decreased brain damage after focal cerebral ischaemia,suggesting an important function for one or more lipidmediators generated by this enzyme in ischaemic neu-ronal injury81.

Mice lacking specific caspases or given caspaseinhibitors show reduced brain damage after stroke82,83.In addition, delivery of neurotrophic factors known toprevent neuronal apoptosis can prevent neuronal deathafter stroke — bFGF, BDNF, NGF and sAPP-α areparticularly effective3. A pivotal role for mitochondrialalterations in stroke-induced neuron death is suggestedby studies showing that lack of mitochondrial Mn-SODexacerbates focal ischaemic brain injury84, whereasoverexpression of Mn-SOD has the opposite effect85. In

8-dependent apoptosis63. In addition, huntingtin can becleaved by caspases, which may promote protein aggre-gation and neurotoxicity64.

Transgenic mouse models of Huntington’s diseaserecapitulate certain aspects of the human disease,including intracellular inclusions of huntingtin anddegeneration of striatal neurons with several features ofapoptosis65. Inhibition of caspase-1 was reported toslow disease progression in one of these models66.Expression of mutant huntingtin in the brains of adultrats using viral vectors results in the formation of INTRA-

NEURONAL INCLUSIONS and cell death67. However, the for-mation of nuclear inclusions containing huntingtinmay not be required for apoptosis; in fact, such inclu-sions may be part of a cytoprotective response68.Moreover, wild-type — but not mutant — huntingtincan protect cells against apoptosis by suppressing celldeath before mitochondrial dysfunction69.

People with amyotrophic lateral sclerosis (ALS) sufferprogressive paralysis resulting from degeneration ofmotor neurons in the spinal cord. Most cases of ALS aresporadic, but some are inherited. This selective degenera-tion of motor neurons involves increased oxidative stress,overactivation of glutamate receptors and cellular calci-um overload70. Production of autoantibodies againstvoltage-dependent calcium channels may contribute tothe pathogenesis of ALS71. Mutations in the antioxidantenzyme Cu/Zn-superoxide dismutase (SOD) areresponsible for some inherited cases of ALS, and expres-sion of genes containing these mutations in transgenicmice results in spinal cord pathology remarkably similarto that of patients with ALS72,73. The mutations do notdecrease antioxidant activity of the enzyme, but result inthe gain of an adverse pro-apoptotic activity that mayinvolve increased peroxidase activity. Through interac-

Figure 4 | Brain tissue section from the hippocampus of a patient who died withAlzheimer’s disease. Examination reveals two prominent abnormalities. First, abnormal,silver-stained (black) accumulations are present in degenerating neurons (arrow). Molecularanalysis of such ‘neurofibrillary tangles’ reveals that they are composed of filamentousaggregates of the microtubule-associated protein tau. Second, spherical accumulations ofamyloid are present and are often associated with degenerated neurites (arrowhead); theplaques contain aggregates of amyloid-β peptide.

INTRANEURONAL INCLUSIONS

Aggregates of proteins thataccumulate in neurons withinthe cytoplasm or nucleus.

INFARCT

Brain tissue surrounding thesite of a stroke in which cellsdie.

ISCHAEMIC PENUMBRA

A region of tissue surroundingthe necrotic core of anischaemic infarct in whichneurons die primarily byapoptosis.



R E V I E W S

addition, treatment of rats with cyclosporin A decreasesthe size of the ischaemic infarct14.

Traumatic brain and spinal cord injury. Brain and spinalcord injuries account for most deaths and permanentdisabilities in people under 40 years old. Trauma initi-ates biochemical and molecular events involving manyof the same neurodegenerative cascades and neuro-protective signalling mechanisms that occur in thechronic neurodegenerative diseases described above.Histological and immunochemical analyses of thebrains from patients who died after traumatic braininjury (TBI) have found apoptosis-related changes inneurons, including the presence of DNA strand breaks,caspase activation and increased Bax and p53expression86. Sensory, motor and cognitive deficits afterTBI in mice are strongly correlated with the numbers ofneurons showing apoptotic nuclear damage87. Suchneuronal deaths are associated with increased expres-sion of p53 (REF. 88) and of the death receptor Fas and itsligand89. Caspases are also thought to be involved — cas-pase-3 activity increases markedly in the cerebral cortexof rats in response to TBI, and intraventricular adminis-tration of the caspase-3 inhibitor z-DEVD-fmk beforeinjury reduces cell death and improves symptoms, indi-cating a central function for caspases in this brain injurymodel90. In addition, mice expressing a dominant-nega-tive inhibitor of caspase-1 show reduced brain damageand free radical production after TBI91. Intraventricularinfusion of NGF in rats, beginning 24 hours after TBI,resulted in improved learning and memory, anddecreased death of neurons compared with controlrats92. Cyclosporin A protects against synaptic dysfunc-tion and cell death in rodent models of TBI, consistentwith a key role for mitochondrial membrane permeabil-ity in the neurodegenerative process93.

Apoptosis, as demonstrated by nuclear DNA frag-mentation and caspase activation, was a prominentfeature in the spinal cords of 14 out of 15 people whohad died three hours to two months after traumaticspinal cord injury (SCI) — apoptosis of OLIGODENDRO-

CYTES in the injury centre and adjacent white mattertracts was particularly prominent94. In rodents, SCIresults in neuronal apoptosis, which can be preventedby glutamate-receptor antagonists95. After SCI inrats96, caspase activation occurs in neurons at theinjury site within hours, and in oligodendrocytes adja-cent to, and distant from, the injury site over a periodof days. Studies of SCI in rats and monkeys showapoptosis of oligodendrocytes involving a progressiveinflammation-like process97. Thus, apoptosis of bothneurons and oligodendrocytes may contribute greatlyto the paralysis of patients with SCI.

Prospects for treatment and preventionMost translational research into neurodegenerativediseases is focused on developing drugs that inhibitneuronal dysfunction and death early in the diseaseprocess. Better understanding of the molecular andcellular underpinnings of neuronal apoptosis has ledto the identification of specific drug targets. One

Cell B

MLP

cGMP

PKG NF-κB

Glucose

Cytoplasm

Na+

Ca2+K+

Cell A

sAPPα

α β

γ

Aβ

Aβ

Aβ

Aβ

Aβ

Aβ

Aβ

APP

Aβ

R

Ins(1,4,5)P3

Ins(1,4,5)P3RYR

PS-1

Ca2+

R

AChGlutamate

Ca2+

APP

Aβ

Aβ

ROS

Ca2+

Ca2+

ER Mitochondria

a

Extracellularspace

b

OLIGODENDROCYTES

A specific type of glial cell,which forms myelinmembranes that insulate axonsof neurons and thereby increaseimpulse conduction velocity.

Figure 5 | Mechanisms underlying the pro-apoptoticactions of altered APP processing and presenilin-1mutations. a | The amyloid precursor protein (APP) can beproteolytically processed in two main ways. Cleavage ofAPP within the amyloid-β (Aβ) sequence by α-secretase (α)releases a secreted form of APP (sAPPα) from the cell sur-face. Secreted APPα activates a putative receptor (R) linkedto cyclic-GMP production and activation of cGMP-depen-dent protein kinase (PKG). PKG can then promote openingof K+ channels, resulting in membrane hyperpolarization,and can also activate the transcription factor NF-κB; theseeffects of sAPPα are believed to mediate its neuron-survival-promoting properties. A second pathway of APP processinginvolves cleavages at the N- and C-termini of Aβ by β-sec-retase (β) and γ-secretase (γ), respectively. This releases Aβfrom cells, which, under suitable conditions (high concentra-tion, oxidizing environment), begins to self-aggregate. Underthese conditions, Aβ induces membrane lipid peroxidation(MLP), which impairs the function of membrane ion-motiveATPases (Na+ and Ca2+ pumps) and glucose transporters.Neurons are thus vulnerable to apoptosis. b | Presenilin-1(PS-1) is an integral membrane protein located primarily inthe endoplasmic reticulum (ER). Mutations in PS-1 perturbER Ca2+ homeostasis, resulting in increased release of Ca2+

through ryanodine receptors (RYRs) and inositol-1,4,5trisphosphate (Ins(1,4,5)P3) receptors. The enhanced Ca2+

release triggers further Ca2+ influx through Ca2+ releasechannels in the plasma membrane, and this altered Ca2+

homeostasis makes neurons vulnerable to apoptosis andexcitotoxicity, and alters APP processing in a manner thatincreases Aβ production.



R E V I E W S

of micronutrient nutrition) is known to extend thelifespan of all mammalian species examined, andreduces development of various age-related diseases.Epidemiological data indicate that low calorie intake isassociated with reduced risk for Parkinson’s disease102.Importantly, the neurons of rodents maintained ondietary restriction are more resistant to apoptosis andshow improved symptoms in experimental models ofneurodegenerative disorders and stroke34,35,103. Themechanism by which dietary restriction benefits neu-rons may involve decreased levels of mitochondrialoxyradical production, or a mild metabolic stressresponse, in which neurons respond to the stress ofreduced energy availability by increasing the produc-tion of stress proteins and neurotrophic factors34,35,103.Developing further means of increasing neuronalresistance through dietary and perhaps behavioural104

manipulations will provide an important complementto drugs that may delay neurodegeneration in patientswho are already showing symptoms of disease.

approach is to block the apoptotic trigger; for exam-ple, by suppressing amyloid-β production inAlzheimer’s disease or glutamate-receptor activationin stroke. Other approaches target early premitochon-drial alterations, such as drugs that scavenge free radi-cals, block calcium influx into neurons or inhibit theactivity of Par-4. Activation of anti-apoptotic path-ways by treatment with neurotrophic factors is anoth-er approach. In several cases, work has progressed toclinical trials in patients; for example, bFGF in stroke,BDNF and insulin-like growth factor-1 (IGF-1) inALS, and vitamin E in Alzheimer’s disease98–101. Thenext wave of trials will probably include caspaseinhibitors, anti-inflammatory drugs and agents thatstabilize mitochondrial function, such as cyclosporinA and creatine, which suppresses mitochondrialoxyradical production and prevents ATP depletion.However, there is always the possibility that such ther-apies will have serious side effects.

Although this is an exciting era in the field of neu-rodegenerative disorders, with genetic and molecularbiological approaches rapidly advancing our under-standing of disease pathogenesis, there are no trulyeffective treatments for any of the disorders describedabove. We therefore need to develop methods of pre-venting or reducing risk for neurodegenerative dis-eases. One such approach is available now that, on thebasis of existing data, is likely to be effective. Dietaryrestriction (reduced calorie intake with maintenance

Links

DATABASE LINKS Bcl-2 | Bcl-xL | Bax | Bad | Par-4 |

caspase-8 | caspase-3 | BDNF | NGF | bFGF | TNF-α |CNTF | LIF | CREB | NCKAP1 | APP | presenilins 1 and 2 |GDNF | IGF-1ENCYCLOPEDIA OF LIFE SCIENCES Alzheimer disease |Apoptosis: molecular mechanisms

1. Wyllie, A. H. Apoptosis and carcinogenesis. Eur. J. CellBiol. 73, 189–197 (1997).Provides a historical overview of the morphologicaland biochemical features of apoptosis.

2. Oppenheim, R. W. Cell death during development of thenervous system. Annu. Rev. Neurosci. 14, 453–501(1991).

3. Mattson, M. P. & Lindvall, O. in The Aging Brain Vol. 2 (edsMattson, M. P. & Geddes, J. W.) 299–345 (JAI Press,Greenwich, Connecticut, 1997).

4. McKay, S. E., Purcell, A. L. & Carew, T. J. Regulation ofsynaptic function by neurotrophic factors in vertebratesand invertebrates: implications for development andlearning. Learn. Mem. 6, 193–215 (1999).

5. Ankarcrona, M. et al. Glutamate-induced neuronal death: asuccession of necrosis or apoptosis depending onmitochondrial function. Neuron 15, 961–973 (1995).

6. Glazner, G. W., Chan, S. L., Lu, C. & Mattson, M. P.Caspase-mediated degradation of AMPA receptorsubunits: a mechanism for preventing excitotoxic necrosisand ensuring apoptosis. J. Neurosci. 20, 3641–3649(2000).

7. Choi, D. W. Excitotoxic cell death. J. Neurobiol. 23,1261–1276 (1992).

8. Wong, P. C., Rothstein, J. D. & Price, D. L. The geneticand molecular mechanisms of motor neuron disease. Curr.Opin. Neurobiol. 8, 791–799 (1998).

9. Sastry, P. S. & Rao, K. S. Apoptosis and the nervoussystem. J. Neurochem. 74, 1–20 (2000).

10. Mattson, M. P. Modification of ion homeostasis by lipidperoxidation: roles in neuronal degeneration and adaptiveplasticity. Trends Neurosci. 21, 53–57 (1998).

11. Beal, M. F. Aging, energy, and oxidative stress inneurodegenerative diseases. Ann. Neurol. 38, 357–366(1995).

12. Duan, W., Zhang, Z., Gash, D. M. & Mattson, M. P.Participation of prostate apoptosis response-4 indegeneration of dopaminergic neurons in models ofParkinson’s disease. Ann. Neurol. 46, 587–597 (1999).

13. Kroemer, G., Dallaporta, B. & Resche-Rigon, M. Themitochondrial death/life regulator in apoptosis andnecrosis. Annu. Rev. Physiol. 60, 619–642 (1998). Detailed review of the role of mitochondria in

apoptosis.14. Matsumoto, S., Friberg, H., Ferrand-Drake, M. & Wieloch,

T. Blockade of the mitochondrial permeability transitionpore diminishes infarct size in the rat after transient middlecerebral artery occlusion. J. Cereb. Blood Flow Metab. 19,736–741 (1999).

15. Pellegrini, M. & Strasser, A. A portrait of the Bcl-2 proteinfamily: life, death, and the whole picture. J. Clin. Immunol.19, 365–377 (1999).

16. Martinou, J. C. et al. Overexpression of Bcl-2 in transgenicmice protects neurons from naturally occurring cell deathand experimental ischemia. Neuron 13, 1017–1030(1994). Direct evidence that Bcl-2 can prevent death ofneurons in vivo during development and in anexperimental stroke model.

17. Guo, Q. et al. Alzheimer’s presenilin mutation sensitizesneural cells to apoptosis induced by trophic factorwithdrawal and amyloid β-peptide: involvement of calciumand oxyradicals. J. Neurosci. 17, 4212–4222 (1998).

18. White, F. A., Keller-Peck, C. R., Knudson, C. M.,Korsmeyer, S. J. & Snider, W. D. Widespread elimination ofnaturally occurring neuronal death in Bax-deficient mice. J.Neurosci. 18, 1428–1439 (1998).Direct evidence that the pro-apoptotic protein Bax isessential in natural cell death during development ofthe nervous system.

19. Green, D. R. & Reed, J. C. Mitochondria and apoptosis.Science 28, 1309–1312 (1998).

20. Chan, S. L. & Mattson, M. P. Caspase and calpainsubstrates: roles in synaptic plasticity and cell death. J.Neurosci. Res. 58, 167–190 (1999).

21. Raoul, C., Pettmann, B. & Henderson, C. E. Active killingof neurons during development and following stress: a rolefor p75(NTR) and Fas? Curr. Opin. Neurobiol. 10, 111–117(2000).

22. Nagata, S. Apoptotic DNA fragmentation. Exp. Cell Res.256, 12–18 (2000).

23. Guo, Q. et al. Par-4 is a mediator of neuronal degenerationassociated with the pathogenesis of Alzheimer disease.Nature Med. 4, 957–962 (1998). Initial description and characterization of theinvolvement of Par-4 in neuronal apoptosis in

experimental models of developmental cell deathand Alzheimer’s disease.

24. Mattson, M. P., Duan, W., Chan, S. L. & Camandola, S.Par-4: an emerging pivotal player in neuronal apoptosisand neurodegenerative disorders. J. Mol. Neurosci. 13,17–30 (1999).

25. Liu, J. P. Studies of the molecular mechanisms in theregulation of telomerase activity. FASEB J. 13, 2091–2104(1999).

26. Fu, W., Killen, M., Pandita, T. & Mattson, M. P. The catalyticsubunit of telomerase is expressed in developing brainneurons and serves a cell survival-promoting function. J. Mol. Neurosci. 14, 3–15 (2000).

27. Fu, W., Begley, J. G., Killen, M. W. & Mattson, M. P. Anti-apoptotic role of telomerase in pheochromocytoma cells.J. Biol. Chem. 274, 7264–7271 (1999).

28. Tamatani, M., Ogawa, S., Nunez, G. & Tohyama, M.Growth factors prevent changes in Bcl-2 and Baxexpression and neuronal apoptosis induced by nitric oxide.Cell Death Differ. 5, 911–919 (1998).

29. Hagg, T. & Varon, S. Ciliary neurotrophic factor preventsdegeneration of adult rat substantia nigra dopaminergicneurons in vivo. Proc. Natl Acad. Sci. USA 90, 6315–6319(1993).

30. Middleton, G. et al. Cytokine-induced nuclear factor kappaB activation promotes the survival of developing neurons.J. Cell Biol. 148, 325–332 (2000).

31. Mattson, M. P. & Camandola, S. NF-κB inneurodegenerative disorders. J. Clin. Invest. 61,134–139(2000).

32. Yu, Z. F., Zhou, D., Bruce-Keller, A. J., Kindy, M. S. &Mattson, M. P. Lack of the p50 subunit of NF-κB increasesthe vulnerability of hippocampal neurons to excitotoxicinjury. J. Neurosci. 19, 8856–8865 (1999).

33. Aschner, M., Allen, J. W., Kimelberg, H. K., LoPachin, R.M. & Streit, W. J. Glial cells in neurotoxicity development.Annu. Rev. Pharmacol. Toxicol. 39, 151–173 (1999).

34. Yu, Z. F. & Mattson, M. P. Dietary restriction and 2-deoxyglucose administration reduce focal ischemic braindamage and improve behavioral outcome: evidence for apreconditioning mechanism. J. Neurosci. Res. 5, 830–839(1999).

35. Lee, J., Duan, W., Long, J. M., Ingram, D. K. & Mattson,



R E V I E W S

M. P. Dietary restriction increases survival of newly-generated neural cells and induces BDNF expression inthe dentate gyrus of rats. J. Mol. Neurosci. (in the press).

36. Berridge, M. J., Lipp, P. & Bootman, M. D. The versatilityand universality of calcium signalling. Nature Rev. Mol. CellBiol. 1, 11–21 (2000).

37. Yano, S., Tokumitsu, H. & Soderling, T. R. Calciumpromotes cell survival through CaM-K kinase activation ofthe protein-kinase-B pathway. Nature 396, 584–587(1998).

38. Hu, S. C., Chrivia, J. & Ghosh, A. Regulation of CBP-mediated transcription by neuronal calcium signaling.Neuron 22, 799–808 (1999).

39. Furukawa, K. et al. The actin-severing protein gelsolinmodulates calcium channel and NMDA receptor activitiesand vulnerability to excitotoxicity in hippocampal neurons.J. Neurosci. 17, 8178–8186 (1997).

40. Furukawa, K., Barger, S. W., Blalock, E. M. & Mattson, M. P. Activation of K+ channels and suppression ofneuronal activity by secreted β-amyloid-precursor protein.Nature 379, 74–78 (1996).

41. Guenette, S. Y, & Tanzi, R. E. Progress toward validtransgenic mouse models for Alzheimer’s disease.Neurobiol. Aging 20, 201–211 (1999).

42. Sathasivam, K. et al. Transgenic models of Huntington’sdisease. Phil. Trans. R. Soc. Lond. B 354, 963–969(1999).

43. Borchelt, D. R., Wong, P. C., Sisodia, S. S. & Price, D. L.Transgenic mouse models of Alzheimer’s disease andamyotrophic lateral sclerosis. Brain Pathol. 8, 735–757(1998).

44. Cummings, J. L., Vinters, H. V., Cole, G. M. &Khachaturian, Z. S. Alzheimer’s disease: etiologies,pathophysiology, cognitive reserve, and treatmentopportunities. Neurology 51, S2–S17 (1998).

45. Haass, C. & De Strooper, B. The presenilins in Alzheimer’sdisease — proteolysis holds the key. Science 286,916–919 (1999).

46. Su, J. H., Anderson, A. J., Cummings, B. & Cotman, C. W.Immunocytochemical evidence for apoptosis inAlzheimer’s disease. Neuroreport 5, 2529–2533 (1994).

47. Masliah, E., Mallory, M., Alford, M., Tanaka, S. & Hansen,L. A. Caspase dependent DNA fragmentation might beassociated with excitotoxicity in Alzheimer disease. J.Neuropathol. Exp. Neurol. 57, 1041–1052 (1998).

48. Suzuki, T. et al. Molecular cloning of a novel apoptosis-related gene, human Nap1 (NCKAP1), and its possiblerelation to Alzheimer disease. Genomics 63, 246–254(2000).

49. Loo, D., Copani, A., Pike, C., Whittemore, E.,Walencewicz, A. & Cotman, C. W. Apoptosis is induced byβ-amyloid in cultured central nervous system neurons.Proc. Natl Acad. Sci. USA 90, 7951–7955 (1993).

50. Weidemann, A. et al. Proteolytic processing of theAlzheimer’s disease amyloid precursor protein within itscytoplasmic domain by caspase-like proteases. J. Biol.Chem. 274, 5823–5829 (1999).

51. Chen, Y., McPhie, D. L., Hirschberg, J. & Neve, R. L. Theamyloid precursor protein-binding protein APP-BP1 drivesthe cell cycle through the S–M checkpoint and causesapoptosis in neurons. J. Biol. Chem. 275, 8929–8935(2000).

52. Lu, D. C. et al. A second cytotoxic proteolytic peptidederived from amyloid precursor protein. Nature Med. 6,397–404 (2000).

53. Guo, Q. et al. Neurotrophic factors (activity-dependentneurotrophic factor (ADNF) and basic fibroblast growthfactor (bFGF)) interrupt excitotoxic neurodegenerativecascades promoted by a presenilin-1 mutation. Proc. NatlAcad. Sci. USA 96, 4125–4130 (1999). Uses knock-in mice to identify the gain-of-functionaction of mutant presenilin-1 as a pro-apoptoticactivity that results from perturbed cellular calciumhomeostasis and can be suppressed byneurotrophic factors.

54. Mattson, M. P. et al. Calcium signaling in the ER: its role inneuronal plasticity and neurodegenerative disorders.Trends Neurosci. 23, 222–229 (2000).

55. Vito, P., Lacana, E. & d’Adamio, L. Interfering withapoptosis: Ca2+-binding protein ALG-2 and Alzheimer’sdisease gene ALG-3. Science 271, 521–525 (1996).

56. Jenner, P. & Olanow, C. W. Understanding cell death inParkinson’s disease. Ann. Neurol. 44, S72–S84 (1998).

57. Polymeropoulos, M. H. Autosomal dominant Parkinson’sdisease and alpha-synuclein. Ann. Neurol. 44, S63–S64(1998).

58. Klevenyi, P. et al. Transgenic mice expressing a dominantnegative mutant interleukin-1β converting enzyme showresistance to MPTP neurotoxicity. Neuroreport 10,635–638 (1999).

59. Gash, D. M. et al. Functional recovery in parkinsonianmonkeys treated with GDNF. Nature 380, 252–255 (1996).

60. El-Agnaf, O. M. et al. Aggregates from mutant and wild-type α-synuclein proteins and NAC peptide induceapoptotic cell death in human neuroblastoma cells byformation of β-sheet and amyloid-like filaments. FEBS Lett.440, 71–75 (1998).

61. Brandt, J. et al. Trinucleotide repeat length and clinicalprogression in Huntington’s disease. Neurology 46,527–531 (1996).

62. Sawa, A. et al. Increased apoptosis of Huntington diseaselymphoblasts associated with repeat length-dependentmitochondrial depolarization. Nature Med. 5, 1194–1198(1999).

63. Sanchez, I. et al. Caspase-8 is required for cell deathinduced by expanded polyglutamine repeats. Neuron 22,623–633 (1999).

64. Wellington, C. L. et al. Inhibiting caspase cleavage ofhuntingtin reduces toxicity and aggregate formation inneuronal and non-neuronal cells. J. Biol. Chem. 275,19831–19838 (2000).

65. Leavitt, B. R., Wellington, C. I. & Hayden, M. R. Recentinsights into the molecular pathogenesis of Huntingtondisease. Semin. Neurol. 19, 385–395 (1999).

66. Ona, V. O. et al. Inhibition of caspase-1 slows diseaseprogression in a mouse model of Huntington’s disease.Nature 399, 263–267 (1999).

67. Senut, M. C., Suhr, S. T., Kaspar, B. & Gage, F. H.Intraneuronal aggregate formation and cell death after viralexpression of expanded polyglutamine tracts in the adultrat brain. J. Neurosci. 20, 219–229 (2000).

68. Kim, M. et al. Mutant huntingtin expression in clonal striatalcells: dissociation of inclusion formation and neuronalsurvival by caspase inhibition. J. Neurosci. 19, 964–973(1999). Provides insight into the pathogenic action ofpolyglutamine-repeat huntingtin in neuronalapoptosis.

69. Rigamonti, D. et al. Wild-type huntingtin protects fromapoptosis upstream of caspase-3. J. Neurosci. 20,3705–3713 (2000).

70. Cookson, M. R. & Shaw, P. J. Oxidative stress and motorneurone disease. Brain Pathol. 9, 165–186 (1999).

71. Smith, R. G. et al. Autoimmunity and ALS. Neurology 47,S40–S45 (1996).

72. Gurney, M. E. et al. Motor neuron degeneration in micethat express a human Cu,Zn superoxide dismutasemutation. Science 264, 1772–1775 (1994).

73. Wong, P. C. et al. An adverse property of a familial ALS-linked SOD1 mutation causes motor neuron diseasecharacterized by vacuolar degeneration of mitochondria.Neuron 14, 1105–1116 (1995).

74. Kruman, I., Pedersen, W. A. & Mattson, M. P. ALS-linkedCu/Zn-SOD mutation increases vulnerability of motorneurons to excitotoxicity by a mechanism involvingincreased oxidative stress and perturbed calciumhomeostasis. Exp. Neurol. 160, 28–39 (1999).

75. Martin, L. J., Price, A. C., Kaiser, A., Shaikh, A. Y. & Liu, Z.Mechanisms for neuronal degeneration in amyotrophiclateral sclerosis and in models of motor neuron death. Int.J. Mol. Med. 5, 3–13 (2000).

76. Li, M. et al. Functional role of caspase-1 and caspase-3 inan ALS transgenic mouse model. Science 288, 335–339(2000).

77. Kostic, V., Jackson-Lewis, V., de Bilbao, F., Dubois-Dauphin, M. & Przedborski, S. Bcl-2: prolonging life in atransgenic mouse model of familial amyotrophic lateralsclerosis. Science 277, 559–562 (1997).

78. Dirnagl, U., Iadecola, C. & Moskowitz, M. A. Pathobiologyof ischaemic stroke: an integrated view. Trends Neurosci.22, 391–397 (1999). Details molecular and cellular mechanisms in theneuronal death in ischaemic stroke.

79. Lipton, P. Ischemic cell death in brain neurons. Physiol.Rev. 79, 1431–568 (1999).

80. Yu, Z. et al. Pivotal role for acidic sphingomyelinase incerebral ischemia-induced ceramide and cytokineproduction, and neuronal death. J. Mol. Neurosci. (in thepress).

81. Bonventre, J. V. et al. Reduced fertility and postischaemicbrain injury in mice deficient in cytosolic phospholipase A2.Nature 390, 622–625 (1997).

82. Schielke, G. P., Yang, G. Y., Shivers, B. D. & Betz, A. L.Reduced ischemic brain injury in interleukin-1β convertingenzyme-deficient mice. J. Cereb. Blood Flow Metab. 18,180–185 (1998).

83. Hara, H. et al. Inhibition of interleukin 1β convertingenzyme family proteases reduces ischemic and excitotoxicneuronal damage. Proc. Natl Acad. Sci. USA 94,2007–2012 (1997).

84. Murakami, K. et al. Mitochondrial susceptibility to oxidativestress exacerbates cerebral infarction that followspermanent focal cerebral ischemia in mutant mice withmanganese superoxide dismutase deficiency. J. Neurosci.18, 205–213 (1998).

85. Keller, J. N. et al. Mitochondrial manganese superoxidedismutase prevents neural apoptosis and reducesischemic brain injury: suppression of peroxynitriteproduction, lipid peroxidation, and mitochondrialdysfunction. J. Neurosci. 18, 687–697 (1998).

86. Clark, R. S. et al. Increases in Bcl-2 and cleavage ofcaspase-1 and caspase-3 in human brain after head injury.FASEB J. 13, 813–821 (1999).

87. Fox, G. B., Fan, L., Levasseur, R. A. & Faden, A. I.Sustained sensory/motor and cognitive deficits withneuronal apoptosis following controlled cortical impactbrain injury in the mouse. J. Neurotrauma 15, 599–614(1998).

88. Napieralski, J. A., Raghupathi, R. & McIntosh, T. K. Thetumor-suppressor gene, p53, is induced in injured brainregions following experimental traumatic brain injury. Mol.Brain Res. 71, 78–86 (1999).

89. Beer, R. et al. Expression of Fas and Fas ligand afterexperimental traumatic brain injury in the rat. J. Cereb.Blood Flow Metab. 20, 669–677 (2000).

90. Yakovlev, A. G. et al. Activation of CPP32-like caspasescontributes to neuronal apoptosis and neurologicaldysfunction after traumatic brain injury. J. Neurosci. 17,7415–7424 (1997).

91. Fink, K. B. et al. Reduction of post-traumatic brain injuryand free radical production by inhibition of the caspase-1cascade. Neuroscience 94, 1213–1218 (1999).

92. Sinson, G., Perri, B. R., Trojanowski, J. Q., Flamm, E. S. &McIntosh, T. K. Improvement of cognitive deficits anddecreased cholinergic neuronal cell loss and apoptotic celldeath following neurotrophin infusion after experimentaltraumatic brain injury. J. Neurosurg. 86, 511–518 (1997).

93. Albensi, B. C., Sullivan, P. G., Thompson, M. B., Scheff, S. W. & Mattson, M. P. Cyclosporine ameliorates traumaticbrain injury-induced alterations of hippocampal synapticplasticity. Exp. Neurol. 162, 385–389 (2000).

94. Emery, E. et al. Apoptosis after traumatic human spinalcord injury. J. Neurosurg. 89, 911–920 (1998).

95. Wada, S. et al. Apoptosis following spinal cord injury in ratsand preventative effect of N-methyl-D-aspartate receptorantagonist. J. Neurosurg. 91, 98–104 (1999).

96. Springer, J. E., Azbill, R. D. & Knapp, P. E. Activation of thecaspase-3 apoptotic cascade in traumatic spinal cordinjury. Nature Med. 5, 943–946 (1999).

97. Crowe, M. J., Bresnahan, J. C., Shuman, S. L., Masters, J. N. & Beattie, M. S. Apoptosis and delayed degenerationafter spinal cord injury in rats and monkeys. Nature Med.3, 73–76 (1997).

98. Ay, H., Ay, I., Koroshetz, W. J. & Finklestein, S. P. Potentialusefulness of basic fibroblast growth factor as a treatmentfor stroke. Cerebrovasc. Dis. 9, 131–135 (1999).

99. The BDNF Study Group. A controlled trial of recombinantmethionyl human BDNF in ALS: The BDNF Study Group(Phase III). Neurology 52, 1427–1433 (1999).

100. Borasio, G. D. et al. A placebo-controlled trial of insulin-likegrowth factor-I in amyotrophic lateral sclerosis. EuropeanALS/IGF-I Study Group. Neurology 51, 583–586 (1998).

101. Grundman, M. Vitamin E and Alzheimer disease: the basisfor additional clinical trials. Am. J. Clin. Nutr. 71,630S–636S (2000).

102. Logroscino, G. et al. Dietary lipids and antioxidants inParkinson’s disease: a population-based, case-controlstudy. Ann. Neurol. 39, 89–94 (1996).

103. Duan, W. & Mattson, M. P. Dietary restriction and 2-deoxyglucose administration improve behavioral outcomeand reduce degeneration of dopaminergic neurons inmodels of Parkinson’s disease. J. Neurosci. Res. 57,195–206 (1999).

104. Ohlsson, A. L. & Johansson, B. B. Environment influencesfunctional outcome of cerebral infarction in rats. Stroke 26,644–649 (1995).

105. Mattson, M. P. & Duan, W. Apoptotic biochemicalcascades in synaptic compartments: roles in adaptiveplasticity and neurodegenerative disorders. J. Neurosci.Res. 58, 152–166 (1999). Reviews the evidence for and implications ofapoptosis-related mechanisms in synapticremodelling and neuronal cell death.

106. Ivins, K. J., Bui, E. T. & Cotman, C. W. Beta-amyloidinduces local neurite degeneration in culturedhippocampal neurons: evidence for neuritic apoptosis.Neurobiol. Dis. 5, 365–378 (1998).

107. Albensi, B. C. & Mattson, M. P. Evidence for thenvolvement TNF and NF-κB in hippocampal synapticplasticity. Synapse 35, 151–159 (2000).



R E V I E W S

“I know a man who grabbed a cat by the tail and

learned 40 per cent more about cats than the man

who didn’t.”

Mark Twain

Until quite recently, scientists could only investigatechemical processes on a bulk level. The forces and stress-es that molecules exert on each other or develop in thecourse of reactions were not directly measurable.During the past few years, this situation has changedrapidly thanks to the development of methods formanipulating single molecules1–4. Methods such as OPTI-

CAL TWEEZERS and scanning force microscopy (SFM)5 aremaking it possible to follow, in real-time and at a single-molecule level, the movements, forces and strains thatdevelop during the course of a reaction. These methodscan be used to measure directly the forces that holdtogether molecular structures. They can also be used toexert external forces to modify the extent and even alterthe fate of reactions in the hope of discovering rules thatgovern the inter-conversion of mechanical and chemicalenergy in these processes. This area of research canrightly be called ‘mechanochemistry’, and includes bio-chemical processes as diverse as protein folding6, DNAelasticity7–9, the protein-induced bending of DNA10, thestress-induced catalysis of enzymes11, the behaviour ofmolecular motors12–15, and even the ubiquitous process-es of protein–protein recognition16.

Here we focus on the current capabilities and limita-tions of single-molecule manipulation methods, andprovide guidelines for choosing the most appropriatemethod for a given problem.

Choosing the appropriate methodAll single-molecule manipulation methods require twobasic elements: a probe, which is usually of microscopicdimensions, that can generate or detect forces and dis-placements; and a way to spatially locate the molecules.As summarized in TABLE 1, the relevant force ranges,minimum displacements, probe stiffness, applicationsand practical advantages of each technique vary signifi-cantly.

Mechanical transducersMechanical force transducers apply or sense forcesthrough the displacement of a bendable beam. The mostcommon examples are SFM cantilevers5 (FIG. 1) andmicroneedles12 (FIG. 2). The spatial control of transducerscan be accomplished efficiently by PIEZO-ELECTRIC position-ers (FIG. 1a). Mechanical transducers have been used toinvestigate systems ranging from protein unfolding6 andcell motility17 to forces generated by motor proteins12.Mechanical transducers possess a linear response over abroad range of displacement and forces. Two importantfactors determine how mechanical transducers interactwith single molecules: their size and stiffness. The effectof these parameters is described below and in BOX 1.

SFM cantilevers. Microfabricated cantilevers are avail-able in a wide variety of sizes, shapes and materials.SFM devices are also commercially available, and someare specifically designed for manipulating single mole-cules18 (FIG. 1b). The advantages of SFM are its high spa-tial range and sensitivity, its throughput (the ability tostudy many single molecules on a surface) and versatility.

GRABBING THE CAT BY THE TAIL:MANIPULATING MOLECULES ONE BY ONECarlos Bustamante*‡§, Jed C. Macosko‡ and Gijs J. L. Wuite§

Methods for manipulating single molecules are yielding new information about both the forcesthat hold biomolecules together and the mechanics of molecular motors. We describe here thephysical principles behind these methods, and discuss their capabilities and current limitations.

*Howard Hughes MedicalInstitute and ‡Departmentsof Molecular and CellBiology and §Physics,University of California,Berkeley, California 94720,USA. e-mails:[email protected];[email protected];[email protected]

OPTICAL TWEEZERS

Focused photon fields.

PIEZO-ELECTRIC

Describes a device that expandsor contracts as a voltage isapplied to an internal crystal.



R E V I E W S

domain (28.4 nm), matches the unfolding rate deter-mined by chemical denaturation experiments (4.9×10–4 s–1). So, mechanical unfolding experiments mea-sure the same interactions as their chemical analogues,and have the potential to follow secondary-structure-unfolding events.

Microneedles. Because of their dimensions (typically50–500 µm long and 0.1–1 µm thick), glass micronee-dles are usually softer than cantilevers (TABLE 1). Thisproperty gives them an advantage over SFM cantileversfor probing delicate biological systems. Microneedles arenot commercially available, however, and the devices todetect their displacement are less standardized than inSFM. Two general approaches for displacement detec-tion have been reported: imaging the microneedleitself 12,25–28; and using a chemically etched optical fibrethat projects light from its tip onto a photodiode29–31 as amicroneedle. This latter method has been used to mea-sure the stretching of twisted DNA31 and to study thebinding of RecA to stretched DNA30.

Microneedles have also been used successfully toquantify the force that myosin exerts on F-actin12,25,27. Inthe first of these experiments, fluorescently labelled F-actin filaments were attached to a microneedle coatedwith monomeric myosin (FIG. 2). This filament wasbrought into contact with a myosin-coated surface. Thesubsequent bending of the microneedle correspondedto the combined force (9.6 pN) exerted by no more than53 interacting myosin heads (at least 0.2 pN per myosinhead). This average force is comparable to that exertedin a muscle during contraction32 (~1 pN). More recentstudies, using complex experimental geometries (FIG. 3a),have determined the step size of myosin on actin to be5.3 nm (REF. 25), and the force generated during the actu-al power stroke to be ~3–5 pN (REFS 21,27).

External field manipulatorsExternal fields provide another approach to themanipulation of single molecules. Examples are HYDRO-

DYNAMIC, magnetic and PHOTON fields. Unlike mechani-cal transducers, external fields act on molecules from adistance. These fields can be used to exert forces on

For example, SFM can be used both as an imaginginstrument and as a manipulation device, as first shownby Müller et al.19 and further exploited by Oesterheltand colleagues20 (see below).

SFM cantilevers have stiffness (κ) ranging from 10–3

to 100 N m–1. Stiffer cantilevers have lower sensitivities,as force is always detected by measuring a displacementthat is inversely proportional to the stiffness. They arenonetheless useful when a given process (conforma-tional changes, for example) requires the application ofhigh forces. Although stiffer cantilevers experience cor-respondingly large force fluctuations owing to thermalmotion (for a typical cantilever with κ = 0.06 N m–1,the root-mean-square force fluctuation is ~16 pN; BOX

1), the signal-to-noise ratio of the measurement isindependent of the stiffness of the cantilever2,21. Inaddition, drift over time caused by the thermal expan-sion and nonlinear voltage response of the piezo-elec-tric crystals can further compromise the control offorce on the sample. This is not a serious limitation,however, as ways to compensate for drift have beendeveloped22.

A promising development in SFM methodology isthe fabrication of smaller, but still soft, cantilevers23.Their small physical dimensions allow them to havehigher sensitivity and faster response times24. Being soft,these small cantilevers allow high spatial resolutionwithout a subsequent increase in force fluctuations.This higher spatial resolution stems from the distribu-tion of thermal fluctuations over a broader frequencyrange, thus decreasing the noise at biologically relevantfrequencies (BOX 1).

The scanning force microscope has been used suc-cessfully to study the mechanism of unfolding in pro-teins. Fernandez and co-workers6 unfolded a proteinmade of repeating immunoglobulin-type domains bypulling it with an SFM cantilever. To obtain refoldingrates, they allowed the protein to refold for a varyingamount of time before it was re-extended. Theseauthors also determined how the unfolding force varieswith pulling speed. Extrapolation of these data to zeroforce yielded a pulling speed of 0.013 nm s-1, which,when divided by the extension required to unfold one

Table 1 | Overview of single-molecule manipulation methods

Methods Fmin–max (N)§ Xmin (m)§ Stiffness (N m–1) Applications Practical advantages

Cantilevers* 10–11–10–7 10–10 0.001–100 Protein/polysaccharides6,64 High spatial resolutionBond strength65,66 Commercially available

Microneedles* 10–12–10–10 10–9 10–6–1 Myosin motor force12 Good operator controlDNA/titin strength26,28 Soft spring constant

Flow field‡ 10–13–10–9 10–8 n.a. DNA dynamics38 Rapid buffer exchangeRNA polymerase36 Simplicity of design

Magnetic field‡ 10–14–10–11 10–8 n.a. DNA entropic elasticity8 Specificity to magnetsTopoisomerase activity41 Ability to induce torque

Photon field‡ 10–13–10–10 10–9 10–10–10–3 Protein motors13,14 Specific manipulationProtein unfolding52 High force resolution

*Mechanical transducers: probes are bendable beams; spatial location is by beam deflection. ‡External field manipulators: probes are microscopic beads; spatial locationis by bead displacement. §These numbers represent only empirical, not absolute limits. (Fmin–max, force range; Xmin, minimum displacement.)

HYDRODYNAMIC FIELD

A force field resulting from themomentum imparted bymolecules in a flowing aqueoussolution.

PHOTON FIELD

A force field resulting from themomentum imparted byphotons in a beam of light.



R E V I E W S R E V I E W S

STOKES’S LAW. For spheres of any size in water, Stokes’s lawremains valid for forces up to ~10 nN, beyond whichturbulence becomes a factor. Forces up to ~10 nN cantherefore be applied reliably.

The advantages of using flow fields for single-mole-cule experiments include the fact that the liquids sur-rounding the tethered macromolecule can be easilyreplaced. This feature is important in many single-mol-ecule studies of enzymes, which require varying bufferconditions. Moreover, the flow can be used to introducenew beads or biomolecules.

To calculate the drag force, the size of the bead han-dles and the actual flow velocity must be known.Furthermore, an accurate drag-force calculationrequires the bead handle to be stationary in the flow,such as in the case of a bead tethered to a surface by apiece of DNA (FIG. 3b). In addition, as single-moleculeexperiments are almost always carried out inside amicrochamber, force determination should also takeinto account the modification of Stokes’s law, owing tothe coupling, through water, between the bead and theboundaries of the microchamber8,33,34. Finally, it shouldbe kept in mind that, because the drag coefficient of anobject scales largely with its longest dimension35, oftenthe friction coefficient of a long polymer such as DNAfor example is comparable to that of the bead handleand cannot be neglected.

The flow-field manipulation technique was demon-strated in the earliest single-molecule study of DNAelasticity8. In that study, biotinylated DNA was attachedby one end to a streptavidin-coated glass surface and itsother end was attached to a magnetic bead (FIG. 3b).With this set-up, different tensions were applied to single

molecules by acting either on the molecules them-selves, or through ‘handles’ such as glass beads, poly-styrene beads or metallic particles attached to the mol-ecules. The various external fields give differentdegrees of control over the magnitude and stability ofthe applied forces.

Flow fields. Flow fields exert forces on objects throughthe transfer of momentum from the fluid to the object(FIG. 3b). In LAMINAR FLOW, the drag force between a mov-ing liquid with viscosity η and velocity v and a station-ary bead handle of radius r can be calculated using

Figure 1 | Applications of the scanning force microscope (SFM). a | The principal SFM components. Laser light is focused onto the back of a cantilever thatends with a nanometre-scale tip. The reflection and corresponding position of the tip is detected by a position-sensitive photodiode. A piezo-electric scannermoves the sample in all directions, enabling the tip to scan topography or to extend molecules attached to the surface. b | Diagrams and force curves showing themechanical unfolding of repeating immunoglobulin-like domains6,64. As the distance between the surface and tip increases (from state 1 to state 2), the moleculeextends and generates a restoring force that bends the cantilever. When a domain unfolds (state 3), the free length of the protein increases, relaxing the force onthe cantilever. Further extension again results in a restoring force (state 4). The last peak represents the final extension of the unfolded molecule before detachmentfrom the SFM tip (state 5).

Laser

Lens

Cantilever

Piezo-electricscanner

Sample

Position-sensitivedetector

5

4

4

3

2

1

3

2

1

20 nm

200

pN

a b

Figure 2 | Using a microneedle to measure the force ofmyosin acting on actin. A bendable microneedle coated inmyosin heads (not shown) catches an actin filament. Thisfilament is brought into contact with a glass coverslip coatedin myosin molecules. In the presence of ATP, the myosindrags the actin filament across the coverslip and generates aforce on the microneedle, which is observable by video-fluorescence microscopy.

Actin filamentMyosin

molecules

Bendable microneedle

Coverslip

LAMINAR FLOW

A flow of molecules in whichneighbouring molecules havelinearly dependent velocities,that is, not a turbulent flow.

STOKES’S LAW

Fdrag

= 6πrηv



R E V I E W S

workers37–39 to characterize the rheological properties ofindividual DNA molecules.

Magnetic fields. Magnetic fields can be used to manipu-late and apply forces to biomolecules that are tethered tomagnetic particles, as most biomolecules have zeromagnetic susceptibility.Very stable and small forces canthen be created with magnetic fields from either perma-nent magnets or electromagnets.

The forces generated with permanent magnets onsmall magnetic beads (r < 3 µm) are usually below 10pN (REFS 8,40). A drawback of magnetic fields is that themagnetic forces often have to be measured indirectly.For example, the forces acting on magnetic beads can becalibrated by determining the velocity attained by thebeads in liquid in a given field and by using Stokes’s law8.However, an elegant and direct measurement of mag-netic force has been implemented by Strick et al.40,41 byusing the equipartition theorem (BOX 1).

Magnetic force has been used to apply torsionalstress to individual DNA molecules. Strick et al.40,41

used the fact that magnetic beads have a preferredmagnetization axis that makes them orientate with anexternal field and rotate when the field rotates. In thisway, individual DNA molecules, torsionally con-strained between a coated glass surface and a magneticbead (FIG. 3c), were under- or overwound to determinethe force–extension behaviour of supercoiled DNA.More recently, this same set-up was used to investigatethe relaxation of a supercoiled DNA by single topoiso-merase II molecules (BOX 2).

Photon fields. Optical tweezers developed rapidly aftertheir power was shown by Ashkin and colleagues42 inthe 1980s. Optical tweezers rely on forces imparted tomatter by scattering, emission and absorption of light.The RADIATION PRESSURE, which stems from the momen-tum change as light refracts off an object43 (FIG. 4a),allows objects (a bead, for example) to be held in afocused laser beam, with which it is possible to generatea spring-like force. As with mechanical transducers, thestiffness of the optical trap is an important parameterfor force and position resolution (BOX 1). In general, thespring stiffness is much smaller than that of a cantilever(TABLE 1). The force exerted on a refractive objectdepends on the power of the laser, the dimensions ofthe object and the difference in the refractive indexbetween the object and its surrounding medium44,45.

To apply forces in the piconewton range with tens ofmilliwatt power, beads are required that are at least onewavelength in diameter because for smaller beads thetrapping force scales as the third power of the beadradius2. Such beads can then be attached to macromole-cules46,47. There are several unique advantages of opticaltweezers over other external field techniques. First, theradiation pressure will only trap a bead near the focus ofthe laser beam, so the photon field does not simultane-ously affect other beads. Second, the momentum trans-fer between the trapped object and the laser beam caneasily be calibrated against displacement and force, thusproviding a method of direct, high-resolution force and

DNA molecules by changing the flow rate in one direc-tion relative to an orthogonal magnetic force. Theseresults have led to a precise description of DNAelasticity7,9. A recent study of transcriptional pausing andarrest of RNA polymerase (BOX 2) used computer-con-trolled fluid flow, which applied force to sub-piconewtonresolution on active single RNA polymerases36. Theflow-field technique has also been used by Chu and co-

Box 1 | Balancing signal, thermal noise and time resolution

Every object in solution isbombarded constantly bysurrounding molecules. As aresult, a spring-like device suchas a cantilever, a microneedle ora bead in an optical trapexperiences a mean-squaredisplacement noise, <∆x2>,proportional to thetemperature, T, and inversely related to spring stiffness, κ (<∆x2> = k

BT/κ, where k

Bis

the Boltzmann constant). This is the so-called equipartition theorem. If the device islinear, its corresponding mean-square force noise is <∆F 2> = κk

BT. Stiff mechanical

transducers such as cantilevers (κ = 0.06 N m–1) experience larger force fluctuations butsmaller displacement fluctuations than do soft transducers such as a bead in an opticaltrap (κ = 10–4 N m–1).

As it happens, fluctuations are not spread out uniformly over all frequencies. Thespectrum of fluctuations of an object is determined by the proportionality that existsbetween its ability to absorb thermal energy and its ability to dissipate it by friction.This result is embodied in the ‘fluctuation-dissipation theorem’:

In EQN 1, <∆x2(ω)>eq

is the mean-square amplitude of fluctuations per unit frequencyof the device at frequency ω, and γ is the friction coefficient of the device. The ‘cornerfrequency’, ω

c= κ/γ, is the frequency above which the system cannot respond to an

external stimulus. The corner frequency sets a limit to the rate at which processes canbe observed and measured experimentally. A 1-µm diameter bead in a typical opticaltrap has a corner frequency of 1,000 Hz, whereas a commercial cantilever, 100 µm longand 10 µm wide, has a corner frequency of 6,000 Hz and thus data can be gathered sixtimes faster. The figure shows the effect of increased corner frequency on fluctuationdistribution. For the same stiffness κ and bandwidth B, the signal-to-noise ratio of themeasurement will be higher for the transducer having the larger corner frequency ω

c2

(that is, the transducer with the smaller dimensions) than for ωc1

.Measurements are often performed in a narrow band (bandwidth B) around the

frequency of the signal (see figure). Assume that a molecule attached to a transducercan generate a force F. Then, the signal-to-noise ratio (SNR) of the measurement, forB << ω

c, is given by EQN 2:

In general, therefore, the SNR can be increased by decreasing the bandwidth of themeasurement (that is, by averaging the signal over longer times). Decreasing thebandwidth reduces the time resolution of the measurements, however, so this approachis limited by the frequency of the biological event of interest. Note that this ratio isindependent of stiffness2,21. Thus, a soft transducer is not a more sensitive detector thana stiff one: as κ decreases, the noise increases exactly as fast as the sensitivity.

The physical meaning of the SNR expression is the following: the total area under thecurves in the figure is a constant equal to k

BT/κ. As the noise spreads over a larger

frequency range for a transducer with a higher corner frequency, the fraction of totalnoise observed in a given bandwidth (shaded areas) can be reduced by increasing ω

c=

κ/γ, that is, by decreasing γ(see figure). Thus, the SNR and the time resolution of forcemeasurements can only be improved by reducing the dimensions of the transducer18,62.

Am

plitu

de2

Hz–

1 Signal

FrequencyB ωc1 ωc2

Area=kBT/κ

Area=kBT/κ

S/N = F/ √2ykBTB (2)

∆x 2 ( ) eq =2kBT

( c2 + 2)

(1)ω

ω ωγ

RADIATION PRESSURE

The pressure on an object thatarises from photon collisionsrather than from bombardingmolecules.



R E V I E W S

NANOTECHNOLOGY

Any technological developmentthat exceeds standard lower sizelimits of modernmicrofabrication techniques(hundreds of nanometres orless).

ments. Exposure to laser light of a wavelength mostcompatible with biomolecules (λ = 835 nm) stilldecreased the active lifetime of this enzyme from morethan 300 s, when not exposed to laser light, to 89 s at90 mW of laser power48. Also, the data throughput isrelatively low because, unlike in flow fields, only onemolecule is handled at a time.

Manipulation of biological systems with opticaltweezers began with relatively large objects, such asbacteria, yeast and mammalian cells49,50. But opticaltrapping, combined with microsphere handles linkedto molecules of interest, has quickly become a princi-pal tool for manipulating and measuring the force-producing properties of various molecular motors,including kinesin moving along microtubules13, acto-myosin complexes28, RNA polymerase36,48 (BOX 2) andDNA polymerase11 (FIG. 4b). Moreover, optical trappinghas also been used both to measure the elastic proper-ties of DNA47 and to characterize the mechanicalunfolding of proteins52,53.

What next?The future of biomolecular manipulation depends onthree factors: the integration and further developmentof single-molecule techniques; progress in the field ofNANOTECHNOLOGY; and the use of high-throughput sys-tems such as MICROFLUIDICS. These factors will facilitatethe application of single-molecule methods to morecomplex problems, in particular to in vivo systems.

Already the power of integrating SFM imaging andpulling has been demonstrated in a study of bacteri-orhodopsin20. Oesterhelt et al.20 first imaged a crys-talline region of membrane-embedded bacteri-orhodopsin, then pulled on the last two helices, F andG, of the seven bacteriorhodopsin transmembranehelices. As these two helices unfolded and left the mem-brane, the polypeptide chain extended and beganpulling on the next helix pair. In this manner, the whole

position measurement (<1 pN and <10 nm,respectively).Last, the applicable force range for photon fields(10–13–10–10 N) is highly relevant for biological systems.

A major disadvantage of optical tweezers is laserdamage to active biological systems. The negative effectof laser exposure on the working lifetime of biologicalcomplexes has been shown in RNA polymerase experi-

Figure 3 | Geometries of typical single-molecule experiments. a | A single myosin head attached to a microneedle moves along an actin filament. The motion ofthe myosin is observed with a laser by total internal reflection fluorescence microscopy, while the forces are detected by observing the displacement of themicroneedle25. b | DNA tethering a magnetic bead to a point on the glass slide8. A magnetic force, which can be determined by means of Stokes’s law (see text), isapplied perpendicularly to a flow force. The latter, and therefore the total resultant force acting on the molecule, can be determined from the known magnetic force andthe angle between the DNA and the magnetic field, θ. The combined magnetic and flow fields can be used to stretch the DNA more than that achievable with themagnetic field alone. c | A rotating magnetic field is used to under- or overwind double-stranded DNA tethered between the glass slide and a magnetic bead40,41. Theresulting supercoils (plectonemes) can be studied by measuring the displacement of the bead perpendicular to the glass slide as a function of the magnetic force.

Rotating magnetic field

Dig–anti-dig connector

Microneedle

MyosinActinfilaments

Evanescentfield

Glass

Laser in Laser out

θ

Magnetic bead

Magnetic bead

FlowforceDNA

Glass

Magneticforce

Biotin–avidin connector

Biotin–avidin connectors

a b c

Box 2 | What can single-molecule manipulation tell us about biology?

Two recent studies havedemonstrated the ability ofsingle-molecule techniques toelucidate new aspects of enzymekinetics. First, Davenport et al.36,by using a single RNApolymerase (RNAP) moleculemoving along a DNA strandattached to a bead in a flow field,determined that RNAP canoperate in at least two modes,one slow and one fast. The figure illustrates the averaged peak rates of single RNAPmolecules, showing that they can be in a slow or a fast transcription state.

Molecules in the slower mode paused more readily than did the faster molecules, and ahigh correlation between the pausing and the complete stopping of individual moleculeswas observed. Together, these results imply that a temporary pause in transcription is thefirst kinetic step towards a complete halt in activity. Furthermore, they indicate thatidentical enzymes can exist in different microstates with distinct functions, opening thepossibility of yet another level of control over gene expression in the cell.

Second, Strick et al.41 examined the activity of type II topoisomerase (topo II) onDNA attached to magnetic beads and supercoiled by the application of a rotatingmagnetic field (FIG. 3c). Topo II was observed to unwind DNA, which was detected asdiscrete 90-nm jumps in the DNA length. This observation supports the claim thattopo II catalyses the relaxation of two supercoils per molecule of ATP hydrolysed63.However, mechanically extending the DNA strand did not accelerate topo II catalysis.In fact, at saturating ATP concentrations, topo II activity decreased by a factor of threeas the force was raised from 0.3 pN to 5 pN, suggesting that the rate-limiting step of thereaction is directly affected by the applied force.

Cou

nt

0

2

4

6

8

10

0 2 4 6 8 10 12

Average peak rate (bp s–1)

5.5±1.3 bp s–1

9.1±1 bp s–1



R E V I E W S

demonstrated54,55, but not fully exploited. Carbon nan-otubes have also been used as mechanical tweezerscapable of grabbing polystyrene particles56. In fact,these atomically thin tubes may prove to be an idealbuilding block for the next generation of single-mole-cule manipulation devices.

Ultimately, the goal of single-molecule manipulationis to access the machinery of a living cell. Although thetask of characterizing molecular machines andorganelles seems daunting, there has been excitingprogress. Researchers have used microneedles to probecells during cytokinesis57,58, and have adapted force-mapping atomic-force microscopy to study the activityof actin under the membrane of living cells17,59.Magnetic beads may also prove useful, in vivo, as themagnetic field will not interfere with other cellularprocesses. Finally, motor proteins for cell motility areperfect targets for in vivo manipulation and single-mol-ecule studies have already yielded information abouttheir force, efficiency and regulation60,61.

In the near future, scientists may come to see eachcell as an individual with its own set of molecularmachinery. By using methods for manipulating singlemolecules, biologists will be able to investigate thenature of molecular machines one by one, and inferfrom their behaviour those properties common to thepopulation and those corresponding to specific sub-states. Indeed, what Mark Twain observed with cats maybe equally true of biomolecules.

protein was pulled out of the membrane, helix pair byhelix pair, revealing details of the attractive forcesbetween helices and the membrane. Finally, the sameregion was imaged to verify that only one bacteri-orhodopsin molecule had been extracted from themembrane. Although this study illustrates the advan-tage of combining single-molecule techniques, the pos-sibilities abound. For instance, combining single-mole-cule fluorescence with optical tweezers will make itpossible to observe spectroscopic signals in response tomechanically induced changes.

Nanotechnology has yet to be effectively applied tosingle-molecule methods. Attaching nanotubes toSFM tips for improved imaging resolution has been

Figure 4 | Geometries of typical optical-trap single-molecule experiments. a | An opticaltrap measuring the force generated during transcription by a single RNA polymerase molecule(RNAP)54. During transcription, an RNAP bound nonspecifically to a glass slide must thread thetemplate and do work against a load applied by the optical trap through a polystyrene beadattached at the end of the DNA molecule. b | A single-stranded DNA molecule, bound with aprimer, connects a bead fixed at the end of a micropipette and a bead held in the optical trap.A feedback circuit is used to keep the DNA molecule at a fixed tension, F. As the DNAPconverts single-stranded DNA into double-stranded DNA, keeping the tension constantrequires the pipette to adjust its position relative to the optical trap by an amount proportionalto the movement of the enzyme over the template11.

a

RNAP

Polystyrenebead

bF

DNAPdsDNA

ssDNA

PipetteLaser light

DNA

1. Bustamante, C., Smith, S., Liphardt, J. & Smith, D. Single-molecule studies of DNA mechanics. Curr. Opin. Struct.Biol. 10, 279–285 (2000).

2. Svoboda, K. & Block, S. M. Biological applications ofoptical forces. Annu. Rev. Biophys. Biomol. Struct. 23,247–285 (1994).

3. Ludwig, M. et al. AFM, a tool for single-moleculeexperiments. Appl. Phys. Mater. Sci. Process. 68,173–176 (1999).

4. Mehta, A. D., Rief, M. & Spudich, J. A. Biomechanics, onemolecule at a time. J. Biol. Chem. 274, 14517–14520(1999).

5. Sarid, D. in Scanning Force Microscopy: With Applicationsto Electric, Magnetic, and Atomic Forces I–XI, 253 (OxfordUniv. Press, New York, 1991).

6. Carrion-Vazquez, M. et al. Mechanical and chemicalunfolding of a single protein: a comparison. Proc. NatlAcad. Sci. USA 96, 3694–3699 (1999).

7. Bustamante, C., Marko, J. F., Siggia, E. D. & Smith, S.Entropic elasticity of lambda-phage DNA. Science 265,1599–1600 (1994).

8. Smith, S. B., Finzi, L. & Bustamante, C. Direct mechanicalmeasurements of the elasticity of single DNA molecules byusing magnetic beads. Science 258, 1122–1126 (1992). This first single-molecule study of DNA elasticitydemonstrates the combined use of magnetic andflow fields.

9. Marko, J. F. & Siggia, E. D. Stretching DNA.Macromolecules 28, 8759–8770 (1995).

10. Erie, D. A., Yang, G., Schultz, H. C. & Bustamante, C. DNAbending by Cro protein in specific and nonspecific

complexes: implications for protein site recognition andspecificity. Science 266, 1562–1566 (1994).

11. Wuite, G. J., Smith, S. B., Young, M., Keller, D. &Bustamante, C. Single-molecule studies of the effect oftemplate tension on T7 DNA polymerase activity. Nature404, 103–106 (2000).

12. Kishino, A. & Yanagida, T. Force measurements bymicromanipulation of a single actin filament by glassneedles. Nature 334, 74–76 (1988). The basic premise of microneedle manipulation isillustrated here in one of the earliest measurementsof the force from biological molecules using amechanical transducer.

13. Svoboda, K., Schmidt, C. F., Schnapp, B. J. & Block, S. M.Direct observation of kinesin stepping by optical trappinginterferometry. Nature 365, 721–727 (1993).

14. Howard, J., Hudspeth, A. J. & Vale, R. D. Movement ofmicrotubules by single kinesin molecules. Nature 342,154–158 (1989).

15. Ishijima, A., Doi, T., Sakurada, K. & Yanagida, T. Sub-piconewton force fluctuations of actomyosin in vitro.Nature 352, 301–306 (1991).

16. Nakajima, H. et al. Scanning force microscopy of theinteraction events between a single molecule of heavymeromyosin and actin. Biochem. Biophys. Res. Commun.234, 178–182 (1997).

17. Rotsch, C., Jacobson, K. & Radmacher, M. Dimensionaland mechanical dynamics of active and stable edges inmotile fibroblasts investigated by using atomic forcemicroscopy. Proc. Natl Acad. Sci. USA 96, 921–926(1999).

18. Viani, M. B. et al. Small cantilevers for force spectroscopyof single molecules. J. Appl. Phys. 86, 2258–2262 (1999).

19. Müller, D. J., Baumeister, W. & Engel, A. Controlledunzipping of a bacterial surface layer with atomic forcemicroscopy. Proc. Natl Acad. Sci. USA 96, 13170–13174(1999).

20. Oesterhelt, F. et al. Unfolding pathways of individualbacteriorhodopsins. Science 288, 143–146 (2000). In this study of a membrane protein, the authorspresent an elegant example of combining force andimaging SFM.

21. Finer, J. T., Simmons, R. M. & Spudich, J. A. Single myosinmolecule mechanics: piconewton forces and nanometresteps. Nature 368, 113–119 (1994).

22. Thomson, N.H. et al. Protein tracking and detection ofprotein motion using atomic force microscopy. Biophys. J.70, 2421–2431 (1996).

23. Almqvist, N. et al. Methods for fabricating andcharacterizing a new generation of biomimetic materials.Mater. Sci. Eng. C 7, 37–43 (1999).

24. Viani, M. B. et al. Fast imaging and fast force spectroscopyof single biopolymers with a new atomic force microscopedesigned for small cantilevers. Rev. Sci. Instrum. 70,4300–4303 (1999).

25. Kitamura, K., Tokunaga, M., Iwane, A. H. & Yanagida, T. A single myosin head moves along an actin filament withregular steps of 5.3 nanometres. Nature 397, 129–134(1999).

26. Kellermayer, M. S. & Granzier, H. L. Elastic properties ofsingle titin molecules made visible through fluorescent F-actin binding. Biochem. Biophys. Res. Commun. 221,

Links

FURTHER INFORMATION SFM overview | SFM in depth |Microneedle research page | Movies of flow fieldsstretching DNA | Theory of optical tweezers | Buildingoptical tweezers | Background of optical tweezers |Microfluidics applications

MICROFLUIDICS

Microscopic channels etchedinto a surface by modernmicrofabrication techniques forthe purpose of transportingsmall amounts of solution fromone place to another.



R E V I E W S

491–497 (1996).27. Ishijima, A. et al. Multiple- and single-molecule analysis of

the actomyosin motor by nanometer-piconewtonmanipulation with a microneedle: unitary steps and forces.Biophys. J. 70, 383–400 (1996).

28. Essevaz-Roulet, B., Bockelmann, U. & Heslot, F.Mechanical separation of the complementary strands ofDNA. Proc. Natl Acad. Sci USA 94, 11935–11940 (1997).

29. Cluzel, P. et al. DNA: an extensible molecule. Science 271,792–794 (1996).

30. Leger, J. F., Robert, J., Bourdieu, L., Chatenay, D. &Marko, J. F. RecA binding to a single double-strandedDNA molecule: A possible role of DNA conformationalfluctuations. Proc. Natl Acad. Sci. USA 95, 12295–12299(1998).

31. Leger, J. F. et al. Structural transitions of a twisted andstretched DNA molecule. Phys. Rev. Lett. 83, 1066–1069(1999).

32. Oosawa, F. Actin–actin bond strength and theconformational change of F-actin. Biorheology 14, 11–19(1977).

33. Lorentz, H. A. Abhändlungen über Theoretische Physik(Teubner, Leipzig, 1907).

34. Happel, J. B., H. Low Reynolds Number Hydrodynamicswith Special Applications to Particulate Media (PrenticeHall, Englewood Cliffs, New Jersey, 1991).

35. Berg, H. C. Random Walks in Biology (Princeton Univ.Press, New Jersey, 1983).

36. Davenport, R. J., Wuite, G. J., Landick, R. & Bustamante,C. Single-molecule study of transcriptional pausing andarrest by E. coli RNA polymerase. Science 287,2497–2500 (2000).

37. Smith, D. E., Babcock, H. P. & Chu, S. Single-polymerdynamics in steady shear flow. Science 283, 1724–1727(1999).

38. Perkins, T. T., Smith, D. E., Larson, R. G. & Chu, S.Stretching of a single tethered polymer in a uniform flow.Science 268, 83–87 (1995). This paper describes how flow fields can be used tostretch DNA and reveal new facets of single-molecule polymer rheology.

39. Perkins, T. T., Quake, S. R., Smith, D. E. & Chu, S.Relaxation of a single DNA molecule observed by opticalmicroscopy. Science 264, 822–826 (1994).

40. Strick, T. R., Allemand, J. F., Bensimon, D., Bensimon, A.& Croquette, V. The elasticity of a single supercoiled DNAmolecule. Science 271, 1835–1837 (1996).

41. Strick, T. R., Croquette, V. & Bensimon, D. Single-molecule

analysis of DNA uncoiling by a type II topoisomerase.Nature 404, 901–904 (2000). Here, to investigate the activity of individualtopoisomerase molecules, magnetic force is used totwist a molecule of DNA and supercoil it.

42. Ashkin, A., Dziedzic, J., Bjorkholm, J. & Chu, S.Observation of a single-beam gradient force optical trapfor dielectric particles. Optical Lett. 11, 288–290 (1986).

43. Gordon, J. P. Radiation forces and momenta in dielectricmedia. Phys. Rev. A 8, 14–21 (1973).

44. Ashkin, A. Forces of a single-beam gradient laser trap on adielectric sphere in the ray optics regime. Biophys. J. 61,569–582 (1992).

45. Wright, W. H., Sonek, G. J. & Berns, M. W. Parametricstudy of the forces on microspheres held by opticaltweezers. Appl. Optics 33, 1735–1748 (1994).

46. Chu, S. Laser manipulation of atoms and particles.Science 253, 861–866 (1991).

47. Smith, S. B., Cui, Y. & Bustamante, C. Overstretching B-DNA: the elastic response of individual double-strandedand single-stranded DNA molecules. Science 271,795–799 (1996).

48. Yin, H. et al. Transcription against an applied force.Science 270, 1653–1657 (1995).

49. Kuo, S. C. & Sheetz, M. P. Force of single kinesinmolecules measured with optical tweezers. Science 260,232–234 (1993).

50. Ashkin, A. & Dziedzic, J. M. Optical trapping andmanipulation of viruses and bacteria. Science 235,1517–1520 (1987). This landmark paper demonstrates the power ofoptical traps to manipulate microscopic objects.

51. Wang, M. D. et al. Force and velocity measured for singlemolecules of RNA polymerase. Science 282, 902–907(1998). By using an optical trap, the authors reveal aspectsof transcription on a single molecule level.

52. Kellermayer, M. S., Smith, S. B., Granzier, H. L. &Bustamante, C. Folding–unfolding transitions in single titinmolecules characterized with laser tweezers. Science 276,1112–1116 (1997); erratum 277, 1117 (1997).

53. Tskhovrebova, L., Trinick, J., Sleep, J. A. & Simmons, R. M.Elasticity and unfolding of single molecules of the giantmuscle protein titin. Nature 387, 308–312 (1997).

54. Wong, S. S., Joselevich, E., Woolley, A. T., Cheung, C. L. &Lieber, C. M. Covalently functionalized nanotubes asnanometre-sized probes in chemistry and biology. Nature394, 52–55 (1998).

55. Cheung, C. L., Hafner, J. H. & Lieber, C. M. Carbonnanotube atomic force microscopy tips: direct growth bychemical vapor deposition and application to high-resolution imaging. Proc. Natl Acad. Sci. USA 97,3809–3813 (2000).

56. Kim, P. & Lieber, C. M. Nanotube nanotweezers. Science286, 2148–2150 (1999).

57. Skibbens, R. V. & Salmon, E. D. Micromanipulation ofchromosomes in mitotic vertebrate tissue cells: tensioncontrols the state of kinetochore movement. Exp. Cell Res.235, 314–324 (1997).

58. Roy, P., Petroll, W. M., Cavanagh, H. D., Chuong, C. J. &Jester, J. V. An in vitro force measurement assay to studythe early mechanical interaction between cornealfibroblasts and collagen matrix. Exp. Cell Res. 232,106–117 (1997).

59. Rotsch, C., Braet, F., Wisse, E. & Radmacher, M. AFMimaging and elasticity measurements on living rat livermacrophages. Cell Biol. Int. 21, 685–696 (1997).

60. Muramoto, K. et al. High-speed rotation and speedstability of the sodium-driven flagellar motor in Vibrioalginolyticus. J. Mol. Biol. 251, 50–58 (1995).

61. Muramoto, K. et al.Rotational fluctuation of the sodium-driven flagellar motor of Vibrio alginolyticus induced bybinding of inhibitors. J. Mol. Biol. 259, 687–695 (1996).

62. Chand, A., Viani, M. B., Schaffer, T. E. & Hansma, P. K.Microfabricated small metal cantilevers with silicon tip foratomic force microscopy. J. Microelectromech. Sys. 9,112–116 (2000).

63. Osheroff, N., Shelton, E. R. & Brutlag, D. L. DNAtopoisomerase II from Drosophila melanogaster.Relaxation of supercoiled DNA. J. Biol. Chem. 258,9536–9543 (1983).

64. Fisher, T. E., Oberhauser, A. F., Carrion-Vazquez, M.,Marszalek, P. E. & Fernandez, J. M. The study of proteinmechanics with the atomic force microscope. TrendsBiochem. Sci. 24, 379–384 (1999); erratum 25, 6 (2000).

65. Florin, E. L., Moy, V. T. & Gaub, H. E. Adhesion forcesbetween individual ligand–receptor pairs. Science 264,415–417 (1994).

66 Lee, G. U., Chrisey, L. A. & Colton, R. J. Directmeasurement of the forces between complementarystrands of DNA. Science 266, 771–773 (1994).

AcknowledgementsWe thank S. Smith and J. Choy for their helpful comments. Thiswork was supported in part by grants from the NIH and the NSF(to C.B.).


Cellular differentiation is the process by which a cellacquires a new phenotype to accomplish specific func-tions, and it is accompanied by activation of a specificsubset of genes and silencing of the remainder. As genesare silenced, the extent of chromatin condensationincreases, and extended regions of DNA are packaged ina transcriptionally inactive form often referred to asHETEROCHROMATIN1. The amount and distribution of con-densed chromatin is similar in terminally differentiatedcells of the same lineage, but it varies in the nuclei ofdifferent cell types, indicating that nuclear organizationmay be cell-type specific2 (FIG. 1).

These observations have led to the idea that thetopological organization of the INTERPHASE nucleus isrelated to the differentiated state of the cell, and thatthis spatial organization is involved in the establish-ment of the tissue-specific pattern of gene expressionduring cellular differentiation. Clearly, many overlap-ping pathways are involved in regulating gene expres-sion3–6. Much research is focused on determining theproteins (trans-acting factors) and DNA sequences(cis-acting elements) involved in the dynamic localiza-tion of genes within the nucleus, and we are just begin-ning to understand the link between the structure and

NUCLEARCOMPARTMENTALIZATION AND GENE ACTIVITYClaire Francastel*, Dirk Schübeler*, David I. K. Martin‡ and Mark Groudine*§

The regulated expression of genes during development and differentiation is influenced by theavailability of regulatory proteins and accessibility of the DNA to the transcriptional apparatus.There is growing evidence that the transcriptional activity of genes is influenced by nuclearorganization, which itself changes during differentiation. How do these changes in nuclearorganization help to establish specific patterns of gene expression?


*Fred Hutchinson CancerResearch Center, 1,100Fairview Avenue North,Seattle, Washington 98109,USA.‡The Victor Chang CardiacResearch Institute, 384Victoria StreetDarlinghurst, Sydney,New South Wales 2010,Australia.§Department of RadiationOncology, University ofWashington School ofMedicine, Seattle,Washington 98195, USA.Correspondence to M.G.e-mail: [email protected]

Figure 1 | Patterns of chromatin condensation in haematopoietic cells. The electron micrographs show haematopoieticcells from normal human bone marrow82, and illustrate the distinct patterns of chromatin condensation in distinct terminallydifferentiated cells. a | Proerythroblast (immature red blood cell). The nucleus contains almost no heterochromatin. b | Lateerythroblast. Heterochromatin in the nucleus is predominant and appears as darkly staining regions. c | The nucleus of amonocyte (right) appears less condensed than that of a granulocyte (left). (Figure adapted with permission from REF. 82 ©Harcourt Brace, Madrid, Spain.)

HETEROCHROMATIN

A condensed form ofchromatin; the degree ofcompaction is similar to that ofmitotic chromosomes. It isusually found around thecentromere.

INTERPHASE

The period between two mitoticdivisions.

R E V I E W S



R E V I E W S

study the putative coupling of chromatin condensationand gene repression during the differentiation of somat-ic cells. The haematopoietic programme is initiatedfrom a stem cell with self-renewal and differentiationcapacities, which can give rise to several distinct lineages.In multipotent haematopoietic stem cells, many genesrelevant to different lineage fates are transcribed beforethe decision to commit to any one lineage8,9. In otherwords, before the commitment decision, a general ‘per-missive’ state exists in which genes specific to several lin-eages may be in an open chromatin configuration.Moreover, the subsequent activation of genes for anyone lineage is probably coincident with closing of thosegene domains associated with other potential cell fates.Indeed, this relationship may be a general feature ofgene regulation during cellular differentiation. Forexample, microscopic analysis of the developing phar-ynx in the nematode worm Caenorhabditis elegansreveals a uniform structure of EUCHROMATIN in the nucleiof those cells committed to the pharyngeal cell fate(FIG. 2a). However, differentiation of these cells is associ-ated with activation of pharyngeal-specific genes andextensive formation of heterochromatin, indicating thatother domains may be inactivated (FIG. 2b).

So gross changes in chromatin structure accompanycellular differentiation. Chromatin condensation andgene silencing correlate with a change in the nucleosomalstructure characterized by the deacetylation of core his-tones10–13 and accumulation of linker histones14,15(BOX 1).For example, differentiation of embryonic stem cells isaccompanied by a progressive deacetylation of histoneH4 in the heterochromatin around centromeres16.Conversely, acetylation of histone tails marks ‘open’(nuclease-sensitive) domains and is also involved in pro-moter activation (FIG. 3).

Although gene potentiation is necessary for tran-scription, it is not in itself sufficient for full activation ofa promoter. For example, the α-globin locus resides in aregion that is constitutively open, but this gene is nottranscribed in non-erythroid cells17. In haematopoieticcell lines, the nuclease sensitivity of potentially activegenes — for example the β-globin locus and c-myc —correlates with their level of histone H4 acetylation,regardless of their transcriptional activity18,19. Therefore,chromatin opening and histone acetylation could repre-sent the first step of a two-step mechanism to achievegene expression — gene potentiation before transcrip-tional activation of the promoters.

Whether activation of globin genes in normal differ-entiating red blood cells is a multistep process is notclear, but there is evidence that it might be. Studies inmouse erythroleukaemia (MEL) cells, comparing themouse β-globin locus before and after induction of dif-ferentiation, reveal that it can exist in two distinct struc-tural states. The first is a potentiated chromatin struc-ture in cells that are committed to the erythroid lineagebut not terminally differentiated. The second is an acti-vated state seen in terminally differentiating cells20. Inthe potentiated state, the β-globin locus is more sensitiveto nuclease digestion than in non-erythroid cells21,where the nucleosomes are arranged in a regular

function of chromatin and large-scale changes in thenucleus itself.

We propose that tissue-specific ENHANCERS preventactive genes from being included in regions of tran-scriptionally inactive condensed chromatin that formduring cell differentiation. This allows a subset of genesto be active in the appropriate lineage, whereas theremainder are silenced, and ensures, for example, thatglobin genes are expressed only in red blood cells, andimmunoglobulin genes only in B cells.

Gene potentiation during differentiationModulation of cell- and stage-specific gene expressionduring differentiation is thought to involve inactivationof those genes that do not need to be expressed in a givenlineage. Conversely, those sets of genes that need to beactivated are placed or maintained in a decondensed —or ‘potentiated’ — state in precursor cells of that lineage.Potentiation is characterized by an ‘open’ chromatinstructure, meaning that a locus is accessible to the tissue-specific factors required for its appropriate expression7.

Haematopoiesis has been used as a model system to

Box 1 | Nucleosome structure and function

The nucleosome is the fundamental subunit of all chromatin, giving it thecharacteristic ‘beads-on-a-string’ appearance. Each nucleosome is composed of about200 base pairs of DNA coiled roughly twice around an octamer of ‘core’ histoneproteins — two molecules each of histones H2A, H2B, H3 and H4. The ‘linker’ histoneH1 can be associated with each nucleosome; it seems to mediate the packing of adjacentnucleosomes during chromatin condensation.

The structure and function of chromatin can be modulated by acetylation (anddeacetylation) of histones. Acetylation is a covalent modification of lysine residues —addition of an acetyl group — in the histone tails. Acetylated histones are found inregions of decondensed chromatin, whereas deacetylated histones are found in regionsof condensed chromatin. Acetylation is mediated by histone acetyltransferases (HATs),and deacetylation by histone deacetylases (HDACs). Because changes in histoneacetylation are associated with changes in chromatin structures, these enzymes areimportant cofactors in the regulation of gene expression.

Figure 2 | Chromatin condensation accompanies cellular differentiation. Electronmicrographs of Caenorhabditis elegans pharyngeal cells a | before and b | after commitment tothe pharyngeal lineage. a | Cells are committed to give rise to pharyngeal cells, but have not yetundergone terminal differentiation. The nucleus seems to contain almost exclusivelyeuchromatin, with a thin rim of heterochromatin at the nuclear periphery. b | In terminallydifferentiated cells, the nucleus appears more compact and condensed chromatin formsclumps inside the nucleus. (Images courtesy of J. Priess, Fred Hutchinson Cancer ResearchCenter, Seattle, USA.)

ENHANCERS

Increase transcription of alinked promoter if placed ineither orientation, upstream ordownstream.

EUCHROMATIN

Chromatin that appears lesscompact than mitoticchromosomes. Active genes arecontained within euchromatin.



R E V I E W S

Changes in nuclear structure and differentiationAlthough the nucleosomal organization of chromo-somes has been the focus of much study, the higher-order structure of chromosomes in the interphasenucleus is only poorly understood. Recent advances influorescence in situ hybridization (FISH) have allowedindividual genes in the interphase nucleus to be visual-ized and topographically analysed (FIG. 4), contributing tothe increasing awareness that spatial context within thenucleus is important in modulating gene expression24–26.

Two main levels of nuclear organization can be dis-tinguished. First, chromosomes and genes are confinedto discrete nuclear zones within the nuclear volume,referred to as ‘chromosome territories’27–29. These terri-tories occupy non-random positions in the interphasenucleus that are specific to certain cell types30–32. Second,many nuclear functions — such as replication, RNAprocessing and transcription — take place in well-defined compartments in the nucleus33.

How might this higher-order organization regulategene expression? One view is that active and inactiveregions of the genome, as well as protein factorsinvolved in activation or repression of gene expression,are compartmentalized within the nucleus. Nuclease-sensitive regions of chromatin and sites of active tran-scription are located in clusters scattered throughoutthe nucleus33–35. Whereas chromosomes with few genesare often associated with the nuclear periphery, gene-rich chromosomes reside in a more internal nuclearposition36. Genes on the same chromosome are furthercompartmentalized into distinct active and inactivedomains within a chromosome territory: non-codingsequences are found in the interior or randomly distrib-uted in the chromosome territory37,38, whereas poten-tially active genes, regardless of their transcriptionalactivity, are preferentially located at the periphery ofchromosome territories. The surface area of the chro-mosomes interacts with the nuclear space between thechromosome territories, the so-called interchromoso-mal domain, where gene transcription and messengerRNA splicing are thought to occur38,39. Moreover, tran-scriptionally active gene clusters can be found on largechromatin loops extending outwards from the surfaceof the chromosome territory, indicating that genomicsequences may be recruited to environments that arepermissive for transcription40.

If active transcription is concentrated in a fraction ofthe nucleus, it is conceivable that RNA polymerase IIand specific transcription factors would also accumulatein these nuclear compartments for active transcription.Indeed, a significant fraction of active transcription hasbeen shown to colocalize with RNA polymerase II orwith basal transcription factors41. Furthermore, nuclearregions containing transcription factors, as well asTATA-box binding protein (TBP), RNA polymerase IIand nascent RNA, have been described42.

However, a large fraction of transcription-factormolecules are not associated with sites of active tran-scription, nor with RNA polymerase II; the function ofthese sites is not yet clear38. One possibility is that theyare storage sites from which proteins can be recruited. In

(phased) pattern along the β-globin gene. In MEL cells,however, phasing is disrupted both before and afterinduction of differentiation, and this correlates with anincreased generalized sensitivity to DNaseI (REF. 22).Moreover, in uninduced cells, the non-transcribedβ-globin gene contains a hypersensitive site in the inter-vening sequence II, indicating the binding of a non-his-tone protein. In induced cells, however, transcription ofthe β-globin gene is associated with the appearance of ahypersensitive site at the promoter and an increase ingeneral sensitivity to nucleases21,23.

How are tissue-specific genes maintained in an openchromatin configuration before promoter activation?Although the molecular mechanisms behind thisdynamic alteration of chromatin are not clear, a facet ofchromatin opening and gene activation that has becomeapparent only in recent years deserves consideration —namely that active genes are localized in nuclear com-partments that are permissive for transcription.

Figure 3 | Chromatin structure and gene expression. a | Organization of chromatin into thetightly condensed 30-nm fibre and ‘beads-on-a-string’ 10-nm fibre. b | Region of DNAcontaining an actively transcribed gene. The region encompassing the transcribed geneprobably adopts the 10-nm configuration, whereas sequences up- and downstream areorganized into a 30-nm fibre. Transcription involves changes in the chromatin structure that canbe detected by digestion with the endonuclease DNaseI. The transcribed regions also containhypersensitive (HS) sites, which may be associated with a small stretch of DNA devoid ofnucleosomes or simply reflect a different nucleosomal structure (positioning or histonemodification). They usually correlate with non-histone proteins bound to the DNA. Regions ofgeneral DNaseI sensitivity correlate with acetylation of the histone tails, defining a structural andfunctional ‘domain’ for gene activity that is characteristic of ‘open’ chromatin. This domain isembedded in regions of higher chromatin compaction that are more resistant to DNaseIdigestion. This compacted structure correlates with the presence of deacetylated histones(blue wavy lines), and is characteristic of ‘closed’ chromatin. (Figure adapted from REF. 83.)



R E V I E W S

which is found mainly around the centromeres, stayscondensed. The position of the centromeres in inter-phase nuclei seems to be cell-type specific, and it canchange during the cell cycle, during differentiation, orwith cellular transformation30,50–52. In mammalian cells,centromeres tend to cluster in so-called CHROMOCENTRES.The specific combinations of centromeres found inchromocentres are different in fibroblasts andhaematopoietic cells52. Moreover, they also seem to bedistinct in mature cells of different haematopoieticlineages. Just as the organization and location of cen-tromeres change with various cellular states, so adynamic positioning of chromosomes, specific chromo-somal domains or trans-acting factors has also beenobserved during cell-cycle progression, differentiationand malignant transformation51–58. For example, duringnerve growth factor (NGF)-induced differentiation ofrat pheochromocytoma cells, changes in gene expres-sion correlate with redistribution of DNaseI hypersensi-tive domains34, and, in hormone-induced differentiationof pre-adipocytes, the trans-acting factor C/EBP is relo-cated into discrete nuclear foci59.

Nuclear compartmentalizationDynamic nuclear architecture could reflect changes inthe physical position of tissue-specific genes during celldifferentiation, which affect the accessibility of genes toregulatory factors and the transcriptional machinery.This, in turn, may facilitate nuclear functions such asgene transcription or silencing.

Experimental evidence supports the idea that posi-tioning of a gene near heterochromatin compartmentspromotes gene silencing. For example, in Drosophilamelanogaster the insertion of a block of heterochro-matin into one allele of the euchromatic brown generesults in the association of both alleles with centromer-ic heterochromatin, leading to transcriptional inactiva-tion of the wild-type allele60–62. During differentiation ofmouse lymphocytes, gene silencing is heritable andassociated with repositioning of genes close to cen-tromeres45,63. In addition, the human β-globin genedomain, in its native context, can be found in distinctlocations in the interphase nucleus relative to the hete-rochromatin compartment19. Its positioning in the ery-throid nucleus correlates with the chromatin and generalacetylation configurations of the locus, but not withtranscription of the β-globin genes19. Together with thefact that suppression of transgene silencing and mainte-nance of its open chromatin configuration require dis-tance away from centromeric heterochromatin, theseresults indicate that stably inherited chromatin openingof a locus might be mediated by its sequestration in apermissive compartment. These findings also argueagainst the idea that nuclear relocation is the conse-quence of transcription.

Other silencing systems — such as the Sir-depen-dent, telomeric and mating-type-locus silencing in yeast— are associated with compartmentalization of thesilenced gene. Telomeric silencing occurs at the nuclearenvelope in compartments where telomeres are clus-tered and the concentration of Sir proteins is high64–68.

this case, they may be specific for particular transcriptionfactors, as most of the factors investigated occupy non-overlapping sites. Interestingly, a subset of transcription-factor-rich zones colocalize in the nuclei of neurons inthe hippocampus, and this may reflect their coordinatedaction on genes involved in neuronal excitability43.

Factors involved in gene silencing have also beenshown to colocalize with each other and with constitu-tive heterochromatin. These factors include proteinsassociated with heterochromatin (heterochromatin pro-tein 1 (HP1), SUV39H1 and Ikaros)44–46, METHYL-DNA-BIND-

ING DOMAIN PROTEINS (MBDs) and histone deacetylases(HDACs). For example, MBD proteins associate withmajor SATELLITES in embryonic stem cells47,48, and HDACs(BOX 1) colocalize with centromeric DNA49.

Current evidence supports the idea that the eukary-otic nucleus is functionally divided into heterochro-matin compartments that repress transcription, andcompartments in which transcription is permitted.How are these discrete compartments and the specificspatial relationship between them established to pro-duce a precise pattern of gene expression?

Some evidence indicates that centromeres and CON-

STITUTIVE HETEROCHROMATIN provide a structural frame-work for this nuclear architecture, and that changes inthis architecture are involved in cellular diversification.During interphase, the constitutive heterochromatin,

Figure 4 | Fluorescence in situ hybridization to studygene localization. a | Chromosome from a metaphasespread. Centromeric sequences are detected with a Texas-red-conjugated probe (red), and a transgene — integrated inthe genome of a human erythroleukaemia cell line — islocalized with a fluorescein isothiocyanate-conjugated probe(green). The transgene is integrated on the long arm of thechromosome distant from its centromere. b | Fluoresence insitu hybridization experiments in interphase cells reveal thatthe transgene is in close proximity to centromericheterochromatin in silent cells (bottom cell), but it is foundaway from this compartment in expressing cells (upper cell).Silencing of the transgene is therefore accompanied by itsrelocalization, in the interphase nucleus, close to theheterochromatin compartment70.

METHYL-DNA-BINDING

DOMAIN PROTEINS

Proteins that bind specifically tomethylated DNA through amethyl-DNA-binding domain.Some of these proteins areinvolved in transcriptionalrepression of methylated DNA.

SATELLITES

Relatively short DNA sequencesthat are highly repeated in longtandem arrays.

CONSTITUTIVE

HETEROCHROMATIN

The fraction ofheterochromatin that stayscompact through the cell cycle.It is mainly composed ofrepetitive sequences (satelliteDNA; see above), and isconcentrated in characteristicregions such as centromeres.

CHROMOCENTRES

Aggregates of constitutiveheterochromatin from differentchromosomes.



R E V I E W S

Role of cis-acting elementsFinally we can ask how modifications of the chromatinstructure are achieved over broad regions71, and whatsequences are involved in mediating an open or closedstructure. Sequences that mediate silencing have beendescribed in yeast and Drosophila, and the observationsindicate that the formation of an inactive structure mayinvolve assembly of proteins that initially bind toSILENCER ELEMENTS. Such a repressive structure may thenpropagate along the chromatin.

Mating-type silencing in yeast involves the nucle-ation of a repressive chromatin structure at silencer ele-ments, with subsequent spreading of this structure.Interestingly, silencing at the mating-type loci is associ-ated with both a silenced chromatin structure and local-ization of silenced regions near the nuclear periphery.This indicates that nuclear compartments and thespreading of specific chromatin structures are notincompatible phenomena but that both define thesilenced state66. Telomeric silencing in yeast involvessimilar spreading of silenced chromatin, mediated bySir proteins, and is associated with histone hypoacetyla-tion25,72. Similar nucleation and spreading effects arethought to account for silencing of genes located in ornear heterochromatin in higher eukaryotes and forPolycomb-mediated gene silencing in Drosophila — asystem required for maintaining the inactive state ofgenes containing a Polycomb response element (PRE).So far, the PRE is the only clear example of a cis-elementresponsible for silencing in higher eukaryotes73,74.

Numerous studies have shown that enhancers can

Furthermore, positioning of a locus to the nuclearperiphery can provoke Sir-mediated silencing69.

In centromeric silencing, by contrast, there is nodirect evidence that association with centromeric hete-rochromatin is the cause of silencing, nor is there evi-dence that localization away from heterochromatin isrequired for transcriptional activity. For example, anactive transgene may be located close to centromeric het-erochromatin. However, transcription and the openchromatin structure of a transgene are unstable in thiscompartment70. Conversely, a silent transgene can belocated away from centromeric heterochromatin70, andtransient gene silencing during differentiation of B-celllines is not associated with relocation close to cen-tromeres63.

So although there is evidence that positioning tospecific compartments may affect transcriptional activi-ty or gene silencing, such compartmentalization doesnot seem to be an absolute requirement. In general,however, stably inherited gene activity and open chro-matin configurations seem to be associated with posi-tioning away from centromeric heterochromatin,whereas stably inherited gene silencing requires posi-tioning close to centromeric heterochromatin. If thereare indeed specific mechanisms to place and keep genesin a silent compartment, the potentiated state would bedetermined by the positioning of a gene in an activecompartment of the nucleus. This, in turn, would allowtissue-specific transcription factors to induce the pre-cise programmes of stable gene expression typical ofdifferentiated cells.

SILENCER ELEMENTS

Cis-acting elements that areinvolved in silencing, mostprobably by directly recruitingrepressive proteins.

Figure 5 | Activation at the β-globin locus: a multistep process? Chromatin opening and transcriptional activation at the β-globin locus can be dissociated80,81, indicating that they are achieved through distinct mechanisms. First, an open chromatinconfiguration, marked by a locus-wide acetylation, is mediated by positioning of the locus away from centromericheterochromatin. The resulting open chromatin structure favours the binding of tissue-specific trans-acting factors to enhancersand the promoter, concomitant with a local hyperacetylation of histone H3 and promoter activation. Although sequences outsidethe locus control region (LCR) are sufficient to promote the first step, gene activation requires an intact LCR. The demonstrationthat the β-globin locus can adopt distinct structural correlates, together with the demonstration that erythroid differentiation in amodel system is a multistep process, supports the idea that activation of globin genes in normal differentiating red blood cells isalso a multistep process. The configuration of the chromatin structure in a multipotent stem cell, as well as the structure andlocalization of the β-globin locus in precursor cells, remains to be determined. (HDACs, histone deacetylases; HP1,heterochromatin protein 1; Pc, Polycomb, HATs, histone acetyltransferases.)



R E V I E W S

positive transcriptional regulators in the heterochromat-ic compartment, and from a high concentration of neg-ative regulators. Perhaps the simplest mechanism foractivation by regulatory elements is one in which tran-scriptional activators bound to regulatory elements dis-rupt local interactions between a locus and heterochro-matin, permitting the gene to move into the activecompartment.

Future directionsThe chromosomal and nuclear position of a gene caninfluence its activity, and the position of a gene within thenucleus can be dictated by the cis-acting sequences linkedto it. Tissue-specific enhancers may prevent a linked genefrom being included in FACULTATIVE HETEROCHROMATIN, whichforms during cell differentiation, therefore allowing it tobe active in the appropriate lineage. If the precise topolog-ical conformation of genes within the nucleus is disrupt-ed, the result may be phenotypic changes or disease. Forexample, cancer involves disruption of stable patterns ofgene expression, and can be accompanied by a global dis-organization of chromosomal positioning in the nucle-us78. Chromosomal abnormalities and ANEUPLOIDY, whichoften affect nuclear architecture, can influence geneexpression not only of the affected chromosome, but alsoof nearby chromosomal regions79. So investigations intolinks between structural and functional aspects of inter-phase chromosomes promise to reveal fundamentalaspects of biology, as well as insights into the developmentof disease.

counteract such silencing events75–77. The demonstra-tion that an enhancer is sufficient for both localizationof a transgene away from centromeric heterochromatinand suppression of transgene silencing indicates thatenhancers may maintain gene expression by preventinglocalization close to the repressive heterochromaticcompartment70 (FIG. 5).

The situation in a multigene endogenous locusseems to be more complex when compared with thesimplified structure of a transgene. The human β-glo-bin LOCUS CONTROL REGION (LCR), although essential fortranscriptional activation, is not required to relocate thenative locus away from heterochromatin19. The apparentparadox between results obtained with the native locusand those with transgenes can be explained by the exis-tence of numerous factor-binding sites throughout thenative locus, which probably alter subnuclear locationand chromatin structure. So cis-acting elements otherthan the LCR, but with similar function, may maintainthe β-globin locus in an open chromatin/acetylated con-figuration, localized away from centromeres.

There are various means by which genetic regulatoryelements, such as enhancers and LCRs, could cause agene to localize away from heterochromatin. Initial tran-scription-factor binding to the enhancer might promoterecruitment of a locus to a nuclear compartmentenriched in chromatin remodelling complexes and his-tone acetyltransferases (HATs; for example, HAT1), aswell as other elements of the transcriptional machinery.

Just as there are transcription-factor-rich sites, thereare regions of the nucleus that are devoid of, or containlow concentrations of, such factors. These areas mightbe the default location of a gene or transgene if tran-scription factors do not bind to its enhancer elements.Alternatively, the lack of a suitable transcription factorcould result in failure of a gene to be targeted to a tran-scription-rich region of the nucleus. The instability ofgene expression associated with localization near hete-rochromatin may result from this low concentration of

Links

DATABASE LINKS Histone H4 | α-globin | β-globin | c-myc | TBP | RNA polymerase II | HP1 | SUVAR39H1 |Ikaros | MBDs | HDACs | C/EBP | brown | HAT1FURTHER INFORMATION Polycomb-mediated genesilencingENCYCLOPEDIA OF LIFE SCIENCES Nucleosomes:detailed structure and mutations

1. John, B. The Biology of Heterochromatin (ed. Verma, R. S.) (Cambridge Univ. Press, Cambridge, 1988).

2. Leitch, A. R. Higher levels of organization in the interphasenucleus of cycling and differentiated cells. Microbiol. Mol.Biol. Rev. 64, 138–152 (2000).

3. Felsenfeld, G. Chromatin as an essential part of thetranscriptional mechanism. Nature 355, 219–224 (1992).

4. Gregory, P. D. & Horz, W. Chromatin and transcription —how transcription factors battle with a repressivechromatin environment. Eur. J. Biochem. 251, 9–18(1998).

5. Kadonaga, J. T. Eukaryotic transcription: An interlacednetwork of transcription factors and chromatin-modifyingmachines. Cell 92, 307–313 (1998).

6. Wolffe, A. P. Transcription: In tune with the histones. Cell77, 13–16 (1994).

7. Kramer, J. A., McCarrey, J. R., Djakiew, D. & Krawetz, S. A. Differentiation: The selective potentiation ofchromatin domains. Development 125, 4749–4755 (1998).

8. Jimenez, G., Griffiths, S. D., Ford, A. M., Greaves, M. F. &Enver, T. Activation of the β-globin locus control regionprecedes commitment to the erythroid lineage. Proc. NatlAcad. Sci. USA 89, 10618–10622 (1992).

9. Hu, M. et al. Multilineage gene expression precedescommitment in the hemopoietic system. Genes Dev. 11,774–785 (1997).

10. Lee, D. Y., Hayes, J. J., Pruss, D. & Wolffe, A. P. A positiverole for histone acetylation in transcription factor access to

nucleosomal DNA. Cell 72, 73–84 (1993).11. Wolffe, A. P. & Pruss, D. Targeting chromatin disruption:

Transcription regulators that acetylate histones. Cell 84,817–819 (1996).

12. Turner, B. M. Histone acetylation as an epigeneticdeterminant of long-term transcriptional competence.Cell. Mol. Life Sci. 54, 21–31 (1998).

13. Wolffe, A. P. Packaging principle: How DNA methylationand histone acetylation control the transcriptional activityof chromatin. J. Exp. Zool. 282, 239–244 (1998).

14. Crane-Robinson, C. How do linker histones mediatedifferential gene expression? Bioessays 21, 367–371(1999).

15. Wolffe, A. P., Khochbin, S. & Dimitrov, S. What do linkerhistones do in chromatin? Bioessays 19, 249–255 (1997).

16. Keohane, A. M., O’Neill L, P., Belyaev, N. D., Lavender, J. S. & Turner, B. M. X-inactivation and histone H4acetylation in embryonic stem cells. Dev. Biol. 180,618–630 (1996).

17. Vyas, P. et al. Cis-acting sequences regulating expressionof the human α-globin cluster lie within constitutively openchromatin. Cell 69, 781–793 (1992).

18. O’Neill, L. P. & Turner, B. M. Histone H4 acetylationdistinguishes coding regions of the human genome fromheterochromatin in a differentiation-dependent buttranscription-independent manner. EMBO J. 14,3946–3957 (1995).

19. Schübeler, D. et al. Nuclear localization and histone

acetylation: A pathway for chromatin opening andtranscriptional activation of the human β-globin locus.Genes Dev. 14, 940–950 (2000).Indicates that positioning away from theheterochromatin compartment may mediate openchromatin configuration and histone acetylation ofthe human globin locus.

20. Reddy, P. M. & Shen, C. K. Erythroid differentiation ofmouse erythroleukemia cells results in reorganization ofprotein–DNA complexes in the mouse β-major globinpromoter but not its distal enhancer. Mol. Cell. Biol. 13,1093–1103 (1993).

21. Cohen, R. B. & Sheffery, M. Nucleosome disruptionprecedes transcription and is largely limited to thetranscribed domain of globin genes in murineerythroleukemia cells. J. Mol. Biol. 182, 109–129 (1985).

22. Benezra, R., Cantor, C. R. & Axel, R. Nucleosomes arephased along the mouse β-major globin gene in erythroidand nonerythroid cells. Cell 44, 697–704 (1986).

23. Sheffery, M., Rifkind, R. A. & Marks, P. A. Murineerythroleukemia cell differentiation: DNase Ihypersensitivity and DNA methylation near the globingenes. Proc. Natl Acad. Sci. USA 79, 1180–1184 (1982).

24. Lamond, A. I. & Earnshaw, W. C. Structure and function inthe nucleus. Science 280, 547–553 (1998).

25. Cockell, M. & Gasser, S. M. Nuclear compartments andgene regulation. Curr. Opin. Genet. Dev. 9, 199–205(1999).

LOCUS CONTROL REGION

Defined by its ability, intransgenic assays, to conferhigh-level, tissue-specificexpression on a linkedpromoter, at all integrationsites.

FACULTATIVE

HETEROCHROMATIN

Fraction of chromatin that iscondensed and inactive in agiven cell lineage, which may bedecondensed and active inanother.

ANEUPLOIDY

The ploidy of a cell refers to thenumber of sets ofchromosomes that it contains.Aneuploid karyotypes are thosewhose chromosomecomplements are not a simplemultiple of the haploid set.



R E V I E W S

26. Marshall, W. F., Fung, J. C. & Sedat, J. W. Deconstructingthe nucleus: Global architecture from local interactions.Curr. Opin. Genet. Dev. 7, 259–263 (1997).

27. Munkel, C. et al. Compartmentalization of interphasechromosomes observed in simulation and experiment. J. Mol. Biol. 285, 1053–1065 (1999).

28. Manuelidis, L. A view of interphase chromosomes.Science 250, 1533–1540 (1990).

29. Zink, D. et al. Structure and dynamics of humaninterphase chromosome territories in vivo. Hum. Genet.102, 241–251 (1998).

30. Manuelidis, L. Different central nervous system cell typesdisplay distinct and nonrandom arrangements of satelliteDNA sequences. Proc. Natl Acad. Sci. USA 81,3123–3127 (1984).One of the first papers showing that the spatialorganization of centromeres is non-random andcell-type specific, and indicating that this couldrepresent specific functional capacities.

31. Manuelidis, L. & Borden, J. Reproduciblecompartmentalization of individual chromosome domainsin human CNS cells revealed by in situ hybridization andthree-dimensional reconstruction. Chromosoma 96,397–410 (1988).

32. Haaf, T. & Schmid, M. Chromosome topology inmammalian interphase nuclei. Exp. Cell Res. 192,325–332 (1991).

33. Sadoni, N. et al. Nuclear organization of mammaliangenomes. Polar chromosome territories build upfunctionally distinct higher order compartments. J. CellBiol. 146, 1211–1226 (1999).

34. Park, P. C. & De Boni, U. Transposition of DNasehypersensitive chromatin to the nuclear peripherycoincides temporally with nerve growth factor-induced up-regulation of gene expression in PC12 cells. Proc. NatlAcad. Sci. USA 93, 11646–11651 (1996).

35. de Graaf, A. et al. Three-dimensional distribution of DNaseI-sensitive chromatin regions in interphase nuclei ofembryonal carcinoma cells. Eur. J. Cell Biol. 52, 135–141(1990).

36. Croft, J. A. et al. Differences in the localization andmorphology of chromosomes in the human nucleus. J. Cell Biol. 145, 1119–1131 (1999).Analyses the influence of the cell cycle and theinhibition of transcription on the morphology anddistribution of specific human chromosometerritories.

37. Kurz, A. et al. Active and inactive genes localizepreferentially in the periphery of chromosome territories. J. Cell Biol. 135, 1195–1205 (1996).

38. Verschure, P. J., van Der Kraan, I., Manders, E. M. & vanDriel, R. Spatial relationship between transcription sitesand chromosome territories. J. Cell Biol. 147, 13–24(1999).Shows that transcription is compartmentalized, andthat sites of active transcription are distinct fromsites of inactive chromatin.

39. Zirbel, R. M., Mathieu, U. R., Kurz, A., Cremer, T. & Lichter,P. Evidence for a nuclear compartment of transcriptionand splicing located at chromosome domain boundaries.Chromosome Res. 1, 93–106 (1993).

40. Volpi, E. V. et al. Large-scale chromatin organization of themajor histocompatibility complex and other regions ofhuman chromosome 6 and its response to interferon ininterphase nuclei. J. Cell Sci. 113, 1565–1576 (2000).This paper finds a correlation between the formationof large external chromatin loops, outside achromosome territory, and the transcriptionalactivity of a gene cluster.

41. Grande, M. A., van der Kraan, I., de Jong, L. & van Driel,R. Nuclear distribution of transcription factors in relation tosites of transcription and RNA polymerase II. J. Cell Sci.110, 1781–1791 (1997).

42. Pombo, A. et al. Regional and temporal specialization inthe nucleus: A transcriptionally active nuclear domain richin PTF, Oct1 and PIKA antigens associates with specificchromosomes early in the cell cycle. EMBO J. 17,1768–1778 (1998).

43. van Steensel, B. et al. Partial colocalization ofglucocorticoid and mineralocorticoid receptors in discrete

compartments in nuclei of rat hippocampus neurons. J.Cell Sci. 109, 787–792 (1996).

44. Aagaard, L. et al. Functional mammalian homologues ofthe Drosophila PEV-modifier Su(var)3-9 encodecentromere-associated proteins which complex with theheterochromatin component M31. EMBO J. 18,1923–1938 (1999).

45. Brown, K. E. et al. Association of transcriptionally silentgenes with Ikaros complexes at centromericheterochromatin. Cell 91, 845–854 (1997).Ikaros, a protein associated with heterochromatinfoci in the nucleus of mouse B lymphocytes, is alsoassociated with transcriptionally inactive genes,indicating a different mechanism for gene silencing.

46. Wreggett, K. A. et al. A mammalian homologue ofDrosophila heterochromatin protein 1 (HP1) is acomponent of constitutive heterochromatin. Cytogenet.Cell Genet. 66, 99–103 (1994).

47. Hendrich, B. & Bird, A. Identification and characterizationof a family of mammalian methyl-CpG binding proteins.Mol. Cell. Biol. 18, 6538–6547 (1998).

48. Lewis, J. D. et al. Purification, sequence, and cellularlocalization of a novel chromosomal protein that binds tomethylated DNA. Cell 69, 905–914 (1992).

49. Kim, J. et al. Ikaros DNA-binding proteins direct formationof chromatin remodeling complexes in lymphocytes.Immunity 10, 345–355 (1999).Ikaros can recruit histone deacetylases andchromatin remodelling complexes to regions ofheterochromatin in the nucleus of mouse Tlymphocytes.

50. Shelby, R. D., Hahn, K. M. & Sullivan, K. F. Dynamic elasticbehavior of α-satellite DNA domains visualized in situ inliving human cells. J. Cell Biol. 135, 545–557 (1996).

51. Vourc’h, C., Taruscio, D., Boyle, A. L. & Ward, D. C. Cellcycle-dependent distribution of telomeres, centromeres,and chromosome-specific subsatellite domains in theinterphase nucleus of mouse lymphocytes. Exp. Cell Res.205, 142–151 (1993).

52. Alcobia, I., Dilao, R. & Parreira, L. Spatial associations ofcentromeres in the nuclei of hematopoietic cells: Evidencefor cell-type-specific organizational patterns. Blood 95,1608–1615 (2000).

53. Manuelidis, L. Indications of centromere movement duringinterphase and differentiation. Ann. NY Acad. Sci. 450,205–221 (1985).

54. Minc, E., Allory, Y., Worman, H. J., Courvalin, J. C. &Buendia, B. Localization and phosphorylation of HP1proteins during the cell cycle in mammalian cells.Chromosoma 108, 220–234 (1999).

55. Kozubek, S. et al. Distribution of ABL and BCR genes incell nuclei of normal and irradiated lymphocytes. Blood89, 4537–4545 (1997).

56. Neves, H., Ramos, C., da Silva, M. G., Parreira, A. &Parreira, L. The nuclear topography of ABL, BCR, PML,and RARα genes: Evidence for gene proximity in specificphases of the cell cycle and stages of hematopoieticdifferentiation. Blood 93, 1197–1207 (1999).

57. Bartova, E. et al. Nuclear topography of the c-myc gene inhuman leukemic cells. Gene 244, 1–11 (2000).

58. Bridger, J. M., Boyle, S., Kill, I. R. & Bickmore, W. A. Re-modelling of nuclear architecture in quiescent andsenescent human fibroblasts. Curr. Biol. 10, 149–152(2000).

59. Tang, Q. Q. & Lane, M. D. Activation and centromericlocalization of CCAAT/enhancer-binding proteins duringthe mitotic clonal expansion of adipocyte differentiation.Genes Dev. 13, 2231–2241 (1999).

60. Dernburg, A. F. et al. Perturbation of nuclear architectureby long-distance chromosome interactions. Cell 85,745–759 (1996).

61. Csink, A. K. & Henikoff, S. Large-scale chromosomalmovements during interphase progression in Drosophila.J. Cell Biol. 143, 13–22 (1998).

62. Csink, A. K. & Henikoff, S. Genetic modification ofheterochromatic association and nuclear organization inDrosophila. Nature 381, 529–531 (1996).

63. Brown, K. E., Baxter, J., Graf, D., Merkenschlager, M. &Fisher, A. G. Dynamic repositioning of genes in thenucleus of lymphocytes preparing for cell division. Mol.

Cell 3, 207–217 (1999).64. Aparicio, O. M., Billington, B. L. & Gottschling, D. E.

Modifiers of position effect are shared between telomericand silent mating-type loci in S. cerevisiae. Cell 66,1279–1287 (1991).

65. Chien, C. T., Buck, S., Sternglanz, R. & Shore, D.Targeting of SIR1 protein establishes transcriptionalsilencing at HM loci and telomeres in yeast. Cell 75,531–541 (1993).

66. Gotta, M. & Gasser, S. M. Nuclear organization andtranscriptional silencing in yeast. Experientia 52,1136–1147 (1996).

67. Laroche, T. et al. Mutation of yeast Ku genes disrupts thesubnuclear organization of telomeres. Curr. Biol. 8,653–656 (1998).

68. Maillet, L. et al. Evidence for silencing compartmentswithin the yeast nucleus: A role for telomere proximity andSir protein concentration in silencer-mediated repression.Genes Dev. 10, 1796–1811 (1996).

69. Andrulis, E. D., Neiman, A. M., Zappulla, D. C. &Sternglanz, R. Perinuclear localization of chromatinfacilitates transcriptional silencing Nature 394, 592–595(1998).

70. Francastel, C., Walters, M. C., Groudine, M. & Martin, D. I.A functional enhancer suppresses silencing of a transgeneand prevents its localization close to centrometricheterochromatin. Cell 99, 259–269 (1999).Stable transgene expression requires positioningaway from the heterochromatin compartment andthe binding of transcription factors to the enhancer.

71. Bulger, M. & Groudine, M. Looping versus linking: Towarda model for long-distance gene activation. Genes Dev. 13,2465–2477 (1999).

72. Lustig, A. J. Mechanisms of silencing in Saccharomycescerevisiae. Curr. Opin. Genet. Dev. 8, 233–239 (1998).

73. Moehrle, A. & Paro, R. Spreading the silence: epigenetictranscriptional regulation during Drosophila development.Dev. Genet. 15, 478–484 (1994).

74. Loo, S. & Rine, J. Silencing and heritable domains of geneexpression. Annu. Rev. Cell Dev. Biol. 11, 519–548 (1995).

75. Aparicio, O. M. & Gottschling, D. E. Overcoming telomericsilencing: a trans-activator competes to establish geneexpression in a cell cycle-dependent way. Genes Dev. 8,1133–1146 (1994).

76. Festenstein, R. et al. Locus control region function andheterochromatin-induced position effect variegation.Science 271, 1123–1125 (1996).

77. Walters, M. C. et al. Transcriptional enhancers act in cis tosuppress position-effect variagation. Genes Dev. 10,185–195 (1996).

78. Linares-Cruz, G. et al. p21WAF-1 reorganizes the nucleusin tumor suppression. Proc. Natl Acad. Sci. USA 95,1131–1135 (1998).Global disorganization of nuclear architecture canbe associated with cancer, but is reversible duringtumour suppression.

79. Qumsiyeh, M. B. Impact of rearrangements on functionand position of chromosomes in the interphase nucleusand on human genetic disorders. Chromosome Res. 3,455–465 (1995).

80. Epner, E. et al. The β-globin LCR is not necessary for anopen chromatin structure or developmentally regulatedtranscription of the native mouse β-globin locus. Mol. Cell2, 447–455 (1998).

81. Reik, A. et al. The locus control region is necessary forgene expression in the human β–globin locus but not themaintenance of an open chromatin structure in erythroidcells. Mol. Cell. Biol. 18, 5992–6000 (1998).

82. Rozman, C., Woessner, S., Feliu, E., Lafuente, R. & Berga,L. Cell Ultrastructure for Hematologists (ed. EdicionesDoyma, S. A.) (Barcelona, Spain, 1993).

83. Alberts, B. et al. Molecular Biology of the Cell 3rd edn(Garland, New York, 1994).

AcknowledgementsThe authors thank Matthew Lorincz and Bas van Steensel for theiruseful comments on this manuscript. This work was supported bya special fellowship to C.F. from the Leukemia and LymphomaSociety, a fellowship from the Deutsche Forschungsgemeinschaftto D.S., and NIH grants to M.G.


PERSPECTIVES

tide bond, to the carboxy-terminal glycine ofanother ubiquitin. This reaction can be con-tinued to form a multi-ubiquitin chain,which is the signal for degradation of the tar-get protein by the 26S proteasome.

Until 1942, biochemists accepted thedogma, laid down by Folin in 1905 on thebasis of urinary metabolites, that poorlyunderstood macromolecules known as‘endogenous proteins’were stable. Folin mighthave understood the error of this suppositionhad he read Lewis Carroll’s Through theLooking Glass1, in which the Red Queen (FIG. 2)

states:“Now, here, you see, it takes all the run-ning you can do, to keep in the same place. Ifyou want to get somewhere else, you must runtwice as fast as that.” Biological systems are

constantly turning over molecules to maintaina steady state, a fact that became apparent forproteins with the discovery of 15N. By measur-ing the uptake of this label into proteins,Schoenheimer2 (working at Harvard) discov-ered that proteins are continuously synthe-sized and degraded in cells (see TIMELINE).

But what are the mechanisms of synthesisand degradation? The discovery of the path-ways by which proteins are degraded in cellswas held back because more emphasis wasplaced on protein synthesis than on proteindegradation — not least because it is mucheasier to measure the incorporation of radio-labelled amino acids into newly synthesizedproteins than to measure the loss of a label ina pulse–chase analysis. It was not until the endof the 1970s that such general approacheswere recognized as being limited: a commentat a Ciba Foundation Symposium by GünterBlobel3 spoke volumes:“We have heard manystudies here measuring trichloroacetic acidsoluble counts, but this is a rather prehistoricmethod of looking at proteins now!”

Discovery of ubiquitin Our story begins only two years before Blobelmade his incisive remarks at the Ciba

It is often the case in biology that researchinto breaking things down lags behindresearch into synthesizing them, and this iscertainly true for intracellular proteolysis.Now that we recognize that intracellularproteolysis, triggered by attaching multiplecopies of a small protein called ubiquitin totarget proteins, is fundamental to life, it ishard to believe that 20 years ago this fieldwas little more than a backwater ofbiochemistry studied by a handful oflaboratories. Among the few were AvramHershko, Aaron Ciechanover andAlexander Varshavsky, who were recentlyawarded the Albert Lasker award for basicmedical research for discovering theimportance of protein degradation incellular physiology. This Timeline traceshow they and their collaborators triggeredthe rapid movement of ubiquitin-mediatedproteolysis to centre stage.

The concept of intracellular proteolysis is avery modern one. Cells have two main mech-anisms for intracellular protein degradation:the endosome–lysosome system, whichdegrades proteins internalized by endocyto-sis, and the non-lysosomal system, whichdegrades proteins from the nucleus, cytosoland endoplasmic reticulum (FIG. 1). Mostnon-lysosomal intracellular proteolysis iscarried out by the ubiquitin/26S proteasomesystem, in which an ε-amino group of alysine residue in a target protein is covalentlyconjugated with the carboxy-terminal glycineresidue of ubiquitin. Lysine 48 of the firstubiquitin can then link, through an isopep-


The meteoric rise of regulatedintracellular proteolysis

R. John Mayer

T I M E L I N E

Figure 1 | The main proteolytic pathways in eukaryotic cells. The endosome–lysosome pathway(green) degrades extracellular and cell-surface proteins, such as receptors and their ligands. Intracellularorganelles also enter this pathway through double-membraned autophagosomes. Theubiquitin–proteasome pathway (red) degrades proteins from the cytoplasm, nucleus and endoplasmicreticulum (ER). Evidence is now emerging that the two pathways cooperate. Finally, the mitochondrion hasits own proteolytic system (blue), similar to that in prokaryotes.

Mitochondrion

Nuclearproteins

Autophagosome

Autophagy

Endosome/lysosome

Cooperation?

Endosome–lysosome system

Ubiquitin–proteasome

system

Mitochondrialproteolytic

systemNucleus

ER proteins

Cytoplasmicproteins

Plasma membrane



P E R S P E C T I V E S

because histone H2A was found to be ubiqui-tylated. Varshavsky and his graduate studentDaniel Finley, in collaboration with AaronCiechanover (who was by now an establishedresearcher), used cells expressing a tempera-ture-sensitive E1 to show that ubiquitin isrequired for protein degradation in livingcells, and that ubiquitin conjugation is essen-tial for cell viability. At the non-permissivetemperature, cells were arrested in the G2phase of the cell cycle, showing that ubiqui-tin-dependent proteolysis is also required forcell-cycle progression9,10. The dual role ofprotein phosphorylation and proteolysis inregulating the levels and activities of cyclins isnow well understood11, but at the time thisobservation was quite unusual.

Ubiquitylation then began to invade otherareas of research (BOX 1).Varshavsky and col-leagues purified some of the individual E2enzymes of the yeast Saccharomyces cerevisiae,and showed that a genetically defined compo-nent of DNA repair pathways, termed RAD6,and another protein with weak homology to it(CDC34), are E2 enzymes12,13. These findingsindicated that E2s have key roles in DNArepair and the cell cycle. It was at this time thatthe cell-cycle community became really inter-ested in the ubiquitin pathway, particularlywhen Kirschner’s laboratory (then in San

in a reaction that requires ATP. The activatedubiquitin is then transferred from E1 to aubiquitin-conjugating enzyme (E2) by trans-esterification. Finally, a ubiquitin–protein lig-ase (E3) catalyses the transfer of ubiquitin tothe ε-amino group of a lysine residue in a tar-get protein to produce a ubiquitin–proteinconjugate5. Furthermore, Hershko et al.6

showed that the activity of the ubiquitin path-way is greatly increased in cells making abnor-

mal proteins. It was independently shownthat there are multiple E2s (REF. 7) and E3s,and that E3s are the arbiters of substrate selec-tion for ubiquitylation. Specific recognition ofproteins for degradation lies at the heart ofthe ubiquitin pathway, and we are only justbeginning to appreciate how different combi-nations of E2 and E3 enzymes can imposeexquisite specificity on substrate selection5.We now realize that mutations in proteinspreventing recognition by E3s, mutations inE3s or sequestration of E3s by viruses8 cancause cancer (BOX 1).

Ubiquitylation is essential for lifeIt took years before the importance of thesebiochemical findings was fully appreciated.The turning point came in 1984 through thework of Alexander Varshavsky (FIG. 3b), then atthe Massachusetts Institute of Technology inBoston. He studied chromatin and geneexpression and became interested in ubiquitin

Foundation. Working at the TechnionUniversity in Haifa, Avram Hershko and hisgraduate students, including AaronCiechanover (FIG. 3a), first showed that a smallprotein ubiquitously expressed in all celltypes, which was subsequently found to beubiquitin4, is a component of an ATP-depen-dent in vitro proteolytic system in rabbit retic-ulocyte extracts. They found that ubiquitin iscovalently linked to protein substrates andthat it was necessary for proteolysis in theirsystem. The linkage of one protein to anotherthrough an isopeptide bond was unusual butelicited little interest, except from the smallnumber of workers in the field. Subsequently,Hershko and colleagues discovered that theaddition of ubiquitin to target proteins (ubiq-uitylation) requires three types of enzyme,termed E1, E2 and E3. First, a ubiquitin-acti-vating enzyme (E1) forms a thioester with thecarboxy-terminal glycine residue of ubiquitin

Figure 2 | Lewis Carroll’s Red Queen. The RedQueen and Alice find themselves running to standstill, a process familiar to cells — steady state(standing still) is achieved by a constant turnoverof molecules with short half-lives.

Figure 3 | The fathers of the field of regulatedintracellular proteolysis. a | Avram Hershko(left) with Aaron Ciechanover (right). b | AlexanderVarshavsky.

Folin states that‘endogenousproteins’ are stable.

Schoenheimer uses15N to show continu-ous protein turnoverin cells.

Lewy discovers his ‘bodies’in the brain stems of somepatients with Parkinson’sdisease.

Hershko andCiechanover dis-cover the processof ubiquitylation.

Hershko and colleaguesidentify enzymes of theubiquitin–protein ligasesystem.

Rechsteiner andcolleagues partial-ly purify the 26Sproteasome.

Varshavsky, Ciechanover and Finleydiscover that ubiquitylation isessential for viability and is neces-sary for cell-cycle progression.

Kirschner and col-leagues discover that acyclin is degraded bythe ubiquitin pathway.

The future — resolve the mechanism bywhich the ubiquitin–proteasome system andthe endosome–lysosome system coordinatelydegrade membrane proteins. Uncover evi-dence for further versatility of the isopeptidebond in cellular physiology. Develop therapiesbased on the ubiquitin–proteasome system.

Several groups discover thecombinatorial control andspecificity of SCF ubiquitinprotein ligases.

Lowe, Landon and Mayerdiscover that Lewy bod-ies are full of ubiquitylat-ed proteins.

1905 1912 1942 1978 1983 1984 1986 1988 1991 1997 2000

Timeline | Regulated intracellular proteolysis

“Now, here, you see, it takesall the running you can do,to keep in the same place. Ifyou want to get somewhereelse, you must run twice asfast as that.”



istry not only revealed ubiquitylated proteinsin intraneuronal inclusions in all the mainneurodegenerative diseases (for example,Alzheimer’s, Parkinson’s and Huntington’sdiseases, and amyotrophic lateral sclerosis)31,

Francisco) found that a cyclin was degraded bythe ubiquitin pathway (TIMELINE)14. Functionsfor ubiquitylation in key cellular processes con-tinue to emerge at an alarming rate, as anyonewho tries to keep up with the literature sur-rounding this field knows only too well (BOX 1).

Piecing together the 26S proteasomeBy 1984, it was known that ubiquitylated pro-teins are targeted for degradation by an ATP-dependent protease15, but the identity of theprotease was a mystery. Marty Rechsteiner andco-workers16,17 in Salt Lake City discovered andpartially purified the 26S proteasome by mon-itoring its ability to degrade ubiquitin–proteinconjugates. Rechsteiner et al.18 also suggestedthat the 20S proteasome shares subunits withthe 26S enzyme. But when the Haifa laborato-ry showed that the large protease is formedthrough the association of three multisubunitcomponents19 and that one of these, CF3, isindeed identical to the 20S proteasome20,21,several independent strands of research beganto intertwine the ubiquitin-dependent prote-olysis pathway and the proteasome. The 20Sproteasome corresponds to a previouslynamed 19S ‘prosome’particle22,23.

The association of a 19S regulator withthe 20S core to form the 26S particle wasshown in several laboratories. The 19S regu-lator was shown to contain six ATPasestogether with a cohort of other proteins,including the first receptor for multi-ubiqui-tylated proteins24.

The essential and non-redundant func-tional roles of the ATPases, together with theactivity of the 26S proteasome in controllingthe cell cycle, was first shown by using elegantyeast genetics25,26. A seminal advance wasmade in Daniel Finley’s laboratory by usingyeast lacking a proteasomal receptor formulti-ubiquitylated proteins. Ion-exchangechromatography of these mutant protea-somes resulted in separation of the so-called19S ‘bases’ and ‘lids’: the former contain thesix ATPases plus two non-ATPase subunits;the latter contain non-ATPases27. The ATPaseshave chaperone activity28 and are presumed tobe involved in unfolding proteins for entryinto the 20S core, which contains the prote-olytic activities. Crystallography has revealedthe elegant three-dimensional structure of the20S core29,30 and emphasizes that the catalyticsites are compartmentalized within the cylin-drical particle, necessitating that unfoldedproteins must be inserted into the centralchamber for proteolysis to occur.

Ubiquitin and diseaseEvidence for malfunction of the ubiquitin–proteasome system is now beginning to

emerge in various unrelated human diseases,from cancer to neurological disorders (BOX 1).But perhaps the earliest observations were inrelation to human chronic neurodegenerativediseases, where ubiquitin immunocytochem-


Box 1 | Ubiquitylation in health and disease

Studies over the past ten years or so have revealed that ubiquitylation is involved in the regulationof a host of cellular processes, and is disrupted in several diseases. The selected examples below givea flavour of this system’s versatility, and also hint at the disastrous consequences whenubiquitylation and protein degradation malfunction.

Signal transductionThe multiple protein-dependent steps in signal transduction pathways offer excellent targets forregulated proteolysis. Examples are numerous and include receptor tyrosine-kinase dependentpathways involving an E3 enzyme36 and the Hedgehog and Wnt/Wingless pathways37.

TranscriptionTranscription factors are short-lived proteins that are eliminated by the ubiquitin system5.Examples abound, but the degradation of IκB has been intensively studied since the discovery thatdegradation of IκB activates NF-κB-dependent gene expression38.

Cell cycleProgression through the cell cycle requires cyclin-dependent kinases together with activators(cyclins) and inhibitors (for example, p27). Ubiquitin-regulated degradation of these and othercell-cycle proteins ensures temporal control of the cycle39.

MHC class I antigen processingPeptides derived from proteins bind to MHC class I molecules to trigger the cytotoxic lymphocyteresponse — proteasomes are a major producer of small peptides for this purpose40. The cytokineinterferon-γ induces a specific set of proteolytic and regulatory subunits41 that produce betterpeptides for class I presentation42.

ER quality controlProteins regularly become misfolded in the endoplasmic reticulum (ER). In yeast, misfoldedproteins can be retrogradely extracted from the ER and degraded by the 26S proteasome43. ERproteins can be similarly extracted in higher eukaryotes44. The ER degradation system, togetherwith the unfolded protein response, controls the well-being of the ER lumen and the secretorypathway.

CancerSeveral oncogenes and tumour suppressors influence the ubiquitin–proteasome pathway; forexample the von Hippel–Lindau tumour suppressor (pVHL) is a component of a ubiquitin proteinligase45, and the liver oncoprotein gankyrin interacts with a regulatory ATPase of the 26Sproteasome46.

Neurological disordersApart from the accumulation of proteins in the main neurodegenerative diseases, a mutated E3 isassociated with Angelman’s syndrome — a developmental neurological disorder47.

Viral infectionViral proteins subvert the proteasome by binding to subunits of both the 20S core and 19Sregulator48 to minimize the MHC class I response and promote viral replication.

26Sproteasome

Cancer

CMV Alzheimer's disease

Juvenile-onsetfamilial Parkinsonism

Angelman's syndrome

EBV

Gankyrin

Ub-X

E3

SCF-TrCP

c-Cbl

pVHLSCF-TrCP

MDM2;E6-AP*SCF; APCSCF

?

Parkin

E6-AP‡

?

*Induced by papilloma viruses. ‡Mutated in Angelman's syndrome.

Process

Signaltransduction

Transcription

Cell cycle

Antigenprocessing

?

?

ER qualitycontrol

Substrate (X)

β-catenin

EGF receptor

HIFIκB

p53

CyclinsCDK inhibitors

MHC class Iantigens

?

?

Misfoldedproteins




27. Glickman, M. H. et al. A subcomplex of the proteasomeregulatory particle required for ubiquitin conjugatedegradation and related to the COP-9-signalasomeand eIF3. Cell 94, 615–623 (1998).

28. Braun, B. C. et al. The base of the proteasomeregulatory particle exhibits chaperone-like activity.Nature Cell Biol. 1, 221–226 (1999).

29. Lowe, J. et al. Crystal structure of the 20S proteasomefrom the archeon T. acidophilum at 3.4 Å resolution.Science 268, 533–539 (1995).

30. Groll, M. et al. Structure of 20S proteasome from yeastat 2.4 Å resolution. Nature 386, 463–471 (1997).

31. Lowe, J., Mayer, R. J. & Landon, M. Ubiquitin inneurodegenerative diseases. Brain Pathol. 3, 55–65(1993).

32. McKeith, I. G. et al. Clinical and pathological diagnosisof dementia with Lewy bodies (DLB). Report of theCDLB international workshop. Neurology 47,1113–1124 (1996).

33. Doherty, F. J. et al. Ubiquitin–protein conjugatesaccumulate in the lysosomal system of fibroblaststreated with cysteine protease inhibitors. Biochem. J.263, 47–55 (1989).

34. Mizushima, N. et al. A protein conjugation systemessential for autophagy. Nature 395, 395–398 (1998).

35. Hicke, L. Gettin’ down with ubiquitin: turning off cell-surface receptors, transporters and channels. TrendsCell Biol. 9, 107–112 (1999).

36. Joazeiro, C. A. P. et al. The tyrosine kinase negativeregulator c-Cbl as a RING-type E2-dependent ubiquitinprotein ligase. Science 286, 309–312 (1999).

37. Jiang, J. & Struhl, G. Regulation of the Hedgehog andWingless signalling pathways by the F-box/WD40-repeat protein Slimb. Nature 391, 493–496 (1998).

38. Palombella, V. J., Rando, O. J., Goldberg, A. L. &Maniatis, T. The ubiquitin–proteasome pathway isrequired for processing the NFκB1 precursor proteinand the activation of NFκB. Cell 78, 773–785 (1994).

39. Huibregtse, J., King, R. W., Deshaies, R. J., Peters, J.-M. & Kirschner, M. W. How proteolysis drives the cellcycle. Science 274, 1652–1659 (1996).

40. Groettrup, M., Soza, A., Kuckelkorn, U. & Kloetzel, P.M. Peptide antigen production by the proteasome:complexity provides efficiency. Immunol. Today 17,429–435 (1996).

41. Gaczynska, M., Rock, K. L. & Goldberg, A. L. Interferonand expression of MHC genes regulate peptidehydrolysis by proteasomes. Nature 365, 264-267(1993).

42. Dick, T. P. et al. Coordinated dual cleavages induced bythe proteasome regulator PA 28 lead to dominant MHCligands. Cell 86, 253–256 (1996).

43. Hiller, M. M., Finger, A., Schweiger, M. & Wolf, D. H. ERdegradation of a misfolded luminal protein by thecytosolic ubiquitin–proteasome pathway. Science 273,1725–1728 (1996).

44. Plemper, R. K. & Wolf, D. H. Retrograde proteintranslocation: ERADication of secretory proteins inhealth and disease. Trends Biochem. Sci. 24, 266–270(1999).

45. Lisztwan, J., Imbert, G., Wirbelauer, C., Gstaiger, M. &Krek, W. The von Hippel-Lindau tumor suppressorprotein is a component of an E3 ubiquitin–protein ligaseactivity. Genes Dev. 13, 1822–1833 (1999).

46. Higashitsuji, H. et al. Reduced stability ofretinoblastoma protein by gankyrin, an oncogenicankyrin-repeat protein overexpressed in hepatomas.Nature Med. 6, 96–99 (2000).

47. Kishino, T., Lalande, M. & Wagstaf, J. UBE3A/E6-APmutations cause Angelman’s syndrome. Nature Genet.15, 70–73 (1997).

48. Ferrell, K., Wilkinson, R. M., Dubiel, W. & Gordon, C.Regulatory subunit interactions of the 26S proteasome,a complex problem. Trends Biochem. Sci. 25, 83–88(2000).

AcknowledgementsI thank Avram Hershko and Aaron Ciechanover (Haifa), AlexVarshavsky (Pasadena), Wolfgang Dubiel (Berlin), Dieter Wolf(Stuttgart), Mark Hochstrasser (Chicago), Peter Zwickl(Martinsreid), Cecile Pickart (Baltimore), Keith Wilkinson(Atlanta), Alan Weissman (Washington), Ron Hay (St Andrews),and Simon Dawson, Michael Landon, Andy Alban and RobLayfield (Nottingham) for help with this article; Rohan Baker(Canberra) for critically reviewing the manuscript; and theMRC, BBSRC, Wellcome Trust and EU Framework IV forsupport of some of the quoted work. Numerous pivotalcontributions have been omitted due to space constraints;many thanks to the ‘unsung heroes’.

1. Carroll, L. Alice’s Adventures in Wonderland. Throughthe looking glass (Penguin Books, London, 1973).

2. Schoenheimer, R. The Dynamic State of BodyConstituents (Harvard Univ. Press, Boston, 1942).

3. Blobel, G. in Ciba Foundation Symposium Vol. 75, 398(Excerpta Medica, Amsterdam, 1980).

4. Wilkinson, K. D., Urban, M. K. & Haas, A. L. Ubiquitin isthe ATP-dependent proteolysis factor I of rabbitreticulocytes. J. Biol. Chem. 255, 7529–7532 (1980).

5. Hershko, A. & Ciechanover, A. The ubiquitin system.Annu. Rev. Biochem. 67, 425–480 (1998).

6. Hershko, A., Eytan, E., Ciechanover, A. & Haas, A. L.Immunochemical analysis of the turnover ofubiquitin–protein conjugates in intact cells. Relationshipto the breakdown of abnormal proteins. J. Biol. Chem.257, 13964–13970 (1982).

7. Pickart, C. M. & Rose, I. A. Functional heterogeneity ofubiquitin carrier proteins. J. Biol. Chem. 260,1573–1581 (1985).

8. Huibregtse, J., Scheffner, M. & Howley, P. M. Cloningand expression of the cDNA for E6-AP, a protein thatmediates the interaction of the human papillomavirus E6oncoprotein with p53. Mol. Cell. Biol. 13, 775–784(1993).

9. Finley, D., Ciechanover, A. & Varshavsky, A.Thermolability of ubiquitin-activating enzyme from themammalian cell cycle mutant ts85. Cell 37, 43–55(1984).

10. Ciechanover, A., Finley, D. & Varshavsky, A. Ubiquitindependence of selective protein degradationdemonstrated in the mammalian cell cycle mutant ts85.Cell 37, 57–66 (1984).

11. Zachariae, W. & Nasmyth, K. Whose end is destruction:cell division and the anaphase promoting complex.Genes Dev. 13, 2039–2058 (1999).

12. Jentsch, S., McGrath, J. & Varshavsky, A. The yeastDNA repair gene RAD6 encodes a ubiquitin-conjugatingenzyme. Nature 329, 131–134 (1987).

13. Goebl, M. G. et al. The yeast cell cycle gene CDC34encodes a ubiquitin-conjugating enzyme. Science 241,1331–1335 (1988).

14. Glotzer, M., Murray, A. W. & Kirschner, M. W. Cyclin isdegraded by the ubiquitin pathway. Nature 349,132–138 (1991).

15. Hershko, A., Leshinsky, E., Ganoth, D. & Heller, H. ATP-dependent degradation of ubiquitin–protein conjugates.Proc. Natl Acad. Sci. USA 81, 1619–1623 (1984).

16. Hough, R., Pratt, G. & Rechsteiner, M.Ubiquitin–lysozyme conjugates. Identification andcharacterization of an ATP-dependent protease fromrabbit reticulocyte lysates. J. Biol. Chem. 261,2400–2408 (1986).

17. Hough, R., Pratt, G. & Rechsteiner, M. Purification oftwo high molecular weight proteases from rabbitreticulocyte lysate. J. Biol. Chem. 262, 8303–8313(1987).

18. Hough, R., Pratt, G. & Rechsteiner, M. in Ubiquitin (ed.Rechsteiner, M.) 101–134 (Plenum Press, New York,1988).

19. Ganoth, D., Leshinsky, E., Eytan, E. & Hershko, A. Amulticomponent system that degrades proteinsconjugated to ubiquitin. Resolution of factors andevidence for ATP-dependent complex formation. J. Biol.Chem. 263, 12412–12419 (1988).

20. Eytan, E., Ganoth, D., Armon, T. & Hershko, A. ATP-dependent incorporation of 20S protease into the 26Scomplex that degrades proteins conjugated to ubiquitin.Proc. Natl Acad. Sci. USA 86, 7751–7755 (1989).

21. Armon, T., Ganoth, D. & Hershko, A. Assembly of the26S complex that degrades proteins ligated to ubiquitinis accompanied by the formation of ATPase activity. J.Biol. Chem. 265, 20723–20726 (1990).

22. Arrigo, A.-P., Tanaka, K., Goldberg, A. L. & Welch, W. J.Identity of the 19S ‘prosome’ particle with the largemultifunctional protease complex of mammalian cells(the proteasome). Nature 331, 192–194 (1988).

23. Falkenburg, P. E. et al. Drosophila small cytoplasmic19S ribonucleoprotein is homologous to the ratmulticatalytic proteinase. Nature 331, 190–192 (1988).

24. Dubiel, W. & Rechsteiner, M. The 19S regulatorycomplex of the 26S proteasome. Adv. Mol. Cell Biol. 27,129–163 (1998).

25. Gordon, C., McGurk, G., Dillon, P., Rosen, C. & Hastie,N. Defective mitosis due to a mutation in the gene for afission yeast 26S protease subunit. Nature 366,355–357 (1993).

26. Ghislain, M., Udvardy, A. & Mann, C. S. cerevisiae 26Sprotease mutants arrest cell division in G2/metaphase.Nature 366, 358–361 (1993).

but also led to the identification of a newneurological illness — dementia with Lewybodies — which is the second most commoncause of cognitive decline in the elderly afterAlzheimer’s disease32. Astoundingly, 76 yearselapsed between the discovery of Lewy bodiesand the observation that they containubiquitylated proteins.

The isopeptide bond does it allThere were early indications that protein ubiq-uitylation is related to the endosome–lysosomeproteolytic system33.This link has recently beenstrengthened by the remarkable findings ofYoshinori Ohsumi and colleagues in Okazaki34,which show that autophagy,an important lyso-some-requiring process for cellular proteindegradation in adaptation to starvation, is con-trolled through conjugation of novel proteinsby a set of enzymes similar to those of the ubiq-uitin pathway (FIG.1).This work emphasizes thefact that the ubiquitous isopeptide bond hasbeen used throughout evolution to control thetwo main systems of intracellular proteolysis.With ubiquitin helping to control the endocyt-ic pathway35, it is conceivable that the mainprotein degradation pathways in cells are con-trolled by isopeptide bond formation — a tri-umph for evolution indeed!

As proteomics blooms, it is salutary tothink that different forms of post-transla-tional modification increasingly emerge ascontrollers of biological processes. Whichmodifications are the front runners? Does itmatter? It is already clear that phosphoryla-tion works hand in glove with proteolysis toregulate many aspects of the chemistry oflife. The current state of the intracellularproteolysis literature indicates that compre-hension of cell physiology may soon be dra-matically changed — selective degradationof regulatory proteins is at the heart of life.Our perception of cell biology, medicineand pharmaceutical intervention is at apoint equivalent to the discovery of platetectonics for geology.

R. John Mayer is in the Laboratory forIntracellular Proteolysis, School of Biomedical

Sciences, University of Nottingham MedicalSchool, Queen’s Medical Centre, Nottingham,

NG7 2UH, UK.e-mail: [email protected]

Links

FURTHER INFORMATION Regulatory subunitsof the 26S proteasome | Ubiquitin index | TheCiechanover laboratory | The Hershkolaboratory | The Varshavsky laboratory | PressRelease on the 2000 Albert Lasker Awards |Nature Medicine commentariesENCYCLOPEDIA OF LIFE SCIENCES Ubiquitinpathway | Protease complexes



through was the idea that body function canbe explained by similar physical laws to thosethat account for the action of artificialmachines. This idea was elaborated on philo-sophical grounds by René Descartes10, anddeveloped into a scientific manifesto by inves-tigators such as Giovanni Alfonso Borelli andMalpighi11,12. As a result, physiology no longerneeded to depend on metaphysical theoriesfor the interpretation of body functions.Instead, like astronomy and new physics, itcould become a ‘true’ science — an investiga-tion that combined experimental study withthe application of the ‘laws of mathematicsand geometry’ to body machines.

One result of this new scientific attitudewas that scientists were discouraged fromsearching for the ultimate causes of ‘vitalprocesses’. This was vividly expressed byMalpighi, in a beautiful passage from hisOpera postuma4:

“The way our soul uses the body in operating

is ineffable, yet it is certain that in the

operations of growth, sensation and motion

the soul is forced in conformity with the

machine on which it is acting, just as a clock

or a mill is moved in the same way by a

pendulum or lead or stone, or by an animal

or by a man; indeed if an angel moved it, he

would produce the same motion with

changes of positions as the animals or agents

do. Hence, even though I did not know how

the angel operates, if on the other hand I did

know the precise structure of the mill, I

would understand this motion and action,

and if the mill were out of order, I would try

to repair the wheels or the damage to their

structure without bothering to investigate

how the angel moving them operated.”

To know how a machine operates, you need toknow its structure. So the idea of ‘organicmachines’prompted anatomical investigations— both classical, descriptive macroscopicanatomy, and a new,‘subtle anatomy’, based onthe use of newly invented techniques (some ofwhich were the precursors of modern histo-logical methods). It is no surprise, then, thatthe basis of the modern microscopic anatomyof animals and plants emerged in the seven-teenth century, owing to the work of Malpighiand many others4,12–17. As had happened withGalileo’s astronomical observations, this newinvestigative attitude was not due simply to theavailability of new technology, but also to thenew cultural climate.

Decline and fallThe climate changed in the eighteenth centu-ry, as interest in microscopic studies dwin-

machines whose performance can beexplained by similar laws to those operating inman-made machines. In the seventeenth cen-tury, this concept was used not only to explainfunctions that obviously reflected those ofmechanical devices (such as skeletal and artic-ular motion or the action of muscles), but alsofor other operations — digestion, sensation,fermentation and production of blood, forexample1–3. To account for these more delicateoperations of animal economy, body machineswere thought to involve tiny components thatcould escape detection by the naked eye. Thisview derived, in part, from a recurrence of thephysicists’ view that the Universe is composedof atoms. In Greek classical science this viewwas advocated by Democritus, and in the sev-enteenth century by the French philosopherand scientist Pierre Gassendi. As MarcelloMalpighi (FIG. 1), one of the greatest seven-teenth-century life scientists1,2, put it4:

“Nature, in order to carry out the marvellous

operations in animals and plants, has been

pleased to construct their organized bodies

with a very large number of machines, which

are of necessity made up of extremely minute

parts so shaped and situated, such as to form

a marvellous organ, the composition of

which are usually invisible to the naked eye,

without the aid of the microscope.”

The rise of machinesUntil the sixteenth and seventeenth centuries,most progress in the life sciences and medi-cine elaborated on classical doctrines, datingfrom Hippocrates, Aristotle and Galen (BOX 1,overleaf). But in the seventeenth century,interest in experimental studies explodedbecause, as had happened in astronomy andphysics, new investigations cast doubt on theinfallibility of the Ancients. In particular, thediscovery of blood circulation, published in1628 by William Harvey5 — and the subjectof some debate at the moment6–8 — ques-tioned the very foundation of classical physi-ology on which the whole body structure wasinterpreted.

In the wave of the scientific revolutionpromoted by Galileo9, a conceptual break-


Although scientific progress is usuallyrepresented as being linear, it may, in fact,have a cyclical character — somediscoveries may be forgotten or lost (at leasttemporarily), and themes may reappearthrough the centuries. Consider, forexample, the concept of ‘molecularmachines’, from the exciting phase ofresearch that flourished in the seventeenthcentury, to the idea of machines that is atcentre stage in modern cell biology.

More than three centuries ago, the birth ofmodern life sciences was marked by the ideathat body function is based on organic

Biological machines: from mills to molecules

Marco Piccolino

T I M E L I N E

Figure 1 | Marcello Malpighi (from the OperaPostuma, 1798 Venetian in Folio edition).Malpighi, a prominent scientist in theseventeenth century, was one of the first toattribute body function to an organized series ofminute ‘organic machines’. The conceptunderlying his metaphor of the ‘angel and themill’ prompted anatomical investigations, whichlaid the foundation for modern microscopicanatomy. (Image courtesy of the library G. Romiti of the Anatomical Institute of theUniversity of Pisa.)




same function in living matter as theNewtonian idea of gravity had in the inor-ganic world. In fact, Haller’s reluctance topropose a mechanistic explanation for irri-tability paralleled Newton’s aversion toproposing hypotheses about the mecha-nism of gravity. The point to emerge fromthese discussions was that an organism’sresponse to a stimulus is not a purely physi-cal consequence of that stimulus, but that itreflects the organism’s internal organiza-tion. In other words, the response (in thecase of irritability, contraction of a muscle)is what the organism is prepared — wewould now say ‘programmed’ — to pro-duce. The energy of the response is‘enclosed’ in the organism, and does notcome from the energy of the stimulus. Sowhat really matters is the informationencoded by the stimulus. In the first half ofthe nineteenth century, a similar idea wasbehind the development, by JohannesMüller, of the doctrine of ‘specific nervousenergy’, according to which the sensationaroused by the stimulation of a sensorystructure does not depend on the character-istics of the stimulus, but on the type ofstructure stimulated20.

Machines revisitedHaller’s ideas laid the foundation for thedevelopment of another fundamental ideain the nineteenth century: that of the ‘inter-nal milieu’, developed by Claude Bernard21

in 1865. Bernard attempted to found medi-cine as a true science, based on the laws ofphysics and chemistry. He studied the char-acteristics of living organisms that seemedto elude physico-chemical principles, suchas their relative independence of the condi-tions of the external environment (milieucosmique). He attributed these characteris-tics to the organizational complexity oforganisms, and often referred to the body orits working components as ‘machines’(although his machines were more opera-tional than structural devices). For example,he discovered the liver’s ability to synthesizesugar not because he studied the morpho-logical structure of this organ, but becausehe used chemical analysis to follow the fateof blood sugar passing through the liver.

This typifies the study of body machinesin the eighteenth and nineteenth centuries— the emphasis was not on knowing theminute structures responsible for physio-logical responses such as contractions orsensations, but rather on studying theiroperation. In part, this was due to the lackof knowledge about the organization of liv-ing tissues. For instance, cellular theory was

es. Different kinds of stimuli (chemical,mechanical or electrical) could excite mus-cle irritability, which was normally broughtabout by the action of a nerve. But Hallerbelieved that the nerve’s influence was notthe real cause of the muscle contraction —instead, it acted only as a stimulating (orexciting) factor that activated the intrinsicirritability of muscles.

Haller’s ideas spread over Europe, caus-ing lively discussions and dividing physiolo-gists into ‘Hallerians’ or ‘anti-Hallerians’.Hallerians claimed that irritability had the

dled. This was partly due to the apparentfailure of the investigative programme,based on mechanistic explanations of bodyfunction, that had dominated the seven-teenth century. Although many discoverieswere made during that time, such as thestructures of blood capillaries and alveoli inthe lungs, the possibilities of explaining lifeprocesses on a simple, mechanical basiswere limited.

So the idea of mechanical body machineswas largely abandoned. Instead, interestmoved towards new forces — particularlyelectricity which, together with the study ofgas (‘airs’) and chemistry, took centre stage.Electricity was particularly attractive as aprinciple for explaining vital processesbecause its application could produce move-ments in paralytic limbs or in animal prepa-rations. As well as muscle contraction it waseasy to evoke an electrical mechanism fornerve conduction, owing to the easy andrapid propagation of electricity, whichseemed to match the rapid flow of sensationalong nerve fibres18.

Sensibility and irritabilityIt was a conceptual advance in the secondhalf of the eighteenth century that sowedthe seeds for the modern idea of machines.This came from the idea of ‘irritability’, con-ceived by the Swiss physiologist Albrechtvon Haller (FIG. 2) in 1752. On the basis ofanimal experiments, Haller concluded that‘sensibility’ (the ability to perceive a stimu-lus) and ‘irritability’ (the ability to react tothat stimulus with a contraction) were dif-ferent properties of living tissue, pertainingtypically to nerves (sensibility) and muscles(irritability)19.

Haller confessed that he could not ascer-tain the mechanism of irritability, but sug-gested that it depended on an essential con-stituent of living tissue (the gluten). Hedistinguished irritability — a vital property— from elasticity, which has purely physicalproperties and is unrelated to vital process-

Box 1 | The four humours

Although anatomy was a part of classical medicine, it was not used to investigate body function.According to the doctrine of four humours — blood, yellow bile or choler, phlegm, and black bileor melancholy — the body and its organs were conceived as the stage where the humoursinteracted (depending on astronomical, atmospheric, climatic or other influences). Health andgood temper resulted when humours were in correct proportions and mixing was appropriate (‘tomix’ in Latin is temperare). Conversely, diseases, or bad tempers, were produced when onehumour was in excess or the mixture was inappropriate. This idea hardly favoured anatomicalinvestigation and certainly did not promote the study of the structure of organs and tissues.Indeed, many organs, including the liver and lung, were considered to consist of effused blood(parenchyma), and so were thought to be devoid of a real internal structure. Similarly, smallanimals, such as insects, were thought to lack an internal structural organization.

Figure 2 | Albrecht von Haller (from the firstedition of his Elementa physiologiae42). Haller, ofSwiss origin, was a leading figure in eighteenth-century physiology. He conceived the idea of‘sensibility’ and ‘irritability’ to explain the body’sreaction to stimulus. In his formulation of theconcept of irritability to account for musclecontraction, he first acknowledged, although in animplicit way, the importance of information flow inbiological systems. (Image courtesy of the libraryG. Romiti of the Anatomical Institute of theUniversity of Pisa.)



patterns in muscle fibres during contraction— did not immediately result in an antici-pation of the ‘sliding theory’ that was even-tually formulated in the 1950s. Many scientistswere neither interested nor confident inmicroscopy, preferring to visualize musclecontraction as the result of shortening of amuscular protein, fuelled by a chemicalprocess akin to those being discovered infermentation reactions27,29. The first genera-tion of biological chemists were more inter-ested in breaking down the cellular andsubcellular components of living tissues tomake them amenable to chemical analysisthan in adjusting their chemical techniques to the complex components of biologicalmaterials.

On the other hand, a biochemicalapproach, combined with physiological andclinical investigation, was fundamental indeveloping the concept of a ‘hormone’ at theturn of the nineteenth century (the word wasintroduced by Bayliss and Starling in 1905).It soon became clear that hormones, togetherwith nervous reactions, were essential forregulating body function and maintainingstability of the internal milieu. This ledWalter Cannon, in 1925, to propose the ideaof ‘homeostasis’30. Through the study of hor-mones and other chemical messengers itbecame clear that, besides being involved inmetabolic processes and in other chemicalactions, molecules can carry importantinformation in biological systems. Moreover,these molecules might relay informationthrough specific receptor and effector sys-tems.

The idea of catalysis also emerged throughchemistry. Biological materials were found tohave specific and highly efficient catalysts,termed ‘enzymes’ by Willy Kühne31 in 1877.The study of enzymes (and of other proteins,as well as large molecules such as nucleicacids) was, in fact, behind the resurgence of interest in the idea of ‘minute machines’during the twentieth century32.

Twentieth-century machinesIt became increasingly clear that the func-tion of enzymes depends not only on theirelementary chemical composition, but alsoon the configuration of their components.For example, effective interactions betweenenzymes, substrates and cofactors dependon the spatial arrangement of the interact-ing elements. This insight led to interest inthe structure of complex molecules. It wasalso evident that the function of enzymesand other biological molecules could beregulated through specific control mecha-nisms. For instance, in 1963, Jacques

developed only around 1839 by MatthiasSchleiden22 and Theodor Schwann23, andmuch time elapsed before there was any realknowledge about genetic laws, the existenceand structure of membranes, the functionsof proteins and enzymes, and the existenceof hormones and other chemical messen-gers. In the absence of adequate knowledge,attempts to devise mechanistic hypothesesof biological phenomena were likely to fail.

First-generation biochemistsAnother reason for the lack of interest in theminute organization of body structure wasthe growing importance of chemistry inbiological studies in the eighteenth century.For example, the discovery by AntoineLaurent Lavoisier, Pierre Simon de Laplace24

and Lazzaro Spallanzani25, that a processakin to combustion occurs in living tissues,had great biological relevance26. In the fol-lowing century it became increasingly evi-

dent that many functions of living organ-isms depend on chemical reactions. Achemical reaction typically occurs in a solu-tion, and involves particles that move by dif-fusion and collide randomly with oneanother. Similarly, within an organism,chemical reactions seemed to require a liq-uid medium, and did not depend on theexistence of particular structures. So it is notby chance that Claude Bernard developedhis idea of a liquid internal milieu just whenbiologists were becoming interested inchemistry.

Interest focused on those reactions thatcould, potentially, produce the energy nec-essary for life. In his book Reflections onMuscle 27, Andrew Huxley remarked that therelative lack of interest in the structuraldetails of biological processes partlyexplains why the observations made around1880 by Theodor Engelmann28 — of charac-teristic changes in the dimensions of band


Box 2 | Modern molecular machines

Today, biology is revealing the importance of ‘molecular machines’ and of other highly organizedmolecular structures that carry out the complex physico-chemical processes on which life isbased. There are many diverse molecular machines:

• The photosynthetic system and complex devices that produce ATP.

• DNA replication and protein translation apparatus.

• Enzymatic cascade of phototransduction.

• The integrated membrane system, involving ionic pumps and channels, that produces ionicgradients and generates electric differences across membranes; this underlies the production ofelectric signals in nerve fibres.

• Machines that convert chemical energy into mechanical energy during muscle contraction orflagellar motion.

• Finely integrated metabolic cycles and networks, including the system involved in antigenrecognition and antibody production, the integrated system of hormones, extracellularmolecules and intracellular messengers that are connected by many control pathways.

4H+4H+

bb

γ

εα

β

β

δa

c

Figure 3 | The metaphor of a ‘machine’, applied to living organisms. Compare an old, manuallyoperated hydraulic machine (left) to the rotary ATP synthase of modern molecular biology (right). Bothmachines are reversible with minor readjustment. In the molecular machine, electrochemical energy in aproton gradient is normally used to produce rotary movement and ATP, but the machine can also work inreverse to produce an electrochemical gradient at the expense of ATP (figure adapted from REF. 43). In theman-made machine, the hydraulic potential energy could be converted into mechanical work that theman could use (from the ‘Stanzino delle Matematiche’ Museo degli Uffizi, Florence; © by ‘Ministero AffariCulturali’ of Italy).




1. Adelmann, H. B. Marcello Malpighi and the Evolution ofEmbryology — 5 Vols (Cornell, Ithaca, 1966).

2. Belloni, L. Opere Scelte di Marcello Malpighi (UTET, Torino,1967).

3. Piccolino, M. Marcello Malpighi and the difficult birth ofmodern life sciences. Endeavour 23, 175–179 (1999).

4. Malpighi, M. Opera Postuma (Churchill, London, 1697).5. Harvey, W. Exercitatio Anatomica de Motu Cordis et

Sanguinis in Animalibus (Fitzeri, Frankfurt, 1628).6. Chau, P. -L. Ancient Chinese had their fingers on the pulse.

Nature 404, 431 (2000).[nature contents page]7. Cheng, T. O. Did the Greeks beat Chinese on blood

circulation… . Nature 405, 993 (2000).[nature contentspage]

8. Prioreschi, P. …or was ‘blood as the river of life’ justpoetic? Nature 405, 993 (2000).[nature contents page]

9. Galileo, G. Discorsi e Dimostrazioni Matematiche Intorno aDue Nuove Scienze (Elzeviri, Leyden, 1638).

10. Descartes, R. De Homine (Leffen & Moyardum, Leyden,1662).

11. Borelli, G. A. De Motu Animalium — 2 Vols (Bernabò,Rome, 1680).

12. Malpighi, M. Opera Omnia (Scott & Wells, London, 1686).13. Stensen, N. Observationes Anatomicae (Chouët, Leyden,

1662).14. Swammerdam, J. Bybel der Natuur (Severinus,

Amsterdam, 1737).15. Leeuwenhoek, A. Opera Omnia (Langerak, Leyden, 1722).16. Croone, W. De Ratione Motus Musculorum (Hayes,

London, 1664) Translated by P. Maquet in On the Reasonof the Movement of the Muscles (American PhilosophicalSociety, Philadelphia, 2000)

17. Mayow, J. Tractatus Quinque Medico-Physici (TheatroSheldoniano, Oxford, 1674).

18. Galvani, L. De viribus electricitatis in motu muscularicommentarius. Bon. Sci. Art. Inst. Acad. Comm. 7,363–418 (1791).

19. Haller, A. De partibus corporis humani sensibilibus etirritabilibus. Comm. Soc. Reg. Scient. Gottingensis 2,114–158 (1753).

20. Müller, J. Handbuch der Physiologie des Menschen(Hölscher, Coblenz, 1844).

21. Bernard, C. Introduction à l’Étude de la MédecineExpérimentale (Baillière, Paris, 1865).

22. Schleiden, M. J. Beiträge zur Phytogenesis. Arch. Anat.Physiol. Wiss. Med. 13, 137–176 (1838).

23. Schwann, T. Mikroskopische Untersuchungen über dieÜbereinstimmung in der Struktur und dem Wachstum derTiere und Pflanzen (Reimer, Berlin, 1839).

24. Lavoisier, A. L. & de Laplace, P. S. in Oeuvres de LavoisierVol. I 528–530 (Imprimerie Royale, 1780, printed in 1864,Paris).

25. Spallanzani, L. in Rapports de l’Air avec les ÊtresOrganisés (ed. Senebier, J.) (Paschoud, Genève, 1807).

26. Keilin, D. The History of Cell Respiration and Cytochrome(Cambridge Univ. Press, Cambridge, 1966).

27. Huxley, A. F. Reflections on Muscle (Liverpool Univ. Press,Liverpool, 1980).

28. Engelmann, T. W. Mikrometrische Untersuchungen ancontrahirten Muskelfasern. Arch. Ges. Physiol. 23,571–590 (1880).

29. Needham, D. M. Machina Carnis (Cambridge Univ. Press,Cambridge, 1971).

30. Cannon, W. B. Organization for PhysiologicalHomeostatics. Physiol. Rev. 9, 399–431 (1925).

31. Kühne, W. Verh. Ueber da Verhalten VerschiedenerOrganisirter und Sog. Ungeformter Fermente. Naturhist.-medic. Vereins Heidelb. 1, 190–193 (1877).

32. Alberts, B. et al. Molecular Biology of the Cell 3rd edu(Garland, New York, 1994).

33. Monod, J., Changeux, J. P. & Jacob, F. Allosteric proteinsand cellular control systems. J. Mol. Biol. 6, 306–329(1963).

led Malpighi to suppose, more than threecenturies ago, that “machines will be eventu-ally found not only unknown to us but alsounimaginable by our mind”38. If we considerthat basically the same molecular deviceunderlies ATP synthesis and bacterial flagellarmotion, we see that modern biologicalmachines correspond to the uniformity ofnature pictured by Malpighi when he said4:

“In its things Nature operates by necessity

always in a uniform way. . . . Even though

they appear disparate, the things of Nature

are not so disconnected that one cannot

observe a concatenation and uniformity in

operating.”

However, in the importance of informa-tion flow, modern biological (and non-bio-logical) machines differ from old machines,and surpass the expectations of the early lifescientists. The old biological machines weresupposed, at a minute level, to be “. . . madeup of cords, filaments, beams, levers, tissues,fluids coursing here and there, cisterns,canals, filters, sieves and similar mecha-nisms”4. Besides the “fluids coursing hereand there”, energy — rather than informa-tion — was thought to circulate throughsuch components. No feedback mechanismsor control processes were predicted. Thelack of an adequate concept of ‘information’explains other difficulties encountered byearly life scientists. For instance, it wasimpossible to come up with a reasonabletheory of body development and the trans-mission of hereditary characteristics14,39,40.

Some modern molecular devices, such asthe rotary mechanism involved in ATP syn-thesis, may visually resemble the artificialmachines that inspired the scientific revolu-tion more than three centuries ago (FIG. 3).However, as well as the intrinsic regulatorymechanisms in what Paul Boyer called thesenew “splendid molecular machines”41, thereare the regulatory actions based on informa-tion flux which, in this case, control thevarious phases of energetic metabolism cul-minating in ATP synthesis. With reference toMalpighi’s metaphor of the angel and themill, we could perhaps say that, besides try-ing to understand the mechanisms of thesebiological wheels, modern scientists havestarted to picture how, by controlling a fluxof signals through information networks,the angel regulates the complex machine ofthe living mill.

Marco Piccolino is in the Dipartimento diBiologia, Università di Ferrara, Via Borsari 46,

44100 Ferrara, Italy. e-mail: [email protected]

Monod, Jean Pierre Changeux and FrançoisJacob33 introduced the concept of allostericregulation — that enzymatic action can beregulated by chemical signals acting on sitesother than the enzyme’s catalytic site. Thishas since provided a reference for interpret-ing mechanisms involving molecules andsystems that differ from those based on typ-ical enzymatic actions; for instance, lig-and–receptor interactions and variousmodulatory actions. An important advancehas been the recognition that complexreceptor assemblies are linked to second-messenger systems through specialized pro-teins34,35 and that there is a flux of biologicalinformation. This information is carried byspecific messengers, which act on systemsthat recognize them and develop specificresponses. Through this complex flux ofinformation, different mechanisms can beorganized in more complex systems, result-ing in highly integrated and efficientprocesses.

The concept of information flux is alsocentral to one of the biggest advances oftwentieth-century biology — recognition ofthe molecular mechanisms responsible fortransmitting genetic information and pro-tein synthesis. These mechanisms involvethe coding of genetic information by nucle-ic acids; transmission of this informationthrough complex molecular devices thatwork at high rates with few errors; tran-scription of this information; translationinto an amino-acid sequence; and finally,post-translational editing of thissequence32,36. Although these devices carryout basically chemical reactions, these reac-tions can no longer be considered as purelychemical processes due to unrestrictedencounter-limited diffusion in a liquidmedium. In fact, cellular compartments canhardly be considered typical liquid media.

The idea now is that ‘structure’ is funda-mental to the operation of modern molecu-lar devices: for example, take the three-dimensional arrangement of individualmolecules; the spatial arrangement of pro-teins in sequential operations; and thearrangement of different proteins in a givenprocess with respect to the membranes sur-rounding subcellular organelles or the cellas a whole. Given the importance of struc-ture, modern biological pathways fullydeserve the names “molecular andsupramolecular machines”36,37.

Ancient versus modern machinesTo an extent, these extraordinary biologicalmachines (BOX 2) realize the dream of the sev-enteenth-century scientists — a dream that

Links

FURTHER INFORMATION Piccolino lab pageENCYCLOPEDIA OF LIFE SCIENCES A. Huxley| Antibody function | Bacterial flagella |Energy cycle in vertebrates | Enzyme kinetics;steady state | E. Starling | History ofbiochemistry | L. Spallanzani | Nervous andimmune system interactions | Photosynthesis| Protein translation initiation | M. Schleiden | T. Schwann | W. Bayliss | W. Cannon | W. Harvey



transport represent the movement ofcytoskeletal and cytosolic proteins at muchslower rates, and the nature of the carrierstructures for these proteins is not known.Proteins that associate with neurofilamentsand microtubules move in slow component ‘a’at average rates of roughly 0.3–3 mm day–1

(~0.004-0.04 µm s–1), and proteins that associ-ate with microfilaments, as well as many othercytosolic proteins, are transported in slowcomponent ‘b’ at average rates of roughly 2–8mm day–1 (~0.02–0.09 µm s–1) (TABLE 1).

No movement en masseIn radioisotopic pulse-labelling experiments,slow components ‘a’ and ‘b’ form unimodalasymmetrical waves, often loosely describedas ‘bell-shaped’, which spread as they movealong the axon towards the axon tip (FIG. 1).Each wave represents the concerted move-ment of many distinct proteins whose indi-vidual waveforms coincide. Early studies onslow axonal transport stressed the coherenceof these transport waves but not the spread-ing, and this gave rise to the idea thatcytoskeletal and cytosolic proteins movealong the axon en masse, that is, in a slow andsynchronous manner1.

The expectation of a slow and synchronousmovement has had a profound influence onthe design of experiments aimed at detectingslow axonal transport. For example, manystudies have used fluorescence photobleachingor photoactivation strategies in which fluores-cent or caged fluorescent cytoskeletal proteinsare injected into nerve cells and then a popula-

34. Sutherland, E. W. Studies on the mechanism of hormoneaction. Science 177, 401–408 (1972).

35. Udrisar, D. & Rodbell, M. Microsomal and cytosolicfractions of guinea pig hepatocytes contain 100-kilodaltonGTP-binding proteins reactive with antisera against alphasubunits of stimulatory and inhibitory heterotrimeric GTP-binding proteins. Proc. Natl Acad. Sci. USA 87, 6321–6325(1990).

36. Alberts, B. The cell as a collection of protein machines:preparing the next generation of molecular biologists. Cell92, 291–294 (1998).

37. Mitchell, P. & Moyle, J. Chemiosmotic hypothesis ofoxidative phosphorylation. Nature 213, 137–139 (1967).

38. Malpighi, M. The Viscerum Structura (Montii, Bologna,1666).

39. Malpighi, M. Dissertatio Epistolica de Formatione Pulli in

Ovo (Martyn, London, 1673).40. Bonnet, C. Considérations sur les Corps Organisés (Rey,

Amsterdam, 1762).41. Boyer, P. D. The ATP synthase — a splendid molecular

machine. Annu. Rev. Biochem. 66, 717–749 (1997).42. Haller, A. Elementa Physiologiae Corporis Humani

(Bousquet, Lausanne, 1757).43. Rastogi, V. K. & Girvin, M. E. Structural changes linked to

proton translocation by subunit c of the ATP synthase.Nature 402, 263–268 (1999).

AcknowledgementsThis article has benefited from discussions with A. Cattaneo ofthe International School for Advanced Studies (S.I.S.S.A.) ofTrieste, and has been made possible by bibliographical help fromL. Iannucci of the University of Pisa. I also thank L. Galli-Resta, A.Pignatelli and B. Pelucchi for critically reading the manuscript.


Efforts to observe the slow axonal transportof cytoskeletal polymers during the pastdecade have yielded conflicting results, andthis has generated considerable controversy.The movement of neurofilaments has nowbeen seen, and it is rapid, infrequent andhighly asynchronous. This motile behaviourcould explain why slow axonal transport haseluded observation for so long.

Neurons communicate with other cells byextending cytoplasmic processes called axonsand dendrites. Remarkably, axons can attainlengths of one metre or more, although theylack ribosomes and Golgi complexes. Axonalproteins and Golgi-derived vesicles are formedin the neuronal cell body and are shippedalong the axon by a process called axonaltransport. This movement is essential for thegrowth and survival of axons, and continuesthroughout the life of the nerve cell.

Studies on axonal transport in laboratoryanimals with radioisotopic pulse labellinghave shown that there are hundreds of axonal-ly transported proteins, but that these proteinsmove at a small number of discrete rates,which can be categorized as either fast or slow.Each discrete rate component represents themovement of a largely distinct subset of pro-teins that are transported together throughouttheir journey along the axon. To explain theseobservations, Lasek and colleagues proposedthe structural hypothesis of axonal transport,which postulates that all axonal proteins moveby association with, or as integral parts of, sub-cellular carrier structures1. According to thishypothesis, each rate component represents

Slow axonal transport: stop and gotraffic in the axon

Anthony Brown

O P I N I O N

the movement of a unique type of macromol-ecular structure(TABLE 1).

The fast components of axonal transportare now known to represent the anterogradeand retrograde movement of distinct types ofmembranous organelles along microtubules ataverage rates of roughly 50–400 mm day–1

(~0.5–5 µm s–1), propelled by the action ofmolecular motor proteins2. Membranousorganelles can therefore be considered to bethe carrier structures for fast axonal transport.In contrast, the slow components of axonal

Table 1 | The moving structures of axonal transport*

Rate class Average rate Moving structures Composition(selected examples)

Fast components

Fast anterograde 200–400 Golgi-derived vesicles Synaptic vesicle proteins,mm day–1 and tubules kinesin, enzymes of (≈2–5 µm s–1) (secretory pathway) neurotransmitter metabolism

Bi-directional 50–100 Mitochondria Cytochromes, enzymes ofmm day–1 oxidative phosphorylation(≈0.5–1 µms–1)

Fast retrograde 200–400 Endosomes, lysosomes Internalized membranemm day–1 (endocytic pathway) receptors, neurotrophins,(≈2–5 µm s–1) active lysosomal hydrolases

Slow components

Slow component ‘a’ 0.3–3 Neurofilaments, Neurofilament proteins, mm day–1 microtubules‡ tubulin, spectrin, tau proteins

Slow component ‘b’ 2–8 Microfilaments, Actin, clathrin, dynein,mm day–1 supramolecular dynactin, glycolytic (≈0.02–0.09 µm s–1) complexes of the enzymes

cytosolic matrix

*Data compiled from REFS 1,41,44. ‡ In some neurons, microtubule proteins are transported in slowcomponent ‘b’ as well as slow component ‘a’.



transport represent the movement ofcytoskeletal and cytosolic proteins at muchslower rates, and the nature of the carrierstructures for these proteins is not known.Proteins that associate with neurofilamentsand microtubules move in slow component ‘a’at average rates of roughly 0.3–3 mm day–1

(~0.004-0.04 µm s–1), and proteins that associ-ate with microfilaments, as well as many othercytosolic proteins, are transported in slowcomponent ‘b’ at average rates of roughly 2–8mm day–1 (~0.02–0.09 µm s–1) (TABLE 1).

No movement en masseIn radioisotopic pulse-labelling experiments,slow components ‘a’ and ‘b’ form unimodalasymmetrical waves, often loosely describedas ‘bell-shaped’, which spread as they movealong the axon towards the axon tip (FIG. 1).Each wave represents the concerted move-ment of many distinct proteins whose indi-vidual waveforms coincide. Early studies onslow axonal transport stressed the coherenceof these transport waves but not the spread-ing, and this gave rise to the idea thatcytoskeletal and cytosolic proteins movealong the axon en masse, that is, in a slow andsynchronous manner1.

The expectation of a slow and synchronousmovement has had a profound influence onthe design of experiments aimed at detectingslow axonal transport. For example, manystudies have used fluorescence photobleachingor photoactivation strategies in which fluores-cent or caged fluorescent cytoskeletal proteinsare injected into nerve cells and then a popula-

action. Science 177, 401–408 (1972).35. Udrisar, D. & Rodbell, M. Microsomal and cytosolic

fractions of guinea pig hepatocytes contain 100-kilodaltonGTP-binding proteins reactive with antisera against alphasubunits of stimulatory and inhibitory heterotrimeric GTP-binding proteins. Proc. Natl Acad. Sci. USA 87, 6321–6325(1990).

36. Alberts, B. The cell as a collection of protein machines:preparing the next generation of molecular biologists. Cell92, 291–294 (1998).

37. Mitchell, P. & Moyle, J. Chemiosmotic hypothesis ofoxidative phosphorylation. Nature 213, 137–139 (1967).

38. Malpighi, M. The Viscerum Structura (Montii, Bologna,1666).

39. Malpighi, M. Dissertatio Epistolica de Formatione Pulli inOvo (Martyn, London, 1673).

40. Bonnet, C. Considérations sur les Corps Organisés (Rey,Amsterdam, 1762).

41. Boyer, P. D. The ATP synthase — a splendid molecularmachine. Annu. Rev. Biochem. 66, 717–749 (1997).

42. Haller, A. Elementa Physiologiae Corporis Humani(Bousquet, Lausanne, 1757).

43. Rastogi, V. K. & Girvin, M. E. Structural changes linked toproton translocation by subunit c of the ATP synthase.Nature 402, 263–268 (1999).

AcknowledgementsThis article has benefited from discussions with A. Cattaneo ofthe International School for Advanced Studies (S.I.S.S.A.) ofTrieste, and has been made possible by bibliographical help fromL. LIannucci of the University of Pisa. I also thank L. Galli-Resta,A. Pignatelli and B. Pelucchi for critically reading the manuscript.


Efforts to observe the slow axonal transportof cytoskeletal polymers during the pastdecade have yielded conflicting results, andthis has generated considerable controversy.The movement of neurofilaments has nowbeen seen, and it is rapid, infrequent andhighly asynchronous. This motile behaviourcould explain why slow axonal transport haseluded observation for so long.

Neurons communicate with other cells byextending cytoplasmic processes called axonsand dendrites. Remarkably, axons can attainlengths of one metre or more, although theylack ribosomes and Golgi complexes. Axonalproteins and Golgi-derived vesicles are formedin the neuronal cell body and are shippedalong the axon by a process called axonaltransport. This movement is essential for thegrowth and survival of axons, and continuesthroughout the life of the nerve cell.

Studies on axonal transport in laboratoryanimals with radioisotopic pulse labellinghave shown that there are hundreds of axonal-ly transported proteins, but that these proteinsmove at a small number of discrete rates,which can be categorized as either fast or slow.Each discrete rate component represents themovement of a largely distinct subset of pro-teins that are transported together throughouttheir journey along the axon. To explain theseobservations, Lasek and colleagues proposedthe structural hypothesis of axonal transport,which postulates that all axonal proteins moveby association with, or as integral parts of, sub-cellular carrier structures1. According to thishypothesis, each rate component represents

Slow axonal transport: stop and gotraffic in the axon

Anthony Brown

O P I N I O N

the movement of a unique type of macromol-ecular structure(TABLE 1).

The fast components of axonal transportare now known to represent the anterogradeand retrograde movement of distinct types ofmembranous organelles along microtubules ataverage rates of roughly 50–400 mm day–1

(~0.5–5 µm s–1), propelled by the action ofmolecular motor proteins2. Membranousorganelles can therefore be considered to bethe carrier structures for fast axonal transport.In contrast, the slow components of axonal

Table 1 | The moving structures of axonal transport*

Rate class Average rate Moving structures Composition(selected examples)

Fast components

Fast anterograde 200–400 Golgi-derived vesicles Synaptic vesicle proteins,mm day–1 and tubules kinesin, enzymes of (≈2–5 µm s–1) (secretory pathway) neurotransmitter metabolism

Bi-directional 50–100 Mitochondria Cytochromes, enzymes ofmm day–1 oxidative phosphorylation(≈0.5–1 µms–1)

Fast retrograde 200–400 Endosomes, lysosomes Internalized membranemm day–1 (endocytic pathway) receptors, neurotrophins,(≈2–5 µm s–1) active lysosomal hydrolases

Slow components

Slow component ‘a’ 0.3–3 Neurofilaments, Neurofilament proteins, mm day–1 microtubules‡ tubulin, spectrin, tau proteins

Slow component ‘b’ 2–8 Microfilaments, Actin, clathrin, dynein,mm day–1 supramolecular dynactin, glycolytic (≈0.02–0.09 µm s–1) complexes of the enzymes

cytosolic matrix

*Data compiled from REFS 1,41,44. ‡ In some neurons, microtubule proteins are transported in slowcomponent ‘b’ as well as slow component ‘a’.




that the extent of bleaching in the photo-bleaching studies on neurofilament proteinswas only partial, reducing the fluorescenceintensity in the axon to 20–50% of its initialvalue6. If the residual unbleached fluores-cence in the bleached region exceeded the flu-orescence intensity of a single neurofilament,then it is likely that the movement of neuro-filaments across the bleached zone could havegone unnoticed. This could also apply to thephotobleaching studies on actin and tubu-lin3–5,9,12,13. For example, in the study of Okabeand Hirokawa12 on tubulin, photobleachingreduced the fluorescence intensity in the axonto 10–40% of its initial value and, in othersimilar studies, researchers have estimatedthat as much as 10–20% of the total tubulinin the axon could have moved through thephotobleached gaps without detection4.Similar detection limits have also been esti-mated for the fluorescence photoactivationtechnique8. The ability of the photobleachingexperiments to detect the movement ofcytoskeletal polymers may also have beenhampered by the short length of the bleachedregions (3–5 µm), and the relatively longtime-lapse intervals (typically five minutes ormore). These considerations suggest that thephotobleaching and photoactivation strate-gies should be capable of detecting the slowaxonal transport of cytoskeletal proteins ifthey could be optimized to enable the detec-tion of single rapidly moving polymers.

in axons has been a vexing problem, but themost likely explanation is that cytoskeletal pro-teins do not move en masse in axons after all.

Neurofilaments move in fits and startsA recent breakthrough in the study of slowaxonal transport has come from observationson neurofilament proteins, tagged with greenfluorescent protein (GFP), in cultured ratsympathetic neurons14,15. These cultured neu-rons contain relatively few neurofilaments andfrequently show discontinuities in their axonalneurofilament array, resulting in short seg-ments of axon that lack neurofilaments14.Time-lapse imaging of these naturally occur-ring gaps in the axonal neurofilament arrayhas enabled the observation of axonal trans-port without the need for photobleaching orphotoactivation approaches. Contrary toexpectations, neurofilaments move rapidly,with peak rates as high as 3 µm s–1, and thesemovements are frequently interrupted by pro-longed pauses14,15 (FIG. 2). The average velocityexcluding the pauses is about 0.2–0.3 µm s–1.Assuming an average transport rate in therange of 0.3–3 mm day–1 (TABLE 1), we can esti-mate that individual neurofilaments spend83–99% of their time pausing during theirjourney down the axon.

Radioisotopic pulse labelling studies in themouse optic nerve led Nixon andLogvinenko17 to propose almost 15 years agothat there are two kinetically distinct popula-tions of neurofilament proteins in axons, onethat moves and one that is stationary.According to this model, neurofilament poly-mers or oligomers exchange between the mov-ing and stationary phase as they move alongthe axon18. On the other hand, Lasek and col-leagues16,19 have challenged this hypothesis,arguing that there is a single population ofneurofilaments in axons that all move relent-lessly, but at a broad range of rates. In princi-ple, the alternating movements and pausesobserved for GFP-tagged neurofilaments incultured neurons14,15 could be regarded astransitions between two distinct moving andstationary phases, or simply as the intermittentmovements of a single population of neurofil-aments that move at a broad and continuousrange of rates. Further studies will be requiredto distinguish between these two possibilities.

Re-evaluating previous approachesWhy have previous attempts6,7 to observe theaxonal transport of neurofilament proteinsusing fluorescence photobleaching failed toreveal movement? One possible explanation isthat those studies were not capable of detect-ing the rapid movement of single cytoskeletalpolymers. For example, it is important to note

tion of these proteins is marked by bleachingor activating the fluorescence in a narrow bandacross the axon [see supplementary figureonline]. In these experiments, a slow and syn-chronous movement should be manifested asa slow translocation of the marked zonetowards the axon tip. However, most studieson tubulin, actin and neurofilament proteinsusing one or both of these techniques showedthat the marked zone does not move3–9.Although gradual recovery of the fluorescencewas observed after photobleaching, it had noobvious directionality and was thereforeattributed to exchange between the bleachedpolymers and diffusible fluorescent subunits.Directional movement of the photobleachedor photoactivated zone was observed in cul-tured frog neurons10–12, but it probably resultedfrom stretching of the axon owing to the rapidgrowth and poor adhesion of these neuronson laminin substrates11,13. The repeated failureof so many efforts to demonstrate slow syn-chronous movement of cytoskeletal proteins

Figure 2 | A neurofilament on the move. Time-lapse images of a neurofilament moving through anaturally occurring gap in the axonal neurofilamentarray of a cultured nerve cell. The neurofilamentswere visualized using green fluorescent protein(GFP)-tagged neurofilament protein M. Thefluorescence images are shown in invertedcontrast for greater clarity. Scale bar= 5 µm.(Figure adapted from REF. 14.) (See movie online).

0

5

10

15

20

25

30

Tim

e (s

)

Anterograde

Figure 1 | Kinetics of slow axonal transport.Diagram illustrating the kinetics of slow axonaltransport, as revealed by radioisotopic pulselabelling. a | Radioactive amino acids injected intothe vicinity of the neuronal cell body produce atransient pulse of newly synthesized radioactiveproteins, which b,c | move together along the axonby axonal transport. After a time interval rangingfrom hours to months, the animal is killed, thenerve is excised and sliced into segments, andeach segment is analysed biochemically to identifythe radioactive proteins. The pulse-labelledproteins form an asymmetrical wave (red) thatspreads as it moves along the axon.

a

b

c



cytosolic proteins move, but also the motorsthat move them, and the substrates that theyinteract with. Potential substrates for motor-driven movements of cytoskeletal polymersin axons include the plasma membrane, theendoplasmic reticulum and other cytoskeletalfilaments21. In principle, neurofilaments couldmove by direct interaction with molecularmotors, or they could ride ‘piggyback’ byattachment to other moving structures.Evidence for a direct interaction has comefrom a recent report that neurofilamentspurified from bovine spinal cord can moverapidly along microtubules in an ATP-dependent manner in vitro37, at peak rates ofup to 1 µm s–1. A similar motile mechanismhas also been described for vimentin fila-ments and their precursors along micro-tubules in non-neuronal cells38.

Slow axonal transport has generally beenassumed to be exclusively anterograde, mov-ing towards the axon tip, but in the observa-tions on GFP-tagged neurofilamentsdescribed above, about 20–30% of theobserved filaments actually moved in a retro-grade direction, towards the cell body14,15.Similarly, in the in vitro study describedabove, neurofilaments were observed to movetowards the minus as well as the plus ends ofmicrotubules37. One possible explanation isthat the retrogradely moving neurofilamentsrepresent a distinct population, as proposedby Griffin and colleagues39 on the basis oftheir studies on the redistribution ofcytoskeletal proteins in transected peripheralnerves. Alternatively, the retrograde move-ments could represent transient reversals of

Why so slow?The rate of slow axonal transport in radioiso-topic pulse labelling experiments is generallyquoted as the rate of movement of the wavepeak, but the spreading of the transport waveindicates that the radiolabelled proteins actu-ally move at a broad range of rates16,19(FIG. 1).The motile behaviour of GFP-tagged neurofil-aments described above suggests a model forslow axonal transport that can account for theslow rate and the spreading of the transportwave. Consider a pulse of radioactive neurofil-ament proteins that assemble into filaments inthe neuronal cell body20. Let us assume thateach neurofilament moves rapidly along theaxon but that the overall rate of movement isslow because the filaments spend a large pro-portion of their time pausing. By chance, orperhaps due to intrinsic differences, some fila-ments move more frequently than others, andthis causes the population to spread out as itmoves along the axon. The frequency withwhich filaments move, or the amount of timethat they spend pausing, could be determinedsimply by proximity to the transport machin-ery or substrate, or by local variations in theresistance to movement, or by some regulato-ry process. Neurofilaments that move mostoften will end up at the leading edge of thetransport wave19, whereas neurofilaments thatmove least often will end up at the trailingedge16. According to this hypothesis, the trans-port wave represents the distribution of manythousands of neurofilaments whose individ-ual movements and pauses are summed overthe days, weeks or months that they spendtravelling down the axon (FIG. 1).

Polymers as carrier structuresThe mechanism of slow axonal transport hasbeen debated for almost 15 years, and most ofthe controversy has focused on the structuralform in which the cytoskeletal subunit pro-teins move21,22. Some studies have concludedthat cytoskeletal proteins move as assembledpolymers23–28, and some have concluded thatthey move as unassembled subunits28–32, butnone has been conclusive. The observations onGFP-tagged neurofilaments describedabove14,15 have shown unequivocally that neu-rofilament polymers do move in axons.Whether actin and tubulin also move in theform of assembled polymers in axons remainsto be determined, although there is clearlyprecedent for such movements in non-neu-ronal cells (for example, REFS 33,34).Microtubule polymers have been observed tomove in growth cones and developing axonalbranches of cultured neurons35, whereasexperiments using fluorescence-specklemicroscopy have not detected any

movement36. If the motility of microtubules isas rapid and infrequent as for neurofilaments,then it is possible that their movement mighthave gone undetected using the specklingtechnique.

Slow axonal transport represents themovement of a myriad of other cytosolic pro-teins in addition to cytoskeletal proteins(TABLE 1). One attractive hypothesis is thatthese cytosolic proteins are transported byforming physical associations with movingcytoskeletal polymers1. The relatively simpleprotein composition of slow component ‘a’indicates that neurofilaments and micro-tubules could be the sole carrier structures forthis rate component; all of the proteins thatmove in slow component ‘a’ are either integralparts of these cytoskeletal polymers or areknown to associate with these polymers invivo. In contrast, the protein composition ofslow component ‘b’ is extremely complex andincludes more than 200 proteins, many ofwhich are traditionally described as ‘soluble’1.The presence of actin indicates that microfila-ments could function as carrier structures forthis rate component. However, given the largenumber of diverse proteins in slow compo-nent ‘b’, it is likely that the carrier structures forthis rate component are complex and maycomprise several macromolecular complexesthat move by direct or indirect associationwith the moving microfilaments.

Motors and substratesTo understand the mechanism of slow axonaltransport, we must identify not only thestructural forms in which cytoskeletal and


Figure 3 | A model for the movement of neurofilaments in axons. In this model, neurofilaments areconsidered to move bidirectionally along microtubules through the action of a plus-end-directed motorsuch as a kinesin-related protein, and a minus-end-directed motor such as dynein. Note that axonalmicrotubules are all orientated with their plus-ends distal, towards the axon tip. Only a small fraction of theaxonal neurofilaments move at any one point in time.

Proximal Distal

Microtubule

Neurofilament

Retrograde motor (minus-end-directed)

Anterograde motor (plus-end-directed)

– +

– +




(1997).22. Hirokawa, N., Terada, S., Funakoshi, T. & Takeda, S.

Slow axonal transport: the subunit transport model.Trends Cell Biol. 7, 384–388 (1997).

23. Terasaki, M., Schmidek, A., Galbraith, J. A., Gallant, P. E.& Reese, T. S. Transport of cytoskeletal elements in thesquid giant axon. Proc. Natl Acad. Sci. USA 92,11500–11503 (1995).

24. Ahmad, F. J. & Baas, P. W. Microtubules released fromthe neuronal centrosome are transported into the axon.J. Cell Sci. 108, 2761–2769 (1995).

25. Yu, W., Schwei, M. J. & Baas, P. W. Microtubuletransport and assembly during axon growth. J. Cell Biol.133, 151–157 (1996).

26. Slaughter, T., Wang, J. & Black, M. M. Microtubuletransport from the cell body into the axons of growingneurons. J. Neurosci. 17, 5807–5819 (1997).

27. Ahmad, F. J., Echeverri, C. J., Vallee, R. B. & Baas, P. W.Cytoplasmic dynein and dynactin are required for thetransport of microtubules into the axon. J. Cell Biol. 140,391–401 (1998).

28. Galbraith, J. A., Reese, T. S., Schlief, M. L. & Gallant, P. E. Slow transport of unpolymerized tubulin andpolymerized neurofilament in the squid giant axon. Proc.Natl Acad. Sci. USA 96, 11589–11594 (1999).

29. Terada, S., Nakata, T., Peterson, A. C. & Hirokawa, N.Visualization of slow axonal transport in vivo. Science273, 784–788 (1996).

30. Funakoshi, T., Takeda, S. & Hirokawa, N. Active transportof photoactivated tubulin molecules in growing axonsrevealed by new electron microscopic analyses. J. CellBiol. 133, 1347–1354 (1996).

31. Miller, K. W. & Joshi, H. C. Tubulin transport in neurons.J. Cell Biol. 133, 1355–1366 (1996).

32. Yabe, J. T., Pimenta, A. & Shea, T. B. Kinesin-mediatedtransport of neurofilament protein oligomers in growingaxons. J. Cell Sci. 112, 3799–3814 (1999).

33. Cao, L.-G. & Wang, Y.-L. Mechanism of the formation ofcontractile ring in dividing cultured animal cells. I.Recruitment of preexisting actin filaments into thecleavage furrow. J. Cell Biol. 110, 1089–1095 (1990).

34. Keating, T. J., Peloquin, J. G., Rodionov, V. I.,Momcilovic, D. & Borisy, G. G. Microtubule release fromthe centrosome. Proc. Natl Acad. Sci. USA 94,5078–5083 (1997).

35. Dent, E. W., Callaway, J. L., Szebenyi, G., Baas, P. W. &Kalil, K. Reorganization and movement of microtubulesin axonal growth cones and developing interstitialbranches. J. Neurosci. 19, 8894–8908 (1999).

36. Chang, S., Svitkina, T. M., Borisy, G. G. & Popov, S. V.Speckle microscopic evaluation of microtubule transportin growing nerve processes. Nature Cell Biol. 1, 399–403(1999).

37. Shah, J. V., Flanagan, L. A., Janmey, P. A. & Leterrier, J.-F. Bidirectional translocation of neurofilaments alongmicrotubules mediated in part by dynein/dynactin. Mol.Biol. Cell (In the press).

38. Prahlad, V., Yoon, M., Moir, R. D., Vale, R. D. & Goldman,R. D. Rapid movements of vimentin on microtubuletracks: Kinesin-dependent assembly of intermediatefilament networks. J. Cell Biol. 143, 159–170 (1998).

39. Glass, J. D. & Griffin, J. W. Retrograde transport ofradiolabeled cytoskeletal proteins in transected nerves.J. Neurosci. 14, 3915–3921 (1994).

40. Koehnle, T. J. & Brown, A. Slow axonal transport ofneurofilament protein in cultured neurons. J. Cell Biol.144, 447–458 (1999).

41. Susalka, S. J., Hancock, W. O. & Pfister, K. K. Distinctcytoplasmic dynein complexes are transported bydifferent mechanisms in axons. Biochim. Biophys. Acta1496, 76–88 (2000).

42. Yabe, J. T., Jung, C. W., Chan, W. K. H. & Shea, T. B.Phospho-dependent association of neurofilamentproteins with kinesin in situ. Cell Motil. Cytoskeleton 45,249–262 (2000).

43. Elluru, R. G., Bloom, G. S. & Brady, S. T. Fast axonaltransport of kinesin in the rat visual system: Functionalityof kinesin heavy chain isoforms. Mol. Biol. Cell 6, 21–40(1995).

44. Dahlstrom, A. B., Czernik, A. J. & Li, J. Y. Organelles infast axonal transport — what molecules do they carry inanterograde vs retrograde directions, as observed inmammalian systems. Mol. Neurobiol. 6, 157–177 (1992).

AcknowledgementsThe author thanks Ray Lasek and Peter Baas for stimulating

discussions.

ity to observe the slow axonal transport ofneurofilament polymers in living axons nowpermits, for the first time, direct analysis ofthe molecular mechanism of this remarkable,and once intractable, motile phenomenon.

Anthony Brown is at the Neuroscience Program,Department of Biological Sciences, Ohio

University, Athens, Ohio 45701, USA. e-mail:[email protected]

1. Lasek, R. J., Garner, J. A. & Brady, S. T. Axonal transportof the cytoplasmic matrix. J. Cell Biol. 99, S212–S221(1984).

2. Hirokawa, N. Kinesin and dynein superfamily proteins andthe mechanism of organelle transport. Science 279,519–526 (1998).

3. Lim, S.-S., Sammak, P. J. & Borisy, G. G. Progressiveand spatially differentiated stability of microtubules indeveloping neuronal cells. J. Cell Biol. 109, 253–263(1989).

4. Lim, S.-S., Edson, K. J., Letourneau, P. C. & Borisy, G. G.A test of microtubule translocation during neuriteelongation. J. Cell Biol. 111, 123–130 (1990).

5. Okabe, S. & Hirokawa, N. Turnover of fluorescentlylabelled tubulin and actin in the axon. Nature 343,479–482 (1990).

6. Okabe, S., Miyasaka, H. & Hirokawa, N. Dynamics of theneuronal intermediate filaments. J. Cell Biol. 121,375–386 (1993).

7. Takeda, S., Okabe, S., Funakoshi, T. & Hirokawa, N.Differential dynamics of neurofilament-H protein andneurofilament-L protein in neurons. J. Cell Biol. 127,173–185 (1994).

8. Sabry, J., O’Connor, T. P. & Kirschner, M. W. Axonaltransport of tubulin in Ti1 pioneer neurons in situ. Neuron14, 1247–1256 (1995).

9. Takeda, S., Funakoshi, T. & Hirokawa, N. Tubulindynamics in neuronal axons of living zebrafish embryos.Neuron 14, 1257–1264 (1995).

10. Reinsch, S. S., Mitchison, T. J. & Kirschner, M. W.Microtubule polymer assembly and transport duringaxonal elongation. J. Cell Biol. 115, 365–379 (1991).

11. Okabe, S. & Hirokawa, N. Differential behavior ofphotoactivated microtubules in growing axons of mouseand frog neurons. J. Cell Biol. 117, 105–120 (1992).

12. Okabe, S. & Hirokawa, N. Do photobleached fluorescentmicrotubules move? Re-evaluation of fluorescence laserphotobleaching both in vitro and in growing Xenopusaxons. J. Cell Biol. 120, 1177–1186 (1993).

13. Chang, S. H., Rodionov, V. I., Borisy, G. G. & Popov, S. V.Transport and turnover of microtubules in frog neuronsdepend on the pattern of axonal growth. J. Neurosci. 18,821–829 (1998).

14. Wang, L., Ho, C.-L., Sun, D., Liem, R. K. H. & Brown, A.Rapid movement of axonal neurofilaments interrupted byprolonged pauses. Nature Cell Biol. 2, 137–141 (2000).

15. Roy, S. et al. Neurofilaments are transported rapidly butintermittently in axons: implications for slow axonaltransport. J. Neurosci. 20, 6849–6861 (2000).

16. Lasek, R. J., Paggi, P. & Katz, M. J. Slow axonaltransport mechanisms move neurofilaments relentlesslyin mouse optic axons. J. Cell Biol. 117, 607–616 (1992).

17. Nixon, R. A. & Logvinenko, K. B. Multiple fates of newlysynthesized neurofilament proteins: Evidence for astationary neurofilament network distributed non-uniformly along axons of retinal ganglion cells. J. Cell Biol.102, 647–659 (1986).

18. Nixon, R. A. Dynamic behavior and organization ofcytoskeletal proteins in neurons: reconciling old and newfindings. Bioessays 20, 798–807 (1998).

19. Lasek, R. J., Paggi, P. & Katz, M. J. The maximum rate ofneurofilament transport in axons: a view of moleculartransport mechanisms continuously engaged. Brain Res.616, 58–64 (1993).

20. Black, M. M., Keyser, P. & Sobel, E. Interval between thesynthesis and assembly of cytoskeletal proteins incultured neurons. J. Neurosci. 6, 1004–1012 (1986).

21. Baas, P. W. & Brown, A. Slow axonal transport: thepolymer transport model. Trends Cell Biol. 7, 380–384

filaments that all eventually move in a netanterograde direction. If there is a distinctpopulation of retrogradely moving neurofila-ments in axons, previous studies indicate thatit represents no more than 5% of the totaltransported neurofilament protein40.

Bidirectional movement of neurofilamentsalong microtubules could be achieved by aplus-end-directed motor such as kinesin or akinesin-related protein, and a minus-end-directed motor such as cytoplasmic dynein(FIG. 3). Dynein, dynactin and several putativekinesin-related proteins have been identifiedin neurofilament preparations by immun-oblotting, and the retrograde movement ofneurofilaments on microtubules in vitro canbe partly inhibited by pharmacologicalinhibitors of dynein and by monoclonal anti-bodies specific for dynein intermediatechains37. A potential role for dynein as a slowaxonal transport motor is also indicated bythe fact that a substantial proportion of thedynein and dynactin in axons moves in slowcomponent ‘b’41. Less is known about thepotential roles of kinesin and kinesin-relatedproteins in slow axonal transport.Yabe et al.42

have reported that conventional kinesin,which is a known vesicle transport motor,associates with axonally transported neurofil-aments, whereas Elluru et al.43 detected little orno conventional kinesin in either of the slowcomponents. Further investigation of theaxonal transport of kinesin and kinesin-relat-ed proteins is clearly required.

Some questions for the futureThe motile behaviour of neurofilaments inaxons lends support to a general model forslow axonal transport characterized by therapid, infrequent and highly asynchronousmovement of cytoskeletal polymers and theirassociated proteins. Proof of this model willrequire identification of both the structuralforms in which other cytoskeletal and cytoso-lic proteins move in slow axonal transport,and of the kinetics of their movement. Forexample, do actin and tubulin also movealong axons as assembled polymers, and docytoskeletal polymers serve as the carrierstructures for cytosolic proteins? The answersto these questions may shed light on funda-mental organizational principles of the cyto-plasm that are applicable to all eukaryoticcells. Many questions also remain regardingthe axonal transport of neurofilaments. Forexample, do these cytoskeletal polymers asso-ciate directly with motor proteins or do theyride piggyback on other moving structures?And what is the significance of the retro-gradely moving neurofilaments in axons?There is still much to be learned, but our abil-

Links

FURTHER INFORMATION Movies of movingneurofilaments | The Brown lab page


nature reviews - molecular cell biology -

Documents