© 2001 Macmillan Magazines Ltd490 | JULY 2001 | VOLUME 2 www.nature.com/reviews/molcellbio

H I G H L I G H T S

Wnt wonderlandWnt proteins — mediators ofcell–cell signalling — are inthe news again (see theHighlight on page 488) sowhat better place to expandyour knowledge than to visitthe Wnt homepage,launched and run by RoelNusse?

The homepage itself is wellpresented, beginning with ashort paragraph introducingthe Wnt proteins. This ispositioned above an overviewof the contents, which isfurther subdivided into colour-coded sections coveringdifferent components of theWnt signalling pathway —Wnts, Frizzled, axin anddishevelled, to name ahandful. Within each sectionyou can find information onspecies alignments, links toPubMed and clearlytabulated information aboutexpression patterns,chromosomal location. Forthose seeking structuralsatisfaction, regions of axin,dishevelled and β-catenin canbe viewed in threedimensions using the Cn3Dprogram (easily obtainablefrom NCBI by following thelinks). Another resource to beparticularly commended isthe ‘Wnt reagents andassays’ section, whichprovides invaluableinformation about sources ofWnt antibodies and proteins.Included in this section arestrategies for generatingantisera and advice on howbest to make bioactive Wntproteins.

As no signalling moleculehomepage would becomplete without a signallingcascade, it is reassuring tosee thumbnails on thehomepage, each of whichoffers the option of anattractive full-size diagram.Click on one of the proteinsand you instantly receive adeluge of information —perhaps even slightly visuallyoverwhelming. Overall, thesite provides extremelyrelevant links, includingupcoming meetings (whichare admirably up-to-date),reviews on Wnts and contactdetails of fellow Wnt workers.

Katrin Bussell

WEB WATCH

Understanding how hair follicles are formed and main-tained has always been a challenge, but recent findingsmight help get to the root of this. Reporting in Cell,Huelsken et al. have found that β-catenin has a dual role inhair follicle formation — first it mediates the formation ofhair placodes (the first indication of the site of the futurehair) during embryogenesis, and second it acts in the dif-ferentiation of adult skin cells.

After observing an increase in β-catenin messengerRNA expression in the epithelial part of developing hairplacodes, the authors studied the function of β-catenin inskin and hair development in more detail using Cre/loxtechnology to create a conditional mutant in which β-catenin expression could be specifically deleted from theskin. Although the resultant mutant mice were viable,patches of hairless skin were seen in eight-day-old mice.Over the next two weeks, the rest of the hair grew normally,but was then lost.

Closer inspection revealed the presence of extendedregions of β-catenin-negative epithelium that lacked hairplacodes in 15-day-old mutant mouse embryos, in con-trast to wild-type embryos, in which developing hair pla-codes expressed high levels of β-catenin. This implicatesβ-catenin in the formation of epithelial placodes. Duringthe active phase of the hair cycle (anagen) in wild-typemice, both hair follicle and epidermal keratinocytesexpress β-catenin. In the conditional-mutant mice, how-ever, both the upper part of hair follicles and the epider-mis were devoid of β-catenin. During the catagen(regression and shortening) and telogen (rest) phases ofthe hair cycle in the mutant mice, hair was subsequentlylost, hair canals widened and small cysts appeared. Nonew hair follicles formed, and the cysts grew and became

surrounded by a multilayered epithelium. It thereforeseems that cells from the hair-follicle-derived cysts havelost their follicular characteristics and show epidermaldifferentiation, implying that β-catenin is needed toinduce follicular differentiation.

One population of stem cells in the skin is known togive rise to keratinocyte precursors of the hair follicle andepidermis. So how does β-catenin influence the differentia-tion of these stem cells? And does it do this as a signallingor an adhesion molecule? In early embryos of β-catenin-deficient mice, plakoglobin (γ-catenin) is thought to com-pensate for the adhesive function of β-catenin.Plakoglobin, however, although present in the mutantmice, isn’t thought to compensate for β-catenin signallingin the skin.

Moreover, β-catenin deficiency had no effect on theexpression of tabby and downless — a ligand–receptor sys-tem that is crucial for early placode formation — butexpression of bone morphogenetic protein 2 (bmp2),bmp4 and bmp7 and sonic hedgehog (shh) was lost in theconditional mutants. β-catenin can therefore be placedgenetically downstream of tabby/downless in epithelialplacode formation, but upstream of bmp and shh. It isunclear whether β-catenin targets bmp and shh directly orindirectly, and one of the next challenges will be to under-stand how β-catenin signalling is coordinated with otherpathways involved in stem-cell differentiation.

Katrin Bussell

References and linksORIGINAL RESEARCH PAPER Huelsken, J. et al. β-catenin controls hairfollicle morphogenesis and stem cell differentation in the skin. Cell 104,533–545 (2001) FURTHER READING Oshima, H. et al. Morphogenesis and renewal of hairfollicles from adult multipotent stem cells. Cell 105, 233–245 (2001)

Hair today, gonetomorrow

D E V E LO P M E N T

Cancer cells can maintain theirtelomeres by reactivating telomeraseor using a poorly characterizedtelomerase-independentmechanism. Aylin Rizki and VickiLundblad now report in Nature thata familiar cause of cancer — lack ofmismatch repair — may contributeto this second mechanism.

One function of mismatch repair isto prevent recombination betweenimperfectly homologous sequences.As telomeres contain imperfectrepeats, the authors suspected that

mismatch repair might inhibitrecombination between telomeres.So, they compared the growth ratesof budding yeast lacking EST2 (est2-∆) — which encodes the proteincomponent of telomerase — withmutants lacking both EST2 and amismatch repair gene, MSH2. Asexpected, after about 100generations the growth rate of theest2-∆ mutants was severelyimpaired, but in the double mutantsthis phenotype was less pronounced.Mutations of other mismatch repair

genes also conferred a growthadvantage in an est2-∆ background,but the effect was reversed bydeletion of RAD52, a gene shownpreviously to promote arecombination-dependent pathwayfor telomere maintenance. Couldthe growth advantage simply be asecondary consequence ofmutations caused by defective DNArepair? This is unlikely becauseintroduction of a DNA polymerasemutation that generates mutationsat a similar rate to mismatch repairmutants didn’t provide a growthadvantage.

Are these results relevant to cancer?The authors introduced twodifferent missense mutations into

Mismatch repair comes to an end

T E LO M E R E S

© 2001 Macmillan Magazines Ltd

H I G H L I G H T S

NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 2 | JULY 2001 | 491

In the Golgi, mannose-6-phosphatereceptors (M6PRs) grab newly synthe-sized lysosomal hydrolases by theirsugars to whizz them to their final des-tination, lysosomes. But how do thereceptors themselves know where togo? Two groups now report in Sciencethat the GGA coat adaptors showthem the way.

The lysosomal sorting signal wasidentified almost a decade ago — anacidic cluster followed by a dileucinemotif in the cytoplasmic tail of theM6PR. The AP-1 adaptor complexwas the prime suspect for bindingM6PRs and assembling the clathrincoat. It now turns out that it is, in fact,the most recently discovered adaptors— the GGAs (for Golgi-localized, γ-ear-containing,ARF-binding) — thatbind to the lysosomal targeting signalsof M6PRs and sort them to lysosomes.

GGAs are made of four domains: aVHS domain; a GTP–ARF-bindingdomain (GAT); a hinge that containsbinding sites for clathrin; and a GAEdomain that interacts with γ-synerginand other regulators of coat assembly.The sorting signals of M6PRs interact-ed specifically with the VHS domainsof GGAs in two-hybrid assays, in pull-down assays and by surface plasmonresonance spectroscopy. Specifically,the cation-independent M6PR boundall three human GGAs, whereas the

cation-dependent M6PR preferredGGA1 over GGA3, and did not bindGGA2. Immunofluorescence studiesshowed that M6PRs and GGA1 colo-calize in the trans-Golgi network(TGN) and in the cell periphery, andtime-lapse imaging of live cellsrevealed that GGA1 associates withtubules and vesicles that bud from theTGN.

Kornfeld and colleagues show thatthere is a correlation between reducedin vitro binding of the cation-inde-pendent MPR to GGA2 and mis-sort-ing of lysosomal enzymes. Last,Bonifacino and colleagues provideevidence for a functional involvementof GGAs in sorting of M6PRs, as over-expression of a dominant-negativeGGA1 construct blocks exit of thecation-dependent M6PR from theTGN.

So GGAs probably function asadaptors between M6PRs, ARF,clathrin and regulators of coat assem-bly. But do they recruit receptors intoAP-1–clathrin vesicles or do they buildclathrin-coated vesicles of their own?

Raluca GagescuReferences and links

ORIGINAL RESEARCH PAPERS Puertollano, R.et al. Sorting of mannose 6-phosphate receptorsmediated by the GGAs. Science 292, 1712–1716(2001) | Zhu, Y. et al. Binding of GGA2 to thelysosomal enzyme sorting motif of the mannose 6-phosphate receptor. Science 292, 1716–1718(2001)

yeast MSH2 that occur in humanMSH2 in hereditary nonpolyposiscolorectal cancer. These mutationsenhanced survival of a telomerase-defective strain to a similar extent asMSH2 deletion. Inactivatingtelomerase and mismatch repair inanother yeast, Kluyveromyces lactis— with a pattern of telomeresequence degeneracy similar to thatfound in human cells — yieldedsimilar results, indicating that theseobservations may indeed extrapolateto events occurring in mismatchrepair-defective human cells.

These experiments raise manyquestions for cancer research. Dohuman telomeres undergo morerecombination in the absence of

mismatch repair? Can mismatchrepair-defective cancer cells get bywithout telomerase? What is therelative importance of mutationversus telomere maintenance inthese cells? And can we developtherapeutics that reduce mismatchrepair-defective cells back to thestatus of mere mortals?

Cath BrooksbankEditor, Nature Reviews Cancer

References and linksORIGINAL RESEARCH PAPER Rizki, A. &Lundblad, V. Defects in mismatch repairpromote telomerase-independent proliferation.Nature 411, 713–716 (2001)FURTHER READING Kucherlapati, R. &DePinho, R. A. Telomerase meets its mismatch.Nature 411, 647–648 (2001)WEB SITE Mismatch repair

Giga-sorting in the Golgi

P R OT E I N S O RT I N G IN BRIEF

Impairment of the ubiquitin–proteasome system byprotein aggregation. Bence, N. F. et al. Science 292, 1552–1555 (2001)

Protein aggregates are a prominent feature of manyneurodegenerative diseases. But are these aggregates a cause or aconsequence of neurotoxicity? The authors show here thatoverexpression of proteins that are prone to aggregation — eitherin pericentriolar aggresomes or in amyloid-like fibrils — inhibitsthe ubiquitin–proteasome system. As proteasomal degradation isessential for cell survival, inhibition of this cellular system byprotein aggregation is highly pathogenic.

Specific targeting of insect and vertebrate telomereswith pyrrole and imidazole polyamides.Maeshima, K. et al. EMBO J. 20, 3218–3228 (2001)

The authors report new cell-permeant compounds — polyamidescontaining two, flexibly linked, hairpin DNA-binding moieties —that specifically target telomeres. Fluorescent analogues of thesecompounds colocalize with the telomere-binding proteins TRF1and TRF2. These drugs provide a convenient alternative tocurrent cumbersome methods for estimating telomere length.Polyamides of this type could become useful tools for telomereresearch, and might even have potential as medicinal agents.

Sensitive detection of pathological prion protein bycyclic amplification of protein misfolding.Saborio, G. P. et al. Nature 411, 810–813 (2001)

The infectious prion protein, PrPSc, is thought to work bycatalysing the conversion of the cellular glycoprotein PrPC into itsinfectious form. So far, although there is evidence that prions arepresent in body fluids, it has proved difficult to detect theinfectious prion in patients before death. Here, Saborio et al. useultrasound to free prions from plaques in vitro, which speeds upthe conversion cycle. This might allow diagnosis of the presence ofinfectious prions in the bloodstream.

G-protein signalling through tubby proteins.Santagata, S. et al. Science 292, 2041–2050 (2001)

Loss of tubby function leads to obesity. In this study, tubby isshown to act downstream of G-protein-coupled-receptors topotentially regulate gene transcription. Tubby localizes to theplasma membrane by interacting, through its carboxy-terminal‘tubby’ domain, with phosphatidylinositol-4,5-bisphosphate(PtdIns(4,5)P

2). However, activation of Gα

q-coupled receptors

leads to PtdIns(4,5)P2 hydrolysis — probably by increasing

phospholipase Cβ activity. Tubby then translocates to the nucleus,where it might influence genes involved in energy homeostasis.

S I G N A L T R A N S D U C T I O N

P R I O N S

T E C H N I Q U E

P R OT E I N D E G R A D AT I O N