© 2001 Macmillan Magazines LtdNATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 2 | AUGUST 2001 | 565

Transforming therapies

O N C O G E N E S

One aim of cancer research is to facili-tate rational drug design by dissectingthe molecular basis of malignancies totry and identify commonalities. Breastcancer is the most prevalent malignan-cy among women, and several geneticdefects that are causally linked to thisdisease have been identified. Reportingin Nature, Qunyan Yu and co-workershave dissected the roles of several mol-ecular lesions at the heart of thismalignancy, opening up the prospectof more specific therapeutic interven-tion in the future.

One of the most frequentlyobserved characteristics of breastcancer is overexpression of cyclin D1.This cell-cycle regulatory protein isone of three D-type cyclins, whichinteract with their kinase partners —cyclin-dependent kinases-4 and -6 —to drive cell-cycle progression byphosphorylating (and hence deacti-vating) the tumour suppressorretinoblastoma.

It is well established that overex-pression of cyclin D1 is sufficient toinitiate breast cancer — at least inmice. The same group had previ-ously deleted the cyclin D1 gene, andfound that it had minimal pheno-typic consequences, restricted to adefect in pregnancy-associatedmammary-gland proliferation. Thisraised the possibility that ablation ofcyclin D1 function might lead toattenuation of breast cancers with-out severe side effects in other pro-liferative tissues.

To address this issue, Yu and co-workers crossed mice deleted for thecyclin D1 gene with mice engineeredto overexpress one of a panel oftransforming oncogenes known toinduce breast cancer — c-Myc, c-Neu, v-Ha-Ras and Wnt-1.Remarkably, these mice were com-pletely resistant to induction ofbreast cancers by the Ras and Neuoncogenes. However, the miceremained sensitive to cancersinduced by deregulated Myc andWnt-1.

The authors explain this strikingresult by showing that tumours arisingas a result of overexpressing Ras andNeu express only cyclin D1. Tumoursfrom Myc and Wnt-1 transgenic mice,by contrast, tend to express cyclin D2as well, which seems to compensate forthe function of cyclin D1 in drivingcell-cycle proliferation.

Yu and co-workers also find thatthe dependence on cyclin D1 fortransformation of mammary epithe-lial cells by Neu and Ras only holdsfor breast tissues — Ras transgenicmice with deleted cyclin D1 are stillsusceptible to salivary adenocarcino-mas, and cyclin D1-negative fibrob-lasts are fully transformable by Neuand Ras in vitro. Consistent withtheir model, the authors found thattumours arising from such fibrob-lasts express compensating cyclin D2and cyclin D3. It is well establishedthat the signalling pathway involvingNeu, Ras and mitogen-activated pro-tein kinases leads to the induction ofcyclin D1 gene expression, and theseexperiments show that cyclin D1expression provides the crucial linkbetween this pathway and the cellcycle.

The requirement for cyclin D1 totransduce breast-cancer-inducing pro-liferative signals from the Neu–Ras sig-nalling cascade raises the possibilitythat cyclin D1 could be a specific mol-ecular target for therapy of the subsetof breast cancers that arise from dereg-ulation of this signalling pathway.Conversely, the data imply that breastcancers involving deregulated Myc andWnt-1 are unlikely to respond to anti-cyclin D1 therapy.

Undoubtedly, similar cases of tis-sue-specific transforming pathwayswill be discovered. This raises thehope that development of drugsaimed at targets specifically deregu-lated in cancers, combined with pre-therapeutic screening to provide mol-ecular fingerprints of cancerouslesions, will provide a highly effectiveset of new clinical tools.

Bernd PulvererSenior Editor, Nature

References and linksORIGINAL RESEARCH PAPER Yu, Q., Geng, Y. &Sicinski, P. Specific protection against breastcancers by cyclin D1 ablation. Nature 411,1017–1021 (2001) FURTHER READING Bartek, J. & Lukas, J. Are allcancer genes equal? Nature 411, 1001–1002(2001)

The sensitivity and selectivity of the human hearing organ relyon amplification of vibrations in the cochlea. Variation of thelength of outer hair cells (OHC) in the cochlea in response tochanges in transmembrane potential is central to soundamplification. But how do OHC change their length? Last year,prestin — a novel member of the family of pendrin-relatedtransporters, found in the lateral membrane of OHC — wasdiscovered to be the ‘motor protein’ responsible for OHCelectromotility. Now, Oliver and colleagues report in Science thatintracellular anions are the voltage sensor of prestin.

When sound travels through the cochlea, vibrations deflectthe cilia on the apical side of OHC, opening mechano-sensitiveion channels. The resulting change in membrane potentialinduces cell lengthening in the case of hyperpolarization orshortening if the cell is depolarized. The motile response of OHCis accompanied by prestin-dependent charge movement, whichcan be measured experimentally.

Oliver and colleagues first assumed that a charged amino acidin prestin is likely to underlie the measured charge movement.They mutated the most likely candidates to non-charged residuesand measured the gating current, but the characteristic bell-shaped electrical signature was not affected by any of thesemutations.

The authors then reasoned that if the voltage sensor is notpart of prestin, maybe it is an extrinsic charged particle. Theytested intra- and extracellular ions and found that the chargemovement is in fact due to the movement of cytosolic Cl– orHCO

3– ions across the membrane in response to changes in

membrane potential.A model emerges from this study. Many prestin molecules are

embedded in the lateral membrane of OHC. When thetransmembrane potential changes, it induces movement ofintracellular anions into, or out of, a pocket in prestin, causing aswitch between a wide and a narrow conformation. If all prestinmolecules are in the narrow conformation, the cell shortens,whereas if they are in the wide conformation, the cell lengthens.From a cell biologist’s point of view, prestin is an unusual ‘motorprotein’, as it doesn’t move relative to an anchorage point butrather changes the volume that it takes up within the plasmamembrane, thereby indirectly changing the surface, and hencethe length, of the cell.

Raluca GagescuReferences and links

ORIGINAL RESEARCH PAPER Oliver, D. et al. Intracellular anions as the voltage sensor ofprestin, the outer hair cell motor protein. Science 292, 2340–2343 (2001) FURTHER INFORMATION The authors’ model

Hear, hear

E L E C T R O M OT I L I T Y