REVIEWS After 15 years of research into Writ genes, initiated by the pioneering work of Nus.~ and Vannus l, recent events and discoveries have infused tile field with ne'*v questions and opl~)rtunities worthy of further pursuit. A brief account of the history and properties of Wm genes (reviewed in Refs 2-4) is necessary to assess recent developments. The first member of this multi- gene family, tV*ztl, was discovered by Nusse and Vamius I as a pmto-oncogene that encodes a secreted glycoprotein of 41-44 kDa, and that leads to mammary tumors when inappropriately expressed in mot,se 2. The discovery that tile Drosophila ortholog is the segment polarity gene wingless immediately suggested that ver- tebrate Writ1 might also display signaling functions2. Consistent with this likelihood, ectopic expression of Wntl in embryonic cells and tissues elicits a range of responses, for exantple: triggering a duplication of the embryonic axis in Xenopus. thereby promoting R)mlation .of two-headed tadpoles5; eliciting an increase in ratio- genie cells in the ventrict, lar region of the mouse spinal cord6; and stimulating metanephric mesenchyme to dif- ferentiate into glomerular and renal tubt, lar epithelia "7. Elucidation of the normal function of Wntl was adlieved by gene disruption in mice. ,,hich leads to pronounced and specific neural defectss. That a range of tissues have the potential to respond to WNT1 signals suggests that WNT signaling pathways can nomlally be used in these cells in diverse processes. The cloning of W,Itl from a range of species allowed the design of degenera~.e PCR primers, which were used to isolate WntI-related genes, now a multigene family of at least 16 genes in nlouse that are likely to encode secreted glycoproreins 1.3. The vertebrate tVlll genes are expressed largely in tire nen'ous system and meso- derm 4.9, and recent data have shed light on a nmge of developmental processes that are influenced by Wnt genes. Taken with major advances in understanding WNT signaling path'*vays, there is no'*'*'the opportunity for describing, at a biochenlical level, how an extracel- lular WNT signal works through proteins in distinct cel- lular compartments to modulate regulation of specific target genes. Evidence for functionally distinct WNTs The patterns of expression of vertebrate Wm genes have been most thoroughly described in the ntouse. largely by McMahon and colleagues. During gastmlation WIzL:Ja, IDztSa, WntSb (Ref. 9 ) a n d Wilt8 (Ref. 10) are expressed in unique, but overlapping, patterns in the mes(xlerm and tile ectoderm. These patterns of ex- pression, as well as the expression of at least eight I|'~)lt genes in tile central nen,ous systenl3. raise questions as to whether some or all WNTs :ire functionally distinct. and how this might affect cell physiological responses when cells are exposed to multiple WNT signals. Two independent ass:,ys support tile likelihood that there is functional diversity among IDlt gene products. In one assay, Wong et al. II relied on tile eady obser- vation that Wntl can transtbml C571ng routine mamman" epithelial cells, and found that three categories of ll>),,t genes could be dc..~ribed based on their transformation activity. Wntl, Wnt3a and Wnt7a'**,'ere highly transform- ing, Writ2, WntSb and IDzt 7b were intemlediate in their WNTs modulate cell fate and behavior during vertebrate development RANDALL T. MOON(rtmoon~u.washington.edu) JEFFREY D. BROWN Corownj~a.washington.edu) MONICA TORRES (moosey~u.washingto~edu) Wntgenes encode a family of secreted gO, coproteias that modulate cellfote and behavior in embryos through activation o f receptor.mediated signaling pathways, Wnt sequences, patterns of expression and activities are high~ conserved in evolatiat~ so it has been possible to galn insights into tbefunctians, and mechanisms of adia~ of the Wnt genes through a synthesis of genetic and cell biological approaches in different organisms. These studies suggest that there are fundionally distinct WNTproteins as assayed by the ability to transfornt cells and by differences in embryonic responses to ectopic WNTsignals. Moreover, gain-of.function and loss-of-function studies both support the involvement of Writ proteins in modulating ceU fote and cell behavior during vertebrate derelopmen~ often through contbinotorial interoctions with other signaling pathways to regulate gene expresstot~ transffmnation activity, and Writ4, WntSa and Writ6 failed to transform these cells. A second set of assays, based on overexpression of different WNTs in Xenopus embryos, distinguishes two categories of Writ genes 1>14, which Rlllow a similar grouping to the cell trartsfonnation assays. Xenopus Wntl (Xwntl), Xu,llt3a, Xwnt8 and Xwnt,S'b R)nn tile WIztl class that. when ectopically expressed in the ventral marginal zone of cleaving frog entbryos, induces duplicatkm of the embryonic axes. Two murine Wm genes with intemtediate abilities to transfimn cells. Wnt2and Wnt7b, I)oth duplicate embry- onic axes in .\'enolms embn.'os, suggesting they have some cap.'tcity to activate the same pathway as tile 1Dztl class (J.D. Brown and R.T. Moon, unpublished). The XwntSa class of IVIzt genes, which includes XwntSa as well as Xwnt4 and Xwntll. do not induce complete axis duplication. Instead. the products of these Writ genes decrease cell adhesion and perturb morpho- genetic nlovements during gastla.llation in Xenoptts embryos, producing a very distinct phenotype from the R'9.,t I dass 13. The cell transformation and Xenopus axis- induction assays should not be over-interpreted, how- ever, as the phenotype in eadl assay is likely to be very dependent upon which WNT receptors are expressed. Overall, these functional comparisons of ll))lt genes are useful in cautioning investigators not to pmsulne that all WNTs activate the same cell physiological responses, through tile same signaling pathway. These contpari- sons have also hinted at tile possibility that distinct WNT signaling pathways can interact at various levels. Specifically. nlemlvers of the tl:'tltSa cla~ can ad in a cell non-aulononlous nlanner to block tile ability of nleln= bet's of the IDztl class to induce a secondan." axis in TIG APRIL 1997 Vot. 13 No. 4