Meeting Review:
The Wnt meeting 1996
Kenneth M. Cadigan and Roel Nusse
Howard Hughes Medical Institute
Department of Developmental Biology
Beckman Center
Stanford University, Medical Center
Stanford, CA 94305-5428
keywords: Wnt, wingless, signal transduction, development, tumorigenesis.
A meeting on Wnt
genes was held at Stanford University, August 9 and 10, organized by Harold Varmus
(NIH) and Roel Nusse (HHMI-Stanford). This meeting was the first formal version of
a series of informal Wnt gatherings held over the past years at various locations.
The previous meetings were always inspired by the importance of Wnt
genes in animal development and tumorigenesis , but also frustrated by the lack of
understanding of the molecular aspects of signaling by these secreted proteins. Much
to the delight of the attendants, the 1996 meeting brought together a surprising
number of novel findings on various aspects of Wnt signaling. Only half a year ago, very
little was known about the Wnt signal transduction pathway, but recently, specific
receptors and various intracellular components have been discovered and many of these
novel findings were reported at the meeting. It seems that the power of Drosophila
and C. elegans
genetics, combined with in vitro assays for Wnt signaling and rapid biological assays
in Xenopus
will lead to rapid progress in understanding how Wnt molecules bring about their
interesting biological effects.
Members of the frizzled gene family are Wnt receptors
For a long time, a large gap in our understanding of the mechanism of Wnt signaling
was the lack of Wnt receptors. One of the highlights of the meeting concerns the
identification of frizzled proteins as Wnt receptors. These proteins are part of
the large family of seven membrane spanning domain receptors (sometimes referred to as serpentine
receptors). A Drosophila
frizzled gene, Dfz
2, was found to be expressed in the embryo in a pattern reminiscent of some segment
polarity genes, such as wingless
(wg
) [1] (Bhanot-Nathans, Johns Hopkins). Interestingly, Dfz
2 is expressed in a Drosophila
tissue culture cell line that is wg
-responsive, but not in a non-responding one (the assay for wg
activity being the stabilization and subsequent accumulation of the b-catenin homolog, armadillo, Arm). After transfection with the Dfz
2 gene, these cells now are able to transduce the wg
signal. In addition, these cells can now bind wg
protein on their cell surface (Brink-Nusse, Stanford). Transfection of cells with
Dfz
2 constructs lacking either the extracellular or intracellular domain of the protein
demonstrated that the extracellular domain was required for binding. Although a
direct interaction between the wg
protein and Dfz
2 is still lacking (the binding assay used is indirect), the data indicate that Dfz
2 can bind to and transduce the wg
signal [1].
Some, but not all, of the identified mouse frizzled
genes are also positive in the wg
binding assay described above (Bhanot-Nathans). These results, and the fact that
there are no Dfz
2 mutants make it uncertain whether Dfz2 is the physiological wg
receptor. However, the proposal of frizzled molecules functioning as Wnt receptors
was strengthened by recent work from nematodes and frogs. In the nematode C. elegans
, there are two genes involved in asymmetric cell divisions of certain cell lineages.
One of these genes (lin
-44) encodes a Wnt
gene [2]; and the other reported at the meeting (lin
-17) encodes a frizzled
family member [3]. The phenotypes of the two mutants are similar but not identical
[3], but in light of the biochemical evidence summarized above, the current working
model suggests a ligand-receptor relationship.
In Xenopus
embryos, injection of Wnt RNA has long been known to induce a secondary body axis.
Some Wnt
genes lack this activity and it has been suggested that they function through a different
signaling mechanism. However, a report at the meeting (He-Varmus, NIH) demonstrated
that if these Wnts are co-injected with the appropriate frizzled gene's RNA, the axis duplication effect is restored. This suggests that the only difference between
the two classes of frog Wnt
genes may be their affinity for the endogenous frizzled receptor.
The frizzled
gene of Drosophila
was the first member of the family isolated and its genetics has been studied extensively.
It was not thought to be involved in wg
signaling because null fz
mutants had no wg
-like phenotype, rather they have a disruption of the ordered polarity of cells in
the wing, notum and eye. However, the fz
gene can confer wg
responsiveness to non-responding cultured cells, as well as wg
binding (Brink-Nusse) and overexpression of fz
can cause phenotypes in fly embryos that may be a result of activation of wg
signaling (Tomlinson, Columbia). One possibility is that fz
acts redundantly with Dfz
2 or other as yet unidentified frizzled proteins to transduce the wg
signal.
The planar polarity phenotype of fz
mutations is very similar to several other genes, suggesting that their protein products
act in a biochemical pathway. One of these genes is dishevelled
, which is also required for wg
signaling. Do wg
and other components of its signaling pathway function in the planar polarity pathway?
Evidence for and against this idea was presented at the meeting. Just as loss of
fz
or dsh
function causes changes in the polarity of the ommatidia in the eye, overexpression
of these two genes in the eye also disrupts the normal polarity (Tomlinson). Overexpression
of wg
and zeste-white 3
(zw3
) causes weaker polarity reversals. While this may be evidence for wg
playing a role in this process, it is not clear whether these phenotypes are generated
by the same mechanism. Moreover, with the exception of dsh
, there is no loss-of-function evidence for wg
signaling playing a role in cell polarity.
The dsh
gene clearly plays a role in both wg
signaling and planar polarity, but examination of dsh
mutants suggests that it dual role may require different activities. The two functions
can be separated genetically, with a mutation in the COOH terminal region of the
Dsh protein losing its polarity but retaining its wg
signaling function (Axelrod-Perrimon, Harvard). Site directed mutagenesis, while
somewhat complicated, is not consistent with both functions requiring the same biochemical
activity. At the current time, it seems that the fz
gene may function in wg
signaling, but we think it unlikely that planar polarity is achieved through activation
of a typical Wnt signaling pathway. However, it does appear likely that some Drosophila
Wnt
gene is the ligand for the polarity function of fz
.
The role of dishevelled
in Wnt signaling.
Several talks addressed the function of dishevelled
(dsh
) in the Wnt signaling pathway. This gene encodes a highly conserved protein - several
vertebrate homologues have been cloned - but with only a few clues a to its function.
One is the presence of a PDZ domain, shared with a number of other signaling molecules. Interestingly, PDZ domains of some proteins have been shown to interact directly
with the S/TXV motif found in some molecules, including members of the frizzled gene
family. That would suggest that Dsh directly binds to frizzled, or Dfrizzled2, a
possibility made more likely by the report from Miller and Moon (University of Washington,
Seattle) that Xenopus
Dsh protein , when co-expressed with rat frizzled family members in a Xenopus
blastomer, translocates from a cytoplasmic pool to a membrane location (Yang-Snyder,
J., Miller, J. R., Brown, J. D., Lai, C.-J. and Moon, R. T. (1996). A frizzled homolog
functions in a vertebrate Wnt signaling pathway. Current Biology, in press). Whether Dsh functions by binding directly to the tail of Fz protein is not a forgone conclusion
however. Hitoshi Sawa from the lab of Bob Horvitz (MIT) reported that a lin-17 variant
in which the S/TXV motif is replaced by GFP is still active in rescuing the lin-17 mutation [3]. Moreover, the Dsh PDZ domain lacks a positively charged residue
that has been shown to be part of the S/TXV binding pocket in the 3D structure analysis
[4-6].
Recently, it was also recognized that Dsh contains a DEP domain, found in a number
of proteins interacting with G-proteins and/or Protein kinase C [7]. The Dsh protein
is phosphorylated in Drosophila
embryos, and its phosphorylation can be modulated by wingless
signaling [8]. It was therefore of interest that Willert (Nusse lab, Stanford) reported
on a protein kinase associated with Dsh, casein kinase 2 (CK2). This well known enzyme
can phosphorylate Dsh efficiently. Yet another protein kinase may be associated with Dsh in Drosophila
embryos (Sun-Williams, UCSF). Like its Drosophila
counterpart, the mammalian Dsh proteins can also undergo phosphorylation in a Wnt
dependent manner (Brown, Cornell).
zw3/GSK
3b
, a key enzyme in the Wnt pathway.
Originally found in the wingless
pathway in Drosophila
as a negative regulator of armadillo
, the protein kinase zw3 (GSK3b in mammalian cells) has been the focus of much attention. Last year, several labs
showed that dominant negative forms can mimic Wnt signaling in Xenopus
embryos, and that hyperactive forms of the enzyme inhibit the Wnt response, providing
further evidence that this enzyme mediates the Wnt pathway in a wide variety of organisms
[9-11]. It had yet to be demonstrated that the enzymatic activity of zw3/GSK3b is in fact modulated by Wnt signaling, but at the meeting, Cook (lab Trevor Dale
, ICR, London) showed interesting data on the inhibition of the mammalian enzyme
[12]. Adding soluble Wg
protein to C3H10T1/2 cells results in a considerable down regulation, using a sophisticated
assay to measure GSK3b activity. This assay was then further explored to show that a G protein and PKC may
be involved in the regulation of GSK3b, using specific inhibitors of these signaling components. Evidently, the role of
a G protein in Wnt signaling would be consistent with the identification of the frizzled
7 transmembrane proteins as receptors.
Two labs presented data linking the well known effects of lithium on development (such
as axis duplication in Xenopus
and Dictyostelium spore cell differentiation) directly to zw3/GSK3b (Klein, University of Pennsylvania-Melton, Harvard; Stambolic-Woodgett, Toronto)
[13]. They showed that this ion is capable of inhibiting the kinase activity, at
concentrations similar to those active in bioessays for lithium.
In specifying cell fate, GSK3b may interact with several cellular targets. Apart from the recent demonstration
of GSK3b binding to b-catenin and Drosophila
APC ([14], GSK3b may have other binding partners as well. A novel GSK3b binding protein was reported by Sakanaka and Harrison (Chiron).
Catenins as signal transducers
How does zw3/GSK3b work? Possibly, its genetic target, Arm/b-catenin is directly phosphorylated and then targeted for degradation. At the meeting,
David Kimelman (University of Washington) presented evidence for the first step,
showing phosphorylation of b-catenin by GSK3b [15]. Steve Byers (Georgetown) presented interesting new data on b-catenin turnover, showing that ubiquitination and the proteasome pathway are part
of the machinery. He suggested that the role of APC in the b-catenin proteolysis is analogous to the E3 enzyme in the ubiquitin conjugation pathway.
Exciting news on the role of Arm/b-catenin as a partner in binding to a transcription factor was reported by Molenaar
(Lab Clevers, Utrecht). She found that the HMG box protein TCF-1/Lef-1 can associate
with b-catenin, and can bind to target DNA as a complex [16]. In Xenopus
, a dominant negative form of Tcf-1 can inhibit axis formation, providing a link to
a well known biological effect of Wnt expression. These findings are of great interest
in view of the role of this transcription factor in many inductive events in embryogenesis [17]. Moreover, they provide a mechanistic explanation for the detection of
the Arm/b-catenin protein in the nucleus of cells [15, 18]. Catenins are versatile molecules,
however, that are also implicated in cell adhesion and in cell migration in conjunction
with their binding partners E-cadherin and APC (Barth-Nelson, Stanford).
Genetic Screens to identify components of Wnt signaling
Many of the new Wnt signaling components described at the meeting were identified
through cell biology and biochemistry, but several genetic screens performed in Drosophila
are also finding new genes involved with wg
signaling. Two genes encoding enzymes involved in heparin sulfate synthesis have
an identical phenotype to wg
when mutated (Lin-Perrimon, Harvard). These results nicely complement another talk
where the removal of heparin sulfate from cultured cells partially blocks their ability
to respond to wg
(Cumberledge, University of Massachusetts).
A novel kinesin-like gene has a mutant phenotype very similar to embryos where wg
is ubiquitously expressed. This phenotype is dependent on functional wingless
. It is not yet clear whether this kinesin-like protein inhibits wg
signaling or blocks the diffusion of the wg
polypeptide (Dalby-Goldstein, UCSD).
Two genes reported to be involved in wg
signaling were previously found to function in processes unrelated to wg
. The arrow gene appears to be allelic to centrosomin (DiNardo, Rockefeller University)
a protein localizing to the centrosome during mitosis and involved in cell division
[19]. Weaker alleles of arrow have embryonic phenotypes similar to wg
. Since specific regulation of cell division is not thought to be required for normal
segmentation, arrow role in wg
signaling is probably not a result of its centrosomal function. Consistent with
this, centrosomin protein is localized to the cytoplasm in non-mitotic cells [19].
Another potential example of a dual functioning gene is the warts/lats
protein kinase (Cadigan-Nusse, Stanford), which was originally identified because
of the dramatic overgrowth of mutant clones [20]. It also is required for wg
signaling in the eye, but not in other tissues examined thus far. The cell growth
inhibitory role that warts/lats
functions in appears unrelated to its role in tissue specific wg
signaling.
Many other less characterized genes have been identified in the screens described
at the meeting. One immediate challenge will be to determine whether any existing
mutants in these collections correspond to the Wnt signaling proteins identified
biochemically.
More biological effects of Wnts in vivo
The role of Wnt
genes in various biological processes continues to grow. Some of the new functions
for Wnts presented at this meeting include regulation of hair length, synapse maturation
and sexual differentiation of gonads and apoptotis. The remarkable conservation between Wnt signaling pathways in flies and vertebrates was a recurring theme at the
meeting and, while less dramatic, there is some evidence demonstrating similar functions
for Wnts between insects and chordates . Bob Riddle (University of Pennsylvania)
presented some of his work demonstrating that Wnt
7a acts as a signal for dorsal identity in the developing chick limb [21]. The primary
target of Wnt
7a is Lmx-1, a LIM homeodomain protein. In the insect wing, a Wnt
(wingless
) and a LIM homeobox gene (apterous) are involved in dorsal ventral identity. Though
the genetic circuitry is not the same, in both species, a Wnt
gene acts to establish asymmetry in a developing appendage.
These and many other exciting presentations at the 1996 Wnt meeting stimulated the
attendants into calling for a sequel in the not too distant future.
references
1 Bhanot, P., Brink, M., Harryman Samos, C., Hsieh, J.C., Wang, Y.S., Macke, J.P.,
Andrew, D., Nathans, J. and Nusse, R. (1996) Nature 382, 225-230.
2 Herman, M.A., Vassilieva, L.L., Horvitz, H.R., Shaw, J.E. and Herman, R.K. (1995)
Cell 83, 101-110.
3 Sawa, H., Lobel, L. and Horvitz, H.R. (1996) Genes & Dev. 10, 2189-2197.
4 Morais Cabral, J.H., Petosa, C., Sutcliffe, M.J., Raza, S., Byron, O., Poy, F.,
Marfatia, S.M., Chishti, A.H. and Liddington, R.C. (1996) Nature 382, 649-652.
5 Harrison, S.C. (1996) Cell 86, 341-343.
6 Doyle, D.A., Lee, A., Lewis, J., Kim, E., Sheng, M. and MacKinnon, R. (1996) Cell
85, 1067-1076.
7 Ponting, C. and Bork, P. (1996) Trends Biochem. Sci. 21, 245-246.
8 Yanagawa, S., Van Leeuwen, F., Wodarz, A., Klingensmith, J. and Nusse, R. (1995)
Genes & Dev. 9, 1087-1097.
9 Dominguez, I., Itoh, K. and Sokol, S.Y. (1995) Proc Natl Acad Sci USA 92, 8498-8502.
10 He, X., Saintjeannet, J.P., Woodgett, J.R., Varmus, H.E. and Dawid, I.B. (1995)
Nature 374, 617-622.
11 Pierce, S.B. and Kimelman, D. (1996) Dev Biol 175, 256-264.
12 Cook, D., Fry, M., Hughes, K., Sumathiphala, R., Woodgett, J. and Dale, T. (1996)
EMBO J. 15, 4526-4536.
13 Klein, P.S. and Melton, D.A. (1996) Proc Natl Acad Sci USA 93, 8455-8459.
14 Rubinfeld, B., Albert, I., Porfiri, E., Fiol, C., Munemitsu, S. and Polakis, P.
(1996) Science 272, 1023-1026.
15 Yost, C., Torres, M., Miller, J.R., Huang, E., Kimelman, D. and Moon, R.T. (1996)
Genes & Dev. 10, 1443-1454.
16 Molenaar, M., Van de Wetering, M., Oosterwegel, M., Petersonmaduro, J., Godsave,
S., Korinek, V., Roose, J., Destree, O. and Clevers, H. (1996) Cell 86, 391-399.
17 Kratochwil, K., Dull, M., Fariñas, I., Galceran, J. and Grosschedl, R. (1996) Genes
& Dev. 10, 1382-1394.
18 Orsulic, S. and Peifer, M. (1996) J. Cell Biol 134, 1283-1299.
19 Li, K.J. and Kaufman, T.C. (1996) Cell 85, 585-596.
20 Justice, R.W., Zilian, O., Woods, D.F., Noll, M. and Bryant, P.J. (1995) Genes
& Dev. 9, 534-546.
21 Riddle, R.D., Ensini, M., Nelson, C., Tsuchida, T., Jessell, T.M. and Tabin, C.
(1995) Cell 83, 631-640.