Induction of a β-catenin-LEF-1 complex by wnt-1 and transforming mutants of β-catenin

Oncogene (1997) 15, 2833 ± 2839  1997 Stockton Press All rights reserved 0950 ± 9232/97 $12.00 Induction of a b-catenin-LEF-1 complex by wnt-1 and transforming mutants of b-catenin Emilio Por®ri1, Bonnee Rubinfeld1, Iris Albert1, Karine Hovanes2, Marian Waterman2 and Paul Polakis1 1 ONYX Pharmaceuticals, 3031 Research Drive, Richmond, California 94806, USA; and 2Microbiology and Molecular Genetics Departments, University of California Irvine, Irvine, California 92697, USA Signal transduction by b-catenin involves its posttranslational stabilization and import to the nucleus where it interacts with transcription factors. Recent implications for b-catenin signaling in cancer prompted us to examine colon cancer cell lines for the expression of LEF-1, a transcription factor that binds to b-catenin. The analysis of several cell lines revealed the expression of LEF1 mRNA and a constitutive association of the LEF-1 protein with b-catenin. In contrast to the colon cells, PC12 and 293 cells did not contain a b-cateninLEF-1 complex, even though both proteins were detected in cell lysates. In these cells, the association of endogenous LEF1 and b-catenin was induced by stimulation with the wnt-1 proto-oncogene. The complex formed following transient stimulation with wnt-1 and also persisted in cells stably expressing wnt-1. Ectopic overexpression of b-catenin in 293 cells also induced the assembly of the b-catenin-LEF-1 complex and activated gene transcription from a LEF-1-dependent promotor. Expression of mutant oncogenic forms of b-catenin identi®ed in cancer cells resulted in higher levels of transcriptional activity. The results suggest that a cancer pathway driven by wnt-1, or mutant forms of b-catenin, may involve the formation of a persistent transcriptionally active complex of b-catenin and LEF1. Keywords: b-catenin; LEF1; wnt-1; cancer Introduction Signal transduction mediated by b-catenin, or its homolog armadillo, drives cell fate determinations in embryonic development of Xenopus and Drosophila (Gumbiner, 1995; Peifer, 1995). In mammalian cells, several lines of evidence link b-catenin signaling to cancer progression. The wnt-1 proto-oncogene, which stabilizes b-catenin in cell culture (Bradley et al., 1993; Hinck et al., 1994), promotes tumor formation when expressed in mouse mammary tissue (Tsukamoto et al., 1988). In colon cancer, hyperactive b-catenin signaling may result from inactivation of the APC tumor suppressor gene, the product of which downregulates b-catenin in cultured cells (Munemitsu et al., 1995). In tissue sections, b-catenin is grossly overexpressed in adenomatous epithelium relative to neighboring normal epithelial cells (Inomata et al., 1996). b-catenin Correspondence: P Polakis Received 23 May 1997; revised 28 July 1997; accepted 28 July 1997 has also been identi®ed by random expression cloning of cDNAs capable of transforming NIH3T3 cells (Whitehead et al., 1995). Finally, missense mutations identi®ed in the b-catenin gene in cancer cells promote the stabilization of the protein and prevent its downregulation by the APC tumor suppressor (Rubinfeld et al., 1997; Morin et al., 1997). Similar mutations activate b-catenin or armadillo in wnt/wg signalling (Yost et al., 1996; Pai et al., 1997). Although these observations make a strong case for b-catenin as an oncogene, it is not clear how transformation is achieved by the stabilization of the protein. The identi®cation of LEF/TCF transcription factors as downstream targets for b-catenin signaling in embryonic development (Behrens et al., 1996; Huber et al., 1996; Molenaar et al., 1996), suggests that transcription of speci®c target genes may constitute an endpoint in this cancer pathway. The LEF-1 transcription factor was originally identi®ed as a T cell-speci®c high mobility group DNA binding protein that binds to a speci®c motif in the minimal TCRa enhancer region and is required for its activity (Travis et al., 1991; Waterman et al., 1991). More recently, the LEF1 cDNA and a Xenopus homolog XTcf-3 were found to interact with b-catenin at a site localized to the extreme amino-terminal region of LEF-1/XTcf-3 proteins (Behrens et al., 1996; Molenaar et al., 1996). Both of the gene products functionally interacted with b-catenin when tested for axis duplication of Xenopus embryos. A Drosophila homolog of the LEF/TCF gene has also been identi®ed and shown to act downstream of the b-catenin homolog armadillo in the wingless signaling pathway (van de Wetering et al., 1997; Brunner et al., 1997). These observations suggest that in cancer, deregulation of b-catenin may result in its hyperactive interaction with a LEF/TCF transcription factor and thereby activate transcription of genes involved in cell growth control. A constitutive complex of b-catenin and LEF1 has been detected in melanoma cell lines expressing either mutant b-catenin or mutant APC (Rubinfeld et al., 1997). Also, an additional LEF/TCF family member, TCF-4, was identi®ed in colon cancer cells and shown to activate transcription from a reporter plasmid in response to b-catenin expression (Korinek et al., 1997; Morin et al., 1997). Here we demonstrate that LEF1 mRNA and protein are also expressed in colon cancer cell lines and that b-catenin is constitutively associated with it. We have also investigated the e€ects of the wnt-1 proto-oncogene on LEF1 complex formation and demonstrate that wnt-1 induces the association of b-catenin with LEF-1. wnt-1 activation of b-catenin E Porfiri et al 2834 Results Based on the evidence implicating b-catenin in cancer we sought to determine if the b-catenin binding protein LEF-1 was present in colon cancer cell lines. Five of the six cell lines examined contained polypeptides, ranging in molecular weight from *50 000 ± 60 000, that were reactive with antibody to LEF-1 (Figure 1a). It was surprising to ®nd LEF-1 in colon cancer cells, as its expression was thought to be con®ned mostly to lymphocytes and lymphoid organs in adults (Travis et al., 1991; Waterman et al., 1991), and we were unable to detect it in RNA from normal adult colon or intestine (Figure 1b). To con®rm its expression, Northern analysis was performed on total RNA puri®ed from the colon cancer cell lines. A LEF-1 cDNA probe recognized LEF-1 mRNA in total RNA puri®ed from SW480, CMT93, colo320, DLD1 and colo205 cells (Figure 1c). A 3.4 kb mRNA, previously detected in Jurkat cells (Waterman et al., 1991), was identi®ed in the colon cancer cells along with a novel 2.7 kb mRNA not seen in the Jurkat cells. We also isolated LEF-1 cDNAs derived from reverse transcription of mRNA puri®ed from SW480 and DLD-1 cells. An entire open reading frame was sequenced and the predicted transcript matched a LEF-1 mRNA isoform containing an 84 bp in-frame deletion (data not shown) previously identi®ed in Jurkat cells (Waterman et al., 1991). To look for an interaction between b-catenin and LEF-1 proteins, b-catenin immunoprecipitates were β-cat. — CM T93 Colo320 Colo205 DLD-1 SW480 Jurkat Hela colo205 DLD-1 SW948 colo320 CM T 93 c SW480 a analysed for LEF-1 by immunoblotting. A b-cateninLEF-1 complex was detected in the lysates from the SW480, colo320, DLD-1 and colo205 cell lines (Figure 2a). The coimmunoprecipitation of LEF-1 with bcatenin was also con®rmed by southwestern analysis. Beta-catenin immunoprecipitates resolved on SDSpolyacrylamide gels were transferred to ®lter membranes that were sequentially probed with a radiolabeled oligonucleotide containing a LEF-1 binding site and then with a LEF-1 antibody. The 32P-labeled oligonucleotide bound to a protein of the same mobility as that recognized by the LEF-1 antibody (Figure 2b). The association of b-catenin with LEF-1 in cancer cells may be conditional and dependent upon protein modi®cations or subcellular location. Some cell lines, such as the Jurkat T lymphocyte, contained ample amounts of both LEF-1 and b-catenin, but no complex was identi®ed (Figure 2c). This was not due to an inability of the LEF-1 from Jurkat cells to bind bcatenin, as LEF-1 was anity precipitated from lysates by puri®ed Glu-Glu epitope-tagged b-catenin (Figure 2c). LEF-1 was also recovered from SW480 lysates by anity precipitation, however the majority of the LEF1 in these lysates was pre-complexed with endogenous b-catenin as determined by its coimmunoprecipitation with b-catenin antibodies. The lack of an endogenous b-catenin-LEF-1 complex in the Jurkat cell suggested b-catenin may not have access to LEF-1. We have previously noted that the cytosolic distribution and stability of b-catenin correlated with 28S — LEF1 LEF1 18S — PBL GAPDH colon intestine ovary testis prostate thym us spleen b 7.5 — 4.4 — 2.4 — Figure 1 Expression of LEF-1 in colon cancer cells. (a) Protein-equivalent amounts of total lysates from the indicated cell lines were subjected to SDS ± PAGE and immunoblotting. The blot was cut horizontally and the sections incubated separately with antibody speci®c to b-catenin (top) and LEF-1 (bottom). (b) Northern blot analysis for LEF1 in RNA from the indicated adult human tissues. The blot was purchased from Clonetech and hybridized to a LEF-1-speci®c cDNA probe. (c) Northern blot of total RNA hybridized to a LEF-1-speci®c cDNA probe (top). To control for RNA levels, the same blot was probed for GAPDH expression (bottom). Migration of ribosomal RNA 18S and 28S is indicated at left of panel wnt-1 activation of b-catenin E Porfiri et al α-β-cat. cont. α-β-cat. cont. cont. αβ-cat. cont. αβ-cat. b cont. αβ-cat. cont. αβ-cat. cont. αβ-cat. αβ-cat. a cont. 2835 β-cat. — LEF1 — LEF1 32P-oligo CM T93 SW480 colo320 SW948 colo205 DLD-1 158K 670K cont. β-cat. A.P. αβ-cat cont. Total Lysate β-cat. A.P. d αβ-cat c α-LEF1 15 17 19 21 23 44K 25 LEF1 Jurkat SW480 Jurkat SW480 Jurkat SW480 size fractionation (β-catenin) Figure 2 Association of LEF-1 with b-catenin. (a) Immunoprecipitates of b-catenin (ab-cat.) from the indicated colon cancer cell lines were analysed for b-catenin and LEF-1 by SDS ± PAGE and immunoblotting. A Glu-Glu antibody was used as a control (cont.). (b) The b-catenin (ab-cat) immunoprecipitates were resolved by SDS ± PAGE, transferred to a PVDF membrane and probed with a radiolabeled oligonucleotide containing the LEF-1 binding site (32P-oligo.). Following autoradiography, membranes were developed using an antibody to LEF-1 and visualized by enhanced chemiluminescence (a-LEF1). Glu-Glu antibody was used as a control (cont.). (c) Lysates from Jurkat and SW480 cells were analyzed directly for LEF-1 by immunoblotting (total lysate) or ®rst subjected to immunoprecipitation with antibody to b-catenin (ab-cat.), or anity precipitation using puri®ed Glu-Glu tagged bcatenin (b-cat. AP) or Glu-Glu tagged C3G protein as a control (Cont.). The puri®ed proteins were recovered on anti-Glu-Glu beads and analysed for associated LEF-1. (d) Jurkat and SW480 cell lysates were subjected to size exclusion chromatography and the fractions were analysed for b-catenin by SDS ± PAGE followed by immunoblotting. Column fraction numbers are shown at the top, arrows indicate the elution position of protein standards a monomeric pool observed on size fractionation chromatography (Munemitsu et al., 1995; Papko€ et al., 1996). This monomeric pool of b-catenin, which is present in colon cancer cells, can be induced by the wnt-1 proto-oncogene in nontransformed cells and is opposed by the APC tumor suppressor. Therefore, this pool of b-catenin may be indicative of a translocation event that allows the interaction of b-catenin with additional protein targets. Accordingly, we were unable to detect a free pool of b-catenin by size exclusion chromatography of Jurkat cell lysates (Figure 2d). By contrast, high levels of free b-catenin were detected in the SW480 cells as previously reported (Munemitsu et al., 1995). The inability to detect this form of b-catenin in Jurkat cells suggested that its subcellular distribution may prevent its interaction with LEF-1. That both b-catenin and LEF-1 are present in some cell lines, but not associated, suggests that special circumstances are required to a€ect their interaction. As wnt-1 is known to stabilize b-catenin and promote its distribution to cadherin-independent pools (Bradley et al., 1993; Hinck et al., 1994; Papko€ et al., 1996), we tested the ability of wnt-1 to promote the assembly of the b-catenin-LEF-1 complex. The embryonal kidney cell line 293 was chosen as we have noted that their transient transfection with wnt-1 stabilizes b-catenin and induces a free pool as determined by size exclusion chromatography (P Polakis, B Rubinfeld; unpublished data). No association of b-catenin with LEF-1 was detected in cells transfected with empty vector, although LEF-1 was anity precipitated from the cell lysates by the addition and recovery of puri®ed bcatenin (Figure 3a). By contrast, when 293 cells were transiently transfected with wnt-1, b-catenin formed a complex with LEF-1 as determined by their coimmunoprecipitation (Figure 3a). It was possible that wnt-1 induced the complex through a mechanism not involving the stabilization or distribution of b-catenin to new subcellular compartments. To test this, we transfected 293 cells with D89b-catenin, an N-terminal deletion mutant which is highly stable and localizes to free pools in the cell (Munemitsu et al., 1996). The D89b-catenin formed a complex with LEF-1 when expressed in the 293 cell (Figure 3a). This suggests that wnt-1 likely enables the association of endogenous bcatenin with LEF-1 by promoting b-catenin subcellular redistribution. It was also possible that wnt-1 induced the LEF-1-b-catenin complex by elevating the levels of LEF-1 expression. However, this was unlikely as the association of ectopically overexpressed LEF-1 with bcatenin was still dependent upon wnt-1 stimulation (Figure 3b). wnt-1 activation of b-catenin E Porfiri et al M yc-LEF1 +Wnt-1 vector cont. β-cat .A.P. αβ-cat. cont. β-cat .A.P. αβ-cat. cont. β-cat .A.P. M yc-LEF1 IP: anti-myc b αβ-cat. w nt-1 vector a ∆N89β-cat. 2836 β-cat. — LEF1 w nt-1 Vector Total lysate ∆N89β-cat. (+) w nt-1 cont. β-cat .A.P. αβ-cat. vector cont. αE-cad αAPC αβ-cat. αE-cad αAPC αβ-cat. d αβ-cat. c β-cat .A.P. M yc-LEF1 — LEF1— Vector w nt-1 (–) w nt –1 Total lysate (+) w nt –1 Total lysate Figure 3 Induction of LEF-b-catenin complex by wnt-1. (a) Transient transfections with empty vector or plasmids encoding wnt-1 or DN89b-catenin were performed on 293 cells and the lysates were analysed for total LEF-1 (total lysate) or subjected to immunoprecipitation (ab-cat.) or anity precipitation (b-cat. AP or cont.) as described in Figure 2. (b) Lysates from 293 cells transfected with vector only (vector) or cotransfected with mycLEF-1 and either wnt-1 or empty vector were subjected to immunoprecipitation using antibody to the myc epitope. Immunoprecipitates were analysed for LEF-1 and b-catenin by immunoblotting. (c) Lysates from PC12 cells carrying empty vector or stably expressing wnt-1 were analysed directly for LEF-1 (total lysate) or ®rst subjected to immunoprecipitation with antibodies to b-catenin, APC and E-cadherin. (d) Lysates from wild type PC12 cells grown on extracellular matrix derived from either wnt-1 expressing PC12 cells (+ wnt-1) or from control (7 wnt-1) cells were subjected to immunoprecipitation with antibody to b-catenin (ab-cat.), or anity precipitation with Glu-Glu b-catenin (b-cat. AP). The precipitates or total lysates were analysed for LEF-1 by immunoblotting b S45Yβ-cat. Vector S45Yβ-cat. S37Aβ-cat. w tβ-cat. Vector S37Aβ-cat. anti-m yc IP Total lysate w tβ-cat. a 116 — — b-cat. — 80 — anti-β-catenin anti-m yc tag LEF1 50 — anti-LEF1 Figure 4 Activation of LEF-1-dependent transcription by wild type and mutant b-catenins. (a) A plasmid expressing a LEF-1responsive reporter was cotransfected into 293 cells with either empty vector (vector) or plasmid expressing wild type (wt b-cat.) or mutant (S37Ab-cat., S45Yb-cat.) b-catenins. Reporter activation resulted in luciferase expression that was assayed in aliquots of lysate normalized for protein. The data represent the mean+s.d. of eight independent determinations. (b) Myc immunoblot of total lysates from 293 cells transfected with vector alone or myc tagged wild-type b-catenin or with the indicated mutant b-catenins (left). Protein-equivalent aliquots of the same lysates were subjected to immunoprecipitation with myc antibody and immunoprecipitates were analysed by immunoblotting for b-catenin and LEF-1 (right) wnt-1 activation of b-catenin E Porfiri et al 2837 To further demonstrate the induction of the bcatenin-LEF-1 complex by wnt-1, we employed the PC12 cell line in which b-catenin, and its homolog plakoglobin (Bradley et al., 1993), localize to new subcellular compartments in response to wnt-1. The bcatenin-LEF-1 complex was readily detected in PC12 cells stably expressing wnt-1, but not in those carrying empty vector (Figure 3c). Immunoprecipitates of two additional b-catenin binding proteins E-cadherin and APC did not contain LEF-1, demonstrating that they are not in the b-catenin-LEF-1 complex. The induction of the b-catenin-LEF-1 complex also occurred, albeit to a lesser degree, in parental PC12 cells when plated on an extracellular matrix derived from the wnt-1 expressing cells (Figure 3d). In the absence or presence of wnt-1, LEF-1 was anity precipitated by puri®ed bcatenin, but the endogenous proteins were only associated in cells stimulated with wnt-1. The above results suggest that the stabilization and redistribution of b-catenin may be required for its association with LEF-1. Recently, we have identi®ed missense mutants of b-catenin in melanoma cell lines that contain the substitutions ser37phe and ser45tyr (Rubinfeld et al., 1997). These b-catenins are highly stable, accumulate to high steady-state levels and are distributed throughout the cytoplasm and nucleus. To determine whether these mutants form a transcriptionally active complex with LEF-1, we tested their ability to activate a luciferase reporter plasmid containing the minimal T-cell receptor a enhancer placed upstream of the herpes virus thymidine kinase promoter. This enhancer contains an LEF1 binding site ¯anked on opposite sides by CRE and Ets-1 binding sites. In 293 cells, expression of luciferase activity was greatly enhanced by cotransfection of b-catenins with the reporter construct, and the mutant b-catenins were more e€ective than wildtype (Figure 4a). This was consistent with the higher steady-state levels of mutant b-catenin and the higher levels of LEF-1 that coimmunoprecipitated with the mutants relative to wildtype b-catenin (Figure 4b). Thus, the mutant bcatenins identi®ed in cancer cells can activate LEF-1dependent transactivation and, due to their increased stability, are stronger activators than wildtype bcatenin. Discussion There is considerable evidence now to implicate bcatenin as a contributing factor in cancer progression. Its stability is a€ected in opposing manners by the APC tumor suppressor and the wnt-1 oncogene (Munemitsu et al., 1995; Papko€ et al., 1996), and missense mutations that stabilize b-catenin have been identi®ed in cancer cells (Morin et al., 1997; Rubinfeld et al., 1997). Although these observations are compelling, how b-catenin alters cell growth remains to be determined. This could come about through its interaction with any of the known b-catenin binding proteins including cadherins (McCrea et al., 1991), fascin (Tao et al., 1996), the EGF receptor (Hoschuetzky et al., 1994), APC (Rubinfeld et al., 1993; Su et al., 1993) or the LEF/TCF transcription factors (Behrens et al., 1996; Huber et al., 1996; Molenaar et al., 1996). Finding b-catenin in the nucleus, functionally coupled to a transcription factor, makes the activation of growth controlling genes an attractive model for a b-catenin cancer mechanism. However, it is unclear which of the LEF/ TCF transcription factors represent the appropriate partner for b-catenin in this pathway. LEF-1, TCF-1, and TCF-4 all contain b-catenin binding sites in their amino-terminal regions (Behrens et al., 1996; Korinek et al., 1997). Based on its expression in colonic epithelium, and its interaction with b-catenin, TCF-4 has been proposed as a target for b-catenin in colon cancer (Korinek et al., 1997). Here we have shown that LEF-1 mRNA is expressed in colon cancer cell lines and that the protein is constitutively associated with bcatenin. Thus, LEF-1 also needs to be considered as a potential target for b-catenin signaling in cancer. We were unable to detect LEF-1 mRNA in normal colonic mucosa consistent with previous reports that LEF-1 expression is largely con®ned to lymphoid organs (Waterman et al., 1991). It is unclear how LEF-1 expression becomes reactivated in colon cancer, but this event may represent another component of tumor progression. The cDNA that we identi®ed in the colon cancer cells was identical to a naturally occurring splice variant previously identi®ed in Jurkat cells (Waterman et al., 1991). This alternatively spliced mRNA codes for an LEF-1 lacking amino acids 214-241. The signi®cance of this deletion has not been determined, but this sequence lies within a modular transactivation domain previously de®ned in LEF-1, and deletion of these amino acids impairs the activity of this domain (Carlsson et al., 1993). It is possible that the products of the LEF1 mRNA splice variants respond di€erentially to their interaction with b-catenin. Indeed, alternatively spliced forms of the TCF-1 mRNA result in proteins that either contain or lack the amino-terminal binding site for b-catenin (Van de Wetering et al., 1996). The identi®cation of the speci®c LEF/TCF forms responsible for oncogenic signaling through b-catenin will require a careful analysis of many candidates. Additionally, it is not clear in what context the LEF-1 DNA binding site must occur to receive the relevant b-catenin-dependent signal. Transactivation by LEF-1 does not occur in test systems employing only multimerized copies of the LEF-1 binding site, but requires adjacent binding sites for additional enhancer proteins (Carlsson et al., 1993; Giese and Grosschedl, 1993). Our results show that expression of b-catenin can promote transcription from the T-cell receptor a enhancer, however, the context of the bcatenin-LEF-1 binding site in colon cancer cells is unknown. Moreover, TCF-4 can activate transcription in a b-catenin-dependent manner without requiring ¯anking binding sites for additional transcription factors (Korinek et al., 1997). In the present study, the association of b-catenin with LEF-1 appeared to be dependent on the status of b-catenin in the cell. We have previously reported that the subcellular distribution and the oligomeric composition of b-catenin are in¯uenced by the APC tumor suppressor, the wnt-1 oncogene, and mutations in bcatenin itself (Munemitsu et al., 1995, 1996; Papko€ et al., 1996; Rubinfeld et al., 1997). The colon cancer cells containing constitutive LEF-1-b-catenin complexes all exhibited a free pool of b-catenin indicative of its wnt-1 activation of b-catenin E Porfiri et al 2838 distribution to noncytoskeletal compartments. In cells containing wildtype APC, the b-catenin can be induced to undergo a similar redistribution when stimulated by wnt-1 (Papko€ et al., 1996). Accordingly, wnt-1 also promoted the association of LEF-1 with b-catenin. Finally, in cells containing wildtype APC, speci®c point mutations or deletions in b-catenin will also promote its distribution into cytosolic and nuclear compartments (Munemitsu et al., 1996; Rubinfeld et al., 1997). Ectopic expression of these mutant b-catenins resulted in their association with LEF-1 and strongly promoted the transactivation of a reporter plasmid containing LEF-1 binding sites. Together the results indicate that the induction of b-catenin by a variety of mechanisms, all result in its interaction with the LEF-1 transcription factor. This interaction is also likely to occur in vivo as excessive staining of b-catenin in both the cytoplasm and nucleus of adenomatous colonic tissue has been observed (Inomata et al., 1996). Hence, the identification of target genes activated by b-catenin-LEF/TCF complexes should greatly extend our understanding of this cancer mechanism. Materials and methods Transfections, immunoprecipitations and immunoblotting All cell lines were from the American Type Culture Collection (ATCC). Human LEF-1 (Waterman et al., 1991) and murine wnt-1 (Fung et al., 1985) cDNAs were subcloned into the pCDNA3 (Invitrogen) expression vector containing the Myc epitope tag. The b-catenin ser45tyr mutant was generated by fragment switching between wild type b-catenin cDNA and b-catenin cDNA isolated from a melanoma cell line containing the ser45tyr mutant (Rubinfeld et al., 1997). Transient transfection of 293 cells was carried out using a modi®cation of the calcium phosphate precipitation method (Munemitsu et al., 1994). For the derivation of stable PC12 cell lines, cells were transfected by electroporation at 500 mF and 230 mV, and grown for 3 days in complete medium before selection in G418 (Loeb and Greene, 1993). Wnt-1-conditioned extracellular matrix was generated by growing PC12 cells stably expressing wnt-1 on collagen type IV coated dishes for 2 ± 3 days. Cells were scraped and the dishes used to culture wild type PC12 cells. Under these conditions, bcatenin was stabilized and distributed to the cytoplasm in the cells grown on the wnt-1 conditioned extracellular matrix (E Por®ri, P Polakis, unpublished data). For immunoprecipitation and immunoblotting, the cells were lysed in Triton X-100 bu€er as described (Rubinfeld et al., 1996). Approximately 1 mg of total cell protein was used for each immunoprecipitation or for anity precipitations using 2 mg of Glu-Glu tagged puri®ed b-catenin or GluGlu tagged C3G protein as control. The polyclonal antibodies to LEF-1 (Waterman et al., 1991) and b- catenin (Rubinfeld et al., 1996) used for immunoblotting (0.2 mg/ml) have been described. The monoclonal antibodies used to immunoprecipitate b-catenin and cadherin were purchased from Transduction Laboratories, those to the Myc and Glu-Glu epitope tag have been described (Evan et al., 1985; Grussenmyer et al., 1985). Northern and Southwestern analysis Northern analysis was carried out as described (Waterman et al., 1991). RNA was puri®ed from the indicated cell lines using the RNeasy kit (Qiagen). Ten mg of total RNA were fractionated on a 1.2% agarose formaldehyde gel and transferred to a Hybond N membrane (Amersham). Blots were probed with a radiolabeled DNA probe representing most of the LEF-1 ORF (bp 821 ± 1824 of LEF-1 cDNA). Southwestern analysis was performed essentially as described (Waterman and Jones, 1990). bcatenin immunoprecipitates were resolved by SDS ± PAGE (Novex) and transferred to a PVDF membrane. Membranes were equilibrated in binding bu€er (20 m M HEPES, pH 7.9, 60 mM KCl, 1 mM DTT, 10% glycerol, 0.01% NP-40) for 10 min and blocked for a further 10 min in 5% non-fat milk in binding bu€er containing 4 mg/ml salmon testes DNA. Blots were rinsed and incubated for 16 h at 48C with 107 c.p.m. of a 32P-labeled double stranded oligonucleotide containing a LEF-1 binding site (GATCTAGGGCACCCTTTGAAGCTCT)10 in 3 ml of binding bu€er containing 4 mg/ml salmon testes DNA. Polynucleotide kinase and 32P-g-ATP was used to label the probe. Following hybridization, blots were washed and autoradiographed. After autoradiography blots were incubated with anti-LEF-1 antibody and developed using enhanced chemiluminescence. For isolation of LEF-1 cDNAs, a cDNA pool was ®rst obtained by reverse transcription of total mRNA (RNeasy kit, Qiagen) using random primers. PCR was then performed on the cDNA pool using a primer set corresponding to bp 1 ± 18 and 1178 ± 1200 of LEF-1 open reading frame, and the PCR products were cloned into pCR2.1 (Invitrogen) and propagated in E. coli. Luciferase assay For luciferase assays, the reporter construct a2tk (1 mg), which contained two copies of the TCRa enhancer upstream of the TATA box of the HSV-1 tymidine kinase promoter and luciferase gene (Waterman and Jones, 1990) was cotransfected with the indicated plasmid (1 mg) into 293 cells. Cells were lysed and luciferase activity was measured in a luminometer using the Luciferase Assay System kit (Promega) and normalized to the protein concentration. Acknowledgements We thank David Lowe for production of recombinant Sf9 cells and Jackie Papko€ for the plasmid expressing wnt-1. References Behrens J, von Kries JP, Kuhl M, Bruhn L, Wedlich D, Grosschedl R and Birchmeier W. (1996). Nature, 382, 638 ± 642. Bradley RS, Cowin P and Brown AMC. (1993). J. Cell. Biol., 123, 1857 ± 1865. Brunner E, Peter O, Schweizer L and Basler K. (1997). Nature, 385, 829 ± 833. Carlsson P, Waterman ML and Jones KA. (1993). Genes Dev., 7, 2418 ± 2430. Evan G, Lewis G, Ramsey G and Bishop J. (1985). Mol. Cell. Biol., 5, 3610. Fung YK, Shackleford GM, Brown AM, Sanders GS and Varmus HE. (1985). Mol. Cel. Biol., 5, 3337 ± 3344. Giese K and Grosschedl R. (1993). EMBO J., 12, 4667 ± 4676. Grussenmyer T, Scheidtmann KH, Hutchinson MA and Walter G. (1985). Proc. Natl. Acad. Sci. USA, 82, 7952 ± 7954. wnt-1 activation of b-catenin E Porfiri et al 2839 Gumbiner BM. (1995). Curr. Opin. Cell Biol., 7, 634 ± 640. Hinck L, Nelson WJ and Papko€ J. (1994). J. Cell Biol., 124, 729 ± 741. Hoschuetzky H, Aberle H and Kemler R. (1994). J. Cell. Biol., 127, 1375 ± 1380. Huber O, Korn R, McLaughlin J, Ohsugi M, Herrmann BG and Kemler R. (1996). Mech. Dev., 59, 3 ± 10. Inomata M, Ochiai A, Akimoto S, Kitano S and Hirohashi S. (1996). Cancer Res., 56, 2213 ± 2217. Korinek V, Barker N, Morin PJ, van Wichen D, de Weger R, Kinzler KW, Vogelstein B and Clevers H. (1997). Science, 275, 1784 ± 1787. Loeb DM and Greene LA. (1993). J. Neurosci., 13, 2919 ± 2929. McCrea PD, Turck CW and Gumbiner B. (1991). Science, 254, 1359 ± 1361. Molenaar M, van de Wetering M, Oosterwegel M, PatersonMaduro J, Godsave S, Korinek V, Roose J, Destree O and Clevers H. (1996). Cell, 86, 391 ± 399. Morin PJ, Sparks AB, Korinek V, Barker N, Clevers H, Vogelstein B and Kinzler KW. (1997). Science, 275, 1787 ± 1790. Munemitsu S, Albert I, Rubinfeld B and Polakis P. (1996). Mol. Cell. Biol., 16, 4088 ± 4094. Munemitsu S, Albert I, Souza B, Rubinfeld B and Polakis P. (1995). Proc. Natl. Acad. Sci. USA, 92, 3046 ± 3050. Munemitsu S, Souza B, Muller O, Albert I, Rubinfeld B and Polakis P. (1994). Cancer Res., 54, 3676 ± 3681. Papko€ J, Rubinfeld B, Schryver B and Polakis P. (1996). Mol. Cell. Biol., 16, 2128 ± 2134. Pai LM, Orsulic S, Bejsovec A and Peifer M. (1997). Development, 124, 2255 ± 2266. Peifer M. (1995). Trends Cell. Biol., 5, 224 ± 229. Rubinfeld B, Albert I, Por®ri E, Fiol C, Munemitsu S and Polakis P. (1996). Science, 272, 1023 ± 1026. Rubinfeld B, Robbins, P, El-Gamil M, Albert I, Por®ri P and Polakis P. (1997). Science, 275, 1790 ± 1792. Rubinfeld B, Souza B, Albert I, Muller O, Chamberlain SC, Masiarz F, Munemitsu S and Polakis P. (1993). Science, 262, 1731 ± 1734. Su L-K, Vogelstein B and Kinzler KW. (1993). Science, 262, 1734 ± 1737. Tao YS, Edwards RA, Tubb B, Wang S, Bryan J and McCrea PD. (1996). J. Cell. Biol., 134, 1271 ± 1281. Travis A, Amsterdam A, Belanger C and Grosschedl R. (1991). Genes Dev., 5, 880 ± 894. Tsukamoto AS, Grosschedl R, Guzman RC, Parslow T and Varmus HE. (1988). Cell, 55, 619 ± 625. Van de Wetering M, Castrop J, Korinek V and Clevers H. (1996). Mol. Cell. Biol., 16, 745 ± 752. van de Wetering M, Cavallo R, Dooijes D, van Beest M, van Es J, Loureiro J, Ypma A, Hursh D, Jones T, Bejsovec A, Peifer M, Mortin M and Clevers H. (1997). Cell, 88, 789 ± 799. Waterman ML, Fischer WH and Jones KA. (1991). Genes Dev., 5, 656 ± 669. Waterman ML and Jones KA. (1990). New Biol., 2, 621 ± 636. Whitehead I, Kirk H and Kay R. (1995). Mol. Cell. Biol., 15, 704 ± 710. Yost C, Torres M, Miller JR, Huang E, Kimelman D and Moon RT. (1996) Genes & Dev, 10, 1443 ± 1454.